Bromcom Computers Plc v United Learning Trust & Anor

Table of Contents

In these consolidated proceedings, the Claimant, Bromcom Computers Plc (“Bromcom”) has brought a procurement challenge against the First Defendant, United Learning Trust (“UL”) alleging breach by it of The Public Contracts Regulations 2015 (“the PCR”). It is a claim for damages only and this trial has been concerned with liability and causation, but not quantum.

The procurement in question was conducted through a “competitive dialogue” procedure under the PCR. It concerned the award by UL of a contract for the supply of a cloud-based Management Information System (“MIS”) over a 5 year period to 57 State Academy schools within UL. Of the 57, 23 were primary schools and the remaining 34 were secondary or “all through” schools. The contract was valued at £2 million.

UL itself is an educational charity that managed 75 state academies as at 31 October 2019, being the date of the publication of the relevant contract award notice (“the Notice”). The Notice was published by the second Defendant, United Church Schools Trust (“UCST”) as the contracting authority. This was in fact an error because UCST only manages the 14 independent academies within the United Learning group of charities. Nothing turns on this and it is common ground that the correct defendant is UL.

After an initial bid process, two companies were shortlisted for the final stage of the procurement exercise. One was Arbor Education Partners Limited (“Arbor”). The other was Bromcom. Arbor won by a narrow margin and it was awarded the contract.

Bromcom alleges that there were numerous breaches of the PCR relating to the scoring of the bids of both parties. It also alleges that Arbor’s final bid submission was itself defective and should have been rejected. Further, it alleges that absent the breaches, Bromcom would have won, not Arbor. It says that these breaches were sufficiently serious to justify a damages claim.

UL denies all of these allegations and also raises a limitation defence which Bromcom, in turn, denies. A point is also taken about the fact that Bromcom covertly recorded a number of telephone calls with UL.

Arbor was already the incumbent provider of MISs to 15 of UL’s academies which themselves were not part of the procurement. Part of Bromcom’s case is that the scoring process undertaken by UL was improperly influenced by Arbor’s role as an existing provider to those schools.

For Bromcom, I heard from the following witnesses:

For UL, I heard from the following witnesses:

All of the above UL witnesses acted as evaluators in at least some of the scoring exercises.

There were no witnesses from Arbor.

In addition, there is a substantial body of relevant contemporaneous documentary evidence including, of course, the documents which constituted or evidenced the scoring process itself.

In order to understand the issues and much of the evidence, it is necessary to explain how the procurement process here worked, how the scoring was done and how the eventual scores were calculated. I set this out below.

As set out in the Notice, UL required the provision and implementation of software on a “software as a service” basis. It stated this at paragraph II.2.4:

“The Trusts requirement is to provide and implement a software-as-a-service (SaaS) based Management Information System (MIS) in order to deliver the Trust’s ‘Shift to the Cloud’ strategy. The MIS will be accessed over the internet and hosted in the cloud. The product must be capable of supporting the operational, educational management and administrative needs of the trusts state sector primary, secondary and all-through academies.”

The email address of the contracting authority and also the review body was given as UL-MIS-Procurement@UnitedLearning.onmicrosoft.com.

On 12 December 2019, UL issued the Invitation to Participate in Dialogue (“the ITPD”) to the then bidders in play, being Arbor, Bromcom and Capita (the latter itself being the incumbent supplier of a non-cloud-based MIS system to the 57 schools). These had been the top 3 bidders following the submission of an Initial Questionnaire after which the others were sifted out. The ITPD contained detailed information about UL’s needs and the procurement process going forward. In particular, by noon on 29 January 2020, the bidders had to submit their ITPD Response known as the “initial solution”. This was divided into the further groups of documents:

It should be noted that these documents were used (as amended), throughout the entire procurement process and they were all evaluated one way or another.

Paragraph 5.3 of the ITPD contained further information in relation to Appendix D and how it would be marked, and paragraph 5.4 listed the services to be provided over the 5 year period, on the basis of fixed costs, exclusive of VAT.

Paragraph 6 stated that the proposals would be evaluated by Dominic Norrish, UL’s Chief Operating Officer, Mr Wilson, a procurement adviser from the Novatia Consultancy (ie Ms Wood) and various representatives from the academies. Their names were given. Those names were all of the witnesses I have referred to in paragraph 9 above with the exception of Ms Dzioba, Mr Rowley and Mr Sharman and with the addition of Deborah Smith. At this stage, the percentage allocation of scores out of a total of 100% was 40% to cost and 60% to quality. The latter was then divided between Product Demonstration, Written Responses (i.e. Appendices A-C) and MIS Specification with each scoring 20%.

The second and final stage of the process would follow the Invitation to Continue Dialogue (“ITCD”) which would be issued to one or more of the 3 bidders following assessment at the ITPD stage.

There was an initial product demonstration by webinar, on 13 December 2019. There was also a first competitive dialogue meeting (CD1) which took place on 8 January 2020 with a feedback webinar meeting on 15 January 2020. On 29 January 2020, the three bidders submitted their ITPD documents as requested.

On 21 February 2020, UL released the ITCD document which indicated that following the sift after the ITPD responses, Arbor and Bromcom were shortlisted, with Capita being eliminated. There is no challenge to the scoring behind this decision. The ITCD prescribed the following events:

On the same day, UL also provided to both bidders a document setting out its responses to requests for clarification made by Bromcom, largely in respect of the ITPD Appendices ("the February Clarifications Document"). The covering email to Bromcom included a separate single clarification response to a clarification request from Bromcom which applied solely to it.

The deadline for the ITCD responses was 20 March. No particular time of day was stipulated. These responses would involve an effective resubmission of all of the Appendices produced at the ITPD stage which now would be scored as part of final tender. In addition, however, there would be two new Appendices: Appendix F, being “Integration with 3^rd Party Products”, and Appendix G, being “Timetable and Flexible Working Agenda”. Section 4 of the ITCD set out a number of rules. First, the participants could seek clarification from UL, but UL was not obliged to respond to any such request which had been received after 16 March. Similarly, UL could seek clarification from either participant in relation to its bid at any time. In assessing the Final Tenders, UL could take into account any information provided by a bidder by way of clarification which was contained in a document or had been set out at a clarification meeting. In addition, paragraph 4.7, headed “Disqualification”, stated that any breach of the requirements of Section 4 by a participant would entitle UL to disqualify that participant. I shall refer further to the provisions of the ITCD in context, below.

On 4 March, UL issued a document called “Revised Program & CPD Guidance for ITCD”. I shall refer to it as “the March Guidance” without any implication as to its meaning or status. On 10 March UL issued a document headed “ITCD Clarification Questions to Participants”. It had 6 general questions for both bidders, and then some specific questions addressed to each of them concerning their respective terms and conditions (“the March Clarification Request”). I shall refer to the bidders responses to the March Clarification Request as their March Clarification Response.

On 10 March, revised documents for the ITCD responses were provided by UL.

The precise timing and mode of Arbor’s and Bromcom’s final tender submissions on 20 March need to be set out in considerable detail, as elements of Bromcom’s challenge relate to them.

This was sent by email at 5:55pm by Ms Marsh. It contained a password protected zip file containing all the relevant documents. It is not suggested that it was incomplete in any way. The password was then sent by separate email at 6.01pm. Mr Wilson uploaded Bromcom’s documents into a folder in UL’s SharePoint platform at 6.08pm.

Earlier that day, Ms Marsh had a telephone conversation with Mr Wilson about the actual deadline, since no time had been stipulated in the ITCD. According to Ms Marsh, Mr Wilson said that close of business was “fine” and that there was “no time on this one”. She understood that any time up to midnight would be acceptable. For his part, he agreed in evidence that it could be any time up to 23.59pm, and he would have accepted it up to midnight.

At the time, UL did not question the timing of Bromcom’s submission following its receipt.

Here, the position is somewhat more complex and controversial on the facts. At 3:26pm on 20 March, Ms Jones of Arbor sent an email to Mr Wilson and Ms Wood saying that “You can download our final ITCD submissions at this link:… Please could you confirm receipt once you have received everything.” In other words, the submissions were not attached to the email itself, but were to be accessed at a drop-box hosted by Arbor, and at all times accessible by it.

Mr Wilson’s evidence was that he had downloaded the documents from the drop-box at 3:30pm, in other words almost immediately after Arbor’s email had arrived. There was no password required here. Mr Wilson placed all of the downloaded files into UL’s SharePoint platform. In the course of uploading Arbor’s documents, he noticed that Arbor’s March Clarification Response was not there.

On 23 March, Mr Wilson had a catch-up call with Ms Jones. He mentioned the missing March Clarification Request response. This was then provided by Ms Jones by email later that day at 4.47pm, along with a revised version of Arbor’s “Uptime Statistics” sheet, which had originally been provided as a supplemental spreadsheet to its Appendix A. The email said:

“…I've attached our answers to your clarification questions - apologies for missing them and thank you for flagging it to us. Whilst going through the answers I noticed a couple of oversights in our uptime spreadsheet (our current SLA % and one of the column headings were inaccurate), so I've updated that in the Box folder and attached the corrected version here as well.”

The differences between the original and amended uptime spreadsheets were twofold. First, the original version included the sentence "Our standing contractual commitment to uptime is xxxxaveraged over the whole year", while the subsequent version read: "Our standing contractual commitment to uptime is xxxx averaged over the whole year". Second, the heading of the table on the right-hand side of the document in the revised version read "System Availability (Rolling 12 Months)." On the original version, only part of the heading was legible. The legible part was "mmitment for all our schools to better reflect the actual service levels schools can ex."

At 10:28am on 23 March, Mr Wilson sent out both submissions to the evaluators with template score sheets. On 24 March, he sent out Arbor’s replacement uptime spreadsheet to the evaluators under cover of this email:

“…Please note:

There is a document in the Arbor response which is referenced from Appendix A around uptime. There was an error in the version that was shared with you. This has now been corrected. I've highlighted this via the file name in the updated version and deleted the incorrect one.”

Over the next few days, other documents were sent to the evaluators which I do not need to refer to in detail at this stage.

However, in response to emails from two evaluators, asking whether they could use information gathered at previous meetings as well as the ITCD responses, Mr Wilson emailed on 25 March to say:

“Absolutely fine to use both sets of evidence… We should discuss during moderation on Friday any additional or higher score you’ve ended up on which isn’t reflective of just what is written in bid responses. Hope this makes sense.”

I refer further to this matter below, since Bromcom complains about it.

Later on 25 March, Mr Wilson sent through to Ms Wood a spreadsheet entitled “UL MIS ITCD Scoring Results”. By this stage, almost all of the evaluators had produced their scores for all the various written responses. Mr Wilson observed that “I’ve had a go at moderating both tabs [i.e. Product and Quality] to see how the land lies - bottom-line - very close on quality so it would seem cost would make the difference.” In a second email, he added that “forgot to say - I’ve adjusted the weighting on the written responses slightly. In other words these were weightings for the scoring of Appendices A-C, F and G. The weightings applied by Mr Wilson were as follows:

The ultimate scores were going to be out of a maximum of 1000 points of which (consistently with the ITPD) 400 were allocated to Cost and 600 to Quality. Of the latter, 240 went to Product Score (comprising Product Evaluation and the MIS Specification) and the balance of 360 went to Appendices A-C, F and G. Converting the percentage weightings into points, this therefore entailed the following maximum points for each Appendix:

The actual scoring here had in fact been done by 3 March, following CD3. There were, in the end, 13 evaluators. Each had to give a single score out of 5 in respect of four separate elements, Portal & Comms, Exams, Reporting and Cover, for each of Arbor and Bromcom. See the Correlated Scores document sent by Mr Wilson to Ms Wood on 3 March. This document set out the total number of points scored by Arbor and Bromcom respectively, for each element. The average score for each element was also shown. However, by the time of the moderation on 27 March, the Product Scores were calculated thus: first, the average score for each particular evaluator across the 4 elements was calculated. Then, an average was taken of all of those averages, as it were. So, as can be seen in one of the slides for the moderation, Arbor had scored an average of 4.12 out of 5 and Bromcom had scored 4.08. In each case, this had to be translated into marks out of 240 which represented the 40% of the 600 points available for Quality. Thus translated, Arbor scored 198 while Bromcom scored 196 (rounded figures). I should add that ultimately, the respective scores became 202 and 193. The reason for the difference was that UL removed the scores of 4 evaluators, being Ms McKnight, Mr Furnival, Ms Viner, and Mr Nield. This was because not all of them had attended all parts of the CD2 and CD3 product demonstrations for both bidders (see p1430 of the Trial Bundle). The removal of their scores (and the averages thereof) then affected the average of the averages, as it were and so the final marks out of 240. There is one aspect of the Product Score which is challenged by Bromcom and is dealt with below.

In fact, the score of 202 for Arbor was overstated. It should have been 201. This calculation error is admitted (by reference to an overstatement of Arbor’s whole quality score) in paragraph 56C of the Defence.

As for the other Quality elements, there were 12 evaluators for each Appendix, save for Appendix B. Here, there were only 6 evaluators due to time constraints and other problems. Of those 6, 3 had participated in the scoring of the other Appendices and 3 had not i.e. they were “new”.

For Appendices A-C and F-G, scoring was done as follows: first, individual scores out of 5 were given for each Appendix by each evaluator, then recorded, and then the average score across all the evaluators for each Appendix was calculated down to 2 decimal points, again, out of 5. That average score was then translated into the appropriate number of points out of the maximum weighted score attributed to that Appendix. So, for example, under Appendix A, Arbor scored an average of 3.92 out of 5. As applied to the total score of 90 it would yield (after rounding) a score of 71. Bromcom, on the other hand, had an average of only 3.75 which then translated into 68 points. Again, all of this can be seen in the slides for the moderation.

Finally, the cost element, reflected in the bidders’ Appendix E together with associated documents, was scored. This was a different process because once the total specified costs were input, a total cost for each bidder’s Static and Growth models would be arrived at mathematically. Both bids were marked out of the available 400 points. The better bidder i.e. the one with the lower total cost figure, was awarded the full 400 points. The other bid was scored in this way: the difference in cost was expressed as a percentage of the lower bid cost. That percentage was then applied to 400 and those points were then deducted from 400 so as to give the bidder with the higher cost a lower score. Only the Growth Model was actually marked for these purposes.

Arbor’s total costs were higher than Bromcom’s so the latter scored 400 points. Arbor’s score was something less, namely 380. However it is common ground that this was mathematically incorrect. The correct figure was 376. In addition, the total quality score for Bromcom should have been 465, not 464. Taking these into account, the total points awarded should have been 874 for Arbor and 865 for Bromcom. Unless these were changed at moderation, therefore, Arbor would win by 9 points.

All of the scores were then expressed in a final scoresheet for the purpose of the moderation, as follows (the word Avan should read Avail.): [SCORESHEET REDACTED]

A moderation meeting took place on 27 March. The scores were altered but only marginally. As already noted, the Product Scores were altered, with the result that Arbor now scored 202 and Bromcom, 193.

The other Quality Scores remained as before. So did the Costs Scores but with the incorrect figure of 380 remaining.

It followed that Arbor had been successful and Bromcom had not.

On 30 March, Mr Wilson emailed Bromcom to say that it had not been successful. On 31 March, he sent a letter, showing that Arbor (not identified) had scored 878 and Bromcom, 862. The difference between those total scores and those shown in the image at paragraph 47 above is explained by the different Product Scores referred to in paragraph 48 above.

There were then two feedback calls with Bromcom on 1 and 3 April. In between them, on 2 April, Mr Wilson emailed Bromcom, explaining the methodology used and the weighting applied.

On 14 April, Bromcom’s solicitors sought clarification from UL on a number of matters.

On 16 April, Bromcom had an internal review meeting to see why it could not have done better. This was preceded by an email from Elizabeth Morgan to Mr Ali Guryel, making three particular points.

On 22 April, a further formal letter was sent by UL to Bromcom in place of the 31 March letter and for the purposes of Regulation 86 of the PCR. The error in scoring 380 and not 376 for Arbor on costs was picked up and the final scores thus became 874 and 862. On 23 April, UL provided answers to particular points for clarification which had been raised by Bromcom on 14 April.

There was further correspondence between the solicitors on 24 and 28 April, and on 1, 7, 12, 13, and 15 May.

On 18 May 2020, Bromcom issued the Claim Form against UL in action numbered HT-2020-000183, and Particulars of Claim followed on 22 May (“the First Particulars of Claim”). These were then amended on 17 March 2021 (“the Amended First Particulars of Claim”). What then happened was that on 11 August 2021, Bromcom served draft Re-Amended First Particulars of claim dealing with allegations which it wished to make following disclosure. Because it was concerned with timing it decided not to wait until UL had consented or otherwise; instead, on 11 August 2021 it issued a new Claim Form in action numbered HT-2021-00301 and a new Particulars of Claim (“the Second Particulars of Claim”) was served on 18 August. Subsequently, these 2 actions were consolidated leading to the production of the Consolidated Particulars of Claim on 9 November 2021. These were subsequently amended on 21 March 2022 (“the Amended Consolidated Particulars of Claim”). There were, of course, Defences and Replies, following these statements of case.

I will deal with further factual matters in the context of my discussion of the particular issues.

As refined in the course of written and oral submissions, the following issues arise for determination:

I propose to deal with the Form of Submission and the Timing of Submission Issues after addressing all the scoring breaches..

The general principles in relation to these forms of procurement challenges were helpfully set out by Morgan J in the oft-cited case of Lion Apparel v Firebuy [2007] EWHC 2179 (Ch) as follows:

In addition, the test to be applied in construing tender documents is that one should adopt the construction which would be adopted by a reasonably well-informed and normally diligent (“RWIND”) tenderer: see Healthcare at Home Ltd v The Common Services Agency [2014] PTSR 1081 at paragraphs 8-12.

I shall deal with other points of law as they arise in the context of particular challenges.

Bromcom claims first that two items of cost were added to its submitted costs by UL wrongly, amounting to a manifest error (“the Bromcom Costs Adjustments”). Second, it contends that certain costs should have been added to Arbor’s costs submissions, and that certain discounts included by Arbor should have been removed. The failure to take these steps is said also to amount to manifest error (“the Arbor Costs Adjustments”). Had these errors not been made, then the gap between Bromcom’s and Arbor’s costs would have been greater leading to Arbor having a score somewhere below 376, which would have affected the final outcome.

It is common ground that the bidders had to price for the cost of F2F training on the MIS for the existing 57 schools. Both did so in their Static Costs model. However, while Arbor also costed for F2F training in its Growth Model, Bromcom did not and of course it was the Growth Model which was evaluated. Rather, in its Appendix E it stated as follows for each year after year one against the rubric “All Training Costs”:

“ we will waive the core training costs as incentive for any new schools to be onboarded on the basis that Webinar training is provided. On-site training can be provided at the special £500 per day rate offered to United Learning”

As for Arbor, it populated each of the year columns in its Appendix E with specific costs and making specific reference to the new schools being added each year.

What UL then did was to calculate Bromcom’s costs of F2F training for the additional 5 schools per year at its day rate of £500. This yielded a total additional cost of £42,000, and there is no issue over the arithmetic. Bromcom contends that there was no basis for UL doing that and the £42,000 should be notionally removed.

Bromcom contends first that the bidders were not in fact required to cost for F2F (or any) training for the growth schools in the first place. Given that the Growth Model was different from the Static Model because it included costs for extra schools in Years 2-5, that is an odd proposition. Moreover, Appendix E itself had columns for training costs to be added for Years 2-5. So it invited figures to be put in the Growth Model in those columns which had obviously been left blank by both parties in their Static Model. But as already noted, Bromcom did not put specific figures in the Growth Model. It obviously considered that there was a need to say something about growth schools, however, otherwise it would have left the cells entirely blank. Indeed, on its case that there was no requirement to quote for training for growth schools at all, the latter would be the logical thing to do.

One point relied upon by Bromcom is that it says that Appendix A was devoted entirely to the question of migration by the existing 57 schools only to the new MIS and therefore this shows that there was no reason to provide F2F training costs in Appendix E. I do not accept this. I agree that Appendix A is headed “Managing the change of systems at school level” and that it required a method statement outlining the support to be given to individual schools moving from one system to another. But I do not think that this is tantamount to saying that there was no need to quote for growth schools in Appendix E. If Ms Morgan of Bromcom thought this was the case, she was, objectively speaking, mistaken.

If one examines Bromcom’s entire response to Appendix A, moreover, it is obvious that it would apply to any school joining up to its MIS - see Core Bundle pages 445-450. Equally, I consider that the 7^th bullet point in Appendix A should not be read as limited to existing schools only.

A further argument made is that in any event, UL did not have the power to add costs onto Bromcom’s bid in the way that it did. In fact, if that proposition was correct, it would undermine Bromcom’s later challenge in respect of the costs it says should have been added to Arbor’s bid (see below). However, the substantive point is that I cannot see why UL did not have this power if exercised so as to effect equal treatment between the bids. I agree that no such general power is expressly given in the ITCD, but I do not think that this rules out an implied power. Equally, I do not think that the express intimation of adding legacy costs in the context of mobilisation (referred to at paragraph 8 of the Mobilisation Guidance) which was by its very nature limited to the 57 schools, is inconsistent with such an implied power.

I then need to consider paragraphs 8 and 13 of the February Clarification Document. Together with the originating questions, these state as follows:

“8.

[Q] What mechanism will the Trust used to evaluate total cost of ownership over the 5-year contract?

[A] In order to ensure a fair evaluation of cost, the Trust will evaluate the total cost of MIS to the Trust over the 5-year contract. We recognise that Bidders may propose different programs of implementation and this may have an impact on cost, particularly where there is an overlap of fees for existing provider contracts with a new providers charging model. A further TCO Worksheet tab has been added to Appendix E-Cost Model to provide a structure for provision of supplementary cost information with instructions provided for its completion. For the purpose of tender, Bidders should assume that the Contract Start Date is 1 May 2020 and the Contract End Date is 30 April 2025. The total cost for tender purposes will therefore include your proposal costs as scheduled, plus any incumbent costs incurred over the full 5-year term from 5/1/22 - 4/30/25…

13.

[Q] Having noted the award criteria to be used for the evaluation of Initial Solutions, can you clarify the formula sitting behind the cost evaluation figure to indicate how it would be applied?

[A] Cost is worth 40%... We will validate that each proposal is inclusive of all costs required for delivery of the contract over 5 years. The Total Overall Cost (Growth Model) over 5 years will be used for the evaluation. Please see Clarification #8 for further detail about the approach to the TCO calculation…”

These two paragraphs should be read together, in my view. I agree that paragraph 8 makes a particular point about incumbent costs which could differ as a result of the bidders’ different mobilisation plans, such that a new TCO (total costs of ownership) worksheet was added in respect of this and other additional costs. However, paragraph 8 also makes clear that UL would evaluate/validate the total cost over the 5 year period.

