Why Isn't My App Getting Downloads?

The question almost always arrives phrased as a quality problem. The app works, the onboarding is clean, friends said it was good, and the download counter has not moved in three weeks. So the founder starts rewriting the description and swapping the icon.

That is usually the wrong end of the funnel. Before touching the listing, find out whether anyone is seeing it at all, because the fix for zero impressions and the fix for poor conversion have almost nothing in common.

Why isn’t my app getting downloads?

Your app is not getting downloads because it is not getting impressions, and impressions in an app store are won rather than granted.

The concentration is severe. Sensor Tower’s analysis of publisher concentration found that the top 1% of publishers captured 81% of all downloads in 2025, along with 92% of in-app purchase revenue. That is not a market where a good app surfaces on merit. It is a market where a small set of incumbents occupy the positions that produce installs, and everyone else competes for what is left.

The market is also not growing its way out of the problem. Downloads across the market grew 0.8% year over year while in-app purchase revenue grew 10.6% to $167 billion, per Sensor Tower’s State of Mobile 2026. The average user spent 3.6 hours a day in apps in 2025, up 1.1%. People are spending more inside apps they already have rather than installing new ones. There is no rising tide.

So the practical diagnosis starts in App Store Connect or Play Console, looking at impressions rather than installs. If a new app is showing a few hundred impressions a week, no amount of screenshot work will produce a meaningful install number, because you are optimising the conversion rate on a volume that rounds to nothing. If it is showing tens of thousands and converting poorly, the listing genuinely is the problem and it is a fixable one.

Founders skip this step constantly, and it is why so much ASO effort produces no result.

Where do app downloads actually come from?

App downloads come overwhelmingly from store search: 65% of iOS discovery starts with someone typing a query, against 18% from browse, 12% from referrers and 5% from App Store ads, according to Searchlab’s 2026 ASO statistics. Google Play is similar at 58% from search.

That statistic gets quoted as an argument for keyword optimisation, and it is half of one. The other half is what a search actually is. A search is a person who already knows they want something in your category, typing a term, and choosing from a ranked list. Ranking on that list is decided by download velocity and engagement history you do not yet have.

Which produces the trap most new apps fall into. Search is where installs happen, search rewards apps that already get installs, and a new app has neither. Optimising keywords harder does not break the loop, because the constraint is not relevance. It is that eight other apps have years of velocity on the same term.

The way out is to create the search rather than compete for it. Someone who watched a creator explain a problem your app solves goes to the store and types your brand name, and a branded search is a query with one credible result on it. That demand is generated outside the store and captured inside it, which is why store analytics chronically undercount paid social: the ad gets the credit for the tracked install, and the branded search that the same ad produced two days later looks organic.

Read your channels with that lag in mind before concluding a campaign failed.

Is it my listing or is it my traffic?

Your listing is only the problem if your page-view-to-install rate sits below your category’s band, and comparing those two numbers settles the listing-or-traffic question in about ten minutes.

The store-wide averages are roughly 25% on iOS and 27.3% on Google Play, but the category spread is wide enough to make the average useless. Screenfast’s 2026 category benchmarks put iOS Utilities apps at 33 to 38% and iOS Finance apps at 18 to 22%, roughly a twofold difference driven by how much consideration the category demands rather than by listing quality. A finance app converting at 20% is performing normally. A utility converting at 20% has a real problem.

Category, iOS	Page-view-to-install band	What a 20% rate means
Finance	18-22%	Normal, look upstream
Store-wide average	around 25%	Slightly below average
Utilities	33-38%	A real listing problem

If you sit inside your category band, your listing is fine and your problem is upstream. Stop editing screenshots. Every hour spent there is an hour not spent on the actual constraint, which is that too few people reach the page.

If you sit well below the band, the listing is worth work, and the part that pays back fastest is visual. Roughly 60% of users never scroll past the first two screenshots, per Digital Applied’s 2026 ASO statistics compilation, which means your first two frames carry nearly the whole decision and everything after frame two is decoration for the minority who kept swiping.

