How public sector tenders are scored

Understanding how your tender will be marked changes how you write it. Public sector evaluation is more mechanical — and more predictable — than most first-time bidders expect. Here is how it works.

The published weightings are the whole game

Every compliant tender states how it will be evaluated, usually as a percentage split, for example:

Quality is then broken down further: individual questions each carry a stated share of the quality marks. This tells you exactly where to spend your effort. A 15% methodology question deserves three times the attention of a 5% question — yet many bidders write the same amount for both.

How individual answers are marked

Most UK buyers score each written answer on a defined scale, typically 0–5 or 0–100 in steps, with published definitions along these lines:

• 5 — Excellent. Fully addresses the question, robust evidence, no reservations.
• 4 — Good. Addresses the question well with minor gaps.
• 3 — Acceptable. Meets the requirement but with limited detail or evidence.
• 2 — Minor reservations. Partial response, weak evidence.
• 1 — Serious reservations. Significant gaps or concerns.
• 0 — Unacceptable. Fails to address the question.

Two things follow from this. First, evaluators mark against the question and the scale, not against their general impression of your company. A brilliant business that half-answers the question gets a 2. Second, the difference between a 3 and a 5 is nearly always specificity and evidence — named processes, real examples, numbers — rather than better adjectives.

Evaluators can only score what is on the page

Evaluation teams are usually required to mark independently and then moderate their scores together, recording the reasoning. They cannot give you credit for things they know about your business but which you did not write down; that would be challengeable by other bidders. Assume the evaluator has never heard of you, because procedurally, they must pretend they have not.

Practical consequences:

• Answer every part of the question. Multi-part questions ("describe your approach, including how you will manage risk and report progress") are scored on every part.
• Do not rely on cross-references to other answers unless the tender explicitly allows it — many forbid it.
• Repeat key evidence where relevant, even if it feels redundant.

How price is scored

The most common formula gives the lowest bidder full price marks, with others scored proportionally — for example, (lowest price ÷ your price) × available marks. Some buyers instead score against a budget or use quality thresholds. The formula is published; read it, because it affects strategy. Under a proportional formula against strong quality weighting, being 10% more expensive is often recoverable through better answers. Under a price-heavy evaluation, it rarely is.

Abnormally low bids can be challenged and rejected — buyers may ask you to prove a low price is deliverable.

What this means for how you write

1. Allocate words by weighting, not by how much you happen to know about each topic.
2. Mirror the question's structure in your answer, using its own language as subheadings where word limits allow. You are making the evaluator's marking sheet easy to fill in.
3. Evidence every claim. For each statement, ask: what example, number or named process proves this?
4. Write for a stranger. No shorthand, no assumed reputation.
5. Check the scale definitions and honestly grade your own draft before submitting. If an answer reads like a 3, it will score like one.

Bid Sense's Tender Workspace extracts every scored question with its weighting, explains what the buyer is really asking, and suggests which evidence from your company profile supports each answer. Combined with a clear bid/no-bid recommendation up front, it keeps your effort pointed at marks you can actually win. Join early access to see it on a real tender.