Bromcom appears to accept that some sort of costs adjustment by UL was contemplated and permitted but this was only in relation to mobilisation and incumbent provider costs. So it is not as if Bromcom alleges that UL had no power to add or change costs figures at all. And it is true that for mobilisation costs, these were necessarily limited to the 57 schools – obviously, since it could not possibly be known in advance what the incumbent provider’s position was in relation to any school, then unknown, which might later join UL. Unsurprisingly, therefore, the cells to be completed in the TCO worksheet itself were limited to the existing schools. However, I disagree that UL was somehow limiting its power to ensure equal treatment to that situation alone. The reference to growth schools and validating that each costs proposal is inclusive of all costs bears this out.

In fact, section 5.4 of the ITCD makes plain that the Growth Model is to be used. It is hard to see why that should somehow exclude quoting costs for growth schools. It is correct that the penultimate paragraph within section 5.4 provides that the Growth Model quotation would allow the bidders to consider applying volume discounts in line with an agreed growth assumption. However, the last paragraph then says that for the purpose of the tender, the bidders would be asked to quote based on add an additional 5 schools joining the Trust per annum with growth projections being provided within the Appendix. I do not agree that this meant that growth schools costs would be limited to whatever (discounted) charges might apply on the assumption that they join. I say that, even though Ms Wood at one point said she thought that was to enable bidders to consider how they might apply discounts. In any event, it is obvious that UL wanted to see a complete picture, as far as possible, in relation to growth schools.

In my judgment, the various documents read as a whole by a RWIND, clearly show a need to put in the costs of F2F for all growth schools as well as existing schools. The proof of that, in my view, is that Bromcom actually did quote on that basis but did not perfect the quote. Using Bromcom’s figures for F2F training costs, all UL really did was to perfect it.

Of course, it is possible that one or more growth schools might already be using, for example, a Bromcom MIS such that their training needs might be less than otherwise. But I do not consider that this possibility means that it was unfair or irrational for UL to have used the quoted day rate applied across all the assumed growth schools, in its evaluative judgment. And while there was some suggestion that had UL reverted to Bromcom on this issue before the submission of the final bids, Bromcom might have quoted a lower rate, I do not consider that there was any real evidence to this effect. Anyway, we know what Bromcom’s “special rate” was - it was the £500 per day.

Both sides then rely on what various witnesses said in relation to this matter. Both rely on what Ms Morgan and Mr Lewin said. Bromcom also relies on parts of Ms Wood’s evidence for example that Appendix E did not require bidders to provide F2F training for growth schools. Actually she did not say that - she accepted that “training” in Appendix E was not defined (see Day 5/132). The lack of a detailed definition here goes nowhere, given what was said about F2F training in Appendix A. In any event, I did not find much of this witness evidence helpful since what the witnesses subjectively thought about the interpretation of the relevant documents is immaterial.

Bromcom also argues that even if it should have fully costed the F2F training, which would have led to a different price, once UL appreciated that it had not costed it, UL should have sought clarification from Bromcom and then the additional price would or might have emerged (obviously, Bromcom does not suggest that it would not have quoted at all). But I do not see any need on the part of UL to seek such clarification when the basis for the figure to be calculated was already there. Nor do I think there was a real prospect of Bromcom coming up with some lesser figure. There was a suggestion in the course of answering questions from me by Mr Lewin that this might have been considered, but the gist of his evidence here (Day 3/86-88) was that any clarification would have served to make clear that F2F training for growth schools was required. Even so, there is then UL’s mathematical calculation on the basis of the quoted rates. So I did not think his evidence really went anywhere.

Finally, it seems to me that the only course that would have been open to UL, other than adding costs to Bromcom’s quote, would be to disqualify it for making a non-compliant bid. But against that drastic remedy, the course in fact taken by UL seems to me to be both fair and rational and did not amount to a manifest error.

For the above reasons, this element of the Price Scoring challenge fails.

This is what became referred to at trial as the “push/pull” issue. Somehow, the data accumulated from schools by each of the bidders MIS’s needed, ultimately, to reside in UL’s own data warehouse. This could be done in two ways. Either UL could, with appropriate software, “pull” it into its data warehouse which would have a cost, or the bidder could “push it” there in which case UL would incur no cost.

In its Appendix B, Arbor stated that there would be zero cost for setting up and maintaining the interface with the data warehouse, describing that is “already set up”. This was because the relevant “push” interface had already been set up in respect of the other 15 schools already using the Arbor MIS. There was therefore little or no cost to adding a further 57 (and indeed growth) schools to the existing interface. Xx xxxxx xxxxx Xxxxx xxxx xxxxx xxxx xxxx xxx xxxxx xxxxxx.x

As for Bromcom, in its Appendix B, it stated this under “Assumptions” at the end:

“Bromcom assumes that United Learning's data warehouse solution has the capability to import data via an 0-Data Feed or via direct SQL queries. However, please be assured that if not, subject to contract award we will assist United Learning in achieving an efficient and effective data access method. We are happy to assess the feasibility and remove any external data integrator need, by transforming and pushing the data from MAT Vision to United Learning's data warehouse solution.”

In addition, in the March Clarification Request under item 11, UL had asked technical details concerning Microsoft Azure direct SQL access which would allow data to be pulled into UL’s data warehouse. Most of Bromcom’s response dealt with this but it then added a final paragraph, saying:

“As part of the technical discussion, we also invited you to share the data transformation requirements with us and are happy to push this data to your Data Warehouse to remove the requirements of dealing with other suppliers or third parties to avoid the delay or cost of the implementation of changes. We are still happy to consider your requirements as part of this contract and let you know where we can help.”

The technical discussion referred to was that which took place at a meeting on 6 March. Precisely what was said about a “push” offering from Bromcom, discussed towards the end of the meeting, is the subject of some debate among the witnesses, to which I refer below.

However, in its Appendix E, there was a rubric called “UL Data Warehouse Interface” which asked for the “cost of setting up and maintaining interface with UL Data Warehouse (if not included above”. All of the cells to the right of this said £0.00. In the line below, the question was “Describe here the way in which you make up your interface fee” to which Bromcom’s answer was “the data feeds used to provision information to a data warehouse are not chargeable”.

UL took the view that while Bromcom had adverted to providing a push solution it had not actually included it in its costing. Accordingly, what UL did was then to add to Bromcom’s costing, its own internal costs of setting up the “pull” interface. The reasoning was that was based on 20 days at £220.26p based on a pro-rata of a salary of £50,000 per year. The calculation itself is not in issue.

However, I should add that there was some uncertainty at UL prior to the £4,405 being added. This arose as follows. Mr Sharman produced his scores on Quality on 25 March. He said that on the criteria he was dealing with, they did not appear to include “the considerable internal costs of developing a solution to move the data from Bromcom’s MAT vision to the data warehouse”. He was not here dealing with the Appendix E costs evaluation. Xxxx xxxx xxx xxxxx xxxx xxxx xxxx xxxxx xxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxx xxxxxx.xx. He then spoke to Mr Wilson who recorded that Mr Sharman’s score of 3, not 4 (on Appendix B) reflected the need to include some internal UL costs. Mr Wilson thought that 20 days of internal time, based on a pro rata salary of £50,000 per year would suffice. Mr Sharman came back with the understandable point (clearly drawn from Appendix B) that Bromcom had said that they could do the “push” and wanted to know if it was included in their costs (in fact, on the face of Appendix E, it seemed that it was). This led Ms Wood to comment that if UL wanted Bromcom to do the work, they needed to raise a formal clarification request to Bromcom and see if the cost of the work was already included in their offer, and if not, what the cost would be.

Ms Wood then quoted the relevant part of Bromcom’s Appendix E to Mr Wilson who replied that this clarified matters. He noted that Mr Sharman had not seen the costs model.

There the email trail ended. However, instead of assuming (as per Appendix E) that the “push” cost was included, or seeking clarification, as Ms Wood had suggested, the internal cost was still added.

In those circumstances Bromcom contends that the £4,405 was improperly added. An initial argument was that UL had no power to add costs at all; however I have rejected that argument in paragraph 71 above.

Second, however, Bromcom says that there was no basis for adding the £4,405 anyway. One argument here was that the statements in Bromcom’s Appendix B and March Clarification Response showed that it was indeed including a “push” solution at no charge anyway. I do not accept that this is how those statements should be read. Obviously they were offering the possibility but only upon some information being provided by UL. However, given the email trail referred to above, there was obviously uncertainty if one looked at Appendix B alone and yet no clarification was sought. It clearly should have been.

A third point (which emerged in the course of trial) was that in any event, Bromcom’s Appendix E - objectively read - did clarify the point to the effect that those costs were included.

At this stage, I need to refer to the evidence of Mr Huseyin Guryel, who was the only witness from Bromcom properly to deal with the point. He had read the version of Appendix B that formed part of Bromcom’s ITCD submissions. He said that in the end (not having had any further communication from UL about a “push” notwithstanding what was said in Appendix B) perhaps UL did not want it after all. He then said that he thought that Bromcom had made it clear that the “push” was included, having previously said that probably they could have done a better job by seeking clarification themselves. In answer to questions from me, however, it was clear that Bromcom would have needed certain technical details from UL before they could actually design and implement the push; however, the cost itself was not a concern. That is the point at which he said he thought costs had been included anyway.

I agree that there is something of a tension between Bromcom’s Appendices B and E and it was open to it to have sought clarification. However, here, it seems to me that the burden was clearly on UL. It knew xxxxxxxxxxxxxxxxxxxx either there remained uncertainty as to whether Bromcom was offering it or it clearly was, by virtue of Appendix E. Yet, for some reason, UL changed tack on these points for which there is no rational explanation. Plainly, UL should have got clarification especially xxxxx xxxxx xxxxx xxxxxx xxxxx xxxxx xxxxx xxxxx xxxxx xxxxxxx xxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx.xxxxxxx. On any view, had clarification been sought they would have been told that the cost of the push was included by Bromcom albeit some technical information was required before it could be done.

For those reasons alone, there was manifest error and/or unfairness in UL is adding the £4,405 to Bromcom’s costs model and it should be deducted.

I would also agree with a separate argument that in this case, it is necessary to deduct the £4,405 to “neutralise” Arbor’s inherent advantage as incumbent provider to the other 15 schools. I explained why, briefly, when I deal with other arguments concerned with Arbor’s inherent advantage, in paragraph 152 below.

There will also be implications for how Bromcom’s Appendix B was scored, again, dealt with below.

Each of the 57 existing schools already had one form of MIS or another. In the main this was provided by Capita. In each case, UL paid an annual subscription in respect of those MISs, but they did not all run from the same month. It also emerged from the evidence of Ms Wood at trial that although they would be billed at the end of each month for the next year’s subscription, that subscription actually started from the beginning of the month in which they were billed. So, for example, a renewal date of 30 September meant that the first month of the renewed subscription was that September, and the last month would be the following August. If a subscription was terminated part way through the year, there was no rebate. Thus, once a subscription was renewed, it had to be paid in respect of the whole year even if not used after a certain point.

All of this was important in relation to each bidder’s Appendix E and an ancillary document called TCO (total costs of ownership). The bidders had to produce mobilisation plans showing, in relation to each school, when their MIS would “go live” which meant that the old MIS subscription would be stopped. It was agreed that not all 57 schools could go live at once. There had to be a phased implementation.

Depending when each bidder proposed a go live date for its MIS, there might be significant legacy subscription costs still to be paid. They would have to be borne by UL but that additional cost was then added to each bidders total cost. The bidders knew what the renewal dates were for each of the schools.

Arbor’s submitted mobilisation plan would cost UL significantly less than Bromcom’s. This was because the start dates for its MISs were earlier than Bromcom’s and in particular, were largely to take place in the last month of the legacy subscription. Bromcom’s, on the other hand, often fell in the middle of a subscription year meaning that the full year of the legacy subscription had to be paid for, even though it would not all be used.

Bromcom’s case here is that Arbor’s bid was not compliant and should either have been rejected outright or amended by UL so that its start dates were later in the year which would lead to the legacy costs involved under Arbor’s bid being significantly higher. As calculated from the bidders respective TCO worksheets, Arbor’s “TCO Incumbent Licence Costs” were £471,623, while Bromcom’s were £548,870.

I should interpose at this stage that UL accepts that it made a manifest error in the calculation of Arbor’s TCO because it failed to include legacy fees for 10 schools in the sum of £7,539 which, accordingly, must be added to Arbor’s figure it here.

Bromcom’s complaint comes from the go live timings for what are referred to as the Wave 1 and Wave 4 schools, which related to summer/autumn 2020 and 2021 respectively. It contends that by reason of the Mobilisation Guidance, Arbor should not have proposed go live date earlier than September 2020 and 2021 respectively. It had largely proposed August go live dates for those years. I should add that there is no overlap with the admitted error in respect of £7,539 which relates to Waves 2 and 5.

In its ITPD Appendix C, UL stated that it wanted to achieve mobilisation of the entire programme over 18-24 months. The schools’ ability to make the transformation was key so the mobilisation could not be undertaken at too fast a pace.

Notwithstanding that, Bromcom produced a very aggressive timetable in its ITPD submissions of only 7 months, being May to November 2020. UL considered that this was not feasible. Arbor’s implementation period was around 18 months in total, with 4 main waves of deployment. Both bidders were given feedback on this aspect of their bids at the CD3 meetings on 2 March. Both bidders were told that UL would publish further guidance on its mobilisation requests. This was to become the Mobilisation Guidance. Mr Wilson drafted it with input from Ms Wood. He had intended it to go out prior to the Costs Call scheduled for 4 March but in the event it was sent out shortly after.

Part of the Costs Call also dealt with mobilisation. As Bromcom had (covertly) recorded this call, there is a transcript. In the Costs Call, Mr Wilson said that the document shortly to be sent was as much about UL’s managing the implementation process and giving more clarity in the tender documents so that UL could compare like with like.

Because of the debate concerning the status of the Mobilisation Guidance, it is necessary to set out almost all of it. It consisted of two pages. The first was headed “Revised Program & CPD Guidance for ITCD-March 2020”. On that page appeared the following under the heading “Guiding Principles”:

“Following competitive dialogue meetings on 02 02 [should read 03] 20, we issue the following guidance by way of clarification to help Participant's responses to Appendix C of our ITCD document:

• Making this transition across the 57 schools in our tender across the originally communicated timescales of 18-24 months remains our intention;

• Whilst cost is a factor the key principles for us are that the programme is well managed and delivered at a manageable pace for the Group and United Learning Schools;

• Communication is high quality at all levels and gives schools enough time to plan for successful change whilst also ensuring their normal business as usual activities;

• There are early pilots within the programme, in all types/ phases of schools, so as we can celebrate successes and build momentum;

• Where there is opportunity to align 'go-live' with new cloud MIS with contract end of incumbent MIS then this should be taken but this shouldn't be the single determining factor;

• Participant offers of other potential cost saving strategies such as free licencing during necessary contract overlap periods which might help the program will improve the total cost of ownership element, which has an impact on the cost score;

• There should be enough flex in the resourcing of the program so that following initial school engagement the numbers in each phase could be increased and so the program accelerated - especially in the primary phase;

• Participants should exercise caution in planning too many secondary school transitions in the Spring Terms of 2021 and if necessary 2022 to avoid critical times with regards to KS4 and KS5 terminal examinations;

• Will still regard geographic grouping of schools into phases as advantageous in terms of schools supporting each other; and

• For the purposes of the tendering exercise Participants should plan for and cost the necessary school-by-school training programs.”

The second page was headed Indicative Program. The rubric which followed said this:

“For the purposes of tendering and following the discussions mentioned in the meeting above the following Indicative Program has been included to provide an additional scaffold for Participants to respond against Appendix C. The original problem statement remains as read, so Participants should still program in each individual school and should still be conscious of contract dates, geographical location and phase of school. The indicative program reflects elements of each Participants' initial response that was deemed desirable but give additional clarity over pace and how United Learning regard this as being successful.”

There then followed a table with this content:

A key issue is whether the Indicative Program was mandatory, in the sense that it had to be followed precisely. I deal with that below.

Each bidder responded to the question of mobilisation in the text of their Appendix C but they also included a mobilisation plan setting out the mobilisation sequence for each phase. All of this would be scored. Then, from a costs point of view each bidder had to complete a TCO worksheet sent to them by UL. For each school there was a column headed “Incumbent Renewal Date” and then “Incumbent Retirement Date”. The latter was defined in the rubric to mean: “the date on which the School no longer needs any access to the incumbent product (e.g. the new MIS is fully migrated and in use)”. Both bidders used the waves set out in the Mobilisation Guidance as described therein.

In Arbor’s Appendix C it referred to “schools going live over Summer” and here it was clearly referring to Waves 1 and 4. It said that an Arbor champion would be involved, coming into the schools towards the end of August or the start of September, for final migration. The workloads would be split over the start and end of the holidays. Xxxxx xxxxx xxxx xxxxx xxxxxx xxxxxx xxxxx xxxxxx xxxxx xxxxxx xxxxx xxxxx xxxxx xxxxx xxxxxx xxxxxxxx xxxxxxx xxxxx xxxxx xxxxx xxxx xxx

Arbor’s Appendix E TCO showed the Incumbent Retirement Date in those waves as August. However, it added a further column called “Go Live”. Go Live was put at September. It identified the wave applicable to each school, although somewhat confusingly, it called the Pilot Wave, Wave 1, and the so the following waves therefore went up to Wave 6.

As for Bromcom, in its Appendix C, it stated that it was now providing two illustrative programme plans, one for 24 months and one for 18 months. It said that these plans met UL’s stated timeframes. It also said that the waves matched those in the Mobilisation Guidance with only minor deviations, to be discussed at the Discovery stage. In the attached mobilisation plan, it referred to two sets of dates (among others). One was “Live migration” and the other was “school go live”. For each of these descriptions, there was a start and an end date. So, for example, on Wave 1 Cluster 1, live migration took place between 19 and 22 October with the school go live being 3 November. It can also be seen that Wave 2 included 9 secondary schools not 10 as per the Mobilisation Guidance. Further, and to accommodate this, Wave 4 had 13 not 12 secondary schools.

In its Appendix E TCO, specific Incumbent Retirement dates were given. For example, Abbey Hey Primary Academy’s Incumbent Retirement Date was 22 October i.e. the end date of live migration in Bromcom’s mobilisation plan. It then showed in blue the cells for the month of the Incumbent Retirement Date and the start of Bromcom’s MIS.

Following their receipt by UL, it put each bidders TCO into the same format. The blue cell was the month of the Incumbent Retirement Date. Any yellow cells were referable to incumbent subscription fees still payable because the next renewal date came later. A comparison between the bidders adjusted TCOs shows far fewer cases of yellow cells following blue cells in Waves 1 and 4 in Arbor’s TCO. This was because of Arbor’s having an Incumbent Retirement Date in August i.e. the last month of the legacy subscription.

As refined at trial, Bromcom contends that the Mobilisation Guidance prohibited a Go Live date of 1 September or earlier. In other words, 2 or 3 September would be acceptable. It did not now contend that the Mobilisation Guidance stated that September should be avoided altogether. That would have been very difficult to maintain since the Mobilisation Guidance never said that, while on the first page, it did say that certain other periods should be avoided.

Bromcom also accepted that there was no prohibition on starting the migration process to the new MIS before 1 September. That is so even if it would overlap with the migrations of the Pilot Wave.

What that meant, as accepted by Mr McGurk in the course of the trial, was that what Bromcom was really complaining about was that instead of proposing a Go Live date in August or perhaps 1 September, Arbor should have put in a Go Live date a few days later. This would have meant (as understood by Mr McGurk and the Court originally) that there would be one month’s extra legacy costs to be added to Arbor’s offer. That probably would not have amounted to very much. xxxxx xxxxx xxxxx xxxxxx xxxxx xxxxx xxxxx xxxxxx xxxxx xxxxx xxxxxx xxxxxx xxxxxxxx xxxxxxx xxxxxxx xxxxxxx xxxxxx.

However, Bromcom’s challenge became substantially more significant when we learned from Ms Wood that in fact the new legacy subscription started on 1 September even though the renewal fee was not due until the end of that month, as noted above. Thus, Bromcom then contended that the only way for Arbor to avoid being caught by a new incumbent subscription would be to propose a Go Live date in August, which is what it did in almost all cases where the renewal date was 30 September. But that, according to Bromcom, was precisely what the Mobilisation Guidance prohibited.

I deal first with the true nature of Arbor’s bid. This made it clear that the legacy MIS would be retired i.e. “switched off”, as it were, by the end of August. It would not thereafter be needed, although Arbor catered for making a backup copy. Accordingly, there was no need for the legacy service at any point in September or later. It would also follow that at any time after the relevant August date of retirement, the new MIS could go live. In fact, as Arbor’s TCO recognises, Go Live would not happen until September. The month of September clearly does form part of the early academic year and there is nothing in the Mobilisation Guidance to exclude any part of it.

It is correct, that there was reference in the CD3 discussion to 1 September not being practical, but in my view, once the Mobilisation Guidance was issued, that is where the focus lay.

Accordingly, if one was to take the Indicative Programme as mandatory, it would be acceptable if a bidder proposed Go Live as at 1 September where the legacy MIS had been switched off the day before. In practice, I doubt whether a 24-hour delay until 2 September would have had much impact when staff would have been dealing with the first day back at school on 1 September.

The key document in my view, certainly from the point of view of the costs assessment, must be each bidder’s TCO. True it is that there are the words in parentheses referred to in paragraph 114 above, but I do not consider that they prevented a switch off on one day and a Go Live the next.

On that footing, the Go Live date implicit in Arbor’s TCO of 1 September or later conforms to the Mobilisation Guidance if applied strictly.

In fact, in my view, it did not need to be applied quite so rigidly. I turn to that point now.

I quite accept that this was an important document. It was, in effect, an additional facet of Appendix C. It was made in order to provide clarity to the bidders (especially in the light of Bromcom’s too aggressive first mobilisation plan in its ITPD). It would ensure also equal treatment.

It is also true that the Indicative Programme was detailed in relation to the number of phases and schools. Nonetheless, it used the word “Indicative” and the title of the whole document said “Guidance”. It nowhere stated that any bid which did not follow it would be regarded as non-compliant or anything of that kind. It did refer to what UL would regard as “successful” but that, in my view, is a reference to scoring, not compliance. It has to be remembered that each bidder’s mobilisation plan was scored as part of its Appendix C as well as being treated as part of the Appendix E costs.

In other words, there could be a departure from the terms of the Mobilisation Guidance if a bidder wished to do that. Bromcom did, in fact, because of the slight alteration of school numbers, referred to in paragraph 117 above. I accept that this might be seen as relatively minor and the effect of course was to slow the pace rather than quicken it. Nonetheless, it is hardly consistent with the evidence given by Ms Morgan at one point which is that she had actually been told by UL that the Mobilisation Guidance must be followed to the letter.

