What a CRO bid defence needs

A bid defence is a ninety-minute meeting in which a sponsor decides whether to believe you. The feasibility section is usually where it turns, because it is the part of the bid that makes falsifiable claims: this many sites, in these countries, enrolling at this rate, finishing by this date.

The mistake is to prepare for that meeting by making the numbers better. The sponsor’s team has seen a hundred enrolment curves and they know roughly where yours should sit. What they are actually testing is narrower and harder to fake: do you know where your numbers came from, and would you notice if they were wrong?

The five questions

Almost every challenge in a feasibility defence is one of five questions wearing different clothes:

1. Where did that number come from? — the source and the record.
2. When did you pull it? — the retrieval date, and whether it has moved since.
3. Why that filter? — the rule. Which statuses, which phases, which window, and why those.
4. What if you are wrong? — the sensitivity. What the plan looks like at half the assumed rate.
5. Who signed it? — accountability, and whether the person who signed is the person who ran it.

Note that only one of the five is about the number itself. Four are about the process that produced it. A team that answers the first three fluently earns enormous credit on the fourth, because the sponsor now believes the sensitivity was computed rather than estimated in the room.

The traceability problem is a plumbing problem

Most feasibility packs lose their provenance in transit, not at the source. The analyst had the citation. It survived the working spreadsheet. It did not survive the summary tab, and it certainly did not survive the slide. By the time the figure reaches the bid document it is a bare integer, and the only way back to its origin is the analyst’s memory — which is fine until the analyst is on holiday, or has left, or is answering about the version from six weeks ago.

The fix is not discipline. It is that the citation has to be attached to the number as a property of the number, so that copying the figure copies its origin. Anything that relies on a person remembering to carry the source forward will fail at exactly the moment it matters, because that moment is always under time pressure.

What a defensible pack contains

Independent of tooling, here is the inventory. If your pack has these, the defence is a conversation; if it does not, it is a test of nerve.

Signature is not a formality

A sign-off that says “the feasibility team approved this” is worth less than one that names a person, a role, a moment in time and the exact version of the document they were looking at. The distinction matters most when the answer changes: if the March version said one thing and the May version says another, an accountable record shows who approved which, rather than leaving the sponsor to infer that nobody was really watching.

The same property that makes a signature meaningful for a regulator makes it useful in a commercial defence — it binds a named individual to a specific, unaltered artefact. Where the record is append-only and each entry references the one before it, a retroactive edit is detectable rather than deniable, which is precisely the assurance a sponsor is trying to buy.

Where the honesty pays

There is a natural fear that a pack full of visible caveats loses to a pack full of confident numbers. In a first-round screen, sometimes. In a defence, almost never — because the sponsor’s team will find the soft spot regardless, and the only variable is whether they find it in your document or in the meeting.

A caveat you raised yourself reads as command of the material. The same caveat, surfaced by the sponsor, reads as something you missed or something you hid. That asymmetry is large, and it is the strongest practical argument for citation that we know of: not that it is more rigorous, though it is, but that it is the only version of the pack where the difficult questions arrive on your terms.