Evidence vs Claims: The Line That Moves Your Score

There is one edit that improves more tender responses than any other, and it is not a writing skill. It is a habit of mind.

Read each sentence describing your capability and ask: if a panel member had to defend this score to a colleague, what would they point at?

If the answer is a page, a project, a drawing, a number or a name — that sentence is evidence. If the answer is "well, the sentence says so" — it is a claim, and a claim has a ceiling.

Why claims have a ceiling

Scoring is not a private act. An assessor's individual score gets tested in a moderation meeting, where the panel works through each criterion and reaches a consensus score that must be written down and justified.

Put yourself in that room. A panel member has scored you 8 out of 10 on organisational capability. A colleague who scored 5 asks why.

If the answer is "they say they have extensive experience in this sector", the 8 does not survive. Everyone said that. There is nothing to argue with, so the score drifts to the middle.

If the answer is "section 3.2 — the Riverside upgrade, same client, occupied site, staged over eighteen months, and the reference contact is listed", the 8 holds. The colleague can look at it, and either they agree or they have to make a specific counter-argument.

Evidence is not there to persuade the assessor reading it. It is there to arm them for the meeting. That reframing changes how you write.

The four things a claim is usually missing

Take a typical sentence: "We have significant experience delivering civil works in live operational environments."

A name. Which projects? Unnamed experience cannot be verified, and unverifiable content gets discounted by experienced panels almost automatically.

A comparison. Why is that experience relevant to this job? "Live operational environment" is doing a lot of work in that sentence and explaining none of it. The relevant detail might be that the facility stayed open, or that services stayed live, or that public access was maintained — and those are three different capabilities.

A number. Value, duration, quantity, headcount, percentage. Numbers are the cheapest credibility available and most responses use almost none.

A source. A reference contact, a client name, a completion certificate, a photograph, an extract from an approved plan. Something outside your own prose.

Rewritten with all four, the same claim becomes: "We delivered the Riverside Water Treatment upgrade for [client] — $4.2m, eighteen months, with the plant remaining fully operational throughout and no unplanned shutdowns. Staging and shutdown coordination were managed under the approach described at section 4.3. Reference details at appendix C."

Same underlying fact. A completely different score.

Evidence a panel treats as strong

Not all evidence carries the same weight. Roughly in descending order of how much it moves a score:

1. Verifiable third-party facts — named client, named project, reference contact, certificate, award, audited figure.
2. Extracts from real documents — two pages of an actual inspection and test plan from a comparable job, an actual staging drawing, an actual risk register entry.
3. Specific commitments — named individual, stated time allocation, stated quantity, stated date.
4. Project-specific reasoning — an explanation that could only have been written by someone who read this tender's documents.
5. General description of your systems and approach — necessary, but on its own it sits at the middle of the band.
6. Adjectives about yourself — "extensive", "proven", "world-class", "unrivalled". These add nothing and, in volume, actively cost credibility.

Most weak responses are made almost entirely of items 5 and 6.

The attachment trap

A common half-fix: the writer knows evidence is needed, so they attach the full quality manual, the full risk register, all fourteen CVs and a folder of certificates.

That is not evidence. It is raw material, handed over without an argument attached.

An assessor scoring criterion 4 will not read a 90-page manual to build your case for you. What scores is the two pages that show the system working on a job like this one, extracted and placed where the claim is made, with a sentence explaining what the reader is looking at and why it matters.

Evidence must be cited from the narrative that relies on it. An attachment nobody was pointed to is an attachment nobody read.

Where honesty outperforms confidence

Contractors sometimes worry that specificity exposes weakness — that naming a $3m project looks small, or that admitting a gap invites a lower score.

In practice the opposite holds. Panels read a great many submissions that claim everything and demonstrate little, and they become sceptical by default. A response that says "we have not delivered this exact asset type, but we have delivered three projects with the same access and staging constraints, and here is how we will manage the difference" reads as credible in a way that "extensive experience across all sectors" never does.

Vagueness is not safety. It is the thing every mid-band response has in common.

Decision Signal 3 of circa 150 — Evidence Over Claims. A claim tells a panel what you assert about yourself. Evidence lets a panel verify it. Because consensus scores must be defended in a moderation meeting, evidence is what a panel member uses to argue your score upward — and a claim gives them nothing to hold.

Finding the claims you cannot see

The difficulty with this edit is that your own claims do not look like claims to you. You know the projects behind them. The assessor does not.

That is what the evaluation is for. ProcureHQ's Digital Evaluation Committee assesses your response against the criteria extracted from your own tender documents and returns findings per criterion — strengths, weaknesses and, specifically, missing evidence — along with the evidence the assessment actually relied on, so you can see which of your sentences did work and which did not.

Your first tender is free: one full evaluation and one complete Smart Submission Blueprint, per organisation.

---

Frequently asked questions

How much evidence is enough for one criterion? Enough that every distinct claim has something behind it, and no more. Two well-explained comparable projects beat ten listed ones. Volume without relevance reads as padding.

Can I use the same project as evidence for several criteria? Yes, and you should — but draw out a different aspect each time. The same job can evidence capability, methodology and management systems if each reference addresses that criterion specifically rather than repeating the project summary.

Do referees actually get contacted? Sometimes, particularly for shortlisted tenderers. Assume they might be, keep contact details current, and tell your referees they are listed.

Is a photograph useful evidence? It can be, where it demonstrates something the text asserts — a completed comparable structure, a site setup, a traffic management arrangement. A photograph with no caption tying it to a claim is decoration.