Accordingly, in my view, a bidder could specify a “ready for Go Live” date earlier than September for example in late August. Obviously, there would come a point where a deviation from the Mobilisation Guidance was so significant that not only would it produce a low score but it might become very difficult in terms of the calculation of legacy costs or would otherwise be unworkable. But that is nowhere near this case.

Xxxxx xxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxx xxxxxxx xxxxxxxxx xxxxx xxxxxxxxxx xxxxxx xxxxAt one point, Bromcom suggested that even if there was no non-compliance by reason of Arbor’s precise timing, there was still non-compliance because Arbor favoured the saving of costs too much. That is a hopeless submission. There was no UL requirement to that effect - the balance between cost and pace was essentially up to the bidder, and UL could use its judgment in evaluating those bids.

Accordingly, in my view, Arbor’s bid was compliant. Its Incumbent Retirement Date of August, with a putative Go Live in September fell within the wording of the Mobilisation Guidance. Even if it did not, it was a very modest deviation in circumstances where the mobilisation Guidance did not require it to be followed exactly.

It follows that Bromcom’s challenge in relation to mobilisation (other than the sum admitted by UL) must fail.

The second set of adjustments to Arbor’s pricing contended for by Bromcom all arise from its position as an incumbent provider to the other 15 schools. Two discounts offered by Arbor are challenged. There is also a challenge based on Arbor’s inclusion of the push to the data warehouse. This is another way of putting the argument in relation to the deduction of £4,405, which I have dealt with in paragraphs 82-97 above and allowed. For the sake of completeness I will deal with that challenge in this context as well.

I will describe the discounts offered before dealing with the law and analysis.

Both discounts can be seen by viewing Arbor’s Growth Model in the Costs Growth tab of UL’s costs spreadsheet for Arbor.

The first discount is described under the rubric “SaaS Annual Charge”. Line 33 refers to

“Price for schools based on phased roll out, xxxxx xxxxx xxxxxx xxxxxx xxxxxxxof licenses are provided free to current UL schools to manage the successful phasing of school go-lives regardless of incumbent license end dates (capped at 180 days per school)”

Line 37 refers to a discount rate of xxxx being applied on the basis of 70+ schools going up to xxxx if there were 80 schools. It will be noted that the only way there could be 70+ schools to begin with is if one adds the 15 Arbor schools, otherwise it would be 57. The 70+ schools would rise to 80 if the anticipated growth schools joined. I shall refer to all of this as “the First Discount”.

The second discount is to be found under the heading at line 50, namely “Saving on Arbor Current Contract”. Line 51 states: “Rebate due to 15 Arbor Schools already on contract xxxxxxxxxx xxxx xxxxx xxxxxxx xxxxxxx xxxxx xxxxxxxx xxxxx xxxxx xxxxxxxxxI shall refer to this as the “Arbor Contract Rebate.”

In short, Bromcom says that both of these discounts, when applied, gave Arbor an unfair advantage over Bromcom by reason of the former’s position as the incumbent supplier to the 15 schools.

Bromcom’s challenge relies upon case-law concerned with neutralising an advantage gained by a bidder who is the incumbent supplier to the contracting authority, so as to ensure compliance with the underlying duty of that authority to treat economic operators equally and without discrimination, and to act transparently.

In this context, I refer first to the decision of the CJEU in Dynamiki v Commission 2008 T-345/03. Here, the specification for the tender for the supply of a software service to support an EU Research body called CORDIS included a requirement that there be a three-month unpaid running-in phase. The tender documents specified that the running-in period served the purpose of enabling non-incumbent contractors who were bidding, to familiarise themselves with the CORDIS service. During that phase, the incumbent contractor would still be paid for the service now to be replaced. The incumbent contractor here was a sub-contractor to the successful tenderer, Trasys. Dynamiki was the unsuccessful tenderer. It argued that the unpaid running-in requirement was discriminatory against it. It said that while it had to include within its tender costs, the costs to it of providing the running-in for free, Trasys, on the other hand, incurred no such costs because, as incumbent supplier (or rather the sub-contractor) it would still be getting paid via the old contract. Accordingly, in Dynamiki the first question was whether the requirement for the unpaid running-in was itself discriminatory. The Court said that it was not. This period was necessary in order for any new provider to become acquainted with the CORDIS system and the software that was being replaced. It was of value to it. During that time, the new provider would not be able to produce a guaranteed level of service.

While it was true that the incumbent provider enjoyed an advantage in this respect, this was not as a result of anything done by the contracting authority. It was simply because that provider was an incumbent. Unless any such provider was excluded from tendering, it was inevitable that there would be an “inherent de facto advantage”. So there was no discrimination.

The next question was whether the inherent advantage in the existing provider should be “neutralised” here by removing from Dynamiki’s tender its costs of providing the unpaid running-in phase. The Court held that there was no general obligation upon a contracting authority to neutralise absolutely all advantages enjoyed by the incumbent provider now tendering. However, it went on to say this:

“73.

It also follows from the case-law cited at paragraph 71 above that the principle that tenderers should be treated equally does not place any obligation upon the contracting authority to neutralise absolutely all the advantages enjoyed by a tenderer where the existing contractor is a subcontractor of that party.

74.

To accept that it is necessary to neutralise in all respects the advantages enjoyed by an existing contractor or a tenderer connected to that party by virtue of a subcontract would, moreover, have consequences that are contrary to the interests of the service of the contracting institution in that such neutralisation would entail additional cost and effort for that institution.

75.

Nevertheless, in order to comply with the principle of equal treatment in this particular situation, a balance must be struck between the interests involved.

76 Thus, in order to protect as far as possible the principle of equal treatment as between tenderers and to avoid consequences that are contrary to the interests of the service of the contracting institution, the potential advantages of the existing contractor or a tenderer connected to that party by virtue of a subcontract must none the less be neutralised, but only to the extent that it is technically easy to effect such neutralisation, where it is economically acceptable and where it does not infringe the rights of the existing contractor or the said tenderer.”

However, here, the remuneration paid to the incumbent provider during the running-in was due to the existing contract. As for the potential new contractor, it was not in a position to provide its services at once. It would be unreasonable here to waive the requirement for a running-in simply because of the existing contractor being a subcontractor to the winning tender. There was therefore no infringement of the principle of equal treatment.

The principles set out in Dynamiki were then applied by the CJEU in Amplexor v Commission 28 June 2018 T-211/17. Here, a differential between the tenderer who was the incumbent contractor and the tenderers who were not, was applied. All tenderers had to provide the relevant tools and systems for the transfer from the old to the new service. In the case of non-incumbent tenderers, they could claim up to 3% of the contract price for this. But the incumbent tenderer could only charge up to 0.3%. The rationale was that, as the incumbent, that tenderer would only have to make minor adjustments to transition to the new service. Hence the lower remuneration. The incumbent (losing) tenderer contended that the same cap should have been applied to all tenders.

Having regard to Dynamiki the Court said that the applicant had inevitably enjoyed an inherent advantage due to its position as an incumbent. What the contracting authority had done was to neutralise that advantage. Such neutralisation was here technically easy to achieve. As to being economically acceptable, that depended on whether there was an acceptable economic justification in the light of the principles of sound financial management. It held that the differentiated provision of a remunerated take-over period was liable to promote healthy and effective competition. Nor was the differential arbitrary or excessive. Indeed the Applicant had previously enjoyed a high remuneration rate when itself tendering where it was not the incumbent contractor. The differential here was therefore economically acceptable. Finally, there was no infringement of the applicant’s right; the measure was itself designed to guarantee equal opportunities.

The broad point here is that Arbor enjoyed a de facto inherent advantage as the incumbent provider to the other 15 schools, though not, of course, to the 57 schools the subject of the procurement. Nonetheless, that fact did not per se render the procurement discriminatory. It might give rise to the need to neutralise in a particular case, but there was no absolute obligation to neutralise, as Dynamiki shows.

I have already found that the £4,405 attributed to Bromcom by UL for the data interface should be deducted for the reasons given in paragraphs 82-97 above. But in my view, this would plainly have been a case for neutralisation. The data interface was already there - there was little more for Arbor to do to apply to the 57 schools. That arose directly as a result of its incumbent role. In this respect, the case is analogous not so much to the facts of Dynamiki but rather to those in Amplexor where a differential was justified. As a method of neutralisation, the removal of the £4,405 was technically easy, it had an economic justification (namely not to deter competition where there was this inherent advantage) and it did not, in my view, infringe Arbor’s rights.

In the case of both discounts, again, the role of Arbor as an incumbent is relevant, but in my view, in a less direct way. Further, according to Bromcom there is the potential involvement of a separate legal provision which is set out in Regulation 67 (2) and (5) of the PCR, namely:

“(2)

That tender shall be identified on the basis of the price or cost, using a cost-effectiveness approach, such as life-cycle costing in accordance with regulation 68, and may include the best price-quality ratio, which shall be assessed on the basis of criteria, such as qualitative, environmental and/or social aspects, linked to the subject-matter of the public contract in question…

(5)

Award criteria shall be considered to be linked to the subject-matter of the public contract where they relate to the works, supplies or services to be provided under that contract in any respect and at any stage of their life cycle, including factors involved in—

(a)

the specific process of production, provision or trading of those works, supplies or services, or

(b)

a specific process for another stage of their life cycle,

 even where those factors do not form part of their material substance.”

I now consider each discount, starting with the First Discount.

It is perfectly true that Arbor sought to apply an immediate xxxx discount on its charges as part of its tender. It rationale was to treat the total number of schools involved as 70+. But that could only be done by including the existing 15 using its MIS. However, that xxxx was going to be applied as part of the contract for the 57 in any event. It could have chosen the same discount without using this rationale. And it was not dependent on the continuation of the contract for the other 15, although I agree that they would be unlikely to disappear should Arbor succeed in the tender.

Nonetheless, Arbor was free to charge whatever price it liked. And it made no difference to UL what the basis for that discount was. Indeed, xxxxxxx xxxxx xxxxx xxxxx xxxxx xxxxx xxxxx xxxxx xxxxx xxxxx xxxxx xxxx xxx

In addition, it is hard to see how there could be an easy neutralisation of this. Bromcom says that the discount should be removed (or that part of it which applied only because there were 70+ schools) but that would be unfair because Arbor might have retorted that it was prepared to charge no more than the xxxx discounted price anyway. Any restriction on that would be unfair, in my view. And if the discount was offered because there were in any event economies of scale to be enjoyed by Arbor, it is very hard to see how they could be disentangled. To neutralise in such circumstances would come close to preferring a bidder because they were a very large company as opposed to one which was very small.

Nor is Reg. 67 (2) and (5) relevant because the First Discount here is in truth applied only to the charges applicable to this contract.

Accordingly, the challenge to the First Discount must fail.

However, the position in relation to Rebate is different, in my view. That is because, although the rebate is a deduction from the price offered in this procurement, it is in fact a rebate on the charges levied for the 15 schools i.e. ones which are not in this contract. The wording itself says so: “saving on Current Arbor Contract”. The rebate is therefore to, or in respect of, those 15 schools and so it is all about a reduction in price under a different contract. The fact that the rebate is applied here and not paid separately to UL in respect of those schools, or deducted from their charges going forward, is neither here nor there. I agree that, having put the rebate in here, it could be said to be part of Arbor’s tender for the 57 schools but in truth it is in respect of the other contract. The issue here is not so much the inherent advantage of the incumbent but the fact that UL is permitting Arbor in truth to offer something that is in substance relating to a different contract that does infringe Reg. 67 (2) and (5). For that reason, the challenge here is allowed.

Bromcom makes the following challenges to the general quality scoring methodology applied by UL to the bidders’ submissions:

The average scores were all computed prior to the moderation meeting on 27 March, as can be seen by the slide presentation. They remained unchanged after it. On the face of it, this gives rise to two connected problems; first, they were unaccompanied by any reasons, and second, this was a function of the fact that there was no debate at the moderation meeting about trying to reach a consensus on a whole number score out of 5 for each Appendix. Had there been a consensus on a whole number score, which would have involved give and take on the part of the evaluators, it would have been possible to agree on an underlying reason for the score.

There was no real evidence of any discussion at the moderation meeting as to an appropriate individual score and in substance, it was simply agreed to adopt the averages. Indeed, in evidence, Mr Wilson emphasised that he and Ms Wood had wanted to be transparent about showing their workings to the evaluators - i.e. the averages. Quite so, but the moderation was meant to be a discussion leading to a consensus, not an adoption of pre-produced average scores. It seems that the presentation of and agreement to the averages presented in the slides only took a few minutes.

At paragraph 13 of her WS, Ms Wood said that she had used averaging in previous procurements and understood it to be a “widespread” practice. However these were very general statements and there was no detailed evidence about it. Further, if averaging, unaccompanied by the absence of any consensus discussion and a score with reasons, is objectionable, the fact that others may have employed it does not assist. In argument, Mr Barrett suggested that a Google search which he had undertaken demonstrated that other contracting authorities had used averaging in situations like this. But again, this does not take the matter any further. Moreover this was not put into evidence and there was no opportunity for any witness to deal with it.

In fact, in the first feedback call of 1 April (pages 3190-3191 of the Bundle) Ms Wood suggested giving some detail, that there had been a moderation where evaluator’s individual scores were discussed including the understanding of any outlying scores - all leading to an agreed single score reflective of the whole panel. That was not true. Nor did Ms Wood say anything about averaging. In evidence she sought to explain this by saying that everyone was at least able in the moderation to talk about the rationale for their scores and then discuss the average scores that they later agreed but there was no evidence that this was really done, as opposed to simply adopting the averaging. There was no note of the moderation meetings which might have assisted here. For his part, Mr Wilson denied that the impression given to Bromcom on 1 April was that there had been a “traditional” (to use Ms Wood’s word) moderation to achieve a consensus. He referred to the fact that in this meeting, he said that everyone’s views had been taken into account in making a “combined decision”. However, he denied that this expression implied a consensus score. Rather, he said that the word “combination” implied combining scores into an average. I do not accept that. I am sure that objectively, the expression “combined decision”, connoted a decision of the whole panel ultimately arrived by a moderation i.e. a consensus.

Moreover, use of averaging in this way would build in any outlying scores so that they remained part of the calculation. In a discussion leading to a consensus score, usually the outlying scores would, with agreement, be disregarded. But if there was in fact some persuasive reason for including them, then they would be taken into account in reaching the final score. But all of this would require discussion.

It is also the case that some evaluators, namely Ms Jadeja, Mr Sharman, Mr Rowley, and Ms Dzioba were not present at the meeting while Mr Norrish, who was not an evaluator, was present.

Mr Wilson agreed in evidence that there was no questioning of any evaluator’s individual scores. He said that evaluators were simply asked if they had any “strong feelings” about Appendix A scores. But that does not constitute a true discussion.

The first question is to determine what aspect of unlawfulness characterises Bromcom’s case in the use of averaging in this way. I agree that it is not a case of irrationality per se. Nor is it a case of manifest error. In my view, and as contended for by Bromcom, it is a breach of UL’s obligation to act transparently which itself requires the giving of its reasons for the scores which it made. And the explanation as to why it did not give its reasons is because in truth there was no reasoning process undertaken - which there would have been had a consensus model being used, and used properly.

Those propositions are borne out by the case-law. First, two decisions of the CJEU involving, again, Dynamiki (but not the case referred to above) namely T-272/6 and T-447/10 make plain the duty upon a contracting authority to give reasons, setting out the matters of fact and law on which the assessment was based. This duty was emphasised by the Supreme Court in Healthcare at Home at paragraph 17 in the judgment of Lord Reed.

I refer also to what McCloskey J observed at paragraph 35 of his judgment in Resource (NI) v NICTS [2011] NIQB 121,

“meetings of contract procurement evaluation panels are something considerably greater than merely formal events. They are solemn exercises of critical importance to economic operators and the public and must be designed, constructed and transacted in such a manner to ensure that full effect is given to the overarching procurement rules and principles.”

It is true that in that case, guidance issued to the evaluator panel members made reference to there being a “consensus assessment”. Also, the issue in that case was not the use of an inappropriate collective decision-making method, but whether the panel fell into manifest error on particular matters. Nonetheless, the observations of McCloskey J set out above are pertinent to the attention that will be paid by the court where there is a moderation meeting. In this case, there was, at least purportedly a moderation meeting. That is how it was described in the slide presentation and individual sessions were referred to in the agenda as “moderation and agreement”.

The above cases were all cited by Stuart-Smith J (as he then was) in Lancs Care NHS v Lancs County Council [2018] EWHC 1598. Here, the unsuccessful tenderer challenged the Council’s award of the contract to the successful tenderer (which award was suspended pending the determination of the challenge). In that case, there was a moderation meeting designed to reach a consensus score which would then be communicated to the bidders. There were 8 separate claims within the challenge, collected under 4 issues. The first claim concerned the quality of the evaluation and was described in these terms:

“1.

As to the quality evaluation generally:

a.

Are the reasons given by the Defendant for the scores awarded to the Claimants and Virgin for the quality evaluation questions sufficient in law?

b.

Did the Defendant in fact apply or depart from the stated award criteria and/or evaluation methodology when evaluating tenders?”

This was the context in which Stuart-Smith J referred to the above cases. The evidence showed that there was considerable discussion and references to individual evaluations and comments at the moderation meeting. However, it was said that in the case of 4 out of 14 scores, the Council could not articulate any consensus reasons and further, it could not establish a complete or accurate account of the consensus reasons given in respect of any question.

Stuart-Smith J observed that a contracting authority which cannot explain why it awarded the scores which it did “fails the most basic standards of transparency”. Having considered the evidence in detail, he found that it was not possible to identify a structure in the notes which revealed the panel’s reasoning process which would explain its scores on any question. The indications given by panel members on particular points were insufficient. He found that other reasons, not reflected in the notes, were also in play.

Further, he added this at paragraph 58:

“It follows that I accept the specific criticisms made by the Trusts in support of this submission. However, in my judgment the deficiencies are not limited to the four questions identified by the Trusts in this part of their submissions. Other examples of general observations that lack content are to be found in the Trusts’ Q1 and Q2 and Virgin’s Q1 and (to a lesser extent) Q7. And, viewed overall, I am satisfied that the notes do not provide a full, transparent, or fair summary of the discussions that led to the consensus scores sufficient to enable the Trusts to defend their rights or the Court to discharge its supervisory jurisdiction. First, there is evidence, which I accept, that other reasons (including some agreed reasons) were in play and are not reflected in the notes. Second, pervasively there is no or no sufficient account of the reasoning and reasons that led panel members to resolve their differences (if they did) so as to arrive at consensus scores.”

The fact that the notes referred to positive and negative points made by panel members was not sufficient to reveal the panels reasoning. And the rationale given for the consensus scores was insufficiently articulated. He therefore upheld this first challenge.

Stuart-Smith J’s decision here is relevant, in my judgment, because it shows the importance of understanding and articulating a reasoning process which leads to the ultimate scores given by the contracting authority itself as opposed to the scores of individual panel members.

However, UL submits that Lancs Care is irrelevant, because it was a case where the published evaluation methodology provided for consensus scoring. I confess that I was not able to find the reference for that, but in any event, this was not why the challenge was upheld. I have set out the nature of the challenge above which was all about the failure to articulate sufficient reasons. Indeed, the next challenge, alleging non-application of the correct criteria and methodology, actually failed.

In this case, the meeting purported to be a moderation but in reality there was none. There was no sufficient discussion of any evaluator’s particular scores. Production of an average of the relevant scores is no more than a mathematical exercise. It does not constitute a discussion of individual scores as part of an attempt-with give and take-to reach a collective consensus view so that there is a reasoned basis for the scores given by the contracting authority itself.

It is correct that there is no provision in the PCR that there needs to be an attempt through moderation to reach a consensus score. Equally there is no provision outlawing the use of averaging. But this misses the point. Inherent in the requirement for the contracting authority to give reasons for what, in the end, were its scores, is the undertaking of a process that can yield such reasons and this, in a context where in the usual case, there will be a number of evaluators who produce different individual scores. That process necessarily involves some form of moderated discussion which leads to agreement as to the overall scores (with or without dissent) for which the essential reasons can then be articulated. That duty may not require the contracting authority to delve into every granular detail of the discussion, but it must at least be in a position to say why a tenderer has scored, for example 3, not 4 (going beyond what the definition of each score is) and not merely that it has so scored. It is hard to see how a contracting authority which does not even produce a note which attempts to undertake that exercise has complied with its duty of transparency.

I agree that the parties here were not specifically informed that there would be a moderation designed to achieve a consensus. But that is not necessary in my view. It is a function of achieving a reasoned decision.

On the other hand, the parties were specifically informed that the scoring would be done in whole numbers. The RWIND would surely expect any ultimate score also to be based on a decision using whole numbers. Whether the scores can then be converted into larger numbers of points thereafter is a different question. Even if they can, it does not alter the fact that no collective whole numbers score was here reached in the first place.

UL has pointed out that it also used averaging in the ITPD stage (although then with rounding). I do not see the relevance of this. It hardly legitimises what came later and of course this was not the final stage of scoring.

For all those reasons, I uphold Bromcom’s claim that UL acted unlawfully in using averaging in the way that it did, to reach an ultimate score. There are subsidiary questions about rounding and translating into points thereafter. These are better addressed when I consider the question of causation below.

It seemed to me that in the end, this complaint was not being pursued. In any event, the short point is that even if the weighting was dis-applied, the resulting scores (disregarding for the moment other complaints like averaging) would have remained the same.

However, on the merits, I should add that no particular weightings were set out in the ITCD. Bromcom could have sought clarification if it wished. Finally, and unsurprisingly in the light of that, even if it assumed equal weighting between the 5 Appendices, there is no persuasive evidence that earlier knowledge of the unequal weightings would have caused Bromcom to alter its bid in any material way. Further, the deviations from equal weighting were relatively modest. Instead of an equal score of 72 for each Appendix (and which Appendix B retained) two appendices had 18 more points, and two had 18 less.

Accordingly, to the extent pursued, the weighting claim fails.

At paragraph 11-16 of its Written Closing, Bromcom alleges generalised unlawfulness on the part of UL because some of the evaluators had no proper training and did not have consistent access to particular materials.

I do not see how such a generalised allegation can be pursued and it is noteworthy that no specific consequences are specified in relation to it. It is not suggested that there was a complete breakdown in the evaluation process such that it should be restarted or annulled.

Further, where there is any particular attribute of or circumstance pertaining to an evaluator which can be shown to have led them into manifest error on a particular matter, that will be addressed below.

Accordingly, as a general allegation, this goes nowhere and fails.

For the same reasons, I cannot see where any general complaint about Arbor’s incumbent provider status influencing evaluators really goes. Bromcom has not suggested how any such influence, insofar as it had an effect on UL, could be eradicated, short of excluding Arbor altogether. Mr Bartlett, for example, was unable to point to any particular measure which he said UL should have taken to deal with this advantage. And of course, Dynamiki does not suggest some generalised obligation of neutralisation. By contrast, when, on a particular matter, there has been an incumbent advantage wrongfully applied by UL, I have said so above.

