App Store Optimization: The Complete Guide for Indie Devs (2026)

If you're an indie dev or a small studio shipping mobile apps, you've probably heard the same advice a hundred times: "do your ASO." The advice is correct. The problem is that almost every guide written about it was written for someone managing 50 apps with a $1,000/month tool budget — not for you, shipping one or two apps with the cheapest plan you can find.

This guide is what we wish someone had written us when we started: the things that actually move the needle on App Store and Google Play in 2026, in the order you should care about them.

What App Store Optimization is (and isn't)

ASO is the discipline of getting your app discovered organically inside the App Store and Google Play. That's it. It's not paid acquisition, it's not retention, it's not viral growth — though all of those interact with ASO in ways we'll get into.

The mechanism works in two halves:

1. Indexation. The store decides what queries your app is allowed to show up for. This is mostly driven by the words you put in your metadata (title, subtitle, keywords field on iOS; title, short description, long description on Play) and the country/language you put them in.
2. Ranking. Among all the apps that could show up for a given query, the store decides what order to show them in. This is driven by metadata relevance, install velocity, retention, rating, and a handful of opaque signals nobody outside Cupertino fully knows.

Most ASO advice is about indexation. Most ranking moves come from the second bucket. Both matter.

The order of operations that actually works

If you only do four things in your first quarter of ASO, do these — in this order:

1. Pick the right keywords for the right country. Not the keywords your competitors use. Not the literal translation of your English title. The keywords real users in that country actually type into the search bar.
2. Make your metadata match your app's real intent. A meditation app whose title says "wellness studio" will rank for the wrong queries and convert poorly when it does rank. The metadata has to describe what the app actually is, in the words users use to describe what they want.
3. Stop wasting characters. iOS gives you 30 + 30 + 100 chars of pure indexation surface. If your title and keywords field share three words, you've burned ~15 of those characters. Most apps do this without realising.
4. Iterate. Your first metadata is wrong. So is your second. Treat ASO as a versioned experiment, not a one-time setup.

Notice what's not in the list: chasing review counts, A/B testing screenshots, paying for backlinks, gaming category placement. Those things matter eventually, but they will not save bad metadata.

How keyword research should actually work

The standard advice — "look at what your competitors rank for" — is a trap when you're new. Established competitors rank for head terms because they have install velocity, not because the keyword is winnable for you. Copying their metadata gets you indexed for the same terms with none of the signal that lets them win.

What works for an indie dev is the opposite: find queries where established apps don't dominate, where there's real demand, and where your app genuinely fits. The signals you want:

• Real autocomplete. If Apple's autocomplete suggests a term in a country, real users in that country are typing it. If the term doesn't appear in autocomplete, you're probably guessing about demand.
• Long-tail intent. Three- and four-word queries have less competition and higher conversion. "constellation map poster" beats "wall art".
• Local language. A Brazilian user types "quadro de mapa estelar", not "star map poster". Translation tools won't get you there.
• Sub-category alignment. Apple and Google do not weight your declared category as heavily as people think, but they do penalise mismatched intent — describing a productivity tool with entertainment-app language costs you ranking authority over time.

What "demand" really means (and why most tools lie about it)

Most ASO tools show you a "demand" or "popularity" score per keyword. If you click into the small print, the score is almost always inferred — usually from the review counts of the top 10 apps that already rank for the term. That's a proxy for "how much competition this keyword has", not "how many people search for it".

The honest signals for demand without paying SensorTower:

• Autocomplete presence. Apple's autocomplete only surfaces terms users actually type. If "yoga sleep" is suggested in the US store, people search it. If not, it's probably noise.
• Search results count. A query that returns 1,200 apps has more demand than one that returns 23. Not a perfect signal — quality varies — but a useful triangulation.
• Cross-store overlap. A term that surfaces in both Apple and Google's autocomplete in the same country is almost always a real query.

We built ASO Pilot to lean on those three signals because they're the ones we trust. Anyone who tells you they have "real search volume data" without an SDK partnership is, at best, modelling.

Metadata mistakes that cost the most

In rough order of how much they hurt you:

›1. Title and subtitle that share words

iOS treats title + subtitle + keywords field as one combinatorial space. If your title is "Posterly — Star Map Posters" and your subtitle is "Beautiful star maps for any room", the word "star" is wasting one of your slots. Apple already indexed it from the title.

