Chapter 53: Twenty-Four Shops
Twenty-four shops wanted an answer, and fairness allowed me exactly twenty-four.
On Thursday, the distribution team opened the seventy-two-hour trial allocation table.
The approved factual card had attracted thirty-one independent expressions of interest. Their timestamps remained visible in the buyer-intent register.
Seven arrived after the matched-shop pool had already been defined.
Adding them would give Saltmere more doors than Larkspur during the measured window.
The extra interest was real and commercially useful outside the test.
It could not be smuggled into a comparison merely because I had earned it.
PageTurn and North Quay displayed the common consumer-reach rule beside both projects.
Each route received the same eligible geography, campaign duration and paid-button visibility. Performance would be measured only inside those locked boundaries.
Store counts differed only where independently qualified buyers had accepted the test terms.
Saltmere's permitted ceiling was twenty-four shops.
The seven additional shops would remain dated enquiries, not active trial doors.
I accepted the cap in the distribution workspace.
The system moved those seven names into AFTER TEST — NO MEASURED TRAFFIC. Their enquiries stayed valid for later ordinary sales discussion.
No promotional link, reservation or consumer impression could pass through that queue early.
My coordinator objected that waiting might lose impatient buyers.
“Then that loss belongs in the cost of comparability,” I said.
The trial table next locked what each active shop could expose.
Every store received the same product page, factual card and purchase rules.
No shop could add the Marian Crowe mailing list, old covers or a legacy badge.
The physical samples proved review conditions; they were not sale inventory. No customer could buy one of the numbered proof copies.
Customer purchases would become verified demand signals for later production approval.
The table included a provisional unit ceiling per shop to prevent one promotion dominating. Those ceilings measured demand and authorised no production quantity.
Unused allowance could not be moved between shops during the seventy-two hours.
Nor could I redirect the seven excluded buyers' interest into a participating store.
At 11:32, North Quay randomised shop identifiers across the reporting view.
I could see aggregate reach and my own approved buyer commitments, not customer identities.
PageTurn would retain payment status, cancellation time and invalid-payment flags.
North Quay would receive only the approved de-identified summary and shop payment basis. PageTurn would not transfer customer profiles or reading histories.
The neutral old-rights administrator had no content or performance role in the trial.
I signed the twenty-four-shop allocation and the common reach schedule. Separate signatures fixed the shops and the consumer-touch rules.
The platform fixed the planned opening for Friday evening.
The seventy-two-hour clock would begin only when both purchase routes opened together.
A preview order placed before that moment would not enter measured conversion.
My marketing coordinator showed me a waiting list already promising early purchases.
The gross number looked like a victory before any card had settled.
Then the trial rules opened the calculation beneath it.
QUALIFIED NET PAYMENT = SETTLED ORDER MINUS CANCELLATION, REFUND OR INVALID PAYMENT.
A placed order was not yet a paid unit in the final measure.
Every cancellation would subtract from the same cohort that produced the attractive peak.
The calculation applied to both projects without a manual publicity adjustment.
I approved it knowing my larger early-interest list created more room for reversals.
The locked rule entered the platform and distribution archives with matching hashes.
At 15:06, the twenty-four active shops turned green.
The seven waiting shops remained visible only as post-test opportunities.
I could not call them lost orders, because no measured order had been offered.
I could not call the twenty-four a print run, because nothing had been authorised to print.
The trial would test what customers actually paid and kept.
On the final preview screen, a warning sat directly beneath the buy button.
GROSS ORDERS ARE PROVISIONAL; EVERY CANCELLATION REDUCES NET CONVERSION.