Finally, insofar as any particular evaluator has been improperly influenced by Arbor’s incumbent status, leading it into manifest error, again, that will be discussed below in the context in which it arises.

Accordingly, this complaint goes nowhere either and must be dismissed.

As noted in paragraph 41 above, the final product scores were 202 for Arbor and 193 for Bromcom. The change in scores from 198 and 196 respectively was pleaded in paragraph 48.10 of the Particulars of Claim. Paragraph 61 of the Defence referred to the removal of certain evaluators because they did not attend all sessions.

At paragraph 163 of his WS, Mr Wilson also explained the change. That course, taken by UL, was not originally alleged by Bromcom as being unlawful. Indeed, it did not feature in its Opening Submissions.

However, as Ms Jadeja confirmed in her oral evidence, her product scores (having attended all elements of CD2 and CD3) were taken into account although she did not participate in the separate evaluation of the final submissions, due to pressure of work.

In closing, Bromcom then submitted that she should have been excluded as she had not attended all elements of the scoring process. However, she was not in fact in the same position as the 4 evaluators whose product scores were removed because that exclusion was based on the evaluators not having attended every part of the product demonstrations.

Nonetheless, Bromcom contended that either Ms Jadeja’s scores should have been excluded or if not, Mr Furnival’s scores should have been counted, he being one of the 4 whose scores were excluded. If the latter was correct, then Ms McKight’s scores at least should surely have come back in although this would not make any difference overall. The point about Mr Furnival’s scores was that he marked Arbor with a 2 and Bromcom with a 5. It would also raise questions about Ms Viner’s and Mr Nield’s product scores coming back in although here, there was a separate problem because they both attended all sessions for Arbor but only two for Bromcom. In oral closing argument Bromcom abandoned the suggestion that Ms Jadeja’s scores should be excluded but on the assumption they were not excluded, Mr Furnival’s scores should come in as well.

However, in my view, it was well within UL’s margin of appreciation to exercise a judgment that said that evaluators who had not seen every element of the product demonstrations should have their product scores excluded while those who did (like Ms Jadeja) should not. The position of seeing all the product demonstrations, and scoring the final bids, are not comparable in this context. Nor do I see how what UL did amounted to a lack of equal treatment between Bromcom and Arbor. It is true that the effect of the removal of the scores of the 4 evaluators was to increase the gap between the two bidders. But (a) UL was entitled to take steps to deal with the fact that some evaluators in the event had not participated in all product demonstrations (whether informing the evaluator panel or not) (b) those steps were themselves neutral as between Bromcom and Arbor so that (c) the fact that the gap between them increased is irrelevant. In my judgment, UL was well entitled to take the course which it did.

Accordingly, the Product Score challenge is rejected.

I now deal with a number of individual matters concerned with particular Quality Scores other than the Product Score element with which I have dealt above. I deal first with generic points made in respect of the scores given by two evaluators namely Ms Geran-Haq and Ms Viner. I then turned to points raised in relation to the marking of various of Bromcom’s Appendix responses by certain evaluators.

Ms Geran-Haq gave no reasons at all for any of her scores. She was a Head Teacher, and due to the Covid pandemic, she did not have time to create a contemporaneous record of the reasons behind her scores. The scores themselves are not challenged as being manifestly erroneous. Rather, Bromcom alleges that the lack of reasons stated at the time means that all of her scores should be excluded.

In her WS, Ms Geran-Haq states that, as she said at the time, she had read all of the bid documents carefully. But her time became limited due to the announcement on 23 March 2020 of the first Covid lockdown to take place on 26 March, and she had to prepare her school for closure.

In paragraphs 18-28 of her WS, she gives a clear account of what, according to her, lay behind her scores which were as follows:

She added that she had made some notes at the time although they were subsequently disposed of in July 2020.

At paragraph 75 and 76 of his judgment in Stagecoach v Secretary of State for Transport [2020] EW 1568, Stuart Smith J (as he then was) stated as follows:

“The duty to give sufficient reasons

75.

It is common ground between the parties that there is a duty to provide reasons for a decision such as the disqualification in the present case and that the obligation to state reasons is an essential procedural requirement. The level of detail which must be given in order to satisfy this duty will inevitably be context and fact specific. The guiding principle, as affirmed by the Supreme Court in Healthcare at Homeat [17] is that:

"The reasoning followed by the authority which adopted the measure must be disclosed in a clear and unequivocal fashion so as, on the one hand, to make the persons concerned aware of the reasons for the measure and thereby enable them to defend their rights and, on the other, to enable the court to exercise its supervisory jurisdiction."

76.

Where the context permits, that reasoning can be in summary form as happened in Case T-l83/00 Strabag Benelux NV. A different context requiring different levels and means of explanation is provided by the facts of Lancashire Care NHS Foundation Trust & Anor v Lancashire County Council[2018] EWHC 1589 (TCC), where I summarised the relevant principles and their underlying rationale at [49]-[55]. It remains my view that a procurement in which the contracting authority cannot explain the reasons for its decision fails the most basic standard of transparency. That said, there is no requirement that the reasons and reasoning must all be contained in one document (whether that be the document conveying the decision or otherwise), though the later the purported explanation, the greater the scrutiny that will be required to ensure that what is being provided is in fact the reasons or reasoning that prevailed at the relevant time and not merely an ex post facto justification.”

In Lancs Care itself, Stuart-Smith J distinguished at various points in his analysis, between reasons stated later as a matter of reconstruction and those which were recollected.

There is no principle of law that automatically excludes a statement of reasons given after the event, even some time after the event, by an evaluator which are unsupported by contemporaneous notes but which the evaluator can recall. It is all a question of weight.

In the case of Ms Geran-Haq, I found her to be an intelligent witness whose evidence was clear and honest. I consider that the reasons articulated in her WS had been genuinely recalled by her. I do not consider that her scores should be excluded.

The real problem, in my view, is not the lack of reasons given at the time by Ms Geran-Haq. It is a more fundamental problem of UL, as a contracting authority, not giving a clear statement of its reasons for the ultimate scores adopted, as I have already found in paragraphs169-184 above.

Bromcom makes the general point that Ms Viner was in fact predisposed against Bromcom because she had come across an earlier product (not in fact its MIS) which she thought was rather basic. I do not think there is anything in this. I accept her evidence that she was actually hoping that Bromcom would deliver a winning package and was somewhat disappointed when it did not. I agree there were occasional comparative references to Arbor in her scoring but not so much as to dislodge the fact of her scoring generally being by reference to the stated criteria. There is an exception so far as Appendix C is concerned but I deal with that individually in context, below.

I see no general basis for disregarding Ms Viner’s scores. Further, had they been the product of some wrongful disposition towards Bromcom, the only fair way to compensate for it would be to ignore her scores altogether, rather than create new scores for her, based on the average scores given.

Before turning to individual scores on the Appendices, I should mention a general objection taken to Bromcom making challenges (on at least some occasions, unpleaded) by reference to the reasons given by particular evaluators for their scores set out in their score sheets and/or in their WSs. It was suggested that this was giving too much scope for individual objections to individual scores. The difficulty with this argument, however, is that UL chose, no doubt in order to explain its evaluator’s positions as far as possible, to adduce detailed WSs from almost all of the evaluators who then went into detail about the scores they gave. Once matters had been opened up in this way, it seemed to me to be fair for Bromcom to be able to cross-examine them on the contents of their WSs, and thereafter make points about it. Indeed, although it reserved its position, UL did not seek to restrict or prevent such examination. In the event, UL was well-able to deal with those points in the re-examination of the relevant witness and/or in its oral and written Closing Submissions.

Accordingly, there was no unfairness or prejudice to UL and I therefore deal with each of the individual points made in respect of the scoring of the relevant appendices.

Bromcom makes a general point to begin with. This is to the effect that Mr Wilson forwarded to all evaluators a document which UL had previously received from Bromcom but which did not form part of the collection of bid documents it sent through on 20 March. That document was Bromcom’s Migration Scope document. Mr Wilson sent it to evaluators, trying to be helpful, because it showed Bromcom’s flexibility and he wanted them to have as much information as possible. Nonetheless, he accepted that this was not a document missing from Bromcom’s bid submissions as such, and he did not check with Bromcom as to whether it wanted evaluators to see it. In that sense, it is said that what he did was inappropriate.

I follow that, but taken by itself, I do not see where this point goes. The suggestion is that if asked, Bromcom might have said the document should not be included; or if it was, it should be amended.

It is also said that this treatment of Bromcom’s document was inconsistent with UL’s treatment of Arbor’s bid documents. That is because UL did go back to Arbor on 23 March to ask for its March Clarification Response which UL perceived as being missing from its bid documents. But that is a different matter. UL did not see the Migration Scope document as something missed out and it already had a copy. It just thought it would be useful to add it to the evaluator’s materials. As far as I can see, no individual score is challenged on the basis that there was a manifest error caused by sight of the Migration Scope document. No other consequence of the additional document is contended for in the counterfactual (see Annex 2 to Bromcom’s Written Closing) and accordingly, I disregard it for present purposes.

I therefore turn to the specific challenges in respect of the Appendix A scores. They relate to those awarded by Ms McKay and Mr Richardson.

As for Ms McKay, she awarded Bromcom a score of 3 and Arbor, 5 on Appendix A. Part of Bromcom’s response here said this:

“Meets local school needs

Bromcom's approach is based on supporting schools to go live with the minimum change first that suits their needs, transitioning across to additional modules as and when they are ready, with no extra cost or technical configuration required. For example, many customers go live with the core MIS before adopting:

Any payment systems;

eTimetable - schools can continue to build timetables in NovaT and then import these into the MIS; and

any apps for communicating with parents (MCAS).

Bromcom's free API programme enables existing bolt-on systems to pull or push data to Bromcom MIS to ensure business as usual from day one, giving schools full control as to when/if they phase out existing systems…”

What this was about was indicating flexibility where schools had “bolt-on” third-party products added to their existing MIS. It was preceded by a specific reference to schools’ existing bolt-ons in the previous paragraph. In addition, the whole question of the addition and integration of such products into Bromcom’s MIS was addressed in Appendix F, dealt with below.

However, Ms McKay’s score indicates that she thought that the underlying MIS systems (i.e. the old and the new) may have to run alongside each other for a period. This was not in fact so, as the mobilisation plans made clear. On one day, the old MIS would switch off and on the next day (if not earlier) the new one went live. The question of a phased introduction of bolt-ons was entirely separate. So this was a significant misapprehension on the part of Ms McKay. In cross-examination she sought to diminish this point but I remain of the view that she had been mistaken at the time. The words from her score “less buy in if incumbent was still available to use…” suggests that the old MIS would still be operating in some way but that was not correct, even if certain bolt-ons would be activated for certain schools after their MIS went live.

The next question is whether the absence of this error would have made a difference. I think it would have done, judging by the fact that other than the observation here, the only negative point made by Ms McKay was that Bromcom had not produced actual testimonials. On the face of it, this may well have prevented the award of a 5 but it would not prevent the award of a 4. Although Ms McKay said that there were other, unexpressed, reasons why she awarded a 3 in any event, I do not think this is sufficient to conclude that there was no manifest error, with consequences for the score here.

Accordingly, there should have been a score of 4, not 3.

As for Mr Richardson, he gave Bromcom a score of 3 and Arbor, 4. He also commented on Bromcom’s “Meet Local Needs” response. What he said on his scoresheet was that Bromcom’s approach here was “based on go live with minimum change first, transitioning to additional modules.” However, this was, in a sense, a negative point because at paragraph 22.4 of his WS he said that he felt that the proposed phased, module, implementation might possibly be more difficult than a “Big Bang” approach to implementation. This was because the phases would need to be implemented and managed at individual school level whereas with the big-bang approach, they could decide what additional modules they wanted and bring them across at the outset. In cross-examination he agreed that his view which espoused the “Big Bang” approach was entirely inconsistent with UL’s Problem Statement at Appendix A which had emphasised the need for flexibility. This was a manifest error since he had misunderstood UL’s own requirements from the bidders.

The rest of his comments on Appendix A were positive-or at least not negative. On that basis, I agree that without this manifest error, he is likely to have awarded a 4, not 3 to Bromcom.

I have already concluded that the £4,405 added by UL to Bromcom’s costs to compensate for the fact that it was not perceived to offer a push solution, was a manifest error and a failure to treat the bidders equally. See paragraphs 82-98 and 152 above.

I agree, having regard to the documents, that UL was not requiring a push solution as such. Xx xxxx xxxxx xxxxxx xxxxxx xxxx xxxx xx xxxxxx xxxxxxxxx xxxxxx xxxxx xxxxxx xxxxxx xxxxxxx xxxx xxxx xxxxxx xxxx xxx xxxx xxxxx xxxxx xxxxxx xxxxx xxxxx at least, having regard to what it said in its Appendix B response, as opposed to its Appendix E, it was unclear if Bromcom was offering it or not.

I further held that UL should have sought clarification from Bromcom (as, at one stage, it intended to). Had it done so, it would have become clear to evaluators that Bromcom was xxxxx xxxxx offering a push.

This means that if an evaluator marked Bromcom down for failing to offer a push, as a matter of causation at least, they would not have done so had the true position been revealed to them. I think that this is the correct analysis rather than saying that each relevant score was a manifest error because, on the face of Bromcom’s Appendix B (which is what we are concerned with here) it plainly was offering a push. The latter would not be correct. However, insofar as an evaluator was marking Bromcom down, specifically because it thought UL had required a push xxxxx xxxx xxxxxx xxxx xxx xxxxxx xxxx that would be a manifest error since, read objectively, Appendix B did not require a push.

I now turn to the relevant evaluators. As for Mr Wilson, he agreed that he had marked Bromcom down to a 3 because it had not included a push in its tender. Mr Rowley agreed in evidence that he took the same position and again marked Bromcom 3 with Arbor 4. As for Mr Richardson, in cross-examination he impliedly accepted that he had marked Bromcom down to a 3/4 (which became 3) because of the absence of a push offer. However, he said that there were other reasons for giving a 3 anyway. Nonetheless, the only specific reason he gave (as opposed to generalised statements which were hard to follow) was that Bromcom’s approach was more cumbersome because it was not a “single shot”. But that was back to the top-down centric approach he advocated in relation to Appendix A which was manifestly wrong anyway. See paragraph 226 above. Accordingly, here, there was a manifest error on his part as well. Without that error, Bromcom would have scored 4, not 3.

As for Mr Sharman, he marked Arbor with a 4 and Bromcom with a 3. Here, in his scoresheet, he said this:

“Will require considerable effort from us to build a solution to move data from MAT Vision into our own data warehouse. However, use of MAT Vision allows us to implement future modifications without costly change requests.… Central standardisation seems to be achieved by guidance or a mysterious “blank Bromcom MIS”. Not clear how this would work. Some vague assumptions around expectations provided.”

In a number of respects, these were odd remarks. First, he actually knew from the technical call on 6 March that Bromcom had at least offered to do the push-according to his own contemporaneous note. Then, of course, there was the query which he had raised on 25 March. Here, he noted that Bromcom had said it could implement the push. However, on the face of the emails (see paragraphs 89-91 above) he was not copied into Mr Wilson’s email which said that Bromcom’s Appendix E had clarified matters.

Even if Mr Sharman assumed that the internal cost for UL to do the pull was included, it seems clear that he marked Bromcom down because it did not offer a push-as Mr Wilson had himself said, after speaking to Mr Sharman. Again, see paragraph 89 above.

In fact, Mr Sharman appears to have given Arbor an overall 4 and Bromcom an overall 3, having split up his marks for different elements of his Appendix B comments. See the bottom of pages 2513-2514 of the Bundle. He ended up with an average of 3.75 for Arbor and 3 for Bromcom. On the push/pull issue he actually gave 5 to Arbor and 3 to Bromcom. If one changed Bromcom’s score here to 4, this would produce an average score to Bromcom of 3.25 - not enough to get a 4 overall. However Mr Sharman also accepted that he was influenced by Arbor’s xxxxXxxxx xxx xxxxx xxxxxxx xxxxxxxConsistent with what I observed in paragraph 152 above, this is another manifestation of Arbor’s incumbent advantage-Bromcom could never compete with an in-house push solution already operating. In my judgment, the only way to neutralise that advantage here would be to make Bromcom’s score match Arbor’s-xxxx xxxxx xxxxx xxxxxxx xxxxxxxxx xxx. In other words, to give Bromcom a 5 in this element of Appendix B. That would make sense, because there was a “con” for Arbor being that future modifications would have to be paid for whereas Bromcom’s MIS would not require costs change requests. On that basis, if Bromcom’s score for this element is moved up to a 5, then the average is 3.50 which, under normal calculation conventions, would make it a 4 overall, albeit from a slightly lower average starting point than Arbor.

In relation to all of this, UL points to the fact that one of Bromcom’s “takeaways” from a debrief meeting with UL on 14 April 2020 was that “… Arbor did a better job with the written responses. Already covered the feeling that we pushed the less than perfect fit data flow with the data warehouse…”. I see that, but I do not think this affects the manifest errors in UL’s assessment which I have referred to above and which had causative impact, in my view.

Equally, my reasoning above is not premised on the view that Bromcom clearly did propose a push solution (which UL challenges). It is based on the consequences of UL’s failure to clarify the position, together with a (compensatable) reliance on Arbor as incumbent provider.

Finally, I turned to the position of Ms Dzioba. She scored a 5 for Arbor and, originally, a 4.5 for Bromcom. As to the latter she said:

“… We would need to create and maintain connections/links to our DW but an offer to assess the possibility of pushing data from MAT Vision to DW is provided.”

That is in contrast to what she said about Arbor which was:

In paragraph 32 of her WS she said that Bromcom’s position was different to Arbor’s proposal xxxxxxxxxxxxxxxx xxxxxxx xxxxxxxxxxx xxxxxxxxx xxxxxxxxxxxxxx xxxxxx xxxxxxxxx xxxxx. In respect of her reference to the offer from Bromcom she said that the wording which had been used by Bromcom (in its Assumptions on page 6 of its submission here) meant that there was no guarantee that they would push the data to UL.

At paragraph 38, she relates that when sending her scoresheet to Mr Wilson, she said that if a 4.5 was not allowed she would want to award Bromcom 4. As set out in her email, this was because she wanted to reflect the fact that Arbor’s response was more technical and their suggested solution would mean less work for UL, but both solutions were meeting requirements on a reasonably equal level.

She was cross-examined about all of this and what had been said at the technical call on 6 March. Although she did not accept it, it is plain from what she said in her WS, quoted above, that she was marking Bromcom downx xxxx xxxx xxxxxx xxxx xxxxx xxxx xxxxx xxxxx xxxx xxxxxx xxxx. There is, in truth, really no other explanation for the lower mark, compared with Arbor’s. In my judgment, all of this was a consequence of the failure to ascertain from Bromcom whether it was prepared to absorb the cost of providing the push to the data warehouse. Had it been asked, it would have said it was.

On that basis, the only inference is that if Ms Dzioba had known that Bromcom would provide the push solution xxxxxxxxxxxx xxxxx xxxxx xxxxxxx xxxxxxx xxxxxx xxxxxx xxxxxxx xxxxx xxxxxxx xxxxxxxx xxxxxxxx xxxxxx xxxxxxx xxxxxxx xxxxxx xxxxxx xxxxxx xxxxxxxx xxxxxx xxxxxxxx xxxx,xxxxxxxxxxxxxx in the counterfactual, Ms Dzioba would have scored Bromcom with a 5.

Bromcom takes one point on Appendix C and it is a score given by Ms Viner, namely a 3. Arbor received a 5. Her notes on Bromcom read:

“Program plans still starting too soon-not been adjusted to current climate or given any mitigation….”

I agree that, to penalise Bromcom’s programme for “starting too soon” does not make any sense. Bromcom’s mobilisation plan was more conservative than Arbor’s, as explained in paragraph 117 above. In evidence, Ms Viner had no real recollection of ever reading or taking into account the Mobilisation Guidance which was an important document here. I also agree that although she did not accept it, the reference to “current climate” was to Covid conditions which Mr Wilson had told her should not be taken into account. In my view, there was a clear manifest error here, without which Ms Viner would have given Bromcom at least a 4.

This related to the fact that schools tended to use third-party products as well as a core MIS.

The response which UL required here was as follows:

“• Please fill in the Bolt-Ons tab within the MIS Specification Worksheet. Please do not alter the

conditional formatting or count formulas at the bottom of the sheet. [Already completed as part

of first ITPD submission - update as appropriate and resubmit]

• With reference to projects of similar scale and complexity, please outline your strategic,

partnership approach to working with United Learning over the life of the contract to assist us

with our aims of:

o Consolidating as many 3rd party systems as possible into MIS functionality;

o Simplification of any remaining 3rd party bolt on systems to MIS; and

o Cost savings

• We would like to be in a position whereby:

o as much 'add-on' functionality as possible is brought into core MIS;

o remaining add-ons are rationalised down to as small a 'recommended list' as possible

across the various types; and

o Consistency in use and cost savings are leveraged across the Group.”

The MIS Specification Worksheet itself constituted Appendix D.

Bromcom’s response on Appendix F was short, at 1½ pages. It did also provide (or update) the relevant tab in the MIS Specification Spreadsheet although it made no reference to it in its Appendix F response. Nor did it refer to any particular third-party products in the text of the response. Arbor’s response was just under 5 pages. It made specific reference to a number of Arbor’s MIS features that could reduce school’s reliance on third-party products. It then set out a list of recommended third-party products by name, that would be maintained but with whose providers Arbor did or would work to assist with their integration. It went on to explain how savings of up to 30% could be achieved. It added that it had agreed a market-exclusive rate for UL schools on TimeTabler licences, being a xxxx discount. It did not make specific reference to its bolt-on tab in the MIS Specification Spreadsheet, but on the other hand, it had already listed a large number of particular third-party products in connection with its recommendation.

The bolt-on tabs in Bromcom’s spreadsheet had a list of very many third party products. Against each product there were then columns to be completed. Those columns were “This is web-based we already have full API integration”, “This is non-web-based but we can integrate”, “we would need to develop integration to this product” and “We believe our MIS system could replace this functionality.” In addition, Bromcom added a yellow column headed “Bromcom invites the provider to work with us using our free API’s”.

Bromcom alleges that there were manifest errors in the scores given by Ms Letts, Mr Hudson and Mr Holmes. I deal with each in turn.

As for Ms Letts, she scored Arbor with a 4 and Bromcom with a 2. Part of her comments here was that:

“cost savings are suggested but only as far as schools can move over to built-in functionality instead of bolt on systems. No explanation of which parts could move. No recommended list of remaining systems.”

