Neither Apple nor Google publishes the actual formula behind app store rankings, and that hasn't changed. What's changed is how confidently both platforms now look past the metadata itself, at what happens after someone installs. A strategy still built entirely around keyword placement is optimizing for half the problem.
Getting found isn't the same problem as getting kept
Keywords still control which searches an app is even eligible to appear for — that part of the mechanic hasn't moved. What's shifted is everything downstream of the tap. Install velocity, day-1 and day-30 retention, uninstall rate and crash-free sessions all now factor into whether an app that gets found also gets ranked well over time. An app that wins the click but loses the user fast tends to lose the ranking it just earned, which makes retention less of a product metric living somewhere else and more of a direct input into the same strategy that governs keywords and creative.
Reviews get read for depth, not just averaged for stars
The signal inside ratings has gotten more specific too. A high star average built on short, generic reviews carries less weight than a slightly lower average built on detailed, feature-specific feedback — both platforms appear to treat review depth as a proxy for genuine, sustained usage rather than a one-time reaction. Actively responding to reviews, particularly critical ones, reads as a maintenance signal in its own right, separate from the rating itself.
The two stores don't work the same way underneath
Google Play indexes an app's full description as searchable text, similarly to how a search engine reads a webpage — up to around 4,000 characters, worth writing with real keyword coverage in mind. The App Store doesn't index the description at all; it only reads the title, subtitle and dedicated keyword field, which means the same block of text is doing two completely different jobs on the two platforms. Metadata written once and copied across both stores is usually leaving one of those two mechanics unused.
What this means for an actual strategy
The practical shift is treating ASO as a continuous testing cycle across two connected tracks rather than a metadata task completed once at launch. One track is visibility — keyword research and placement, revisited after major algorithm shifts or competitor moves. The other is conversion — screenshots, preview video and Custom Product Pages tested against specific keyword segments, since a page that converts well for one search intent often underperforms for a different one landing on the same listing. Both tracks feed the same retention numbers that increasingly decide whether early ranking gains hold or quietly erode.
There's no fixed formula to plug numbers into here, on either platform, which is exactly why this stays a strategy rather than a checklist. What holds up across both stores is the same discipline: test one variable at a time, give changes real time to show in the data, and treat what happens after the install as seriously as what happens before it.