September took two real hits — a mispriced lifetime SKU on Apple and an ad format that quietly stopped firing — and is recovering from both. Below is what changed, what it cost, what's already fixed, and what's still open.
The week of Sep 14 was the floor — the week Apple's lifetime purchase went to effectively zero sales after a price update didn't apply cleanly. Two weeks of recovery since.
*Partial week — still collecting. Not a projection.
August was the best month recorded. September is projecting a real step down — both incidents landed in the same three–week window.
*Projected — straight-line from 26 of 30 days. June shown as gross (net breakdown not modelled for that month).
| Month | From app sales | From ads | To account |
|---|---|---|---|
| July | $16,291 | $4,834 | $21,125 |
| August | $17,240 | $5,079 | $22,319 |
| September (proj.) | ~$14,800 | ~$4,130 | ~$18,900 |
Raising the lifetime price to align with competitors (still cheaper than every real alternative — see below) landed cleanly on Play but missed the US and 13 other regions on the first attempt, and Apple's product briefly stopped selling entirely after the update. Both are now selling again at the new price. Issue 346.
| Units/day | 3.23 → 1.17 |
| Revenue/day | $166 → $84 (−50%) |
| Units/day | 2.46 → 1.17 |
| Revenue/day | $118 → $70 (−40%) |
Fewer buyers at the higher price, as expected. But new subscription starts rose at the same time — Apple +50% revenue/day, Play +60% — on both stores, in the same window. Total new-buyer revenue (lifetime + new subscriptions combined) is down 15%, not the 40–50% the lifetime number alone suggests. Some of what was lost on lifetime shows up as subscriptions instead.
A September app update (v11.0.0) deleted the ad trigger on the component screen — the highest-traffic screen in the app, and the one carrying the single most valuable ad format (interstitial, $17–22 per thousand views vs $0.69 for a banner). It was removed on purpose, to stop ads interrupting mid-read; nothing was built to replace it. Issue 317.
Interstitial earns 25× a banner impression and lost two-thirds of its volume in the update. That mix shift — not falling ad prices — is the whole drop: no ad network changed what it pays.
| Diagnosed loss, mid-Sept, pre-fix | ~$1,500/mo |
| Two server-side fixes, live now | already clawed back ~$570/mo |
| Remaining gap today | ~$930/mo, measured Aug→Sep |
A new non-interrupting format (App Open ads) is designed but not built — Feature 49, Flutter work. Its ~$3,000/mo is a projected gross figure for the format itself (no eCPM data exists yet; estimated conservatively at $6), not the incremental amount it would recover against today's ~$930/mo gap. Those are two different claims and shouldn't be read as one.