In cross-examination it was pointed out to her that in the bolt-ons tab, Bromcom had indicated (as had Arbor) what third-party products were already integrated and what would have to be integrated. She could not recall if she had this tab open in front of her at the time. It was also suggested that Bromcom’s yellow column did show the recommended products. She replied that this was not quite so, since it did not say that explicitly. Perhaps, but had she actually looked at it at the time, she would, I think, have made the connection. Her broad point was that she had expected to see a list of the relevant products in the response itself. However, she accepted in the end that her mark of 2 was potentially “extremely harsh”. I do think overall that the lack of her reference to the bolt-on tab did have an effect. If in fact, she did not have the bolt-on tab in front of her (whether due to an oversight by her or by UL) or she had it but did not read it (it must be one or other) that did lead to a manifest error. The cure would be to say that her score should have been a 3.

As for Mr Hudson, he gave a 4 to both Bromcom and Arbor. He said in paragraph 79 of his WS that he did not award Bromcom a 5 for essentially three reasons: first, it did not give a recommended list of remaining add-ons; second, it did not set out specific costs savings; third, while Bromcom said it had significant experience of working with MATs to rationalise the number of systems in use, it produced no evidence of the outcomes.

As to the absence of a list of recommended products in the response, it was put to Mr Hudson (as it had been to Ms Letts) that a definitive list of recommended products could not in fact be given because UL was conducting a parallel rationalisation exercise in respect of third party products. However, UL did ask for a recommended list of products. He was taken to Bromcom’s 5^th column in the bolt-on tab where, he agreed, it showed that Bromcom’s MIS could replace the third party products’ functionality. He agreed that such a column was the best Bromcom could have done at that stage-“if you [ie Mr McGurk] say so”. He still maintained that he would have given Bromcom a 4 anyway because there were other separate (and in my judgment good) reasons for not awarding it a 5. In my judgment, there is no basis for his score to change notwithstanding the error he made.

I finally turned to Mr Holmes. He gave a 4 to both Bromcom and to Arbor. His scoresheet read:

“…Bromcom’s approach seemed to be much like Arbor’s insomuch as they think they have full integration within their product and can work with any 3^rd party supplier. As with Arbor I am concerned about what this means for schools with several third-party suppliers at the point of implementation.”

Indeed he had said much the same in respect of Arbor:

“…I am a little bit concerned by the approach that all things should move to Arbor as this does increase the initial stress of the MIS move for schools especially if they use lots of unsupported 3^rd parties. In a very large school with low parental engagement then this could also cause communication issues.”

At paragraph 49 of his WS, he said that the reason he did not award a 5 to Bromcom was because it did not demonstrate an understanding of the stress which could be caused to school staff by changing the MIS and third-party products at the same time. He said much the same about Arbor - see his paragraph 45.

However, in evidence, Mr Holmes accepted that the 4^th bullet point on the second page of Bromcom’s Appendix F response specifically said that there would usually be a degree of uncertainty after the sites go live as to which modules will be adopted and when. Bromcom therefore configured unlicensed everything from the outset to allow schools to explore the additional modules in their own time. In other words, Bromcom did offer the flexibility which he said it did not. At the end, he said he would still have maintained a score of 4 because of the impression gained at CD2 and CD3 and “everything”. He said he was not in a position to remember enough of what was said at the time to change in score. However, on this point, he had given a highly specific reason for not scoring Bromcom with a 5 which was plainly mistaken on the face of its Appendix F response. It seems to me that without that mistake, he would have scored it a 5.

I agree, overall, that Bromcom itself said internally later on that it had done a crude job of addressing third-party add-ons and had promoted its one-stop shop too much. Nonetheless, I do not think this deprives the points I have just made of their force, leading to a changed score.

With this section of the judgment, I have concluded my analysis of the particular individual challenges raised by Bromcom. I now turn to deal with the challenges concerned with the form and timing of Arbor’s ITCD bid documents.

It is common ground that all of the submissions here had to be sent by email. This is to be inferred from the giving of a procurement email address at paragraph VI.4.1 of the Notice, paragraph 3.4 of the ITPD and Mr Wilson’s email to all bidders dated 29 November 2019 that all submissions should go to the stated email address.

Moreover, Regulation 22 (1) provides as follows:

“.—(1) Subject to paragraphs (3), (5), (8) and (10), all communication and information exchange under this Part, including electronic submission, shall be performed using electronic means of communication in accordance with the requirements of this regulation.”

Here, there was no other electronic form of communication available other than email.

The above, is, in my view common ground and in any event established. However, Bromcom says that what Arbor was not permitted to do was to submit its bid by an email which did no more than provide a link to a drop-box and one actually hosted by Arbor itself, at that. The drop-box submission has been described at paragraphs 31-32 above.

Regulation 22 (16) provides as follows:

“Technical etc requirements for tools and devices

(16)

Tools and devices for the electronic receipt of tenders, requests to participate and, in design contests, plans and projects, must at least guarantee, through technical means and appropriate procedures, that—

(a)

the exact time and date of the receipt of tenders, requests to participate and the submission of plans and projects can be determined precisely;

(b)

it may be reasonably ensured that, before the time referred to in paragraph (12), no-one can have access to data transmitted under the requirements in this paragraph;

(c)

only authorised persons may set or change the dates for opening data received;

(d)

during the different stages of the procurement procedure, access to all data submitted, or to part of such data, must be possible only for authorised persons;

(e)

only authorised persons may give access to data transmitted and only after the time referred to in paragraph (12);

(f)

data received and opened in accordance with the requirements in sub-paragraphs (a) to (e) must remain accessible only to persons authorised to acquaint themselves with the data;

(g)

it must be reasonably ensured that any infringement, or attempted infringement, of the access prohibitions or conditions referred to in sub-paragraphs (b) to (f) are clearly detectable.”

Bromcom claims that the prohibition on the use of a drop-box in this way was either inherent or was precluded by Regulation 22 (16). See paragraph 29A.2 of the Amended Consolidated Particulars of Claim and paragraph 16 DA of the Amended Consolidated Reply. It is correct that the reference to Regulation 22 is only given in the latter statement of case although paragraph 29A.2 did state that use of the drop-box was nonconforming because it precluded scrutiny of when exactly the documents were in their final form submitted. I consider that Bromcom’s challenge here is sufficiently pleaded.

Bromcom then alleges that the only possible consequence of this non-compliance was rejection of Arbor’s bid. If that is correct, Bromcom would be the only bidder left and it would have won the tender.

A considerable amount of time was taken up at trial investigating if the contents of the drop-box had themselves changed following the submission of Arbor’s email. In the end, there was no evidence that they had; a change in metadata did not necessarily indicate a change in content as opposed to for example, the moving or opening of a document.

In my judgment, there was nothing in the Notice, tender materials or communications from anyone at UL expressly to prohibit the submission of documents which involve the use of a drop-box. Nor do I accept that emailing a link to the drop-box was itself not a conforming email for the purposes of the submissions.

The only possible argument, in my view, is that this method of email delivery of submissions did not comply with Regulation 22 (16) (a) and/or (b) set out above.

It is a fact that Arbor did continue to have the ability to access its drop-box after the email had been sent, although there is no evidence that it did so, save perhaps to add its amended uptime document (see below). Of course, as soon as Mr Wilson downloaded the documents from the drop-box to UL’s SharePoint platform, any continued ability of Arbor to access the drop-box would become redundant - figuratively speaking, the documents had already “gone” to UL. However, Bromcom contends that this is no answer, first, because in theory a contracting authority might not access the drop-box immediately and anyway, here, there is always the risk that early downloading (i.e. before the deadline) could constitute “examining the content of tenders” prior to the deadline which is itself prohibited by Regulation 22 (16) (e). There is no evidence that Mr Wilson did actually look at any document which he was downloading but I can foresee problems if downloading or moving the files caused them to be opened along the way, as it were.

In my judgment, to permit the use of a drop-box to which the bidder still had access did involve a breach of Regulation 22 (16) (b) on the part of UL. The way to reasonably ensure that the tenderer did not continue to have access was surely either to use a secure portal or to require expressly that all tender documents had to be submitted as attachments to the underlying email.

I also consider that there was a breach of Regulation 22 (16) (a). Of course, the exact timing of the receipt of the email itself could be established as it was here. But this rather begs the question of whether the relevant time is then when the contracting authority actually accesses the drop- box to download the documents. If this involves no examination of them, it should be done as soon as possible. But one runs into difficulties if, for example, the link does not work or the contracting authority cannot deploy it and so on.

It seems to me that Regulation 22 (16) (a) requires some simple method of ascertaining a single point in time when it can be said that the bidder’s submission has been electronically filed. That, in my view, would exclude the use of a drop-box and link, or at least a drop-box as used by Arbor here. None of this should cause a problem to bidders or contracting authorities. All that has to be done is to require documents to be attached to the covering email or submission via a secure portal as already mentioned. In the case of major tenders, no doubt the latter would be the preferred option.

However, in my view, UL’s infringements of Regulation 22 (16) (a) and (b) here go nowhere. This is because in any counterfactual, the scenario would be not that UL would disqualify Arbor but rather that it would putatively have told Arbor immediately that its method of providing its tender documents was non-compliant and that it should send them as attachments. No doubt Arbor would then have done this within a matter of minutes.

Since, in my view, the time for submission of documents on 20 March did not expire until midnight (see paragraph 280 below), it could not possibly be said that Arbor could not have provided its documents in a compliant fashion in the time allowed. Indeed, even if the deadline was deemed to be 6pm (and Bromcom does not contend for any earlier time) the same would apply.

Accordingly, this part of the claim must fail.

Bromcom contends that both Arbor’s March Clarification Response and its amended uptime sheet had to be submitted by the deadline on Friday 20 March. For these purposes, I accept that the deadline expired on midnight that day in the light of the evidence referred to at paragraph 29 above. Of course, the documents in question were not sent until the afternoon of Monday 23 March.

Bromcom further contends that the only course open to UL (which it did not follow) was to disqualify Arbor by reason of these late submissions because it had no power to extend the time limit.

I deal first with the law and will then consider the position in relation to each document.

Part 2 of the PCR contains rules which implemented Directive 2014/24/EU on public procurement and which itself had repealed Directive 2004/18/E. I shall refer to the former as the Public Contracts Directive. The provisions of the PCR to which I now refer, therefore, have their origin in parallel provisions of the Public Contracts Directive.

The starting point is Regulation 30 (15) to (18) which provides as follows in relation to Competitive Dialogue procurements:

“Final Tenders

(15)

Having declared that the dialogue is concluded and having so informed the remaining participants, contracting authorities shall ask each of them to submit their final tenders on the basis of the solution or solutions presented and specified during the dialogue.

(16)

Those tenders shall contain all the elements required and necessary for the performance of the project.

(17)

Those tenders may be clarified, specified and optimised at the request of the contracting authority.

(18)

But such clarifications, specification or optimisation, or any additional information, may not involve changes to the essential aspects of the tender or of the public procurement, including the needs and requirements set out in the contract notice or in the descriptive document, where variations to those aspects, needs and requirements are likely to distort competition or have a discriminatory effect.”

It is clear that the power in the contracting authority to request and receive clarifications etc. concerns clarifications in relation to the finally submitted bids.

I then turned to Regulation 56 which is headed “General Principles in Awarding Contracts”. It provides, among other things, as follows:

“(1)

Contracts shall be awarded on the basis of criteria laid down in accordance with regulations 67 to 69, provided that the contracting authority has verified in accordance with regulations 59 to 61 that all of the following conditions are fulfilled:—

(a)

the tender complies with the requirements, conditions and criteria set out in the contract notice or the invitation to confirm interest and in the procurement documents;..and

(4)

“Where information or documentation to be submitted by economic operators is or appears to be incomplete or erroneous, or where specific documents are missing, contracting authorities may request the economic operators concerned to submit, supplement, clarify or complete the relevant information or documentation within an appropriate time limit, provided that such requests are made in full compliance with the principles of equal treatment and transparency.”

It is clear that Regulation 56 (4), drawn from Article 56 (3) of the Public Contracts Directive, is dealing with the position where (or at least including where) the final bids have been submitted. Professor Arrowsmith, at paragraph 7-161 of the 3^rd Edition of The Law of Public and Utilities Procurement commented that Article 56 (3) confirms the possibility for authorities to contact tenderers and for various types of correction to what has been submitted, which could lead in some cases to non-conforming tenders being brought into conformity. It also confirmed the applicability of general principles of transparency and equal treatment. Finally, she suggested that the extant case-law would remain applicable in this area to determine what equal treatment and transparency permitted, as well as determining what was required by the principle of proportionality. Bromcom has raised a point as to the true scope of Regulation 56 (4) which I deal with below.

I next deal with a number of cases that were referred to me in relation to late or incomplete tender submissions.

Here, a state-owned Danish company conducted a procurement exercise to select building contractors. One tenderer’s bid was held to be non-compliant because it violated Condition 3, Clause 3 of the general conditions which stipulated that the price offered had to be based on the fact that the tenderer had to undertake the actual design of the project and would take full responsibility for its planning and execution. This provision was held to be a “fundamental requirement of the tender conditions” because it specified how the offer price was to be calculated.

The Court held that for this (and another) reason, Denmark had failed to fulfil its Community obligations by accepting this tender bid nonetheless. Again, this was a case where there was held to be a non-compliant bid by reference to a fundamental condition of the tender.

This case concerned a claim by an unsuccessful tenderer against the Commission as contracting authority, for damages, on the basis of various items of unlawful conduct in respect of the tender. The Court of First Instance had to determine, first, if the Commission had acted unlawfully but then, whether such conduct actually caused the loss pleaded. In the event, the Court held that the Commission had acted unlawfully in numerous respects, but assuming the absence of such conduct, it was not shown that the Commission would or should have awarded the - or a - contract to the applicant, Scan Design.

One matter considered was the fact that the winning tenderer, Frezza, had submitted its tender 4 days after the deadline. It transpired that Frezza had only sought an extension by a letter sent 3 days after the deadline which only arrived 7 days thereafter. It was at this point, or later, that the Commission agreed an extension so that the tender could be received out of time.

The Court held that “accordingly” the Commission had acted unlawfully in accepting Frezza’s late tender. This was so even though the Commission had in fact sent the tender documents to the tenderers Italian rather than Belgian branch. That said, the judgment suggests that this did not cause any real prejudice to Scan Design.

This decision is, in my view, of limited assistance since it did not state the precise basis for the unlawfulness nor whether the position would have been different had an application for an extension been made prior to the deadline. Finally, of course, this was a case where the bid as a whole was late.

In this case, the claimant tenderer submitted its tender documents 3 hours before the deadline of 3pm. While doing a final check at 2:45pm it realised that a set of 4 case studies which were (to quote Richards J (as he then was)) “an integral part of each tender” had not in fact been submitted. This was an error and they had in fact been finalised in the morning of that day. They were ultimately sent through at 3:26pm. The contracting authority Council rejected the tender on the basis that the complete tender had not been submitted on time. The claim made was that the Council acted unlawfully (and in particular, disproportionately) in so doing. Richards J dismissed that claim.

The first point to note is that there was no dispute but that the missing documents were indeed formally part of the tender documents. A number of factors were suggested as to why, exceptionally, the Council should not have rejected the bid. These included that the error was unintentional, the relevant material had been prepared beforehand, the claimant had in fact tried to upload the case studies before 3pm but could not do so, and the rest of the tender had been on time. Nonetheless, Richards J held that the Council was entitled to reject the bid. He made the following points:

“67.

As well as the deadline, the other key elements of submitting tenders, such as the requirement for a single submission and the lack of provision for changes to submitted tenders, were clear and well understood by the claimants, as their witnesses evidence made clear. Fairness to all tenderers, as well as equal treatment and transparency, required that these key features should be observed.

68.

There may be circumstances where proportionality will, exceptionally, require the acceptance of the late submission of the whole or significant portions of a tender, most obviously where, as noted by Professor Arrowsmith, it results from fault on the part of the procuring authority. But in general, even if there is a discretion to accept late submissions, there is no requirement to do so, particularly where, as here, it results from a fault on the part of the tenderer. In addition to the considerations already mentioned, the particular facts on which the claimant relies to characterise its case as exceptional, would require investigation and determination by Devon CC and I do not see that it was required to undertake those tasks. In my judgment, the decision of Devon CC to reject the claimant’s tender was well within the margin of discretion given to contracting authorities.”

It is worth noting that in paragraph 68, he is referring to a situation where “the whole or significant portion of a tender” might still be accepted late.

The issue here concerned the contracting authority’s rejection of a tender on the basis that there was non-compliance with the tender conditions and also that the price offered was abnormally low. In neither case, did the authority revert back to the tenderer for clarification or otherwise, before rejecting the bid.

Here, the Court had to consider the effect of the EC Directive 2004/18/EC, being the predecessor to the Public Contracts Directive. It held that the authority had no obligation to revert back to the tenderer before rejecting the bid. However, it added this:

“40.

Nonetheless, Article 2 of that directive does not preclude in particular the correction or amplification of details of a tender where appropriate, on an exceptional basis, particularly when it is clear that they require mere clarification or to correct obvious material errors, provided that such amendment does not in reality lead to the submission of a new tender.…

41.

In the exercise of the discretion thus enjoyed by the contracting authority, that authority must treat the various tenderers equally and fairly, in such a way that a request for clarification does not appear unduly to have favoured or disadvantaged the tenderer or tenderers to which the request was addressed, once the procedure for selection of tenders has been completed and in the light of its outcome.”

Here, one of the documents to be submitted by the tenderers was stated in the contract notice to be its most recent balance sheet. Two out of the 10 tenderers did not include it by the deadline. Two weeks later, the contracting authority asked the two tenderers to submit their balance sheets, which they then did. Those tenderers were successful and the complainant, Manova, was not. The national court raised the question whether a contracting authority in the position of the authority here was permitted by the equal treatment principle to do what it did.

The Court adopted the reasoning and observations of the Court in Slovensko. It then added this:

“39.

Accordingly, a contracting authority may request the correction or amplification of details of such an application, on a limited and specified basis, so long as that request relates to particulars or information such as a published balance sheet, which can be objectively shown to pre-date the deadline for applying to take part in the tendering procedure concerned.

40.

However, it should be explained that this would not be the case if the contract documents required provision of the missing particulars or information on pain of exclusion. It. The contracting authority to comply strictly with the criteria which it has itself laid down…”

The Court did not rule specifically on whether the course taken by the contracting authority was permitted but instead it declared that the authority could, after the deadline, ask a tenderer to provide documents describing its situation which could be objectively shown to pre-date that deadline so long as it was not expressly provided that without such documents the application would be rejected.

Here, the Court summarised the effect of the case law in this way:

The thrust of the case-law, in my view, is that, as contemplated by Regulation 56 (4), there is a limited extent to which a contracting authority can obtain clarifications or supplementary or otherwise missing information once the bid has been submitted. However, cases like Klaipedos have put something of a gloss on the equivalent provision in the Public Contracts Directive which suggests that this power cannot extend, in effect, to permit the tenderer to provide out of time a document which had to be submitted as part of the tender. Even here, however, the position may not be quite so rigid, given the decision of the European Court in Manova.

Moreover, that approach does not entirely square with paragraph 68 of the judgment of Richards J in Leadbitter.

However, what is entirely clear is that all of these cases were dealing with where there was a missing or incomplete required bid document and, for the most part, an important one at that. They were not dealing with the case of a non-mandatory document.

In the light of my findings below, it is not necessary for me to deal conclusively with the question of the precise scope of Regulation 56 (4). Indeed, this was not a matter on which I had extensive or complete submissions which is perhaps not surprising given the many other issues in this case.

The first critical question is what the status of this document was, and in particular whether it was itself part of the tender documents required to be submitted by the deadline.

First, in my view, it was not stated to be such by the ITCD.

I refer first to parts of paragraphs 3.4 and 3.5 which I consider to be relevant to this analysis:

“3.4

…As and where required, further written clarifications will be supported in the period between the Dialogue Meeting and the deadline for submission of Final Tenders. Participants are advised to use the guidelines for clarifications outlined in Section 4.

3.5

Following submission of Final Tenders, the Trust may request a Participant to clarify, specify or optimise a Final Tender, but such clarification, specification or optimisation shall not involve changes to the essential features of the Final Tender or the public procurement (including the needs and requirements set out in the Contract Notice and Descriptive Document) if such changes are likely to distort competition or have a discriminatory effect.”

This latter provision, therefore, confers an express power on UL to seek clarification after the bids have been submitted but without changing their essential features. It seems to me to be reflective of what the courts said in Slovensko and Manova in particular and in Regulation 30 (15)-(17).

I turn next to Section 4, which is headed “Rules in respect of the ITCD Stage”. Paragraph 4.1 deals with clarifications sought by tenderers of UL. It adds that participants must not lobby or unduly influence UL or its members in respect of the procurement process.

Paragraph 4.2 provides for clarification requests to be made by UL of the tenderers. In this regard it says that:

“The Trust shall be entitled to take account of any subsequent information provided as

clarification in any Participant's responses to:

• written queries from the Trust; and/or

• queries from the Trust at any clarification meetings.”

Paragraph 4.3 deals with material changes in a tenderer’s economic and financial standing or its technical or professional expertise, and for the notification to UL of any such changes.

Paragraph 4.4 deals with freedom of information, paragraph 4.5 deals with copyright and paragraph 4.6 imposes confidentiality obligations on the tenderers. Paragraphs 4.8 and 4.10 confers certain powers on UL, and paragraph 4.9 states that UL will not be responsible for any costs. Paragraph 4.11 imposes an obligation on tenderers not to create a conflict of interest and to notify UL if such a conflict of interest has arisen. However, any such declaration would not result in automatic disqualification.

Finally, and importantly, paragraph 4.7 provides as follows:

“Disqualification

Any breach of the requirements of this Section 4, or the commission of any offence under the Bribery Act 2010 by a Participant or anyone employed by it or acting on its behalf (whether such breach or offence is with or without the knowledge of the Participant) shall entitle the Trust to disqualify the Participant.”

It is plain from the language and context of paragraph 4.7 that this is an express power to disqualify, but only in relation to any breaches of Section 4. It does not therefore confer an express power to disqualify on the basis of a late or non-compliant bid.

I now turn to Section 5 thereof which is headed “INSTRUCTIONS FOR SUBMISSION OF FINAL TENDERS”. Paragraph 5.1 states as follows:

“5.1

Summary of Response Requirements

The following table outlines the full set of response requirements for Final Tenders. These are

broken into three sections:

• Written Prompts - these are written responses to questions based on scenarios

provided.

• MIS Specification - this is the line item response to the requirements document used at

ITPD stage, with opportunity to update any original responses via clarification.

• Cost Model - this is your quote, based upon the structure provided…”

This is then followed by references to the descriptor rules for Appendices A-C and F and G and then the work sheet instructions for Appendices D and E.

Paragraph 5.2 then refers to the written responses required for Appendices A-C and F and G, in these terms:

“5.2

Written Responses

Please refer to the following Appendices which provide the context and format requirements for the response to each of these 5 prompts:

Appendix A: Localised Customisation

Appendix B: Data Flow to Central Office

Appendix C: Service Transfer and Programme Management

Appendix F: Integration with Third Party Products

Appendix G: Timetabling and the Flexible Working Agenda

In dialogue with Participants, the Trust may make revisions to the written prompts to best address details required. The Trust urges Participants to ensure they have full clarity on the requirements of each prompt to ensure the highest quality responses.”

