Bid evaluation in public procurement: the official guidance note, explained

The key facts

• Published in May 2021, the note builds on chapter 10 of the Sourcing Playbook and collects good practice for practitioners who design and run bid evaluation processes.
• Evaluation must comply with the Public Contracts Regulations 2015 and the core principles of transparency, non-discrimination, equal treatment and proportionality.
• Contracts are awarded to the most economically advantageous tender (MEAT), which is not necessarily the lowest priced bid.
• Bids should be independently scored by at least two trained evaluators, with the same evaluator scoring all responses to a given question, and an independent moderator leading consensus scoring.
• Social value must be explicitly evaluated in central government procurement where relevant and proportionate, with a minimum 10 per cent weighting under the Social Value Model.
• Relative price scoring should be treated with caution and used only where there is a specific business reason approved by the commercial lead and project SRO.
• A Should Cost Model shall be produced for all complex outsourcing projects, and its use within the evaluation methodology must be disclosed to bidders.
• Any bid more than 10 per cent below the average of all bids or the Should Cost Model must be referred for central assurance and scrutiny via the Cabinet Office Commercial Continuous Improvement Team.

What this guidance note is and who it applies to

The Bid Evaluation guidance note, published in May 2021, builds on chapter 10 of the Sourcing Playbook. It is a collection of good practice aimed at practitioners responsible for designing and running robust bid evaluation processes, and a useful introduction for senior stakeholders unfamiliar with the subject. It is explicitly not a substitute for suitably qualified and experienced practitioners, supported by legal advice, designing and overseeing the process.

Bid evaluation is defined as the process of assessing bids to identify the most economically advantageous tender (MEAT) for a project, determined against published award criteria. A good evaluation is not only about the final award decision: it is about the design and execution of the whole process leading up to that decision, properly documented so it can stand up to internal and external scrutiny. The note covers assessment against award criteria only, not selection stage assessment, though many principles apply to both.

Getting the foundations right

The note puts heavy emphasis on planning. The procurement timetable should reflect every evaluation activity with contingency built in: agreeing resources, issuing evaluator guidance and conflict of interest forms, developing and testing the evaluation model before advertising, evaluating at each stage where procedures such as competitive dialogue are used, raising clarifications, moderation meetings, and collecting all notes before any award decision. Published timetables should be stated as indicative, with the right to change reserved.

Resourcing rules are specific. Bids should be independently scored by at least two evaluators with sufficient knowledge of the criteria they are assessing. The same evaluator should score all responses to a particular question to keep scoring consistent. Finance models should be evaluated independently of the technical specification. An independent moderator, who is not an evaluator, supports the group in reaching consensus, ideally with a separate note taker. All evaluators and moderators must be trained, and the training records retained as part of the audit trail.

Record keeping is treated as fundamental rather than administrative. Records must justify why each bidder received each score, support full and helpful debriefs to unsuccessful bidders, and stand up in the event of legal challenge. At the end of the process departments should produce an evaluation report, which may form part of the wider report required under Regulation 84 of the Public Contracts Regulations 2015. Where scores change in moderation, the reason and evidence for the change must be noted.

Designing the model: price, quality and social value

The evaluation model should reflect the business objectives, outputs and outcomes of the procurement, built on criteria that are proportionate, linked to the subject matter of the contract, not based on price alone, and clear enough that all reasonably well informed and diligent bidders interpret them the same way. The model should be developed iteratively, tested with the market during early engagement, and scenario tested with dummy bids, including low bids, before being finalised and published. Once published, it must be followed and evidenced.

Government policy is to award contracts on value for money: the best mix of quality and effectiveness for the least outlay over the period of use, not the lowest upfront price. Weightings should reflect the characteristics of the service and be tested with the market before being fixed. An 80/20 quality to price split signals that quality matters far more than price, while a separate social value criterion, for example 60/10/30 across quality, social value and price, makes both priorities explicit. Social value must be evaluated in all central government procurement where relevant and proportionate, using the Social Value Model with a minimum 10 per cent weighting.

