THE PRICE RISE SCORECARD — The Price Rise

The week after a rise, someone will read the conversion rate and declare victory or defeat. Both are premature. New visitors have no reference price. The stock-up window moves orders around the date. And customers who leave over price just don’t come back: in Anderson and Simester’s experiment the lost orders showed up over 28 months Published.

Build the comparison first

You need something to compare with, decided before the rise. Three workable options, best first:

Staggered SKUs. Raise one group of SKUs now and a similar group a quarter later. Until the second group moves, it’s your comparison.
Last year’s cohort. Customers active in the 90 days before the rise, against customers active in the same 90 days a year earlier, followed for the same number of days. It controls for season, not for everything else.
Affected against unaffected buyers. Customers whose usual products went up against customers whose usual products didn’t. Useful when only part of the catalog moves.

Randomly holding some customers at old prices gives the cleanest answer, and creates what chapter 9 warns about: two customers paying different prices. If you do it, keep the group small, give it an end date, and have counsel review it. The Honest Test covers price tests.

The page

Number	Defined as	When	What it catches
Volume against break-even	Units per SKU, four-week average after the stock-up window, against break-even (chapter 3)	Weekly	A rise losing more than it can afford
Contribution per SKU	Monthly contribution against the same months before	Monthly	Whether the rise earns what it should
90-day repeat rate	Share of customers active before the rise who reordered within 90 days, against the comparison	Day 90	The early read on existing customers
180-day repeat rate	The same, at 180 days	Day 180	The verdict
Subscriber survival	Share active through the first two renewals at the new price	Each renewal	Subscribers leaving quietly
Price cancellations and tickets	Cancellations and tickets citing price, per 1,000 orders	Weekly	A story that isn’t landing
Tier mix	Share of orders in each tier	Weekly	Buyers sliding down the ladder (chapter 7)
New-customer conversion and first-order value	Against the four weeks before	Weekly	A guardrail on acquisition, not the verdict

Conversion tells you how strangers took the price. Retention tells you how your customers took the rise.

Reading it

At 90 days, compare each SKU’s volume loss with its break-even and the repeat rate with the comparison. Volume loss well inside break-even and a repeat rate close to the comparison is a rise that worked; confirm it at 180 days. If repeat rate falls further behind than the break-even allows, the fix is usually in chapter 9 or chapter 10, not a price cut. Rerun the tool in chapter 8 with the loss you saw before the next rise.

Do this

Before the rise, choose the comparison and save the baseline: customers active in the 90 days before, and last year’s equivalent group. Put the day-90 and day-180 reads on the calendar with an owner (queries in Appendix A).

This is one chapter of The Price Rise, which is free and readable in full on a single page with no form in front of it.