RFP evaluation criteria and weighting: how buyers turn your proposal into a score

Every RFP contains two documents pretending to be one. The first is the brief: what the buyer wants, the requirements, the questions you have to answer. The second is the marking scheme: the evaluation criteria, the weightings, the scoring scale. Most teams give the first one their full attention. The second one tends to get less love, and that is understandable. The requirements are where the work is, and the criteria section often reads like procurement boilerplate.

But the criteria are the only part of the document that describes how you win. A proposal that answers every requirement but is written without the weighting in mind can lose to one that answers fewer requirements and lands the points where they count.

This article explains what evaluation criteria are, how weightings turn individual scores into a decision, and how to read the criteria before you write a word. If you already know how evaluators arrive at a score on each criterion, the mechanics of the scorecard are covered in the proposal scoring framework. This is the layer above: how those scores become a winner.

What evaluation criteria actually are (and what they are not)

It helps to separate three things that RFPs often blur together: requirements, questions, and criteria.

A requirement is something the buyer needs delivered. “The supplier must provide support 24 hours a day, 7 days a week.” A question is how the buyer asks you to demonstrate it. “Describe your support model, including escalation and out of hours cover.” An evaluation criterion is how your answer will be judged. “The extent to which the proposed support model demonstrates resilience and meets the stated service levels”, weighted at 10 per cent of the total score.

The distinction matters because you can satisfy the requirement, answer the question, and still score poorly against the criterion. If the criterion asks for resilience and your answer describes a rota, you have answered a different exam question from the one being marked.

Pass or fail gates come first

Not everything in the criteria carries a score. Most RFPs start with a set of pass or fail conditions: mandatory requirements, minimum insurance levels, required accreditations, financial standing, submission rules. In UK public procurement these often sit at the selection stage, which the Procurement Act 2023 calls conditions of participation, and are assessed before the award criteria are applied at all.

Gates earn you no points. They simply decide whether your proposal is read. A missing certificate or a late upload removes you from the process regardless of how strong the rest of the document is, which is why the pre-submission checks in why experienced bid managers never submit on deadline day matter as much as the writing.

Scored criteria decide the winner

Once you are through the gates, the scored criteria take over. These typically cover quality or technical merit (your approach, methodology, team, experience), price, and in UK public sector work, social value. Central government has required a minimum 10 per cent social value weighting since 2021, and from 1 January 2027 that rises to a minimum of 20 per cent on central government contracts worth £5 million or more under PPN 026, published in August 2026. Other public bodies set their own weightings. Larger criteria are usually broken into sub-criteria, each with its own weight, so a 40 per cent quality criterion might be split across four questions at 10 per cent each, or unevenly, with one question worth 20 and the others worth less.

Whether those sub-weightings are published varies. In public sector tenders they usually are. In private sector RFPs the buyer may publish headline weightings only, or none at all, and you have to infer where the emphasis sits from the questions themselves. That inference is covered later in the article.

How weighting works: the maths that decides the winner

Weighting is the mechanism that converts a set of individual scores into a single ranked result. Each criterion is scored, usually on a scale such as 0 to 5. That score is converted into a percentage of the marks available for that criterion, and the marks are added up. The bidder with the most marks wins.

Written down, it sounds obvious. In practice, the weighting changes outcomes in ways that are easy to underestimate until you see the numbers.

A worked example

Take two bidders for the same contract. The buyer has published a 60 per cent quality, 40 per cent price split (an illustrative split, not a typical one; real weightings vary widely and are set per procurement).

Bidder A submits a strong proposal that averages 4 out of 5 across the quality criteria, at a price of £520,000. Bidder B submits a solid but less developed proposal averaging 3 out of 5, at £450,000.

Price is most commonly scored by giving the lowest compliant price full marks and scoring the others in proportion: your marks equal the lowest price divided by your price, multiplied by the marks available. It is the method used in the example below. Variants exist. Some buyers deduct a fixed number of marks for each percentage point above the lowest price, some give full marks to every bid within a band of the lowest, and some use absolute methods such as price per quality point. The formula is always published in the tender documents, so you can model what a given price scores before you commit to it.

Bidder A wins by a clear margin despite being £70,000 more expensive. The quality weighting has more than absorbed the price difference.

Now keep both proposals exactly as they are and change only the weighting to 40 per cent quality, 60 per cent price.

Bidder B now wins, by a tenth of a mark. Nothing about either proposal changed. The weighting did.

Pro tip

work out what one point on the scoring scale is worth on each criterion before you start writing. On a criterion worth 20 marks scored 0 to 5, every point is worth 4 marks. On a criterion worth 5 marks, every point is worth 1. That single calculation tells you where an extra hour of effort pays and where it does not.

The quality and price split

The headline split between quality and price is the single most important number in the criteria, because it tells you what kind of competition you are in. A heavily price weighted evaluation rewards a compliant, efficient proposal and a sharp commercial offer. A heavily quality weighted evaluation rewards depth, evidence and a methodology that scores 4s and 5s.

In England, Wales and Northern Ireland, the Procurement Act 2023 frames this as identifying the most advantageous tender. Scotland and the EU retain the most economically advantageous tender test. In both cases the buyer sets and publishes the weighting in advance, and the scoring framework article explains how evaluators then arrive at each score within that structure.

One more thing the example shows. Bidder B did not lose the first evaluation on price. They lost it on quality, because at a 60/40 split a one point gap on the scoring scale across the quality criteria costs 12 marks, and no realistic price advantage recovers that. Reading the weighting tells you which of those two conversations you need to have internally before you write.

