How bids are scored: award criteria and MAT

Winning a public contract is not a matter of persuading a buyer that you are the best firm. It is a matter of scoring more marks than everyone else against a scheme the buyer has to publish before the competition opens. Every above-threshold tender tells you exactly how it will be marked: which questions carry which weight, what scale the evaluators use, and how your price converts into points. If you read that scheme properly and write to it, you are competing on the buyer's terms. If you write a good bid that ignores it, you can lose to a worse firm that did not.

Quality and price: the split that decides everything

Public buyers award to the Most Advantageous Tender, and they do it by dividing the total marks between quality (sometimes called the technical or method response) and price. The split is a deliberate signal. A 60/40 quality-to-price weighting says the buyer will pay more for a better answer; a 30/70 split says price will usually decide it and your quality answers mainly have to clear a bar. Read the split before you read anything else, because it tells you where to spend your effort.

The full set of award criteria, quality and price together, is published with the weight attached to each. A typical services tender might break down like this:

• Service delivery and mobilisation: 25 per cent
• Quality management and continuous improvement: 15 per cent
• Social value: 10 per cent
• Price: 40 per cent
• Other criteria (for example sustainability or safeguarding): 10 per cent

The arithmetic matters more than instinct. A question worth 5 per cent and a question worth 25 per cent may look similar on the page and take similar effort to answer, but one is worth five times the other. Map every question to its weight before you write a word, and put your best people on the criteria that carry the most marks.

Scoring scales and what moves a score up a band

Quality answers are almost always marked on a banded scale, most commonly 0 to 4, 0 to 5, or 0 to 10, with each band defined by words. A 0 to 4 scale usually reads something like: 0 unanswered or non-compliant, 1 poor, 2 satisfactory, 3 good, 4 excellent. The evaluator picks the band your answer fits, and that band, multiplied by the question weight, becomes your points.

The single most useful thing you can do is read the band definitions, because they tell you what a top mark requires. The gap between "satisfactory" and "good" is rarely about writing more; it is about evidence. Bands typically climb like this:

• Satisfactory answers describe what you will do. They are compliant and complete but generic.
• Good answers describe what you will do and show how, tailored to this contract, with a method the evaluator can follow.
• Excellent answers do all that and prove it: named evidence, worked examples, measurable outcomes from comparable contracts, and a clear grasp of this buyer's specific risks.

Most bids sit at satisfactory because they answer the question asked without evidencing the claim. Moving a 20 per cent question from a 2 to a 4 on a 0 to 4 scale is worth 10 percentage points of the whole competition. That is where bids are won.

How price is scored: the relative models

Price is rarely scored on whether it feels reasonable. It is scored by formula, and the formula is usually relative, meaning your score depends on the other bids, not just your own number. Two models dominate.

The lowest-price model gives full price marks to the cheapest compliant bid, and scores everyone else in proportion. If price is worth 40 marks, the formula is commonly: (lowest price / your price) × 40. Bid 10 per cent above the cheapest and you lose roughly 10 per cent of the available price marks, not all of them.

The mean or average model scores bids against the average of all submissions, sometimes penalising bids that sit too far below the mean as well as above it, on the logic that an implausibly low price signals risk.

Two consequences follow. First, you cannot calculate your price score in advance, because you do not know what others will bid; you can only decide how much quality advantage justifies how much price premium. Second, small price differences usually cost fewer marks than bidders fear. A modest premium, defended by genuinely better quality answers, frequently wins under a 60/40 split. Do the sum for the actual weightings in front of you rather than assuming you have to be cheapest.

From MEAT to MAT: what the change of emphasis means

Under the Public Contracts Regulations 2015, buyers awarded to the "most economically advantageous tender", or MEAT. The Procurement Act 2023, in force since 24 February 2025, renamed the test the Most Advantageous Tender, or MAT, for new procurements in England, Wales and Northern Ireland.

The change is one of emphasis rather than substance, and the government has been clear that it does not alter the underlying policy. Buyers could already weigh quality alongside social, environmental and strategic factors under MEAT, but the word "economically" invited a narrowly financial reading in which everything had to trace back to value for money. Dropping it reinforces that a tender need not be awarded on the lowest price, provided the award criteria still relate to the subject matter of the contract and are published in advance. In practice the mechanics are the same: published criteria, weightings, a banded quality assessment and a price formula. What the new wording underlines is that buyers may give real weight to things beyond the lowest lifecycle cost, and that they must justify their award against the priorities they set out. For bidders, the lesson is to take the non-price criteria seriously as scoring opportunities rather than boxes to tick.

Where social value fits in the scoring

Social value is not a separate exercise bolted on at the end; it is an award criterion with its own weight, marked on the same banded scale as any other quality question. Central government procurements have applied a minimum 10 per cent social value weighting since 2021, and many councils and NHS bodies now match or exceed it against their own published priorities.

Because it is scored like any other criterion, the same band logic applies. Generic pledges to "support the local community" score at the bottom because they carry no evidence and no commitment. Answers climb the bands when they are specific, measured and tied to this contract: a stated number of apprenticeships or local jobs, a named local supply chain, a carbon reduction figure with a method behind it, all proportionate to the contract's size and deliverable by the team actually doing the work. Social value is one of the most reliable places for a prepared small firm to gain marks, precisely because so many bidders answer it badly.

Moderation and consensus: how the mark becomes final

Your score is not one person's opinion. Above-threshold competitions are marked by a panel, and individual evaluators first score independently against the published bands to avoid influencing each other. The scores are then moderated: the panel meets, compares marks question by question, and agrees a consensus score for each answer, supported by a written rationale that has to stand up if the award is challenged.

Two things follow for the bidder. Evaluators can only credit what is on the page, because the moderation record has to cite your text, so nothing implied or assumed will count. And the consensus process rewards answers that are easy to score: clear structure, headings that map to the question, and evidence an evaluator can point to when defending the mark to colleagues. Make the marker's job easy and you make a high band easy to award.

Writing to the scheme

Treat the scoring scheme as the brief. Before drafting, build a simple table: every question, its weight, its scale, and the wording of the top band. Write each answer to hit that top band, lead with the method and the evidence rather than background about your firm, and mirror the language of the criterion so the evaluator can match your answer to their marking sheet at a glance. Answer the question that was asked, in the order it was asked, and put your strongest people on the heaviest-weighted criteria. A bid that is engineered to score will beat a bid that is merely well written, every time.