Paragraph 5.3 then says this:

“5.3

MIS Specification

An MIS Specification was provided at ITPD stage for response. The requirements included in this document remain unchanged. We ask Participants to re-submit their response to the MIS Specification, making any changes to their submission to reflect clarifications provided via Dialogue. A final version of this document will become an appendix to the Contract.”

Paragraph 5.4 deals in detail with Appendix E and is not relevant here.

It seems to me that the reference to “clarifications provided via Dialogue” is a reference to clarification information given by UL. That is supported by the extract from paragraph 3.4 cited above.

However, even if paragraph 5.3 and that part of paragraph 3.4 refer to the opportunity for a bidder to amend their submission so as to reflect clarifications which it had provided, it still does not mean that any separate clarification document itself provided or to be provided by the bidder, is or becomes a mandatory document.

It is noteworthy that under Question 11 of the March Clarification Request, Bromcom was asked as follows, in relation to “Bromcom initial response to ITPD Appendix B - Data Flow to Central Office”:

“Your response states:

"Microsoft Azure with direct SQL access.The MAT Vision database is replicated in a live instance to a Microsoft Azure SQL database instance, which is available for direct SQL access if you so wish. This is protected behind a secure firewall."

Please provide additional technical detail in your response at the ITCD stage as to how this is working.”

In other words, the intent was that Bromcom should put its answer to this question in its actual tender response as opposed to in its March Clarification Response. In fact, Bromcom answered that question within its March Clarification Response. What it then did was to add some further information about pushing the data to UL’s data warehouse. See paragraph 85 above. But here, it also said something similar in its Appendix B - see paragraph 84 above. Otherwise, all the questions to both Bromcom and Arbor in the March Clarification Request concerned contractual matters and in particular their terms and conditions.

Bromcom relies on a number of matters emerging from Mr Wilson’s cross examination here. First it refers to what he said about the March Clarification Request in connection with the substantive push-pull issue. He said that he was expecting Bromcom to put the answer in its ITCD response and I agree with that as stated above. However he agreed that what Bromcom said in its March Clarification Response was a response to the question and that is obviously correct. He also agreed that UL could not properly mark Bromcom’s Appendix B response without sight of this document. In a sense, that does not go very far in the context of what I have said above, since it was Bromcom’s case that it had positively agreed to do the push (because of what it said in its Appendix E) despite what is said here. And again, in substance Bromcom did say in its Appendix B something similar to what it said in answer to Question 11 of the March Clarification Request.

Of more potentially direct relevance is the cross examination which then occurred later on, at day 4/147-177. At paragraph 52 (b) of Bromcom’s Written Closing at pages 29-31, many references to Mr Wilson’s answers in cross examination were cited so as to support the notion that in truth he accepted that the March Clarification Response was mandatory and formed part of the ITCD bid. I disagree. He did say that a number of the questions asked related to a particular Appendix. But that is hardly surprising since, on any view, if clarification was being sought, it had to be about something in the putative bid. So that does not prove very much. More importantly, he said that his understanding at the time was that all the questions in the March Clarification Request were really due diligence questions arising out of the parties terms and conditions. That was entirely the case with the Arbor version of the March Clarification Request. It was also true of 10 out of the 11 questions asked of Bromcom and yet again, question 11 specified that the answer there should go into the ITCD response. From Mr Wilson’s point of view, this was now a “parallel exercise” to the receipt of the ITCD responses. The questions were directed to matters to which UL - obviously - wanted answers, but in the context of contractual negotiations which would follow on from the decision to award the contract to one of the bidders. I agree with UL that this was a separate exercise and that the question of the terms to be agreed (or not) is a different matter from the evaluation of the ITCD tender itself.

This is why, according to Mr Wilson, the March Clarification Responses were not themselves sent to the evaluators - because they did not form part of the carefully described documents to be evaluated. Bromcom then turns this evidence on its head, in my view, to say that this made matters worse because in fact, the March Clarification Response should have been sent to the evaluators. But there were never any criteria set for how to mark the March Clarification Responses and Bromcom neither suggested what they might be or that the evaluators criteria were defective because they did not provide for the marking of this document, which would amount to essentially a review of the parties’ terms and conditions. All of that is unrealistic in my view. I reach that conclusion, even though I accept that the March Clarification Responses were included in the Slide Presentations for the moderation meeting.

Mr Wilson did agree that on some matters, UL would need to know whether a bidder would commit to certain important terms and this would need to be known before appointing that bidder as the preferred one. This, of course, was a somewhat abstract question. What Mr Wilson did not go on to accept was that the March Clarification Request was regarded by UL as very important and that there had to be a commitment on the various terms and conditions before deciding on a preferred bidder. He said, instead, and yet again, this was a parallel due diligence exercise. In addition, the question asked of him here actually assumed that the answers to the March Clarification Request had been sought “as part of Arbor’s final bid response” which rather begs the question.

So, in my judgment and for the reasons given above, the March Clarification Response was not a document required as part of the bid by reason of the ITCD document itself.

In the alternative, Bromcom submitted that in effect, and by its very issue on 10 March, it became a required document. First, Bromcom relies upon paragraph 18 of UL’s Defence which said that the contract was to be awarded according to the tender documents and “guidance”. Quite so, at least in certain respects, but that does not assist on the status of the March Clarification Responses. Second it says that the ITCD was issued before CD2 and CD3 and the other calls took place. Again, I agree, but that does not take the matter any further nor does the specific issue to the bidders of the March Clarification Request, as addressed above.

Then Bromcom relies upon the email of 10 March which enclosed the March Clarification Request it stated:

“Please find attached a list of clarifications (some bidder specific) captured over the recent dialogue meetings, calls and our further due diligence. Please provide answers to these with your final ITCD response.”

Mr Wilson agreed that he was telling the bidders to provide their March Clarification Responses at the same time as the tender response. He was asked what the position would have been if Arbor had actually refused to provide the document at all. He said that UL could not have contracted with “whoever was the winning bidder” without considering their March Clarification Response. However, again, this elides the question of making the ultimate contract and deciding who is the successful bidder according to the ITCD bid documents. Indeed, the question put to him assumed that the preferred bidder had already been appointed. Mr Wilson did not agree that the absence of this document would have made the bid non-compliant. In fact, he said that the bid would be compliant and (again) that these questions were all going to the next stage-see his evidence at day 4/176-177.

It is then said, correctly, that as soon as Mr Wilson saw that the March Clarification Response from Arbor was missing, he chased it and he accepted (as do I, especially in the light of Arbor’s email of 23 March cited at paragraph 33 above) that Arbor must only have completed the document on that day, once alerted to its absence. It may seem a subtle point, but in my view, the fact that Mr Wilson (a) asked for the document to be sent at the same time as the ITCD response and (b) chased Arbor when he found it was not there, does not entail the fact that it always was a mandatory document or that it became one.

Bromcom then submits that UL still should not have accepted Arbor’s March Clarification Response late, because it was not a clarification of an already submitted bid document which is what Regulation 30 (18) was contemplating, as was the last part of paragraph 3.5 of the ITCD. I agree and of course, that must follow since the March Clarification Response had been sent out before the bid documents had gone in.

But that does not mean that a contractual authority is powerless to receive a non-mandatory document after the deadline. In my view, and subject to its general obligations of equal treatment and transparency, it clearly can. And subject to compliance with those principles, it is then a matter for its judgment with the usual margin of appreciation.

If necessary, it seems to me that Regulation 56 (4), cited at paragraph 286 above, would apply so as to entitle UL to accept Arbor’s March Clarification Response when it did. In this context, I do not accept that the provision is only concerned with the clarification of an already submitted document. Moreover, the qualifications on the use of Regulation 56 (4) suggested by some of the cases were all stated in the context of its use as a way to accept late a bid or part of it i.e. its required elements.

For the reasons given above, I conclude that it was not unlawful for UL to receive Arbor’s Mandatory Clarification Response late.

However, I do consider that Bromcom should have been told that this had been done, so that UL complied with its duty of transparency. But had it done so, I do not think anything would have flowed from it. As a result of being given this information, there was nothing within Bromcom’s own bid which would need to be changed and there is no real suggestion that it could have used some extra time granted by UL (in the exercise of its duty of equal treatment) to perfect the bid it had already submitted. I suppose that Bromcom could have complained about the late arrival of the March Clarification Response but I do not think for one moment that, had it been told, it would, for example, have sought an immediate injunction to prevent the competition going further on the basis of Arbor’s non-compliant bid. Yet further, I have found that the bid itself was compliant.

So there is no viable claim in respect of the late submission of Arbor’s March Clarification Response.

In its Written Opening, Bromcom did not really focus on the separate question of the amendment of Arbor’s uptime document apart from saying that this was irregular and not a permitted clarification. The focus was principally on Arbor’s March Clarification Response.

For its part, in its Written Opening, UL contended that the revision was minor and made Arbor’s position look less and not more attractive, and that it was at best a permitted clarification or correction.

However, at trial, more was made of this point, including a line of questions on the basis that the drop in percentage xxxxx xxxxx xxxxxwas in fact material. This seemed to be allied to a point that in fact, Arbor’s position was so markedly changed that this specific matter should have been alerted to the evaluators. The amended uptime list was of course sent to the evaluators in the terms of the email referred to at paragraph 35 above. But I do not see why the evaluators needed to be alerted to a change in Arbor’s figures. After all, had the revised uptime document accompanied Arbor’s Appendix A when its ITCD response was submitted on Friday 20 March, it is not suggested that UL needed to take any particular action then. Indeed, because it was sent separately to the evaluators on 24 March on the basis that it was an amended document, it could be said that attention was specifically drawn to it anyway. The point is made that, perhaps UL and some or all of the evaluators did not appreciate the significance of an (only) xxxx uptime rate. But if that was right, this would have formed part of a challenge made much earlier and it was not. In any event, no specific consequences are said to flow from this.

There remains, however, the core point that the uptime document was part of the required bid documents and UL permitted an amended version of it to be submitted late. However, in my judgment, this did not mean that the bid was itself non-compliant or that UL had no power to accept the amended document.

First, all parts of the compliant bid were submitted on time. So this was not a case where, absent some action on the part of the contracting authority, the bid could have been rejected. Compare Denmark, Leadbitter and Slovensko above.

Second, this was in truth a correction but not, in my view, going to some important aspect of the bid documents notwithstanding Bromcom’s attempt now to characterise it as such. I consider that UL had the power to receive the correction under Regulation 56 (4), or indeed under Regulation 30 (17). I appreciate that these provisions refer to cases where it is the contracting authority which raises a point of clarification or about a missing document or incomplete information. However, applying a purposive approach, these must surely also cover a case where the contracting authority is alerted to an error which it would then ask to be corrected. In effect, that is what happened here, save that the process was shortened by Arbor which simply provided the correction itself.

In my judgment, it would be absurd to suggest that a contracting authority had no power whatsoever to receive a correction of any kind once the final bid had gone in. Indeed, I would take that view, even absent provisions like Regulations 56 (4) and 30 (17). That absurdity would be all the more obvious where the correction on its face, made the tenderers position worse not better. Yet further, it seems to me that in such circumstances, the tenderer would be obliged to notify the contracting authority of the error so that there was no risk of the assessment proceeding on a false basis.

Finally, a point was made here that UL had reserved to itself an express power to disqualify for a late or non-compliant bid. Even if it had such a power, I do not accept that this would alter the position or that it was obliged to exercise it. But in fact, the express power in Section 4 relied upon by Bromcom is irrelevant since it only operates in the case of a breach of that section – see paragraphs 316-317 above.

For all those reasons, UL was entitled to accept the corrected uptime document, in its discretion. The exercise of that discretion cannot be said to be unlawful here.

However, again, I do consider that UL should have told Bromcom that it had permitted the other bidder to lodge an amended document, perhaps even an amended uptime document, but of course it could not disclose its contents. Having been informed of this, it is extremely unlikely that Bromcom would have wished to do anything about it.

Accordingly, this element of the Timing Issue also fails.

I should add that UL says that even if it would be liable for the breach in respect of the timing of Arbor’s submission, the court needs then to take into account that Bromcom’s own bid was technically late. Therefore UL in fact should have disqualified Bromcom as well. In which case the counterfactual against UL goes nowhere.

There is nothing in this point. First, I have not found any causative effect on the issue of accepting Arbor’s documents late. Second, in my judgment, the deadline on Friday 20 March was midnight not, for example, 6pm. Therefore, on any view, Bromcom’s bid documents were not late. Third, even if the deadline was 6pm and the provision of the password was a necessary element of its bid, I consider that providing it at only 6.01pm is truly de minimis.

In the light of the procurement law breaches which I have found, I now set out what I believe to be the relevant counterfactual so as to reflect their causative effect. However, the parties should of course check the maths.

So far as Price Scoring is concerned I begin with Bromcom’s score. I have found, in paragraph 97 above, that £4,405 should be deducted. Bromcom’s original cost was put at xxxxxxxxx by UL. This becomes xxxxxxxxx.

As for Arbor’s costs, first it is agreed that £7,539 should be added. See paragraph 105 above. Second, I have found that the Rebate discount should be removed. See paragraph 160 above. In its Annex 2 at paragraph 146 (iii) of its Written Closing, Bromcom calculates the necessary deduction as xxxxxxx, being the total, over 5 years, of the figures at line 51 of Arbor’s growth costs model in the spreadsheet. That calculation appears correct to me. Adding those 2 items together, xxxxxxxx needs to be added to Arbor’s costs. These had originally been stated as xxxxxxxxx. That figure will therefore increase to xxxxxxx. That figure exceeds Arbor’s price by xxxx % which I round to xxxx. Xxxx xxxx xxxxxx xxxx xxxxxxArbor’s Price Score now becomes 340 instead of the original 376. I note that if this reduction in points is carried through without making any other adjustments for other infringements, Arbor would now have 838 points as against Bromcom’s 862. So Bromcom would have won. (I should add that if one applied the percentage as being xxx% exactly, Arbor’s Price Score points only rise by one, to 341).

I now turn to individual quality scores. I decided that certain scores should change. See paragraphs 225, 227, 232, 236, 239 246, 254 and 260 above. There is then a question as to how to factor those changes in, given the averaging challenge which I upheld in principle - see paragraph 185 above.

Bromcom suggests that this is done by adjusting the averages so as to reflect the scoring breaches, even though averaging is impermissible, and then rounding them. If this was done, each bidder would have scored 288 points. This would not only be an increase for Bromcom, which would be expected, but a decrease for Arbor even though its individual quality scores had not changed. This is simply a function of the averaging not being translated directly into points but being rounded first. The results of a counterfactual of this kind will be seen in Counterfactual Table 1 below.

This shows that now, in total, Bromcom would have won by 52 points. If the changes resulting from the Individual Quality score breaches are taken into account without the Price Scoring changes, the final result would have been 865 for Arbor and 881 for Bromcom, although I accept that some of the Individual Quality scoring breaches flowed from the pull/push issue and the £4,405.

On the other hand, if the averages were maintained right across to the points, Arbor would have retained its original score of 296. However, Bromcom would now score 287 and not its original 269. The total points awarded would be 837 for Arbor and 884 Bromcom. If one disregarded the Price Scoring breaches, the ultimate results would be 874 for Arbor, as before and 884 Bromcom. See Counterfactual Table 2 below.

A final alternative would be to avoid the use of averaging altogether and attempting to reach a counterfactual consensus score. The only way this could be done, as a hypothetical, would, in my view, be to take, for each Appendix, the whole-number score given to Bromcom by most evaluators. Using Bromcom’s Annex 2, but with the adjusted quality scores I have determined, this would produce a 4 for each Appendix. There is a potential problem with Appendix G because here, four evaluators gave Bromcom a 4, and another four evaluators gave it a 5. However, there were 3 scores of 3 and one score of 2. Accordingly, the correct approach would be to say the consensus would be 4 and not 5. If this exercise was carried out, the results would be as in Counterfactual Table 1.

I accept that this was not a scenario pursued at trial, but I cannot see any other way to estimate what the score would have been, absent averaging. It would be impossible now to create a moderation meeting which is any more nuanced.

However, the reality check is that however one tries to hypothesise on the Quality Scoring alone, Bromcom would have improved sufficiently to give it a higher score than Arbor, even disregarding the Price Scoring changes.

Once the Price Scoring breaches are taken into account, then, on any view, in the counterfactual, Bromcom would have won the bid by a much more significant margin than it had lost by, originally.

It is common ground between the parties that before any award of damages for the breaches of procurement law that I have found, it must be shown that they are sufficiently serious. This is what is known as the second “Francovich” condition whose application was confirmed by the Supreme Court in Energy Solutions v NDA [2017] 1 WLR 1373.

It is also common ground that I should deal with this issue now, although, of course, when the parties made their closing submissions after the trial, they could not know which, if any, breaches I would find. Both, nonetheless, made helpful submissions as to the approach I should take given the general circumstances of this case including the relative positions of Bromcom and UL and the nature of the procurement exercise in question.

Finally, I did not understand there to be any material disagreement in the parties closing submissions as to the legal principles to be applied.

Apart from the cases cited to me in the opening and closing submissions, on 20 September 2022, I was referred by UL’s solicitors to the decision of Mr Alexander Nissen KC, sitting as a Deputy Judge of the High Court, in Braceurself v NHS England [2022] EWHC 2348. I was invited by them to consider if I should receive further submissions from the parties in the light of that decision. In the event, I did not, and do not consider that to have been necessary. I will return to this case below.

The starting point for the assessment of sufficient seriousness consists of the 8 (non-exhaustive) factors identified by Lord Clyde in R (Factortame) v Secretary of State [2000] 1 AC 524, who noted also that in the end, the ultimate question is a matter of fact and circumstance. Those 8 factors were then summarised by Richards LJ at paragraph 36 of his judgment in Delaney v Secretary of State for Transport [2015] 1 WLR 5177 as follows:

“i)

the importance of the principle which has been breached; (ii) the clarity and precision of the rule breached; (iii) the degree of excusability of an error of law; (iv) the existence of any relevant judgment on the point; (v) the state of the mind of the infringer, and in particular whether the breaches were deliberate or inadvertent; (vi) the behaviour of the infringer after it has become evident that an infringement has occurred; (vii) the persons affected by the breach, including whether there has been a complete failure to take account of the specific situation of a defined economic group; and (viii) the position taken by one of the Community institutions in the matter.”

That approach was adopted by Mr Nissen KC in Braceurself. He also observed that at first-instance in Delaney, Jay J had said of Factortame at paragraph 84:

“What it is important to recognise at this stage is that: (i) the test is objective (p.554D) (if a government acts in bad faith that is an additional factor which falls objectively to be considered); (ii) the weight to be given to these various factors will vary from case to case, and no single factor is necessarily decisive; and (iii) the seriousness of the breach will always be an important factor.”

and noted that in the Court of Appeal, it was accepted by the parties that Jay J’s approach had been the correct one.

This “8 factors” approach was also adopted by Fraser J in Energy Solutions v NDA [2016] EWHC 3326 (TCC). There, he had found a number of breaches in the underlying procurement exercise and he then considered in a separate judgment whether they were sufficiently serious. He did so while the appeal to the Supreme Court on the question as to whether that test was to be applied (see above) was pending.

I was also referred to the judgment of Coulson LJ in Ocean Outdoor v Hammersmith and Fulham [2019] EWCA Civ 1462. At paragraph 83, he summarised the 8 factors. At paragraphs 84 and 85 he rejected a submission that any breach of the procedural elements of the Regulations (not in issue here) did not entail automatically that the breach was sufficiently serious. None of this adds to the principles to which I have already referred.

Equally, I was also referred to the case of R (Negassi) v SSHD [2013] EWCA Civ 151, which arose in the very different context of unlawful acting by the Secretary of State in relation to a claim made by an asylum seeker. Here, Maurice Kay LJ adopted the 8 factors test enunciated by Lord Clyde. Having carefully considered all those factors in the case before him, an exercise which he described as finely balanced, he held that the breach was not sufficiently serious. But none of that assists in the exercise I have to perform here. Nor does the particular decision by Patterson J in another asylum case, AD v Home Office [2015] EWHC 663. Indeed, she observed that the exercise of determining sufficiently serious is case-sensitive and can only be adjudicated upon in the light of all the evidence. See paragraph 65 of her judgment.

Two further points arise; first, where (as in this and many cases) a number of breaches have been found, the Court should look at the position having regard to the breaches collectively. It would be highly artificial, in my view, to take each breach separately. I did not understand either party to suggest otherwise. That does not mean that in the overall assessment the court should not consider the nature and extent of those breaches individually and their potential significance.

Second, there is the question as to whether there is now a tension between the decisions of Fraser in Energy Solutions and that of Mr Nissen KC in Braceurself. To explain this, I need to say something more about the facts of each case.

In Energy Solutions, Fraser J had found multiple breaches of a procurement process which had disastrous financial consequences for the unsuccessful bidder. But for the breaches, that bidder would clearly have won. Fraser J held that the breaches were sufficiently serious. In so doing, he answered the following question at paragraph 72 of his judgment:

“1.

Whether a failure to award a contract to the tenderer whose tender ought to have been assessed as the most economically advantageous offer, is in itself a sufficiently serious breach of the contracting authorities obligations to warrant an award of damages.

Answer: Yes”

He had previously found that the breach of this overarching obligation to award the contract to the most economically advantageous tenderer (based on the numerous defects in the procurement process he had found) itself engaged the first and second Factortame factors so clearly that there was, in effect, no need to go further.

Unsurprisingly, Bromcom relied upon that conclusion to contend that if, in the case before me the breach or breaches would have changed the outcome of the tender, that was enough to make out the sufficiently serious condition.

I now turn to Braceurself. In his previous judgment on liability, Mr Nissen KC found one minor breach based on two manifest errors, but a breach none the less, in a procurement exercise run by the NHS for the provision of orthodontic services for those who had a disability. He explained them as follows:

“11.

Overall, I was generally impressed by the careful way in which the evaluators had tried to carry out their functions. I also concluded that the procurement itself was carefully planned and well organised. Nonetheless, of the very many complaints made by the Claimant in these proceedings, I found one to have been justified. In particular, I found the Defendant made a manifest error in its scoring of question CSD 02 which led to it awarding the Claimant a score of 3 (good) in respect of that question rather than, as I found it should have done on the evidence before me, a score of 4 (excellent). In summary, CSD 02 was concerned with Clinical and Service Delivery. One aspect of this multi-faceted criterion concerned accessibility to the premises. The Claimant’s premises were on the first floor which meant that its bid needed to cater for those patients who could not use the stairs to access the service. In addressing this part of the Claimant’s bid, the Defendant made two errors. The Claimant had proposed to use a device called a stair climber. Mistakenly, the Defendant evaluated the Claimant’s bid on the basis that it was proposing to install a stair lift. The Defendant also mistakenly thought that, by way of partial solution, the Claimant was offering services at alternative premises at least to those patients who could not use the stairs. The suggestion was that the equipment at the alternative premises would not be of the same standard. In fact, the Claimant was only making an offer to use alternative premises as a result of a flood or fire rendering its primary site unusable. I concluded that these mistakes had a causative impact on the Defendant’s scoring and were material to the outcome. Whilst I acknowledged that the Defendant was generally entitled to a margin of appreciation in its scoring of the criterion as a whole, I concluded that such margin was not relevant in determining whether a straightforward misunderstanding of the bid had taken place since that was not a question of judgment or assessment.

