Skip to main content

Benchmark reality check

App Store Conversion Rate by Category

A good App Store conversion rate is your category's median in your own App Store Connect Benchmarks tab, not a number off a chart. Be skeptical of the charts: almost every "2026 benchmark" on the web is recycled AppTweak H1 2024 data, and the sites copying it do not even agree on the numbers.

TL;DR

  • 1.The only current, trustworthy number is your own category percentile in App Store Connect, Analytics, Benchmarks. Aim at or above your category median (p50).
  • 2.Almost every public "2026 benchmark" is AppTweak's H1 2024 data relabeled. It is roughly two years old.
  • 3.The recyclers contradict each other: the same H1 2024 source shows Music at 30.1% on one site and 47.1% on another.
  • 4.The freshest primary number is AppTweak's 2025 overall: iOS 8.56%, Google Play 16.15% (first-time installs). No current by-category table is published anywhere.

The only number that is current and yours

Apple shows every developer their peer group's conversion distribution privately, in App Store Connect under Analytics, Benchmarks. You see where you sit against the 25th, 50th, and 75th percentile of apps like yours, updated on Apple's own data. "Good" means at or above your category's 50th percentile. That number is live and specific to you. Everything below is older public context, useful only for rough orientation.

Why the "2026 benchmark" tables are not 2026

Search for App Store conversion benchmarks and you get a dozen confident tables. We read them. Nearly all trace to one dataset: AppTweak's first-half-2024 benchmark. One "2026" page states its numbers are "per AppTweak's H1 2024 baseline." Another, published in March 2026, is "based on AppTweak's H1 2024 US market data." A third 2026 post uses the same figures. None ran fresh 2025 or 2026 data.

Worse, they do not copy it the same way. The Music category shows up as 30.1% on one site and 47.1% on another, both citing the same H1 2024 source. When the copies disagree, the table is not a benchmark, it is folklore. Treat any public number as roughly two-year-old, directional context, and weight your own App Store Connect figure above all of it.

The numbers you can actually date

Overall conversion rates that come straight from a primary vendor page, each stamped with its period and how it is measured. Watch the top two rows: the same vendor, same year, swings 3x purely on whether redownloads count.

SourcePeriodScopeRate
AppTweak2025
iOS, US
First-time installs / page views (redownloads excluded)
8.56%
AppTweak2025
Google Play, US
First-time installs / page views (redownloads excluded)
16.15%
AppTweakH1 2024
iOS, US
Installs / impressions, redownloads included (the widely-quoted figure)
~25%
AppTweakH1 2024
Google Play, US
Installs / impressions, redownloads included
~27.3%
SplitMetrics2015
iOS + Android, all apps
Downloads / page views, 300+ experiments
26.4%

Google Play's higher listing rate does not mean higher value, iOS typically monetizes more per install. And there is no published 2025 or 2026 by-category iOS table from any primary source, which is exactly why the next section is from 2015.

By category: the one primary table is from 2015

SplitMetrics, 2015, page views to downloads. This is the most-cited self-consistent by-category dataset that exists, and it is a decade old, which tells you how rarely this data gets refreshed. Use it for the relative ordering (games convert low, music and travel high), not as a current rate.

CategoryMedian page-view to download
Music32.46%
Travel21.91%
Health & Fitness18.52%
Social Networking11.36%
Entertainment8.6%
Education6.75%
Games4.47%

Why the public numbers never line up

  • Denominator. Impressions to install, page view to install, and search-direct install are three different rates. AppTweak counts redownloads, SplitMetrics does not, which alone moves the iOS number from 25% to 8.56%.
  • Recency. A 2015 table and a 2024 table are not comparable, and most "2026" tables are the 2024 one in disguise.
  • Transcription. Even the copies of the 2024 data disagree by category, so a single decimal is false precision.
  • Funnel stage. Listing conversion and subscription trial-to-paid are different events. Mixing them produces nonsense.

So treat every public table as directional. The number that decides whether your screenshots are pulling their weight is your own category percentile in App Store Connect.

Below your category median? Start with the first three frames.

When your conversion rate trails your peer group, the cheapest lever is almost always your first one to three App Store screenshots, the ones that show in search results. You will not know which version converts until you see a few options side by side, so describe your app, watch the options stream in, and refine from there.

Try the screenshot builder

Related: what counts as a good conversion rate and the ASO screenshot tool.