Fix: make title and subtitle complementary. Title carries the brand + primary keyword. Subtitle carries a different keyword that pairs with the title.

›2. Generic modifiers

"app", "best", "free", "easy", "amazing" — these words rarely help you rank and almost never help you convert. They're polite. They're meaningless. Replace each one with a category or intent word.

›3. Translating instead of localizing

Translating "star map poster" to "póster de mapa estelar" gives you a Spanish phrase that almost no Spanish speaker actually types. Real Spaniards search "lámina mapa estelar" or "lámina constelaciones". Real Mexicans search "póster mapa estelar" or "cuadro estelar". One word, three different markets, three different ranking outcomes.

›4. Empty or thin Google Play long description

Play indexes its long description aggressively. An empty long description is a ranking dead zone. A 200-character description with one keyword repeated four times is worse — it'll get flagged as spam.

The right move: 1500–2500 chars, structured into 3–5 sections, each leading with a different benefit. Repeat your primary keyword 2–3 times across the whole thing, never more.

›5. Letting your iOS description rot

iOS description is not keyword-indexed. It's still load-bearing, because it converts the visitor who clicked through. The first 150 characters before "more" tap is the make-or-break surface. Lead with the concrete benefit. Do not say "Welcome to..."

Why most A/B tests don't help indie devs

Apple's Product Page Optimization and Google Play's Store Listing Experiments both let you A/B test screenshots, icons, and text. The dirty secret: at indie-dev install volumes, you don't have the statistical power to reach significance on most variants. You'll see a 4% lift in week one, the lift will disappear in week three, and you'll have spent six weeks on a non-result.

A/B testing earns its keep when you have ~5,000 store visitors per week per variant. If you don't, prioritise getting metadata right via reasoning instead of testing — and re-run keyword research quarterly.

The retention loop nobody talks about

Every ASO tool will tell you "improve your retention". None of them tell you why.

Apple and Google both have access to the install-and-then-uninstall signal. Apps with high week-one churn get suppressed in rankings — sometimes silently, sometimes brutally. If your app converts a search visitor into an install but they uninstall in 48 hours, the store learns that your metadata was misleading and will rank you lower for that query.

The implication: never optimize metadata for installs at the expense of fit. A clickbait title that wins installs from the wrong audience will hurt your rank within a week or two.

When to bring in a tool, and which kind

You can do basic ASO with free tools — google-play-scraper, the iTunes Search API, manual autocomplete checks. It's tedious and slow, but it works for a single app in 1–2 markets.

Where a paid tool earns its keep:

• You're shipping in 5+ markets and need to keep localized metadata in sync.
• You want continuous rank tracking (not point-in-time).
• You want prescriptive optimization suggestions, not just diagnostics.
• You want to compare your metadata to competitors across multiple keywords at once.

The market splits into roughly three tiers:

• Enterprise (SensorTower, AppTweak, Mobileaction). $79–$1,500/month. Built for ASO agencies and big publishers. Real install data via SDK partnerships, deep competitor intelligence, but heavy and overkill for one app.
• Mid (AppFollow, ASOMobile). $30–$200/month. Decent rank tracking, basic keyword research, light AI assistance.
• Indie-first (us, a few others). $0–$50/month. Localization-first, opinionated, prescriptive. Built for solo devs and 2–10 person studios.

We wrote ASO Pilot vs AppTweak and ASO Pilot vs SensorTower for the honest breakdowns.

What to do this week

If you're starting from zero:

1. Read How App Store ranking actually works. It's the mental model the rest of this stuff rests on.
2. Audit your current metadata against the five mistakes above. Most apps fail at least three of them.
3. Pick your top 3 markets (don't try to do 10 at once). For each one, find 5 keywords that show up in autocomplete and 3 that don't show up but are obvious to a human.
4. Rewrite your title + subtitle + keywords field for one market. Ship it. Wait 2–3 weeks for the store to re-index. Measure.
5. If you're not seeing movement after 4 weeks, the issue isn't your keywords — it's intent mismatch, rating, or retention. Don't keep tweaking metadata. Diagnose deeper.

ASO isn't fast. The first quarter of getting it right typically returns 30–60% more organic installs. The compound interest after a year is real.

If you want a tool that does this with you instead of for you, give ASO Pilot a try. One free analysis a month, three markets, real autocomplete signals, prescriptive fixes. It's what we use ourselves.