12.

The Court was well placed to reach a conclusion as to the appropriate score which ought to have been given to the Claimant. It was clear that the issue about access had, in fact, impacted negatively on the score which the Claimant received in respect of CSD 02 and the Court was in a good position to reach its own conclusion about the appropriate score in light of the evidence as a whole. The consequence of changing

the score from 3 to a 4 in respect of CSD 02 was to increase the Claimant’s total bid score by 2.5% in circumstances where the difference between the two bidders had been 2.25%. But for the manifest error, the Claimant would therefore have been awarded the contract, having scored 0.25% higher than the other bidder. This was not, therefore, a loss of a chance case.”

This was therefore a case where the difference between the winning and losing party had originally been very small. The same, of course, could be said about this case.

In assessing the 8 factors, Mr Nissen KC first rejected the notion that it does not always follow, where the outcome would have been different absent the breaches, that the sufficiently serious condition is made out. At paragraph 44 he said this:

“Having said that, its importance when set against the other factors remains a question of fact and degree. Although the Claimant has submitted that the failure in this case to award the contract to the operator offering the most economically advantageous tender, without more, constitutes a sufficiently serious breach entitling it to damages, I reject that submission. The mere fact that the principle breached is important cannot, of itself, be determinative. The same can be said of the second factor considered immediately below. As was clear from Factortame and Delaney, no single factor is decisive.”

For my part, and at least for present purposes, I am content to adopt that position. The findings made by Fraser J in Energy Solutions were made in the circumstances of a very different case and he would not have had in mind what I regard to be a very extreme case going the other way, namely Braceurself. Indeed, this is effectively what Mr Nissen KC said, in paragraph 88 of his judgment:

“Thus, the Claimant submits that since here an individual breach has altered the outcome of the competition, it is a sufficiently serious breach on the facts of this case, as it was in Energy Solutions. The Claimant argues that both of the first and second Factortame factors are in its favour and no others displace it, as in Energy Solutions. This is an attractive argument but, in the end, I do not agree with it. Firstly, I have found some factors operate in favour of the Defendant. Secondly, it is right to point out that Fraser J was not articulating a proposition of law that on every occasion in which a single breach has affected the outcome of the competition, the breach is necessarily sufficiently serious, irrespective of other factors. Indeed, Mr Holl-Allen accepted that Fraser J was not articulating a proposition of law to that effect. In any event, at [43], Fraser J positively excluded from his consideration a single breach case which had a powerful effect on the final score, which is just this case. I quite accept that the fact that the outcome of the competition would have been different but for the breach is a highly material one, which I take into account, but it cannot necessarily be determinative. In my judgment, it must also be relevant to consider, and weigh in the balance, the fact that the competition was very close so that even a small change had a significant effect on the outcome. The whole point about an evaluation of the sufficiency of the seriousness of the breach is that one should be able to take account of extent and degree. A breach, or series of breaches, which impacted upon the outcome of the competition (i.e., produced a different winner) where the score was increased by a mere 0.25% may be treated differently from a breach which impacted upon the outcome where the score was increased by something significantly higher than that.”

In the event, Mr Nissen KC found the sufficiently serious test was not satisfied. This was because there was a single breach in a very close competition which happened to have a powerful impact on the outcome, with the two errors being due to misunderstandings. The breach was inadvertent and in good faith. The purpose of the procurement was laudable, because of the publicly funded orthodontic services to be offered to those with disabilities. The procurement exercise was generally well-planned and well-organised. While Braceurself might well suffer financial loss as a result of not being awarded this contract, this was not an existential impact. The (wrong) outcome of the procurement exercise had no material impact on the wider public access to dental treatment, compared with the position had Braceurself won. Finally, it was far removed from the multiple breaches concerned with a multi-billion pound contract for nuclear decommissioning which was the subject of Energy Solutions.

That is a helpful example of the sort of reasoning process to be undertaken when applying the 8 factors. But of course, precisely how much weight is to be attributed to each factor and how that affects the overall question of sufficiently serious is highly dependent on the facts and circumstances of each case, as they are assessed and taken into account by the court in question. In the end, it is a balancing exercise, in my view.

I should add that in this case (as also in Braceurself) the contract was awarded not simply to the most economically advantageous tender in a narrow sense. That is because there was a significant quality component and indeed, from a pure price point of view, in this case, Bromcom’s price bid was always more advantageous. But all this means is that the question of the ultimate outcome is better described as the bidder which has scored the most points overall. I did not understand either party to be submitting that reframing the question in this way affects the exercise. And of course, in a broader sense, the quality criteria also went to the question of financial advantage since they were concerned with the ease and efficiency with which the existing schools could migrate to and thereafter and use the new MIS.

In the light of the above, I now determine the question of sufficiently serious here.

The Importance of the Principle which has been breached

This can be examined at two levels, in my view. First there is the ultimate principle of awarding the contract to the tenderer with the most points being a blend of price and quality. As I have already said, this should be accorded the same importance as the notion of awarding the contract to the economically most advantageous tender. The importance of that principle, which is a function of the importance of the role of fair competition, is clearly very significant. Indeed, on this factor in Braceurself, Mr Nissen KC held that without more, it operated in Braceurself’s favour since on the counterfactual, it would have (just) won.

But there is a second level which is concerned with the nature and extent of the individual breaches and their contribution to the outcome. This was not something which Mr Nissen KC had to consider since, in effect, there was only one breach. It perhaps does not matter where within the 8 factor exercise this is done (in Energy Solutions, Fraser J dealt with them in the context of considering their individual seriousness) so long as it is done, since it is plainly relevant.

Here, the first point is that there were, in my view, four separate types of breach although not all had a distinct or any adverse outcome as a matter of causation.

First, there were the Price Scoring breaches. Taken by themselves, they affected the outcome (see paragraph 357 above). Indeed, just the Rebate error would have done this and of course there was in addition the admitted £7,539 error.

Second, there were the Individual Quality scoring manifest errors. As to those:

Third, there was the averaging approach which entailed (or at least contributed to) the underlying breach of the important principle of transparency. The fact is that (a) there was no moderation meeting in a real sense, (b) as a result, the evaluators never attempted to reach a consensus score and (c) no clear reason could be proffered for any particular score. The lack of these factors has made it very difficult to determine a counterfactual, shorn of all other breaches. But doing the best I could (see paragraphs 359-363 above) in the absence of averaging there would have been some alteration in the outcome.

However, in my judgment, the real point is that this was a serious infringement of an important, indeed fundamental principle. It was not helped by Ms Wood’s failure to disclose it to the parties when they first asked about it. See paragraph 165 above.

I appreciate that it was said that the averaging approach was thought to be a good idea at the time because it might then provide for more nuanced differences between the bidders on scoring where they were so close on the whole number scores. But I do not think that goes very far given the effect it had, and given that the exercise was to achieve whole number scores, save that overall, I do not consider that UL was acting in bad faith here in adopting the averaging procedure.

Fourthly, there are the breaches concerned with the form and timing of Arbor’s submissions. Although they went nowhere in terms of causation, they should be taken into account because there was an obvious step that should have been taken as a matter of fair process which was to inform Bromcom of what had happened. It would not have taken very long and in the event would not have derailed or even affected the procurement process going forwards. Overall, then, both at a higher and lower level, this first factor operates very strongly in Bromcom’s favour.

The Clarity and Precision of the Rule Breached

Again, I approach this on two levels. At the higher level, the principle of awarding the contract to the winning tender is obvious and underlies the whole procurement process. There is no difficulty understanding that. This itself would be in Bromcom’s favour though it does not add much given the outcome of the first factor.

The real point here is at the lower level. Like the exercises undertaken in Energy Solutions and Braceurself, I need to consider the position by reference to the individual breaches or defects.

As to Price Scoring, I consider first the data-flow breach involving the £4,405 addition. This error arose from failing to act in relation to a material uncertainty in Bromcom’s bid which was picked up explicitly at the time as a matter requiring clarification and then taken no further. See paragraphs 89-91 above. So it cannot be said that the underlying “rule” was in some way complex or difficult to follow. And while the sum involved was only £4,405, it of course had a knock-on effect in relation to the individual quality scores on Appendix B, as already noted.

Of course, the much larger item in terms of cost was that relating to the Rebate. This might legitimately be regarded as a somewhat more subtle matter. It involves the proper approach of the rules relating to incumbent advantage and ensuring that a bid related only to the contract in question, which I accept are perhaps not straightforward. All one can say here is that in a general sense, UL obviously had to apply at least some caution to Arbor’s bid given that it was the incumbent supplier to the 15 other schools.

As for the individual Quality Scoring errors, as these arose from a misunderstanding of the documents or information provided, there is not much more to be said. I think the evaluators who went wrong here should have understood what they were actually being provided with, but again, I do not suggest here that they were being wilfully obtuse or anything of that kind.

The averaging error is more problematic for UL because of the duty to give reasons which is so basic and which is itself simple to understand. It meant that UL had, in my view, some difficulty explaining to Bromcom some of the score differentials. Moreover, for the purpose of the trial, UL then felt it necessary to call a large number of witnesses to explain their individual thinking. This would not have been necessary had there been a clearly adopted (and minuted) consensus score for each appendix.

Equally, it should have been obvious to Mr Wilson that he should have alerted Bromcom to what he had done in accepting Arbor’s March Clarification Document (even if it was not mandatory) and its revised uptime figures, on Monday 23 March.

Overall, I think that the underlying rules were clear though not universally so. So this works in favour of Bromcom, or at least not against it.

The Degree of Excusability of an error of law

Here, I take the same approach as Mr Nissen KC in Braceurself who said this at paragraph 54:

“Whilst the very concept of an “excusable breach” is a slightly odd one, it is clear that this factor should, in a given case, have regard to why the breach occurred and whether, whilst it was nonetheless a breach, it is at least understandable why it occurred. Mitigating factors for the occurrence of the breach can be considered.”

To some extent, I have considered this under factor (ii) above. But there are other matters of potential relevance. The first is that some of the evaluators may have lacked rigorous training although I have not upheld a general breach in that regard. But if that is so, this cannot be held against Bromcom.

The second is that in any event, they were busy teachers volunteering to do this important work and in the middle of the exercise, the Covid lockdown rules came into effect. UL is not some large local or central government department. I, of course, have considerable sympathy for the position in which UL found itself.

But I do not see how this, in truth, adds anything to the exercise here, especially as there was no real evidence adduced that the Covid (or other situation) in which the evaluators found themselves was responsible for any errors they made. Nor was there evidence that the position of UL itself as an institution contributed to any errors. Perhaps that is understandable where, of course, UL’s case was that it made no errors at all. Nonetheless, that is the position. So I do not take it into account. Moreover, it has to be remembered that it did engage a professional consultancy to advise and assist on the procurement through the involvement of Ms Wood. But even if I were to take UL’s situation generally into account it would not be of such weight to make any difference to the ultimate outcome on the question of sufficient seriousness.

So it is not possible to compare this case to one in which there was one minor breach which was excusable - i.e. Braceurself - even if the breaches were not as numerous as in Energy Solutions.

Overall, this factor operates in Bromcom’s favour, or at least not against it.

The Existence of any Relevant Judgment on the point

This is a difficult factor to apply, once one seeks to go beyond judgments which affirm the underlying principles of fair competition, transparency and equal treatment. I do not think this factor takes the matter any further for either party. It is neutral.

The State of Mind of the Infringer

Obviously, if a breach was deliberate and/or in bad faith, that would count significantly against the contracting authority. None of the breaches before me are of that character. So this does not reflect adversely on UL. But on the other hand, I do not see why it should count positively in favour of UL. In Energy Solutions, Fraser J concluded that this factor simply did not apply where there was no bad faith.

In Braceurself, Mr Nissen KC took a different view but was clearly affected by an additional factor. This was that the purpose of carrying out the scoring in the NHS’s procurement exercise was itself to maximise public access to dental services for those with disabilities which was a laudable intended outcome. See his paragraph 69. While that is true, I am not myself sure how that plays into any evaluation of the question of sufficiently serious breach because it is not really concerned with the state of mind of the infringer or any other of the 8 factors.

However, in any event, the case before me does not have the same public policy features. Of course, the procurement exercise here is all about state schools and how they can operate more efficiently and proactively by using an improved cloud-based MIS Software. But that does not really add anything, in my view. I should point out that here, Bromcom actually relies upon this feature of the procurement as a point against UL. It says that although the sums involved are relatively modest (a total revenue of £10 million for the successful bidder over 5 years) the project to introduce new software to 57+ schools was the biggest of its kind.

In truth, I do not think that the aim or the particular features of this exercise add anything either way. One is therefore left with the absence of bad faith. I consider this to be neutral.

The behaviour of the infringer after it has become evident that an infringement has occurred

There is nothing of relevance here, in my view. It is a neutral factor.

The persons affected by the breach, including whether there has been a complete failure to take account of the specific situation of a defined economic group

As for Bromcom, the result will obviously have some financial impact upon it. But it is not said to be existential. In other words, there is nothing to suggest a financial or other impact beyond what will often be the case when a bidder loses the competition. So I do not attach much weight to this factor.

There is then the question of any wider community interests being affected. This would only arise if it could be said that the award of the contract to Arbor will in fact produce a material lesser result for UL and the schools acquiring the new software, than if Bromcom had won. There is no evidence that this would be so. Given that the original scores of Arbor and Bromcom were close, it would be surprising if it were otherwise.

The position taken by one of the Community institutions in the matter.”

This does not arise.

I would make the following further observations. I accept that Bromcom made a number of challenges which I have rejected altogether. This was perhaps inevitable where the scores were so close because it would obviously prompt an enquiry into the detail of the evaluator of process. But on the existence of sufficient seriousness, it adds nothing. I am not concerned with the breaches which were not established but rather with those which were.

Second, the counterfactual shows that the “new” outcome would reveal that Bromcom would have been the clear winner here and by many more points than it had lost by in the actual outcome. That is relevant, in my view.

Finally, this is not a case where I could say categorically (as Mr Nissen KC did) that this was a generally well-run procurement exercise. It was, in many respects, but not all, by reason of the breaches I have found, whether causative or otherwise.

I have looked above at each individual factor and explained how it affects one party or other, or neither. But I then have to weigh up all of the factors and the circumstances to determine the outcome of this exercise, which I have done.

In my judgment, the outcome clearly lies with Bromcom and establishes that, taken together, the breaches I have found were sufficiently serious. This follows from the analysis I have undertaken and reasons given, above. In summary, and at the end of the day, there were numerous breaches and collectively they would have meant that the result would have gone the other way by some margin. There were, in truth, no real factors in favour of UL such that the outcome of the sufficiently serious analysis should be in its favour.

The relevant limitation period is 30 days, set out as follows, in Regulation 92 (2):

“(2)

Subject to paragraphs (3) to (5), such proceedings must be started within 30 days beginning with the date when the economic operator first knew or ought to have known that grounds for starting the proceedings had arisen…”

Although there is provision for the grant of an extension of time, none was sought here. Accordingly, since proceedings were started on 18 May 2020, if Bromcom had the relevant knowledge prior to 18 April, the relevant claims would be time-barred.

I deal further with the law as to the nature and extent of the knowledge required, in paragraphs below.

Regulation 86 provides as follows:

“86.

—(1) Subject to paragraphs (5) and (6), a contracting authority shall send to each candidate and tenderer a notice communicating its decision to award the contract or conclude the framework agreement.

Content of notices

(2)

Where it is to be sent to a tenderer, the notice referred to in paragraph (1) shall include—

(a)the criteria for the award of the contract;

(b)the reasons for the decision, including the characteristics and relative advantages of the successful tender, the score (if any) obtained by—

(i)the tenderer which is to receive the notice; and

(ii)the tenderer—

(aa)to be awarded the contract, or

(bb)to become a party to the framework agreement,

and anything required by paragraph (3);

(c)the name of the tenderer—

(i)to be awarded the contract, or

(ii)to become a party to the framework agreement; and

(d)a precise statement of either—

(i)when, in accordance with regulation 87, the standstill period is expected to end and, if relevant, how the timing of its ending might be affected by any and, if so what, contingencies, or

(ii)the date before which the contracting authority will not, in conformity with regulation 87 enter into the contract or conclude the framework agreement…”

Regulation 87 states that where Regulation 86 applies, (as it did here) the contracting authority must not enter into the contract awarded to the successful tenderer until the end of the standstill period. Where a Regulation 86 notice was sent by electronic means, the standstill period ends at midnight on the 10^th day after the day on which the notice was sent.

This case is somewhat unusual in that there has already been one detailed decision by a judge, while these proceedings were well underway, to the effect that none of the claims made were time-barred. This was the interlocutory decision of Judge Eyre in Bromcom v UL [2021] EWHC 18 (TCC), given on 7 January 2021 in relation to an application made by UL to strike out or summarily dismiss the claims on the basis that they were time-barred. He held that they were not. As I will show, the materials for consideration by Judge Eyre were essentially the same as those which I have been invited to consider, although there was some live witness evidence as well before me.

Of course, I am not strictly bound by Judge Eyre’s decision, but given that I am effectively being asked to conduct the same analysis on the basis of essentially the same material, his decision is at least a significant starting point, in my view. Where I refer to paragraph numbers in relation to his decision, they are the paragraph numbers of his judgment.

I should now set out in more detail the relevant materials.

This was a brief letter. It told Bromcom that it had scored 862 while Arbor (though not yet named as such) scored 878. It added a breakdown showing the respective scores out of 400 on cost and out of 600 on Quality. The latter was broken down into the 360 possible points for written questions and 244 the product scores and the respective scores out of those totals were also given.

I agree with Judge Eyre who referred to this letter in paragraph 59 as “sketchy”.

This first feedback call had been offered in the email of 30 March and it was taken up. In it, Ms Wood said that they had added up the total costs of ownership in respect of both parties bids and in the case of Bromcom, the total came to £548,870. She said that costs were added for the setting up of the data interface at £4,405 which was small, being the cost of UL setting up the interface for use at its end. Mr Lewin asked if those costs had been added to Arbor’s bid as well. Mr Wilson said that the two bids proposed “different solutions” so UL did not do the same for Arbor’s bid but they considered the total cost of the implementation of the interface for both bidders. Ms Wood then went on to say that they added costs at £500 per day in respect of FTF for growth schools and Mr Lewin said “okay, I understand what you’re saying”. Later, Ms Wood said that these costs came to £42,000 and again referred to the adjustment for £4,405 and to the incumbent licence costs of £548,874.

The discussion then moved onto other matters, including the scoring process for quality. This is where Ms Wood described the moderation in which the panel sought to achieve a consensus which, as already noted, was not a correct description.

Later, Ms Wood said that the purpose of the call had been to go through the strengths and weaknesses of Bromcom’s bid but they needed another conversation. This was scheduled for 3 April.

An email sent by UL to Bromcom on 2 April set out the weightings of the marks for Appendices A-C and F-G. It also set out a “Costs Evaluation Procurement Clarification”. This document pointed out that the TCO worksheet had required the bidders to fill in details of their charges and that UL would add incumbent licence costs as instructed on the worksheet. It said that this process had been applied fairly to both bidders.

During this call, there was a further reference to the interface with the data warehouse when Mr Ali Guryel asked if UL’s approach had been to “pull the data to data warehouses”. Mr Wilson agreed, saying it was one part of a smaller strategy, but the discussion did not develop from there. There was then much discussion about scoring on the written responses.

On 7 April, Mr Lewin wrote to Mr Wilson, asking in particular for the other bidder’s MIS charges. He said it was highly unusual to have hidden costs elements added to each bidder’s price. He was asking for more details of Arbor’s bid. UL responded on 8 April but did not actually provide any new information on each bidder’s costings. In its reply to that, Bromcom again sought information as to the amounts added to each bidder’s bid. UL sent a further response on 9 April (though the letter is actually dated 8 April) declining to provide any further information on the basis that to do so would enable Bromcom to ascertain Arbor’s raw price.

On 14 April, Bromcom’s solicitors wrote to UL for the first time. As this is an important letter, relied upon by both sides, I quote almost all of it here:

“We have seen the previous correspondence between United Learning (‘UL’) and Bromcom. In response to your letter of 9th April 2020 to Bromcom, we understand that during the debriefing session in relation to the procurement exercise on 1 April 2020, UL informed Bromcom of the extra charges that UL made to Bromcom’s bid costs (in relation to the deployment proposal submitted) before UL completed the overall cost evaluation. This comprised of three elements of deployment costs, namely:

1.

an extra charge for training because Bromcom did not include an on-site training charge for the additional schools that UL included in their Growth COST model;

2.

an extra charge for implementing the data warehouse interface that UL staff would incur; and

3.

an extra charge for the SIMS licences required for the period May 2020 until the schools are deployed according to Bromcom’s deployment proposal. This plan being developed to meet the guidance set out in the requirements given by UL to bidders.

During the same meeting, UL informed Bromcom that the extra charge it added to Bromcom’s bid for the SIMS licences required (element 3 above) was a very significant figure. Bromcom was told by UL the extra charge was a total of £548,870.00, which makes it 26.7% of the 5 year total bid figure used to evaluate Bromcom. This was significantly higher than expected by Bromcom and it has a significant impact on the overall cost figure used in the cost evaluation. Bromcom, therefore, would like to understand why the figure is significantly higher than expected and how it compared to the preferred bidder. As the margin between the overall scores achieved by both Bromcom and the preferred bidder is so small, these extra charges in relation to the SIMS Licences added to cover the deployment period could have had significant impact on the overall outcome of the procurement exercise. Bromcom’s experience and ability would have allowed them to deploy its product faster at a lower extra charge, but this approach was rejected at the previous ITPD stage of the bidding process, and Bromcom was asked to extend the deployment over a longer period. As UL gave guidance and set out requirements to bidders that were quite prescriptive for deployment, Bromcom would expect that a similar surprise would be experienced by the preferred bidder in relation to the deployment costs.

As shown above, the deployment costs may have had a meaningful impact on the overall cost evaluation and outcome of the procurement exercise. Could UL please provide the following:

•

exactly how UL evaluated and attributed the extra deployment costs, how the extra charge for the SIMS licences was calculated and notes of the evaluation meetings;

•

the extra deployment costs attributed to Bromcom;

•

the extra deployment costs attributed to the preferred bidder; and

•

the relative characteristics or advantages of the preferred bidder’s deployment plan.

