FunnelHound / Guides / Update cadence
How often should you update your app? Cadence, ratings and the freshness myth
Let's retire the myth first: there is no freshness bonus. Shipping weekly to look alive doesn't rank you higher. Updates move your funnel through side effects (fixed crashes, fresh ratings, re-indexed metadata), which means the right cadence is "when you have one of those to ship", not a calendar superstition.
What an update actually does to your store presence
- Crash fixes repair retention signals. Crashes drive deletions and one-star ratings; both feed the behavioral tier of ranking. The most powerful "ASO update" is often just a stability release.
- A release invites fresh ratings. Recent ratings weigh on the visible average, and a good update is the natural moment users say so (ask at the win).
- Metadata changes re-index. New subtitle or keywords enter the 2–4 week settling window: the real reason an update can move rankings, and it's the metadata, not the update.
- Release notes talk to humans. Not indexed, but read by upgraders and by product-page visitors judging whether the app is alive. "Bug fixes and improvements" forever reads as a shrug.
Metadata-only vs binary: know the cheap path
Not every change needs Xcode. Subtitle, keyword field, screenshots and promo text ship with a version submission, but a metadata-only update carries no new binary and clears review fast; the app name edits under App Information. Practical consequence: batch your metadata experiments deliberately: each one starts the indexing clock, so dribbling changes weekly keeps you permanently inside turbulence you can't attribute (the rename lesson, generalized).
The ratings-reset decision
Apple lets you reset your ratings summary with a version release. Default answer: don't. The average and count are the social proof your tap-through runs on; wiping a 4.6 with thousands of ratings to chase a clean slate trades an asset for a cold start. The one honest exception: an average so wrecked (deep threes or below) that realistic rating velocity can't dig it out, and the underlying complaints are genuinely fixed. Reset with the fix, never instead of it: a reset on a broken app just rebuilds the bad average with fresher dates.
A sane indie cadence
Ship features when they're ready. Ship fixes fast. Batch metadata deliberately. In practice that lands many indies near monthly-ish binaries with metadata experiments on their own six-week rhythm, but the schedule is an output of having something to ship, not a target. Use phased release for binaries (staged rollout catches a bad build before it meets everyone), and date every release as a milestone: an update is three experiments in a trench coat (stability, ratings, metadata), and the funnel afterward is how you learn which coat it was wearing.
Every release is an experiment. Treat it like one.
FunnelHound marks each release as a milestone and shows installs, deletions and purchases since, from App Store Connect, on your iPhone.
Get FunnelHoundData notes: metadata edit paths, ratings-reset option and phased release mechanics per Apple's App Store Connect documentation as of 2026. The no-freshness-bonus position reflects the absence of credible replicated evidence, not an Apple statement. The folklore file stays open.