Chapter 36: Old List, Locked Door
The email to two hundred thousand old readers was twelve minutes from making the whole trial useless.
At 11:48 on Saturday, the North Quay mailing console showed SEND SCHEDULED — 12:00. The recipient estimate read 200,000.
The subject line belonged to Marian Crowe.
Inside, two draft buttons linked to Saltmere and Larkspur. The links had been inserted by a legacy marketing template, not either project team.
Neither project had authorised the old name to introduce it.
The marketing desk called the mistake “balanced” because both links appeared. Each button occupied the same width beneath the old cover strip.
Equal contamination was still contamination.
PageTurn joined the emergency call with both natural-return baselines already paused. No visit generated during the uncertainty would enter the fourth-episode test.
The neutral contract administrator joined through North Quay's separate legacy channel.
The administrator did not open either new manuscript.
Only the mailing schedule, shared-name licence and blocked-list identifiers appeared on that screen. Reader addresses remained inside North Quay's mailing system.
“Can you remove both links and send the old-book news?” marketing asked.
The administrator checked the schedule history.
The draft had been created as one campaign against the Marian Crowe list.
Changing content now would create a replacement mailing, not preserve a neutral old notice. It would also require a fresh dual approval the clock could not supply.
Oliver asked whether the two links could be sent in equal positions.
“No new project has permission to use this list,” the administrator said.
The answer did not prefer mine. It treated my majority drafting history as irrelevant to a list licensed under two signatures.
It removed the largest audience available to both. The projected reach exceeded our two clean email layers several times over.
Nine minutes remained.
The administrator opened the C2 legacy-resource restriction and the later blocked-list schedule.
Both required no future-brand use without the missing shared approval.
I confirmed that I gave none. My refusal entered the same cancellation record rather than a public statement.
Oliver did not supply the second signature the common licence required.
Pippa's project role gave her no authority over the old list. Her accurate new credit neither expanded nor reduced the shared licence.
At 11:53, the administrator selected CANCEL ENTIRE CAMPAIGN.
North Quay's system demanded a second internal reviewer. The administrator could initiate cancellation but not certify it alone.
The reviewer matched the list identifier, schedule and zero-send state.
At 11:55, the campaign changed to CANCELLED — NO RECIPIENTS. The zero-send counter carried a server timestamp.
The send clock disappeared.
PageTurn required both project teams to accept three consequences.
No replacement email, even under our individual names.
No compensating platform exposure, because losing prohibited traffic was not a delivery failure.
No credit for the cancelled audience in later reports.
I clicked ACCEPT.
The opposing summary turned accepted under the same wording. I saw compliance, not their internal discussion.
North Quay exported the cancellation log and zero-recipient report.
PageTurn cross-checked them against the exposure ledger.
The old list contributed exactly zero visits to either fourth episode. That figure entered the future retention audit as an exclusion check.
My clean-room account retained only the signed summary, not two hundred thousand addresses.
The cancelled campaign would have dwarfed my launch discovery pool.
Losing it felt less principled when I saw the number written in full. Two hundred thousand chances had disappeared with one correct click.
Some old readers were already asking on the public series page why no new-project notice had arrived.
I could not answer by handing them a link the agreement prohibited. Their disappointment remained a real commercial cost rather than evidence of unfairness.
The administrator closed the legacy channel without commenting on our stories or prospects.
PageTurn restarted the natural-return baselines at 12:01.
Both projects now had to earn episode-four visits without the shared-name mailing.
My budget dashboard displayed the next threshold.
HIGH AUDIO TEST BAND — COMPOSITE RETENTION AT LEAST 45%.
With two hundred thousand familiar readers locked outside, Carys alone had to bring enough strangers back.