By Ryan Richardson · Published 8 October 2026
Repricing on evidence, and expecting to do it more than once, is step 57 (Part IX), marked moderate difficulty, almost nobody does this, manual tooling (repricing is a judgement call, not something to automate away).
$16.57 was the own-funnel average order for months, against a floor needing something closer to $20 to $40. The first repricing raised order value modestly and was read as partial credit rather than the finding itself. The second raised it again, with traffic and creative also changing around it, muddying the read a second time. The third raised it to about A$31; the daily line crossed break-even and held, on 19 August 2026, with nothing else changed that week. The fix wasn't finding better customers. It was no longer underpricing the ones already there.
Pricing feels like a decision made once, at the start, when the least is known about the business. It then becomes a fact about the business instead of a variable, and every later problem gets treated as a traffic or creative problem instead, because those feel like things that can be worked on. A price change is legible in a way a reworded headline isn't: move the price and the result gets worse, and that's harder to blur as 'the algorithm shifted.'
Move one price at a time, add-ons before the front end, since add-ons carry more of the order value than expected and are less likely to change who buys; the front end changes who's in the audience too, not just the revenue per buyer, so it moves last. Never reprice during a week when something else also changed. Set a trigger: a ladder review whenever average order value moves more than a stated percentage against the trailing four weeks, owned by one named person who isn't the person who built the automation, and never run in the same week as a page or offer change.
Repricing in a period when something else also changed, visible afterwards as two change-log entries on the same day. Prices living on the page rather than the server, so a reprice becomes a deployment and happens less often than it should. Reading a thin, early sample after a reprice as the verdict, when a small sample will swing on noise before it's telling you anything real.
| Claim | Value | Source |
|---|---|---|
| Average order before fix | $16.57, against a floor needing $20-40 | Measured in Real Money, Field Manual |
| Date the third repricing crossed break-even | 19 August 2026, order value near A$31 | The Sixty Steps manuscript |
| Ladder review trigger | fires when average order value moves more than a stated percent against the trailing four weeks | The Sixty Steps manuscript |
| Matrix status for this step | Moderate difficulty, almost nobody does this, manual tooling | The Sixty Steps matrix |