By Paul Emanuelli

This article is an excerpt from The Art of Tendering: A Global Due Diligence Guide, which is available for purchase.

In its August 2018 decision in BAM Civil Limited v. The Department for Infrastructure, the High Court of Justice in Northern Ireland determined that the government’s evaluation process was flawed due to hidden evaluation factors. The case dealt with a Request for Proposals (RFP) for three road-widening contracts in Belfast, Northern Ireland. The plaintiff challenged the process, arguing that the evaluation committee improperly scored its proposal, and obtained a standstill order to prevent the award of the contracts to its competitor until the dispute was resolved.

As the Court noted, the plaintiff’s score of 91.89 was very close to the 93.56 scored by the winning bidder:

In terms of the overall result, there is a helpful Table which the court will replicate.
Court Overrules Figure
On the basis of the above, it can be seen that the difference between the winning score and the plaintiff’s score was small.
It would only take a change in the quality marks for the plaintiff of a small dimension for there to be a significant impact on the overall outcome.  At paragraph 129 of the plaintiff’s opening skeleton argument it is suggested that the effect of an additional one point for the plaintiff in the overall context of Section C of the quality assessment would result “in the plaintiff scoring highest in the competition”. This has not been contested by either the defendant or notice party and appears to be correct. Indeed, it was acknowledged by one of the defendant’s witnesses when giving evidence.
The proceedings in this case were issued by the plaintiff within the standstill period (as extended).  This has meant that none of the contracts at issue in this procurement competition have, to date, been made pending the outcome of these proceedings.  This is in accordance with the Regulations.

Given the close scores, the plaintiff challenged the assessment on the “Phasing of the Works” category, wherein it the evaluation panel (referred to in the decision as the EP) awarded it four out of a possible five points:

The issue in respect of question 3.01 relates only to how the plaintiff’s response to the first element was assessed.
In answer to 3.01(a) the plaintiff provided a response which can be related sequentially to each of the indicators listed at 3.01(a).  The indicators referred to for this particular aspect were:
  1. Phasing of the Works.
  2. Construction activities in close proximity to residential housing (Little George Street).
  3. Environmental health restrictions on working hours.
  4. Traffic management requirements for the A12 Westlink.
  5. Diversions of statutory undertaker apparatus and existing motorway communication equipment.
  6. Measures to ensure community liaison.
In terms of the plaintiff’s challenge to the evaluation, its complaint related to how the EP assessed that part of the response which dealt with the first indicator “phasing of the works”.  This arises because of the way in which the debrief has been composed by the defendant.
As regards the answer given to question 3.01, the debrief states:
“Very good response with all indicators adequately addressed with the majority in detail. (a) North Queen Street – Indicator 1 – Lacks detail of component steps associated with activity phasing eg erection of bridge beams and parapets, completion of bridge widening.”
The overall score given by the EP for this question was a 4.  It therefore appears to be the case that the only criticism which could give rise to a score of 4 (as against a 5) was that recorded by the EP and referred to above which relates alone to Indicator 1 in connection with that part of the response dealing with North Queen Street.

As the Court noted, the evaluators used the following scoring guide to allocate the scores:

The Scoring Interpretation Table
There is a scoring interpretation table in respect of each question which has been devised to guide scoring in respect of the answer to each question.  It is in much the same terms for question 3-01 and question 3-02, though the language in respect of 3-02 slightly varies from the language in respect of 3-01.
In essence, the top score of 5 accompanies an assessment which is viewed as excellent.  A score of 4 is reserved for a very good answer; 3 for a good answer; 2 for a satisfactory answer; 1 for an unacceptable answer and 0 for a fail.

The plaintiff argued that a reasonably well-informed and diligent tenderer (RWIND) would not have anticipated the level of detail required to score five out of five in the category in question:

Counsel argued that the issue of what the question was seeking had to be placed in the context that the question itself which sought “an outline methodology”, which he interpreted, adopting some of the expressions used by the defendant’s witnesses, as something “skeletal” or “non-specific” or “generic”.  In his submission “a RWIND tenderer viewing the question would not have anticipated that [it] was asking for such intricate and diverse details of the works as ‘the delivery and access of plant’, ‘impact on residents’, ‘closure of footpaths’, ‘widening abutments to the bridge’, ‘siting of cranes’ ‘debris falling onto traffic during bridge widening’, ‘noise issues’, ‘working outside normal hours’ and ’the impact of the increased student pedestrian traffic with the new University’”, details referred in evidence to by witnesses for the defendant.  The court, counsel continued, should take into account that the four members of the scoring panel had expressed different views as to the sort of detail which was appropriate in the context of the answer to the question.

After a detailed interpretive analysis, and after considering the witness testimony of each of the evaluators, the Court concluded that the case turned on whether the evaluation factors were made sufficiently clear to the bidders to enable a uniform interpretation of the requirements:

There are other aspects which, to the court’s mind, require consideration. The principal one relates to the consistency of the EP’s approach to assessment with the established requirements of EU law and, in particular, with the requirement of transparency. Transparency arises in the context of the RWIND tenderer’s need to understand what sort of response was expected but it goes beyond this and requires the evaluation itself to conform to a standard of effective and fair competition in the interests of equal treatment. The criteria have to be applied by the contracting authority – here, the EP – in a way which can be viewed as objective and uniform, as between tenderers, and which allows for a comparative assessment of the level of performance of each tenderer. What must be avoided is a situation in which the criteria are applied in a way which confers on the contracting authority an unrestricted freedom of choice.

In considering these standards, the Court ultimately concluded that the evaluation panel failed to meet the standards of transparency since they failed to apply the evaluation criteria in a manner consistent with how those criteria were identified in the RFP.

As this case illustrates, the scoring of evaluation panels can be put under extreme scrutiny, particularly in close-scoring situations. Purchasing institutions would therefore be wise to clearly draft their evaluation criteria and disclose the factors that will be relied on in scoring those criteria.