App Store Optimization is the practice of improving how your app ranks and converts inside the App Store and Google Play. It is the only acquisition channel where cost per install falls over time, because you are earning visibility rather than renting it.
This is the working guide: what the stores actually index, which levers move rankings, what good looks like in 2026, and how long changes take to register. Benchmarks are cited. The operating notes come from running ASO programmes for app companies.
How Do App Stores Decide What to Rank?
Apple and Google index different things, which is why the same listing copy cannot serve both.
| Element | Apple App Store | Google Play |
|---|---|---|
| App name / title | Indexed, heavily weighted | Indexed, heavily weighted |
| Subtitle / short description | Indexed | Indexed |
| Keyword field | Indexed (100 characters, hidden) | Does not exist |
| Long description | Not indexed | Indexed, keyword density matters |
| Ratings & reviews | Ranking factor | Ranking factor |
| Retention & engagement | Ranking factor | Ranking factor |
The practical consequence: on iOS you optimise a 100-character hidden keyword field and never repeat words across fields. On Play you write a description that reads naturally while covering your keyword set. Teams that copy-paste between stores lose ranking on both.
What Does Good ASO Look Like in 2026?
| Metric | Weak | Median | Strong |
|---|---|---|---|
| Store listing conversion rate | under 20% | 25-30% | 35%+ |
| Keywords ranked top 10 | under 15 | 40-80 | 150+ |
| Day-1 retention | under 20% | 25-30% | 40%+ |
| Rating | under 4.0 | 4.2-4.5 | 4.6+ |
For the paid side of store visibility, Apple Ads cost per tap runs from $0.24 (Religion) to $14.41 (Sports), with a global median of $0.92 and a top-15-category average of $2.25. Sources: AppTweak, Adapty, Business of Apps. Strong ASO lowers that cost, because better listing conversion improves ad relevance.
The Four Levers, In Order of Impact
1. Creative (screenshots, icon, video)
Conversion is where most of the money is, and screenshots do most of the converting. The first two screenshots are visible without scrolling; treat them as the entire pitch. Test one variable at a time, because changing icon, screenshots and subtitle together teaches you nothing about which one worked.
2. Keyword footprint
Metadata determines which searches you appear in at all. Build the target set from real search volume and competitor gaps rather than intuition, and score candidates against your current authority. Ranking 4th for a 200-volume term beats ranking 40th for a 20,000-volume one.
3. Ratings and reviews
Both a ranking factor and a conversion factor. Prompt for reviews after a success moment in the product, never on launch. Responding to negative reviews measurably lifts ratings over time because some users revise them.
4. Retention
The lever most teams forget is ASO at all. Both stores weight engagement, so an app that keeps users climbs, and one that leaks users sinks regardless of metadata. This is why buying junk installs actively damages organic ranking.
How Long Does ASO Take to Work?
- Metadata changes: 2-6 weeks to index and settle. Reverting after five days is the most common way teams destroy a winning change.
- Creative tests: statistically readable in 1-2 weeks at reasonable traffic, longer for low-volume apps.
- Rating improvements: months, because you are moving an average.
- Retention work: a full 30-day cohort before day-30 numbers mean anything.
Anyone promising ranking movement in seven days is describing normal fluctuation or buying installs.
Mistakes That Cost the Most
- Repeating keywords across iOS fields. Apple indexes title, subtitle and keyword field as one set. Duplication wastes characters you cannot spare.
- Optimising metadata, ignoring creative. Metadata earns impressions; creative converts them. Most teams over-invest in the first.
- Chasing head terms too early. At low authority you will not out-rank incumbents on generic category terms. Win the long tail first, then climb.
- Treating ASO as a launch task. It is a loop, not a checklist. Store algorithms change and competitors move.
How Do You Build an ASO Keyword Set?
Start with 20 to 30 realistic terms rather than a list of 200 aspirational ones, and score each against your current authority rather than its raw volume. Ranking fourth for a 200-volume term delivers more installs than ranking fortieth for a 20,000-volume term, and it is achievable this quarter rather than theoretically. Build the set from three sources: terms you already rank for on page two, terms your closest competitors rank for that you do not, and long-tail phrases that describe what your app actually does in the words users would use.
The mistake that wastes the most effort is targeting category head terms early. At low store authority you will not out-rank incumbents on a generic term, and the attempt consumes the metadata characters that could have won you five long-tail rankings.
What Makes Screenshots Convert?
Creative does the converting, and the first two screenshots carry almost all of it because they are visible without scrolling. Treat them as the entire pitch rather than the opening of a sequence.
| Element | What works | Common mistake |
|---|---|---|
| Screenshot 1 | Single clear benefit, large legible text | A raw UI shot with no context |
| Screenshot 2 | The second-strongest reason to install | A feature tour nobody scrolls to |
| Caption text | Readable on a phone at thumbnail size | Copy sized for a desktop mock |
| Video preview | Value shown in the first 2 seconds | A logo animation intro |
How Does Retention Affect ASO?
Both Apple and Google weight engagement and retention when ranking apps, which makes retention an ASO lever rather than a purely product concern. An app that acquires users and loses them sends a negative signal that metadata cannot compensate for. This is the mechanism that makes bought or incentivised installs actively harmful: they inflate the install count your board sees while degrading the exact signal that drives free organic visibility. If day-1 retention sits below 20%, fixing that will move rankings more than any keyword change.
Frequently Asked Questions
Is ASO better than paid user acquisition? They compound together. ASO improves listing conversion, which lowers the effective cost of every paid install. Running paid against an unoptimised listing means overpaying for every tap.
How many keywords should I target? Start with 20-30 realistic terms rather than a list of 200 aspirational ones. Depth of ranking beats breadth of targeting.
Does ASO work for a brand-new app? Yes, and it should be done before launch, because metadata takes weeks to index. Pre-launch ASO is the cheapest growth work available.
Can I do ASO myself? The mechanics, yes. What agencies add is keyword data at scale, testing discipline, and connecting store performance to paid and retention.
Where to Take This Next
For the hands-on version of everything above, see our ASO services. If paid and store visibility need to move together, that is the app marketing agency engagement, and the paid half specifically is mobile user acquisition. For tooling, we compare the options in ASO tools compared.