Want to know where your proposal stands before the evaluator does? ReqFit scores your draft against every requirement in the RFP in minutes. Your first review is free, no credit card required. Get your free review

Where the criteria hide in the RFP

The criteria are always in the document. They are not always where you would expect, and they are not always labelled.

Public sector: published, but scattered

In UK public procurement the award criteria, weightings and scoring methodology have to be published before bids are submitted. That is the good news. The less good news is that they are rarely in one place. The headline quality and price split may sit in the invitation to tender, the sub-criteria and their weightings in a separate evaluation schedule, the scoring scale and its definitions in an appendix, and the price formula in the commercial instructions.

The practical move is to pull all of it into one page before you plan the response. Every scored question, its weighting, the marks available, the scale it is scored on, and any minimum threshold. If any of those is missing from the documents, ask for it through the clarification process. You are entitled to know how you will be marked, and a question submitted early is answered for every bidder.

Private sector: inferred from the signals

Private sector RFPs are under no obligation to publish criteria, and many don’t. Some give headline weightings. Some say only that proposals will be “assessed on quality, value and fit”. Some say nothing at all. That does not mean there are no criteria. It means the buyer has them in their head, or in an internal scoring sheet you will never see, and you have to infer them from the document.

The signals are usually there. The questions the buyer asks first and asks most about. The topics that get their own section rather than a line. Word or page limits, which tell you how much the buyer expects to read on each point and therefore how much it matters to them. Phrases such as “demonstrate”, “provide evidence of” and “describe in detail”, which tell you where a bare statement will not score. And the buyer’s own introduction, which almost always names the two or three outcomes they actually care about.

Take an agency pitch brief that opens with a paragraph about the client’s frustration with slow turnaround from their previous supplier, then asks eight questions of which three concern process, timescales and account management. The brief never mentions weighting. It does not need to. A response that spends most of its length on creative credentials has misread the criteria, however good the credentials are.

Pro tip

when the criteria are not published, write your own. Draft the scoring sheet you think the buyer is using, with weightings, and review the proposal against it. You will be wrong in the detail and right in the emphasis, which is what matters.

Reading the criteria to plan your response

Once the criteria are on one page, they stop being a compliance section and become a plan. Four decisions follow directly from them.

Put the effort where the marks are

Weighting tells you where an extra hour pays. If one question carries 25 per cent of the quality score and another carries 5, they do not deserve equal attention, equal evidence, or equal review time. Yet proposals are often written in document order, with the same care applied to each question until the deadline arrives, which means the last questions get the least attention regardless of their weight.

Reverse it. List the scored questions by marks available, not by their order in the document, and assign your strongest writers and your best evidence to the top of that list. A sales manager coordinating a response across a team can hand out the questions on that basis rather than by who has capacity.

Treat minimum thresholds as gates in disguise

Many tenders set a minimum score on individual criteria, commonly 2 or 3 out of 5, below which the bid is excluded regardless of its total. This changes the calculation. A low weighted question you were planning to answer briefly can still remove you from the competition if the answer scores below the threshold. Weighting tells you where to concentrate. Thresholds tell you where you cannot afford to be thin.

Answer the criterion, not just the question

A question asks for information. A criterion describes what a good answer contains. When the criterion says the response will be assessed on “the extent to which the approach demonstrates an understanding of the risks and includes credible mitigation”, the evaluator is looking for three things: named risks, evidence of understanding, and mitigation that holds up. An answer that describes a methodology in confident general terms may satisfy the question and score a 2 against the criterion.

The most reliable technique is to structure each answer around the components of the criterion, in the criterion’s own language, so an evaluator working through the scoring sheet finds each element where they expect it. It reads less elegantly than free prose. It scores better, and it is the single most common gap between good proposals and winning ones, as explored in why proposals fail.

Model the price before you set it

Because the price formula is published, you can calculate what any given price scores against a range of competitor assumptions. If the split is 70/30 and you expect to score strongly on quality, you can work out how far above the lowest likely price you can sit and still win. If the split is 40/60, the same calculation may tell you the commercial offer is the whole competition. Either way, the price should be set with the total weighted score in view, not the price marks alone.

Scoring yourself against the criteria before they do

Everything above comes together in one habit: before you submit, score the proposal the way the buyer will.

It takes less time than it sounds. Take the one page criteria sheet you built. For each scored question, read your answer against the criterion and the scale definitions, and give it an honest mark. Not the mark you hope for. The mark an evaluator with twenty proposals to read and no prior knowledge of your company would give it. Then multiply by the weighting and add it up.

Three things come out of that exercise, reliably. First, the total, which tells you whether you are competitive or hoping. Second, the answers scoring below any minimum threshold, which have to be fixed whatever else happens. Third, and most useful, the cheapest points available: the heavily weighted questions sitting on a 3 that could be a 4 with one more piece of evidence or one clearer paragraph. That is where the last day of a bid should go.

The exercise works best when the person scoring did not write the answers. Authors read what they meant to say. Evaluators read what is on the page. If nobody else is available, leave the proposal overnight and come back to it with the criteria sheet and nothing else. The gap between the two readings is exactly what the buyer will see.

Evaluators are not trying to catch you out. They are working through a scoring sheet, criterion by criterion, looking for the evidence to justify each mark. Give it to them where they expect to find it, in the weight and order the criteria define, and the proposal does the rest.