Skip to content

Writing a winning bid

Most losing proposals are well written. They describe an excellent company instead of answering the question being scored. The fix is structural, not stylistic.

Proposal writer working through evaluation criteria

An evaluator is not deciding whether your company is good. They are awarding points against criteria, one at a time, and writing a justification that has to survive review. A winning proposal makes that easy: it puts the evidence for each criterion where the evaluator is looking for it, in the language the criterion uses. That is a discipline about structure and evidence, and it can be learned.

Answer the requirement, not the question you prefer

The most common failure is a strong answer to a different question.

When a criterion asks how you will manage schedule risk on a distributed team, it is not asking about your project management philosophy, your certifications, or your company’s track record in general. It is asking what you will do, on this contract, about that specific risk. An answer that describes your methodology in the abstract can be excellent and still score badly.

Read each criterion as a question and write the direct answer as the first sentence of your response. Everything after that is support. Evaluators reading under time pressure reward this heavily, and reviewers checking a scoresheet can find the justification immediately.

Use the buyer’s vocabulary. If the solicitation says transition period, do not write onboarding phase. Terminology mismatches force the evaluator to make a connection on your behalf, and under a strict evaluation they may not be permitted to.

Write for the person scoring the section

They are reading your section in isolation, against a rubric.

Evaluation teams usually split the work. One person may score only the technical approach across every proposal received, comparing your section 3 against six competitors’ section 3. They may never read your management volume, and they cannot credit evidence that sits somewhere they are not reading.

So each section has to stand alone. Repeat the key fact rather than cross-referencing to another volume. Where a page limit forces a choice, cut the background and keep the evidence.

Formatting is a scoring aid, not decoration. Headings that mirror the criteria, a short summary at the top of each section, and tables where the answer is genuinely tabular all reduce the effort of finding your evidence. Dense unbroken prose does the opposite, and the evaluator’s fatigue is a real factor by the fifth proposal.

  • Mirror the solicitation’s headings and numbering
  • Lead each section with the direct answer
  • Make every section stand alone
  • Repeat key evidence rather than cross-referencing

Evidence beats adjectives

A number with a source outscores a paragraph of assurance.

Past performance evidence being matched to evaluation criteria

Every bidder claims to be experienced, reliable, and quality-focused. Those words carry no information and cannot support a score. What supports a score is a specific fact: the client, the scale, the dates, the outcome, and ideally a number that can be checked.

Compare two versions of the same claim. One says the team has extensive experience delivering similar services to government clients. The other says the team delivered the same service to a named department across three sites for four years, sustaining a stated availability level, with the same project manager proposed here. The second gives the evaluator something to write down.

Where you cannot be specific because of confidentiality, be specific about everything you can: the sector, the scale, the duration, the measurable result. Vagueness reads as absence of evidence, and evaluators are trained to treat it that way.

Structure that survives compliance review

Build the skeleton from the solicitation before writing a word.

Start with a compliance matrix. Pull every shall, must, will, and required statement from the entire document, including the annexes and the clauses, and map each one to a proposal section with a named owner. That matrix is your outline, your writing assignment sheet, and your final check.

Then write to the outline rather than reorganising as you go. Proposals that get restructured late are where compliance breaks, because a requirement that lived in a deleted section quietly stops being answered. Re-run the matrix after the final edit, and specifically after any cut made for page count.

Keep a separate checklist for the mechanical requirements: forms, signatures, amendment acknowledgements, file naming, volume separation, page limits, font and margin rules. These are pass or fail and have nothing to do with the quality of your solution.

  • Extract every requirement before outlining
  • Assign each to a section and an owner
  • Write to the outline, do not restructure late
  • Re-run the matrix after every cut

Price is part of the argument

A price without a rationale invites the wrong conclusion.

How price is treated depends on the basis of selection. Under a lowest price technically acceptable evaluation, quality above the acceptable threshold earns nothing, so effort should go into clearing the bar efficiently and pricing sharply. Under a best value trade-off, a higher price can win if the proposal justifies it, and the technical response has to make that justification explicitly.

Where the buyer will assess realism, an unexplained low price is a risk, not an advantage. It can be read as a misunderstanding of the requirement and can cost you technical points or the award. If your price is low because of a genuine efficiency, say what the efficiency is.

Make the price traceable to the technical response. Staffing levels, hours, and assumptions in the price volume should match what the technical volume describes. Evaluators do check, and a mismatch undermines both volumes at once.

Protect time for review

The last 48 hours decide more than the first two weeks.

Schedule backwards from the deadline with real time for a compliance check, a fresh reader who has not written any of it, and the submission itself. A reviewer who has not seen the document is the only person who can tell you whether an answer is actually findable.

Run two separate reviews. One checks compliance mechanically against the matrix and the format rules. The other scores the proposal as an evaluator would, criterion by criterion, using the published criteria and nothing else. If a reviewer cannot find the evidence for a criterion in a minute, an evaluator will not either.

Then submit early enough to survive a portal problem. Upload failures at the deadline are common, and no buyer has discretion to accept a late submission because the system was slow.

Keep the useful signals

Captivaq watches procurement sources against your capabilities, certifications, locations, and past performance. It helps you spend time on opportunities that have a credible fit.

Build a matching profile

Related resources

Questions people ask

How long should a proposal be?

As long as the page limit allows and no longer. Where there is no limit, length should follow the evaluation weighting rather than the size of the scope. Extra pages that do not support a scored criterion add reading time without adding points.

Do proposal graphics help?

Where they carry information an evaluator would otherwise have to assemble from prose, yes. A staffing chart, a transition timeline, or a process diagram can earn their space. Decorative graphics consume page count without supporting a score.

Should you bid if you do not meet every requirement?

Not if a mandatory requirement is unmet, because the bid will be set aside without scoring. Where the gap is in a rated criterion, it becomes a judgement call about the total score and the cost of writing. Teaming with a partner who fills the gap is often the better answer.

How early should proposal work start?

Ideally before the solicitation is published, using forecasts and market research notices to prepare capability material, references, and staffing. Once the document is out, the clock is usually too short to build anything that did not already exist.