Moving a chart to slide one will not repair a claim the chart cannot support. Before rearranging a presentation, choose one sentence you want a reader to believe. Find the evidence behind it. Then check whether both are answering the same question.
That is the useful part of a proof-first pitch: a reader can inspect the basis for an important claim without having to infer what was measured, under which conditions or with what limitation. It is a proposed editing method, not a rule about how every investor reads.
Give the result enough context
A technical result needs a reason to be on the slide. Name the intended application, the problem and the decision the result informs. A recovery measurement may matter to a process question while leaving a buyer's formulation question unanswered.
There is no need to pretend that all credible pitch guidance puts a chart first. Sequoia's March 2019 guide starts with company purpose and the customer's problem before moving to the solution; market potential also has a place. That is historical guidance from one investor organisation, not evidence that a particular sequence causes funding success. Read Sequoia's guide.
Use enough context to make the evidence intelligible. Then put the relevant result close to the sentence it supports. The market argument still needs to be made; a technical chart cannot make it on your behalf.
Pair a claim with its actual boundary
For each important statement, complete this card:
| Field | Your entry |
|---|---|
| Decision | What is the reader being asked to assess? |
| Claim | The exact sentence, including scale and conditions |
| Evidence | The original result, record or document, with date and version |
| Scope | What was tested or observed, and against which comparator or requirement? |
| Gap | What the evidence does not establish |
| Next step | The test, information or decision needed to address that gap |
Keep the strength of the sentence within the scope of the record. A bench result alone does not establish performance at another scale. A paid evaluation alone does not establish repeat orders. These are limits on what follows from the evidence described, not predictions that the business will fail.
Where a source is confidential, identify its nature and the agreed access route. Do not imply that a reader has inspected a record that remains unavailable to them.
Try two claims
Both examples below are invented for this exercise. They are not company results. Rewrite each sentence before reading the suggested answer.
The narrower wording still communicates progress. It also gives the reader a clear place to start the next conversation.
Check the sequence with another reader
Ask someone who has not helped write the deck to read the revised section. Without coaching, can they identify the decision, the supporting record and the unresolved question?
If the answer is no, first establish what they could not understand. They may need a clearer account of the application before the result. They may have read an ambition as an achievement. Or the evidence itself may be too thin for the claim.
Revise the wording or order accordingly, then repeat the check. This is a qualitative test of clarity, not a forecast of investment. Keep the completed cards with the deck, and revisit them when a new result or customer record changes what you can responsibly say.
Adapted from the author's proof-first deck argument. For the broader diligence questions, read Why investors pass on food biotech companies.
