App Store Optimization for New Apps
How to choose keywords, design screenshots, and structure a listing when you have no ratings and no ranking history.
Optimising a listing with no history
Most app store optimisation advice assumes an app with ratings, install history, and enough traffic to run experiments. A new app has none of these, which changes both what is achievable and where the effort should go.
The main consequence is that you cannot compete for broad category terms. Those are held by apps with years of accumulated ratings and install velocity, and no amount of keyword selection overcomes that. Recognising this early redirects effort toward the terms you can actually win.
Specific terms, then history, then breadth
Narrow phrases describing exactly what the app does are winnable, convert better because the intent is precise, and — importantly — generate the ranking history that makes larger terms contestable later.
This is a sequencing insight rather than a permanent constraint. Apps that win narrow terms first accumulate the signals needed to compete for broader ones; apps that target broad terms immediately accumulate nothing.
Two screenshots carry the listing
Most users never scroll past the second screenshot. Those two are the advertisement, and presenting raw interface without context is the most common wasted opportunity in app listings.
Captions explaining what the user gets outperform bare screens consistently, because a screenshot of an interface communicates almost nothing to someone who has never used the app.
Write for the collapsed description
Store descriptions are truncated after roughly a line, with a tap required to expand that most people never make. That first sentence carries the entire weight of the description and should state plainly what the app does — not set up a narrative that pays off in paragraph three.
The icon is displayed smaller than you designed it
It appears in search results, on home screens, and in category grids at a size far smaller than the canvas it was made on, always surrounded by competitors. Legibility and distinctness at actual display size matter more than detail.
Testing it in a grid alongside competing icons reveals problems that no amount of examining it alone will.
Localise the listing first
Translating just the store listing is inexpensive and frequently produces installs from markets you never targeted. It is a considerably better first internationalisation step than translating the entire product on the assumption that demand exists.
Then change one thing at a time. Store experiments need traffic and patience, and simultaneous changes teach you nothing.
The playbook
Target terms you can plausibly rank for
A new app with no ratings will not rank for a broad category term. Specific, lower-volume phrases describing exactly what the app does are winnable, convert better, and are how you accumulate the ranking history needed to compete for anything larger.
Design the first two screenshots as an advertisement
Most users never scroll past the second screenshot. Those two need captions explaining the benefit rather than showing bare interface. Raw screenshots without context are the most common wasted opportunity in app listings.
Write the first line for the collapsed view
Store descriptions are truncated after roughly one line before a tap to expand, and most people never tap. That first sentence carries the description's entire weight and should state what the app does, plainly.
Treat the icon as the highest-leverage asset
The icon appears everywhere, at small sizes, alongside competitors. It must be legible and distinct at the size it is actually displayed, which is far smaller than the canvas it was designed on. Test it in a grid against competing icons.
Localise the listing before localising the app
Translating the store listing is cheap and frequently produces installs from markets you were not targeting. It is a far better first internationalisation step than translating the entire product on speculation.
Change one element at a time
Store experiments need meaningful traffic and time to produce a trustworthy result. Changing the icon, screenshots, and keywords simultaneously means learning nothing from a change that may have helped in one place and hurt in another.