You will find published uplift percentages for designed screenshots and preview videos. We are not quoting them, because the ones in circulation disagree with each other by a factor of two and every one of them is published by a company selling screenshot design. Test your own first two frames against your current ones and use that number instead. The store gives you the tooling for it.

Does running paid social ads fix a download problem?

Paid social fixes a demand problem, which is the download problem most new apps actually have. It does not fix a retention problem, and it does not fix a product nobody wants.

Paid social works here because it reaches people before they know to search. App Store ads account for 5% of iOS discovery per Searchlab’s 2026 data, so buying visibility inside the store reaches an audience already looking for something like you and is a small channel besides. Meta and TikTok reach the far larger group who have the problem but have never typed a query about it.

The cost is real and worth modelling before you commit. iOS cost per install averaged $5.84 in Q1 2026 against $1.92 on Android, a 3.0x gap, with iOS up 19% year over year and Android up 8%. Two things drive it: heavier bidder competition on iOS and the measurement uncertainty left by App Tracking Transparency. The full breakdown of what drives app acquisition cost matters more than the headline gap, because the gap has causes you can partly control.

Two conditions decide whether this works. First, retention has to be adequate before you scale spend, because paid installs on an app that loses users in week one converts budget into churn at a predictable rate. Second, you need creative volume rather than one good ad, since install campaigns fatigue quickly and a single winning asset carries an account for weeks rather than quarters.

Founders who treat paid social as a switch to flip get an expensive month. Treated as a testing system with weekly output, it becomes the demand engine that store search then harvests.

What should you fix first?

Fix impressions first, then retention, then the listing, then demand generation, and resist working on whichever piece is most fun.

Start with impressions, because everything downstream is a rate applied to that number. Pull 30 days of store impressions. Under a few thousand a week, your entire problem is demand, and listing work is premature.

Second, check retention before spending anything. Day one and day seven retention against your category tells you whether buying users is investment or waste. This ordering is unpopular because it can mean pausing acquisition, and it is still correct: an app that cannot hold users converts every marketing dollar into a temporary chart position.

Third, fix the listing only if your conversion rate sits below your category band. Screenshots one and two first, then the title and subtitle keywords, then everything else. Developers who refresh keywords every four to six weeks outperform those who optimise once and stop, per Digital Applied’s 2026 compilation.

Fourth, generate demand outside the store. That is paid social, creator content, and anything else that puts your product in front of people who were not searching. Expect part of the return to arrive as branded search rather than as attributed installs.

The sequence matters because the steps have wildly different costs. Reading impressions is free. Checking retention is free. Rebuilding screenshots is cheap. Buying installs is not, and it is the only step that gets more expensive when the three before it were skipped.

So where should you actually start?

Start by pulling 30 days of store impressions, because an app with no downloads almost never has a listing problem. It has an impressions problem, in a market where the top 1% of publishers took 81% of installs in 2025 and total downloads grew 0.8% year over year, both per Sensor Tower. Check impressions, check day one and day seven retention, then fix the listing only if your page-view-to-install rate sits below your category’s band. After that, the job is creating demand outside the store so that store search has something to capture.

The practice worth abandoning is the ASO-first launch, and it is close to the default. Rewriting the description and swapping the icon feels like progress because it is the only lever a founder controls alone, and on an app showing a few hundred weekly impressions it changes nothing, because a conversion rate applied to almost no traffic is still almost no installs. Searchlab found that 65% of iOS discovery comes from store search, and that gets read as a reason to optimise for search. It is better read as a warning: search is a ranked list, ranking is decided by download velocity you do not have, and keyword work does not manufacture velocity.

Where this stops applying: if you are showing tens of thousands of weekly impressions and converting below your category band, the listing genuinely is the constraint and the order above inverts. That is a real situation, it is just much rarer than the number of founders working on screenshots would suggest.