2026-04-23 · 4 min read
The trigger happened on a Thursday night at 9:40 PM in our Shanghai office. QA sent a message: one App Store screenshot had an old onboarding screen, and German text was clipped on another.
The build was good. The screenshots were not. We were ready to ship, but the store listing was now the blocker.
Our old process was textbook manual work. Capture on a simulator, drop files into a shared folder, open two design files, replace layers, fix wrapping, export again, then rename everything for App Store and Google Play.
Every release looked different in product, but screenshot work looked the same. That was the problem. We were repeatedly paying setup cost instead of only paying change cost.
That weekend I built a crude prototype at my kitchen table, around 7 AM on Saturday. It only automated capture and naming, but even that tiny version saved enough time to prove we should keep going.
Then we set three non-negotiable rules. Rule one: freeze screenshot scope 72 hours before release. If scope changes later, it needs explicit approval from product and growth.
Rule two: one copy sheet is the source of truth. No direct text edits in design files. If copy changes, it changes in the sheet, then propagates forward.
Rule three: no manual pixel edits after capture starts. You can tune template settings, but you do not patch random files in Finder at 1 AM.
Once those rules were in place, the same release that used to take us half a day of screenshot churn dropped to roughly 90 minutes, including review and final packaging.
Store Screenshot Studio came from that exact pain. It was not a side project idea. It was release pressure turned into product.