Evidence & experiments · Essay 060
Build an evidence ledger
The Accidental Salesman Editorial Desk · 2 min read · Published 15 October 2026

A growing firm accumulates claims in proposals, presentations and website copy. Over time, the same result may appear with different dates, metrics or explanations. An evidence ledger creates a single place to check what can responsibly be said.
For each material claim, record the wording, source, period, owner, permission status and limitations. Link to the underlying document rather than copying a number without context. Distinguish direct measurement from estimates and client recollection.
The ledger should be small enough to maintain. Start with claims used repeatedly in sales: outcomes, customer counts, experience statements and comparative assertions. Add new entries when a claim becomes important enough to influence a buyer's judgment.
Assign a review date where information can become stale. A capability statement may change when a key specialist leaves. A product comparison may change when the competing product changes. Historical outcomes remain historical and should retain their original context.
The tool also improves writing. When evidence is weak, the editor can narrow the statement, seek better support or present it as a hypothesis. This is more efficient than discovering the problem after the claim has spread across several channels.
An evidence ledger does not make every statement scientific. It makes the basis of important statements visible. That is a practical standard a small professional firm can sustain without building a research department.
Put it to work
A practical next step
Create entries for your ten most consequential sales claims. Remove or qualify any statement whose source cannot be found. Assign one owner to maintain the record after publication.
Proposed practice, not a completed experiment. Illustrations are not client testimony.
Sources & limits
This essay is editorial interpretation from the supplied manuscript, not an empirical research report. No external study is cited for its recommendations.
Source: The work of winning trust, editorial review edition, 15 October 2026, p. 74. No publication history is inferred from its sequence dates.