On quality, criteria should be objective, proportionate, relevant and unambiguous, developed by subject matter experts and peer reviewed. Open questions differentiate bids best, while closed questions suit mandatory yes or no commitments. Scoring bands should allow genuine differentiation, avoiding bid bunching through devices such as non-linear thresholds and even numbers of bands. On price, whole-life cost should be considered, and alternatives to relative price scoring, such as price per quality point or benchmark price scoring, provide an absolute standard. Relative price scoring carries known distortions and needs commercial lead and SRO approval before use.

Guarding against low cost bids

The note is blunt about the dangers of a bias towards low cost bids. If the true cost of delivery is not understood, contracts run into trouble as suppliers cut costs after award, seek contract changes, manage strictly to the letter of the contract in a crisis, or require injections of further funding to keep critical services running, changes which may even amount to a substantial modification in breach of the regulations.

The defences are set out as questions to answer before commencing procurement: has pre-market engagement tested that suppliers can deliver at an affordable cost, has a Should Cost Model informed the methodology and the whole-life cost calculation, is any minimum quality threshold appropriate, and do the criteria and scoring reward bids offering high value and low delivery risk. A Should Cost Model shall be produced for all complex outsourcing projects, and if it is used as a comparator within the evaluation itself this must be disclosed to bidders. Where a bid appears abnormally low the department must ask the bidder to explain the price, and may reject it only if the explanation is unsatisfactory. Any bid more than 10 per cent below the average of all bids or the Should Cost Model must be referred to the Cabinet Office Commercial Continuous Improvement Team for central assurance.

Applying the model: clarification and moderation

Bids must be evaluated strictly in accordance with the published criteria and methodology. Clarification of bids is permitted where responses appear incomplete or erroneous, but it is not an opportunity for bidders to improve their bids, and questions must go through the procurement or commercial team in writing rather than directly from evaluators, with responses shared with all evaluators covering that area.

Moderation exists to reach a single justifiable consensus score, which should not simply be an average. The note lists warning signs a moderator should challenge: comparisons to other bidders, reliance on matters outside the bid, copy and pasted criteria, inconsistencies between score and justification, or thin and incoherent reasoning. The moderation process should not close until all evaluators and the moderator are satisfied, with a clear note of every decision and every score change.

How eSourcing Data helps

Almost every failure mode this guidance describes is an evidence failure: undocumented scores, inconsistent criteria, clarifications handled loosely, moderation decisions nobody wrote down. eSourcing Data runs evaluations inside a structured platform where published criteria, weightings, individual scores, comments and consensus outcomes are captured as they happen, giving you the audit trail Regulation 84 reporting and debriefs demand without a reconstruction exercise.

The platform supports independent scoring by multiple evaluators and keeps clarification exchanges with bidders in one written, timestamped channel, which is precisely the discipline the note requires. Reporting tools then let commercial leads evidence how the published model was applied, bid by bid and criterion by criterion, when internal reviewers, auditors or unsuccessful bidders ask.

Because the same workflows cover below-threshold procurement, smaller contracting authorities can apply the same rigour proportionately, with evaluation records, notices and award documentation held together for every procurement rather than only the largest ones.

What to do about it

1. 1Build the evaluation timetable early, with contingency, covering training, conflict of interest declarations, model testing, clarifications and moderation.
2. 2Appoint at least two evaluators per question plus an independent moderator and note taker, train them, and retain the training records.
3. 3Design criteria that are proportionate, relevant, clear and never price alone, then scenario test the model with dummy bids before publishing it.
4. 4Set price, quality and social value weightings deliberately and test them with the market before fixing them, remembering the 10 per cent social value minimum.
5. 5Interrogate low bids against your Should Cost Model and refer any bid more than 10 per cent below the average or the model for central scrutiny.
6. 6Route every bid clarification through the commercial team in writing and share responses with all relevant evaluators.
7. 7Document every score, every moderation discussion and every change of score with reasons, and compile the evaluation report at the end.