App Store Impressions vs Conversion Rate: Four Diagnoses

Every few weeks someone sends me a screenshot. Two numbers from App Store Connect: impressions fell, conversion rate climbed. Is this good or bad?

From those two numbers alone, nobody can tell. They measure different things. The dashboard just puts them side by side, so they look like two readings of the same dial.

The short version

Impressions tell you how many people saw you. Conversion rate tells you how many of those people installed.

The first gate sets how many chances you get. The second sets how much of that chance you waste. "App Store impressions are up" and "conversion is up" are not two versions of the same good news, even though they look identical in a report.

What helps is crossing them into four cells: four diagnoses, each pointing at a different layer to fix. Watch the two numbers separately and you will keep changing the wrong thing.

Impressions are two numbers wearing one label

Impressions in App Store Connect is a single total. Two very different things are mixed inside it.

Search impressions: someone typed a query, your icon appeared. Tied to the words in your metadata, and you can compete for it.

Browse impressions: charts, editorial features, "you might also like", placements on other apps' pages. The store hands these out. Short of pitching for a feature, you do not control them.

Average them together and the conversion rate lies to you. One feature placement and browse impressions jump; browse converts worse by nature, because a searcher arrives with intent and a browser is walking past. Total conversion drops. You think the page broke; the denominator changed shape.

So split first. Analytics breaks impressions into search and browse; Play Console splits exploration sources even further. If yours cannot be split, mark the days a chart run or feature was live and read them separately.

Conversion rate does not divide by impressions

Most people get this wrong, ASO veterans included.

Apple's conversion rate divides by product page views, not impressions. There is a step in between: impression, then tap into the page, then install.

High impressions with few page views means your icon, title and rating lost the fight on the results screen. Fix how you look in a list: icon legibility, whether the title carries the query, what the subtitle adds, the star rating, a top review snippet.

Many page views with few installs means people arrived and were not convinced. Fix the page: whether the first two screenshots answer "what is this" and "why trust you", the opening seconds of the preview video, the first two description lines, how subscriptions are explained.

One conversion number covers two different gates. Fix one and the other keeps leaking.

Four combinations, four diagnoses

Split each metric high or low and you get four cells. Measured against your own eight-week median, not an industry average — reason further down.

**High impressions, high conversion.** Nothing is wrong. Words right, page holds up. The only move is to expand: copy the playbook into another market, or push deeper into longer tails of the same cluster. Do not redesign a page that is winning.

**High impressions, low conversion.** The most common misdiagnosis. The reflex is to swap screenshots, but the order is wrong — you do not yet know which gate is failing.

**Low impressions, high conversion.** The page works and nobody sees it. Of the four, this is the one worth spending on, because it already proved it converts. The fix sits at the door: wider keyword coverage, chart placement, paid coverage.

**Low impressions, low conversion.** Usually not an ASO problem. Stop editing metadata and go look at positioning. You may be sitting under a phrase nobody searches, or in a category with demand you cannot serve. Optimising here is repainting the sign on a shop in the middle of nowhere.

High impressions, low conversion: do not touch the screenshots yet

This is the combination that burns the most budget. Three checks, in order.

**One: split the source.** Search and browse separately, each with its own conversion rate. If the weakness sits in the browse column, it is not a page problem but a denominator one — a single feature placement diluted your average.

**Two: read the keyword report.** It shows which queries produced those impressions. If the top ones have little to do with your app, the page is not the problem — the words are. You are feeding yourself low-quality impressions from irrelevant searches, and that traffic fails to convert and muddies the store's own read on whether you belong under those queries.

**Three: check whether Apple Search Ads is running.** Paid installs count toward product page conversion. With a campaign live, the organic conversion rate you are reading is inflated by it. To judge the page, subtract the paid portion and recalculate.

Only when all three come back clean — search share normal, words relevant, no campaign — do the screenshots become the suspect. Then change the first two, not the set. One answers "what is this", the other "why trust you". Give it a week before looking again.

Low impressions, high conversion: the page is fine, the door is small

Start with coverage. How many phrases with real volume does your keyword field carry? Store search matches at word level, so a phrase you never wrote is a context you are absent from. This is not stuffing; it is covering the ways people name one thing — 安卓商店, "Android store", "Google Play" are three audiences heading for one destination. Carry only one and the other two searches never reach you.

Then paid coverage. Apple Search Ads holds brand and category terms so you show up first. It is the most direct way to buy impressions, but whether it pays depends on whether you can catch the traffic, not on cost per install. A converting page plus ads is acceleration; a weak page plus ads is acceleration into a leak.

App Store and Google Play numbers are not comparable

Both dashboards say impressions. Both say conversion rate. They do not mean the same thing.

Apple divides by product page views. Google Play Console reports store listing visitors and acquisitions, with a different calculation and a different attribution window, and Play splits exploration sources more finely. So "App Store converts at one rate, play store conversion rate at another, Android users must convert better" is not a comparison — it is two accounting systems.

Compare a platform against its own past, or two page versions on the same platform. Across platforms, compare direction only, never absolute values. Check the current official guidance before you ship a decision built on these numbers; both companies change their reporting more often than you would like.

Someone else's benchmark is not your benchmark

"Average app store conversion rate" and "what counts as good" get searched every day. The answers are close to useless.

Too many variables move it: category, market, free versus paid, ads running or not, on a chart or not, whether the brand earns direct searches. Change any two and the baseline shifts widely. A borrowed number gives you anxiety and nothing else.

The only benchmark that means anything is your own: open Analytics, take eight weeks, read the weekly curves, take the median. Judge everything against that line. Do not read daily conversion. Read weekly.

When this matrix lies

Four situations where the four cells will mislead you.

**The first weeks after launch.** No baseline, and the store hands new apps exploratory impressions that behave nothing like later traffic. The quadrant has no value yet.

**During a big release or event.** A redesign, a price change, a feature, press coverage — any scrambles both numbers at once, and you cannot tell whether your change worked or the event did.

**During a promotion or bought installs.** Limited-time free, in-app discounts and incentivised installs all pollute conversion — incentivised traffic especially: those users came for the reward. Conversion looks great, retention is a wreck.

**Tiny-volume markets.** At a few hundred views, daily conversion is noise. Read weekly.

Underneath it all: if retention is bad, every optimisation above is just accelerating the leak. Conversion can be bought. Retention cannot.

What to do next

1. In App Store Connect, split impressions into search and browse, calculate conversion for each, compare against your own four-week baseline. This alone kills half the misdiagnoses. 2. Pull the keyword report for thirty days, walk the top twenty queries against what your app actually does, drop the ones that clearly do not belong. Irrelevant impressions are not an asset. 3. If Apple Search Ads is live, subtract paid installs from the organic figure and recalculate the page's conversion before touching any creative. 4. Put your current state into one of the four cells and fix only the layer it points at. Change one thing at a time, or next month you will be guessing again.