# The seven-day cliff: a new app's ranks after launch

Sam · Published November 4, 2026 · Canonical: https://rankaso.app/blog/seven-day-cliff-app-store/


A new app, and to a smaller degree a new version, gets a short boost in App Store search: ranks arrive higher than the listing has earned, hold for about a week, then fall back to where the words and the installs actually put them. Developers who read their day-three ranks as the truth are shocked on day ten; developers who expected the cliff use the week to find out which terms the listing can hold, and change the listing before the fall. Here is the shape, and how to track it so the week is not wasted.

![The Keywords screen in RankASO with movement over the days after a release](https://rankaso.app/assets/device-01.png)

## The shape, day by day

Days one to three: the app appears for many of its terms at once, often far higher than a new listing should. Days four to seven: the positions hold or drift slightly. Days eight to fourteen: the terms where the listing is genuinely strong stay; the rest slide, some out of the top 200 entirely. By the end of the second week the ranks are what the name, subtitle and keyword field earn on their own, plus whatever installs the launch brought.

A new version shows a softer version of the same curve on the terms its metadata touched. Our own app [Escape](https://escapethegrip.com/) shows it with each release: terms that moved in the first days after a version and then settled over the next ten.

## Why it happens

The store gives new and updated listings a chance to prove themselves. Impressions arrive before the listing has a history, and ranks that do not turn into taps and installs are withdrawn. The cliff is the store's verdict on the page, not a penalty. A page that converts keeps more of its launch ranks than a page that does not.

## Make the week count

1. **Track from day zero.** Every term from the name, subtitle and field, plus the store's completions for your category word, in your two main storefronts. The daily check is what makes the cliff visible rather than remembered. RankASO's free plan covers thirty terms in two storefronts.
2. **Read day three, do not act on it.** Note which terms you appeared for. This is the list of terms the store considers you relevant to; it is wider than the one you wrote.
3. **Read day ten.** The terms that held are the listing's real strengths. The terms that fell are the ones where the top 10 outranks you on ratings and name words. The field report shows which words carried the holders.
4. **Change the listing with the version after launch.** Double down on the words that held, drop the words that fell, and add the completions you appeared for by accident. Baseline, ship, review at two weeks.

## What not to do

Do not rewrite the listing on day four because a term dropped two places. Do not read the day-three ranks on your own phone in the App Store app; it personalises, and the boost makes the illusion worse. Do not buy installs to prop up the launch ranks; the store withdraws what the page cannot hold, and the cliff just arrives later.

## After the cliff

The ranks you have on day fourteen are your baseline. From there, ASO is the weekly loop: one listing change per version, a baseline, a review, and a rival list so you know whether a drop was yours or theirs. The launch week told you where the listing is strong; the loop makes it stronger one word at a time.

RankASO on the App Store: https://apps.apple.com/app/id6830350753?ct=blog-en-30&mt=8 (Free to start. No account. Data Not Collected.)

## Questions

### Why did my app's ranks drop a week after launch?

New listings get a short boost while the store tests them. After about a week the ranks settle to what the name, subtitle, keyword field and installs earn on their own.

### How do I keep my launch ranks?

You keep the ones the page deserves: terms where the listing's words and your conversion hold up against the top 10. Track daily from day zero, and use the first version after launch to lean into the terms that held.

### Does a new version also get a boost?

A smaller one, mostly on the terms whose words you changed. Track it the same way and judge the version at two weeks, not in the first days.