Every public sector bid you submit is marked by people following a manual. The manual is public and it is short. The Government Commercial Agency publishes it as “How to evaluate bids in procurement” in its Procurement Essentials series, first published on 17 January 2022 and last updated on 21 May 2026, and it now covers the Procurement Act 2023 alongside the regulations the Act replaced. It tells evaluators how to score you, what they are not allowed to do for you, and what paper trail they must leave behind. Read backwards, it is a specification for your next answer.
Who the Government Commercial Agency is
If the name is unfamiliar, that is because it is new. On 1 April 2026, Crown Commercial Service and several Cabinet Office central commercial teams merged into a single executive agency, the Government Commercial Agency. The guidance CCS used to publish now sits under the GCA banner. If a guide you rely on still says CCS, it was written before the merger; our own April procurement update guide, written days afterwards, covers the Act’s mechanics without mentioning the new agency at all. The MEAT evaluation guide’s May 2026 update covers the Act alongside the regulations it replaced, and it is worth more to a bidder than most of what has been written about the Act since.
The eight stages, seen from the other side of the table
The guide walks a buyer through eight stages: choose your evaluators, keep accurate records, allocate scores against the criteria, clarify bids, moderate scores, finalise the award, observe the standstill period, complete the evaluation report. None of it is secret and none of it is optional in spirit, because the records have to survive a legal challenge. Four of those stages decide your score, and each one tells you something concrete about how to write.
You are scored against the criteria, not against the other bids
Evaluators are instructed to assess each submission against the published evaluation criteria only, and to score it on its own merits rather than by comparing it with the bid next to it in the pile. The statute backs this up. Under section 19(2) of the Procurement Act 2023, the “most advantageous tender” is the one the authority considers satisfies its requirements and best satisfies the award criteria when assessed by reference to the assessment methodology and the stated relative importance of the criteria. Section 23(2) requires those criteria to be “sufficiently clear, measurable and specific” and a proportionate means of assessing tenders.
The consequence for your writing is blunt. Being better than your rivals earns you nothing. A brilliant answer to a question the criterion does not ask scores like a bad answer. We see this constantly in bid reviews: a submission full of genuine capability that never maps itself onto the marking scheme, from a bidder who assumed the evaluator would make the connection. The evaluator is instructed not to. Put the criterion’s own language into your answer, deal with every element of it, and evidence each one, because the person scoring you needs to justify the mark in writing against that wording and nothing else.
Clarification will not rescue a thin answer
The guide is explicit that clarification “is not an opportunity for bidders to improve their bids and it must not provide a competitive advantage to a particular supplier”. A buyer can ask you, in writing, to explain an ambiguity or supply something missing in form. What they cannot do is let you strengthen your answer after the deadline, however obvious the fix.
Bidders lose more to this rule than they realise. If you spot the gap an hour after submission, it is already too late, and hoping the buyer will ask the right clarification question is not a strategy. It also cuts the other way: when a clarification question does arrive, answer exactly what was asked and no more, because an answer that smuggles in improvement can be disregarded.
Moderation is where vague answers lose marks
Scores do not go straight from the first evaluator to the award decision. The guide has bids scored individually first, then moderated. It advises a small evaluation group rather than one person, and asks that at least one member of the team, usually the procurement lead, is well trained on the rules that apply. A moderator who was not part of the original evaluation team then reviews those independent scores and works with the evaluators towards a consensus score. The moderator is told to look for vague or incomplete score justifications, and for inconsistencies between the marks awarded and the reasons given, so a score with a thin rationale behind it gets reopened in that room.
This explains a pattern that frustrates strong bidders: an answer you know is good comes back at 7 out of 10 with a comment like “insufficient evidence”. The first evaluator may well have believed you. The moderation stage exists to strip out belief. A claim without evidence cannot be defended in moderation, so the consensus score drifts down to what the words on the page prove. Write for the moderator, not the optimist: name the contract, the dates, the numbers, the person, so the evaluator who liked your answer can win the argument about it.
The paper trail is your debrief, so use it
buyer evaluation criteria are told to keep records showing why each bidder got each score, sufficient to give what the guide calls “a full and helpful debrief”, and to retain them for at least three years from contract award. Under the Act you no longer depend on the buyer’s goodwill to see any of this. Section 50(3) requires an assessment summary to be provided to every supplier that submitted an assessed tender, before the contract award notice is published, and section 51(2) then gives you a mandatory standstill period of eight working days from the day that notice is published before the contract can be signed.
Be careful here, because the guide has not caught up with the Act on this point. Under its standstill heading it still describes a mandatory 10 day period and applies it to awards under the Act. The statute governs, and section 51(2) says eight working days. Ten days is the old Public Contracts Regulations 2015 position, where the clock ran to midnight at the end of the tenth day after the award decision letter was sent electronically, or the fifteenth day if it went by other means. The Act replaced those regulations for procurements started on or after 24 February 2025. Section 51(3) also takes framework call-offs, awards through a dynamic market and light touch contracts outside the mandatory standstill altogether, so check which route your bid sat under before you start counting days.
Eight working days is not long, but it is enough to read your assessment summary against your submission, put focused questions to the buyer, and decide whether anything needs escalating while there is still time for it to matter. Losing bidders who treat the debrief as a formality throw away the only free evaluation data they will ever get.
Mark your own bid before the buyer does
The cheapest use of the guide is to turn it on your own draft before submission. Score each answer the way an evaluator must: criterion by criterion, against the published wording, giving credit only for what the words in front of you prove. No credit for reputation, no credit for what you meant. If the score embarrasses you, better to find out a week before the deadline than in the assessment summary.
That marking exercise is, in honesty, most of what a good bid review is. Ours draw on a team of specialists that includes APMP-certified writers and ex-evaluators, and split into a compliance audit, a persuasion gap analysis and actionable feedback with prioritised recommendations you can use before the deadline. If you keep scoring second with feedback letters that say “good submission”, the buyer’s own manual says exactly where those marks went, and a reviewer who has sat on the scoring side can usually find them within a day.