Could UL please provide the above by 20th April 2020?”

The next correspondence from UL was by way of its letters dated 22 and 23 April. The email attaching the letter of 22 April was said to be “issued to you so as to comply with the requirements of [the PCR] and in particular Regulation 86.” It added that a separate letter would be written to explain the financial evaluation process in more detail, noting the sending of the 14 April letter.

The letter of 22 April was much fuller than that of 31 March and it said that it was a substitute for the latter which should be disregarded. It was also said to be issued so as to comply with the requirements of the PCR and Regulation 86. These references to Regulation 86 must be to the need to set out reasons as well as provide a (now-revised) standstill period. This letter also now identified Arbor as the winning tenderer.

Unlike the 31 March letter, the 22 April letter included the Appendix Scores. It now also corrected Arbor’s Price Score from 380 to 376. There followed a detailed analysis of the comparative merits of each bidder’s response to the relevant Appendices and also their Product Demos.

The separate letter of 23 April ran to 24 pages, dealing with price scoring and stating that it was a response to what had been requested in the 14 April letter. It then referred to information given at the calls on 1 and 3 April. Here, however, all the figures and reasoning were set out more clearly and coherently. It was also a further explanation because, for example, it included a worked example for incumbent charges using Bromcom’s TCO worksheet, with such an example being added in enlarged form to the end of the letter. Then, the F2F issue was set out, explaining why, according to UL, Bromcom’s bid was non-compliant, while Arbor’s was, in respect of F2F for the growth model. The make up of the £42,000 total was then given.

As for extra deployment costs being added to Arbor’s bid, of course, as we know, there were none. Nor was UL prepared to state what those costs might have been. But it did give a worked example of how one could ascertain Arbor’s overall price by reference to the scores which had been given and which implied, for Arbor, a xxx% uplift on Bromcom’s cost. Then a notional negative adjustment of £500,000 was applied to Arbor’s bid price which would now reduce it to below Bromcom’s. The letter concluded that it had explained the relative advantages of Arbor’s mobilisation plans.

Bromcom accepts that following the 22 and 23 April letters, it had sufficient knowledge for limitation purposes - but not before. It is not necessary, therefore, to consider the correspondence after 23 April and prior to the issue of the Claim Form.

Here, I can do no better than set out Judge Eyre’s conclusions on the law, seen since they seem to me, with respect, to summarise it correctly (a view also expressed by Coulson LJ when refusing permission to appeal):

“12.

In Sita UK Ltd v Greater Manchester Waste Disposal Authority [2011] EWCA Civ 156, [2011] 2 CMLR 32 the Court of Appeal set out the approach to be taken in determining both the matters of which knowledge is required for the purposes of regulation 92(2) and the degree of knowledge required. The majority of the Court approved the test formulated by Mann J at first instance namely: “the standard ought to be knowledge of the facts which apparently clearly indicate, though they need not absolutely prove, an infringement”.

13.

I derive assistance in applying that test from the context of its approval and adoption by the Court of Appeal. Thus at [22] and [23] Elias LJ referred to the range of degrees of confidence which a potential claimant may have in its prospects. It may know variously that it has “an arguable case, a reasonably arguable case, a strongly arguable case, or even a certain case”. At [30] Elias LJ rejected the contention that time did not run until a claimant knew that it had “a real likelihood of success”. He drew a distinction

between knowledge of “the detailed facts which might be deployed in support of the claim” and of “the essential facts sufficient to constitute a cause of action” indicating that knowledge of the latter rather than the former would be sufficient to start the 30 day period. It was in that context that Elias LJ drew attention to “the principle of rapidity” which is “at the core” of the timetable laid down by the Regulations.

14.

It is also important to note that the members of the Court of Appeal were agreed that the starting point was the test laid down by the ECJ in Uniplex (UK) Ltd v NHS Business Services Authority [2020] 2 CMLR 47 at [30] – [31] namely that time runs from when a claimant has: “…come to an informed view as to whether there has been an infringement of the applicable provisions and as to the appropriateness of bringing proceedings.”

15.

The division between the members of the Court of Appeal was as to whether any elaboration or clarification of that test was needed. Arden LJ took the view that elaboration was not required. The other members, however, concluded that the test as formulated by the ECJ did not of itself explain “what degree of knowledge is sufficient to provide that informed view that a legal claim lies” (per Elias LJ at [23]) or “how well informed the informed view has to be” (per Rimer LJ at [92]). So it is to be remembered that the test formulated by Mann J and approved by the majority of the Court of Appeal is a test for the purpose of determining the degree of knowledge necessary to form that informed view.

16.

It follows that what is needed is knowledge of material which does more than give rise to suspicion of a breach of the Regulations but that there can be the requisite knowledge even if the potential claimant is far from certain of success. Answering the question whether the facts of which a potential claimant was aware were such as to “apparently clearly indicate” a breach of duty by the contracting authority will require consideration of the nature of the procurement exercise; of the nature of the particular breach alleged; and of the nature and extent of the particular factual material.

17.

At [36] – [38] Elias LJ explained that a breach by a contracting authority of its duty of transparency will not prevent the start of the thirty-day period if the potential claimant has sufficient knowledge notwithstanding that breach. Conversely if the withholding of information by the contracting authority means that a potential claimant does not have the requisite knowledge then time will not begin to run. In that regard and generally it is to be remembered that the focus is to be on what the potential claimant knew at the relevant time “and not on what it did not know” (per Elias LJ at [75]).

18.

At [88] – [89] Elias LJ concluded that it was not necessary in Sita to resolve the question of whether the effect of the decision in Brent LBC v Risk Management Partners Ltd [2009] EWCA Civ 490 was that the Regulations imposed only a single duty (namely to comply with the required procedure). At [89] he addressed the situation where there are multiple allegations and drew a distinction between the situation where the allegations are of breaches of the same duty and that where “a number of distinct duties can be

spelt out of the procurement obligations”. If the allegations are on proper analysis different breaches of the same duty then a potential claimant has the requisite knowledge when it knows or ought to have known of facts clearly indicating a breach of that duty. The time period is not extended simply by the potential claimant learning at a later stage of further separate breaches of the same duty even if they occurred “before or after the breaches already known”. If, however, the potential claimant learns of facts indicating a breach of a different duty then it may be the position that time begins to run anew in respect of a claim alleging a breach of that duty.

22… There is a difference between cases where challenges are made to separate decisions taken at different stages in a procurement exercise and those where a challenge is mounted to a single decision (typically the outcome of the exercise). The passages in Riverside Truck Rental at [39] – [42] and [60] – [61] to which Mr. Bates referred were contemplating the circumstances in which there could be separate challenges to the various decisions taken at a number of stages in a procurement process. As Fraser J explained in SRCL Ltd v National Health Service Commissioning Board [2018] EWHC 1985 (TCC), [2019] PTSR 383 the nature of the procurement process is typically such that there will be distinct acts or omissions at different stages in the process each of which could be susceptible of a separate challenge under the Regulations. That is why it is possible to have a number of claims issued in relation to the same procurement exercise. In such circumstances time will begin to run at different dates and the period under regulation 92 (2) will expire at different dates in relation to the different claims. That is not the position where the challenge is to a single decision or to a single aspect of the process even if it involved a number of related acts or omissions over a period of time. Thus in Sita various allegations were made as to acts or omissions at different dates but as Elias LJ noted “they all went to the failure to reopen the bidding process.” In those circumstances the effect of Elias LJ’s judgment is that the particular breaches are, at the very lowest, unlikely to be separate causes of action but will normally fall to be seen as particulars of the same infringement even if occurring at different times. The court has to consider whether distinct breaches of separate duties are alleged. However, if the allegations are in reality of breaches of the same duty in respect of the same decision (as Elias LJ’s words indicate they are likely to be) the effect of Sita is that time does not start to run afresh when a potential claimant learns of further breaches of the same duty in relation to the same decision. Here the challenge is to a single aspect of the procurement exercise namely the evaluation process which led to the contract being awarded to Arbor rather than to the Claimant and that is the context in which the assessment of whether the allegations are of breaches of separate duties or of a number of breaches of the same duty in relation to the same matter has to be undertaken.

23.

As in Sita it is not necessary for me to determine whether as a matter of law there can only ever be a single duty under the Regulations in relation to a single decision. It is to be remembered the matter is before the court on the Defendants’ application for the striking out of the claim alternatively summary judgment. If at this stage the breaches alleged are on a proper analysis tenably put forward as breaches of separate duties then it will be appropriate to consider the date of knowledge separately in respect of each such duty.”

54…iii) Reference was made to the decision of Coulson J in Roche Diagnostics Ltd v The Mid Yorkshire Hospitals NHS Trust [2013] EWHC 933 (TCC). Mr. Bates relied in particular on [20(a) and (b)] where Coulson J said that the “broad principles” governing applications for early specific disclosure in procurement cases included:…

(b)

That this should be the general approach is confirmed by the short time limits imposed by the Regulations on those who wish to challenge the award of public contracts. The start of the relevant period is triggered by the knowledge which the claimant has (or should have) of the potential infringement. As Ramsey J said in Mears Ltd v Leeds City Council [2011] EWHC 40 (QB), ‘the requirement of knowledge is based on the principle that a tenderer should be in a position to make an informed view as to whether there has been an infringement for which it is appropriate to bring proceedings’”

It follows from what Judge Eyre stated at paragraphs 12-15 that, while the observations of Mann J were adopted by the majority in Sita as helpful clarification of the degree of knowledge required, the nature of the knowledge in question remains as set out in Uniplex, that is to say knowledge as to (a) whether there has been an infringement and (b) the appropriateness of bringing proceedings.

As for the question as to whether the challenges in this case are, in my view, all concerned with breaches of the same underlying duty or relate to breaches of different duties, for reasons which will become apparent, this is not a matter which I need to determine.

In its submissions following trial, UL made the following key points:

Bromcom says there is nothing in any of these points, with which I now deal.

I should add that UL does not take any limitation point in relation to the separate challenges made by Bromcom as to the form and timing of Arbor’s ITCD submissions dealt with in paragraphs 280 - 354 above.

First, I agree with Judge Eyre that there is no legal impediment to the relevant knowledge being acquired from oral conversations as opposed to documents. However, the form of the communication of information can be significant, as he noted in his paragraph 57.

As to the meeting on 1 April, Judge Eyre concluded that, among other things, the individuals present on 1 April were sometimes talking over each other and the transcript makes this clear. I also heard it on the audio recording to which I listened in its entirety. Judge Eyre also inferred that it was a “heated” meeting from the fact that Ms Morgan had made a reference to it as a “passionate” session. I would not call it passionate in the sense of very heated but certainly, Bromcom representatives were questioning UL quite hard and often the answer was not particularly clear. Indeed, both sides expressed the view at the end of the call that UL had not been able to respond as methodically as it should. Added to that, was Ms Wood’s perspective that it had been a difficult meeting with shouting from the Bromcom side. That was not really borne out by the audio recording but this does not matter. I accept that this was Ms Wood’s honest perception of the call when she had to relive it in the witness box which she clearly found a stressful experience in this regard. It adds to the notion that I think neither side felt it was a very productive session. This is important because it was in this meeting that references were given to the figures outlined above. See further the observations of Judge Eyre in his paragraphs 57, 58 and 61 generally.

As for the call on 3 April, except in one brief respect, it did not really go back to the question of costs adjustments.

I attach particular importance to the fact that UL plainly considered it necessary to give a proper account of its reasoning not through the somewhat ad hoc and unstructured format of meetings where, on any view, I think there was a degree of tension; this is obviously why UL wrote the letter of 22 April as well as the letter of 23 April responding to Bromcom’s solicitors letter of 14 April.

I accept that the requirements of Regulation 86 are not dealing with the level of knowledge needed for the purposes of Regulation 92 (2). Further, if there is a challenge to some intermediate stage of the procurement process (where it is not yet complete) Regulation 86 is irrelevant. Nonetheless, where (as here) it is relevant, it is at least of some evidential value when the contracting authority considers when it has set out its reasons formally so as to comply with Regulation 86.

Likewise, a tenderer is surely entitled to see the information put in a clear and structured form, not least to enable its lawyers to give it timely advice, even though, in these instances, Bromcom had in fact recorded the telephone calls.

Of course, unlike Judge Eyre, I have had evidence about the calls. In particular, Ms Marsh said in her WS at paragraphs 30-31 that she had found the 1 April call somewhat confusing on the question of added costs and thought she had not received a clear breakdown. She also said that she was confused because she thought Bromcom had in any event agreed to provide the interface. In her 3^rd WS, she said that she felt that Bromcom had not got any further on 3 April and things were still unclear, especially about the addition of “sunken costs”.

In cross-examination, Ms Morgan said that she had understood from the 1 April call that the costs of F2F at £42,000 had been added, but she could not say if that meant they could have complained about it as from then. She did say in her WS that she thought UL was not prepared for the 1 April call, and did not seem to have any real confidence in the answers it was giving.

As for Mr Lewin, I agree that in the 1 April meeting he had expressed an understanding of what UL were doing when adding the £500 per day (see paragraph 435 above) on F2F costs. In cross-examination it was put to him that if Bromcom wanted to say that there was no power in UL to add any costs to its tender, then any such challenge could have been made from 1 April onwards. Mr Lewin replied that he had not understood at that point the justification for adding the costs and why they were there. He also said that he did not understand if these costs were being added to both bidders tenders. While he accepted that he understood the maths of the £4,405 and £42,000 adjustments, he did not accept that he had all the information necessary to make a complaint that there was no power to do this.

As to all of this, UL contends that Bromcom had sufficient knowledge for limitation purposes in relation to these complaints; questions like the justification for adding the costs were irrelevant. I think this is too narrow an approach. Bromcom was in my view entitled to see the deductions expressed clearly, as it were, in a document and there was at least some confusion as to what was added, if anything, to Arbor’s costs. I say that even if the putative challenge (which was which was in fact made along with others) was that there was no lawful power to add anything at all.

In this regard, I also agree with Judge Eyre (see his paragraphs 55 (v) and 63) that on a fair, objective reading, the 14 April letter (set out at paragraph 441 above) reads as it purports to be, namely a detailed request for further information following the 1 and 3 April telephone calls. It does not purport to be nor does it read as if it is, a letter before action, as it were. This is important because it is an indication as to where Bromcom thought it had got to in terms of the making of any possible claim against UL.

I accept that in relation to the cost scoring challenges, Bromcom did not have a sufficiently informed view as to whether there had been an infringement and whether proceedings would be appropriate, until after receipt of the 22 and 23 April letters.

I also consider that where, as here, a possible infringement might go nowhere in terms of causation of any loss, part of any consideration undertaken by Bromcom had to involve the totality of the infringements as they appeared to them after 22 and 23 April.

For all those reasons, I do not consider that the cost scoring challenges were time-barred. That being so, there is no consequential limitation effect for any other challenges which might be said to be merely further breaches of the same underlying duty.

I then turn to UL’s limitation point in relation to the question of weightings. As we know, the application of weightings here in fact made no difference to the outcome and that was clear as a matter of maths before the trial commenced. It is correct that paragraph 48.8 of the Consolidated Particulars of Claim (which first appeared in the Second Particulars of Claim i.e. in 2021) made a complaint about weighting in the context of various emails passing between Mr Wilson and Ms Woods and after at least some evaluator’s scores had already been done. At that point, it was suggested that the weightings had a bearing on the outcome. However, the information relied upon here had not been provided prior to 22 and 23 April and indeed not for some time afterwards. Instead, it came out of disclosure. On any view, the mere information as to the giving of weightings, set out in the document of 2 April says nothing about any impact that might have had. And in cross-examination, Mr Wilson agreed that there was no suggestion in the feedback that the weightings made any difference to the outcome.

Again, therefore, I do not accept that Bromcom had the appropriate knowledge as from 2 April instead of at some point after 22 and 23 April.

I take the same view in relation to this allegation which first emerged in paragraph 48.6 of the Second Particulars of Claim, where specific reference was made to what Ms Wood had said in an email to UL about the need to see more “granular differentiation” which had obviously been provided on disclosure, not initially.

As for the allegation that some of the evaluators at the ITCD stage had not been named as such in the ITCD, UL alleges that the fact of their role was known, or should have been known, since CD2 and CD3 in which they participated. I do not think that this follows and it is noteworthy that again, this part of paragraph 48.1 was only pleaded in March 2022 after further disclosure. Certainly, I cannot see how the information relied upon at paragraph 48.1.6 could have been gleaned from CD2 and CD3 themselves.

Finally, I deal with the allegation made at paragraph 64C of the Consolidated Particulars of Claim. This first emerged in the Second Particulars of Claim. It is a general point about manifest errors in the scoring process although it was then expanded in the Amended Consolidated Particulars of Claim. The first iteration was said to be based on the disclosure inspection. UL is in fact directing itself to the amended iteration because at paragraph 222 of its Written Closing, it refers to an allegation about the drafting of Appendix B - not itself pursued at trial. In any event, all UL here says is that the claim is time-barred without any detail. I fail to see how any relevant knowledge could have been obtained prior to 22 and 23 April, indeed prior to a much later date.

For all the reasons given above, in my judgment, Bromcom did not have knowledge of facts which clearly indicated relevant infringements and that proceedings were appropriate in respect of any of its challenges, prior to 22 and 23 April 2020. Nor do I consider that in the circumstances it ought to have had that knowledge. Indeed, UL’s essential allegations concern Bromcom’s actual knowledge, and in truth none of Bromcom’s witnesses were cross-examined in this latter sense.

As foreshadowed above, it is not been necessary, in the light of my conclusions above, to explore whether Bromcom’s challenges were all, in truth, instances of breach of the same underlying duty or not. Judge Eyre considered that, for the allegations he was dealing with, they were. This was because they were seen to arise out of one evaluation process leading to a particular outcome. See paragraph 48. It is not clear to me that the duty not to make manifest errors, or indeed the duty to act fairly, are all part of the same duty to treat economic operators equally without discrimination and in a transparent and proportionate manner, as required by Regulation 18. After all, at least as pleaded, the duties set out at paragraph 28 of the First (and subsequent) Particulars of Claim are said to be drawn not only from the PCR but the EU public procurement directive, and general principles of EU and domestic law in this field, including fairness and good administration, among other sources. However that interesting question is not for me to decide here and indeed Bromcom did not seek to answer the limitation challenge by reference to this sort of distinction. Accordingly, I say no more about it.

In the event, the limitation defence fails.

Under regulation 98 (2) (c), the court “may award damages to an economic operator which has suffered loss or damage as a consequence of the breach”. UL contends that, by reason of the making of a number of covert recordings of telephone calls with UL (six in all), Bromcom acted in a way which constitutes “grave professional misconduct which renders its integrity questionable” - see Regulation 57 (8) (c). That being so, UL had a discretionary power to exclude Bromcom from the procurement exercise anyway, and would have done so in a counterfactual which assumed that UL would have found out about these calls.

UL has referred to the case of Forposta C;2012:801 where the CJEU explained that this grave professional misconduct would cover all wrongful conduct which has an impact on the professional credibility of the operator at issue and not merely the violation of professional ethical standards.

Two outgoing calls made by Ms Marsh to Ms Wood on Bromcom’s internal phone were recorded. She said that she was not aware of this at the time and it was an automatic function. Mr Wilson agreed that she could not be expected to seek consent for those recordings if she did not know that the calls were being recorded. However, the difficulty is that although Ms Marsh made detailed notes of her conversation with Ms Wood on 21 February, she also circulated emails later the same day which contained detailed comments from Ms Wood and which appear to have been taken from the transcript which is in the bundle. See, in particular, B/1094-1102. Indeed, her email to others at Bromcom of 21 February refers specifically to the core transcription “from last Friday’s conversation with Alex Wood”. This suggests that her evidence that she was not aware of the recording at the time could not be right. However, I should add that detailed questions in relation to the transcript and the particular documents relied upon were not put to Ms Marsh in terms but rather to Mr Lewin, whose evidence came later. See day 3/124-126, although I permitted Mr Lewin to deal with those questions. But the relevant questions should have been put directly to Ms Marsh. Nonetheless, the documents say what they say.

In fact, when consent was sought to make audio recordings of formal meetings with UL (for example on 8 January) UL did consent albeit with conditions attached principally that such recording was to be done to assist with note-taking.

Calls were also recorded by Mr Lewin. He said this was done to help him take notes, given his dyslexia. That evidence was not challenged and I accept it. Mr Lewin also denied that the recordings were made for the benefit of Mr Guryel. Mr Wilson also accepted that had he been asked by Mr Lewin to permit the recordings, given his dyslexia, he would absolutely have said yes.

There was then the one April debrief meeting which Ms Marsh did consciously record on any view. UL has in fact made use of this transcript, along with Bromcom, for the purpose of the trial. But more importantly, it is hard to see why UL would have refused its consent, had it been asked, on the basis that the purpose was to assist in note-taking.

Mr Wilson did add that he did not accept that there was nothing in the telephone calls point. All he could say, was that he had not experienced this in 30 years in the industry and it was “purely a part of, as I say, professional courtesy and respect”.

In addition, Ms Morgan accepted in evidence that potential clients might not want to work with a company that recorded calls without telling them or asking for their consent. Likewise, Mr Lewin was asked if many or some people would feel that their trust and confidence was being significantly undermined if others were recording them without telling them or asking for their consent. He said yes, he understood that could be a consequence.

Clearly, in relation to any call that was recorded without the knowledge or consent of UL, its permission should have been sought. It was wrong of Bromcom not to have done so. Moreover, I think that Ms Marsh’s evidence about only using the recordings later on appears to be unsatisfactory given the relevant emails, even though she did not have these matters specifically put to her for her comment.

Nonetheless, for the most part, there was good reason to record the calls because of Mr Lewin’s dyslexia. Moreover, had UL been asked, I am quite sure that it would have given its permission, just as it did on those occasions when it was asked.

I agree with Mr Wilson that Bromcom’s conduct showed a lack of professional courtesy and respect. However, I do not accept at all that this amounted to “grave professional misconduct”.

Moreover, had, in a counterfactual, all of these matters (including Mr Lewin’s particular need for recording) been disclosed to UL at the time, I do not believe that UL would then have excluded Bromcom, notwithstanding Mr Wilson’s statement at paragraph 170 of his WS dated 8 April 2022 that UL would have had no hesitation in doing so. Of course, at that point, the evidence of Mr Lewin’s dyslexia had not emerged. In any event, of course, there would have to have been a demonstration by UL of grave professional misconduct which would not have been established.

For all those reasons, I am quite satisfied that there is no basis for me not to exercise my discretion and award damages to Bromcom, which, in principle, I now do.

It follows that Bromcom’s claim has succeeded on liability and causation in respect of the breaches of procurement law which I have identified above. The next stage will be to assess quantum.

I am very grateful to counsel for their helpful oral and written submissions, and assistance during the trial.