ASO is the most overrated underrated discipline in iOS. Founders skip it for two years, then panic about it the week before launch. This is the working playbook we use across eight live Tappa apps in 2026.
1. Title and subtitle still do most of the work
The top of your listing — name, subtitle, first screenshot — does about 80% of the conversion work. Spend a disproportionate share of your time here. Test variants in App Store Connect. Move on once one wins by more than 10% with confidence.
2. The keyword field is not dead
Apple keeps the 100-character keyword field meaningful. Do not waste it on words that already appear in your title. Do not duplicate. Do not pluralise — Apple does that for you.
Use a tool that pulls keyword volume by region and update the field every release. We update on every shipping cycle.
3. Screenshots are advertising
The screenshots in your listing are not screenshots of the app. They are advertisements that happen to use the app's UI. Each screenshot should make one promise. The first three are the only ones most people see.
We ship a new screenshot set roughly every six months on every app. The data on this is unambiguous: fresh, on-message screenshots beat stale ones every single time.
4. Reviews are a flywheel
Four-point-eight is the modern minimum. Below that, you lose conversion at the listing. To get there:
- Ask for reviews after a clear win in the app, never on launch.
- Cap to one prompt per user per six months. The system does this automatically; do not work around it.
- Reply to every one-star and two-star review you can. The reviewer rarely changes their mind. The next reader does.
5. Custom Product Pages — use them
Introduced years ago and still under-used. Make a custom product page per acquisition channel and read the conversion difference. Our paid-ad pages convert roughly 30% better than our default page across the portfolio. Free upside.
6. Featured story is a nice-to-have
A Today-tab feature is a great week and a forgettable month. Build for the long tail, not the spike. We have been featured. It was a fine week. The rest of the year mattered more.
7. New in 2026 — the LLM channel
A real share of new app discovery is now happening inside ChatGPT, Claude, Perplexity, and Google's AI Overviews. "Suggest me a good currency converter" is a query a model answers. Whether your app is in that answer depends on whether the model has read about it.
This is part of why we built The Forge — long, indexable articles where each app is described in human language a model can summarise. Treat your blog like a press kit a robot will read.
What we no longer spend time on
- App preview videos for utility apps. The data does not support the cost.
- Heavy localisation for apps where the audience is mostly English-comfortable in our target regions. Localise where the data says yes; do not localise everywhere by default.
- Vanity featurings on small media that do not move installs.
If you want a second pair of eyes on your listing, write us. We do free 30-minute teardowns when there is a slot.