By Ryan Richardson · Published 8 October 2026
Every receipt gets five lines: the number, the denominator and window most people skip, the context, what changed because of it, and the era, meaning which pricing, offer and channel were live when it happened. Discard anything missing one of the five, however good the story. It's an anecdote wearing a receipt's clothes.
A number pooled across two different eras is really two distributions wearing one mean. Note which era produced each receipt, or a reader checking your arithmetic will quietly average across a change you made on purpose.
Payment records give you money, dated, recorded accurately because it had to be, but tell you what happened, never why. Old reports and dashboards give the denominator and the window, but retention expires, so sweep before rows age out. Decisions you reversed give you the change itself, which most receipts are missing, usually documented only in the reversal. Your own message history gives you the reasoning at the time, unedited by hindsight, which is exactly why it's uncomfortable and exactly why it works.
Sweep in that order. Doing it out of order usually means the earlier sources have already aged out by the time you reach them.
In one documented sweep, a single day of going back through records produced thirteen usable receipts. Two of them had first been remembered in the author's own favour, and were only caught by going back to the original emails and messages. Rebuilding a receipt from memory arrives with the embarrassment sanded off, and the embarrassment is usually the part most likely to be true.
Apply the five-field test to your best stories first, not just your failures. Search your own draft for 'share,' 'most,' 'many' and 'typically,' and read what sits beside each one. That's usually where an unquantified hedge is hiding, dressed up as a sentence rather than flagged as a gap.
Every number you plan to publish has all five fields filled in, including the era. Any number missing a field has either been dropped or sent back for the missing field to be found, not published with a placeholder.
| Claim | Value | Source |
|---|---|---|
| Usable receipts produced from one day of records sweep | 13 | The Sixty Steps manuscript |
| Receipts in that sweep first misremembered in the author's own favour | 2 of 13 | The Sixty Steps manuscript |
| Matrix difficulty/coverage for this step | Moderate / Almost nobody does this | The Sixty Steps matrix |