How to Rank on the App Store

For 15-minute non-fiction book summaries of best-selling books, check out sumizeit.com.

TL;DR

App Store ranking is won mostly in three places: the text fields Apple indexes (app name, subtitle, keyword field), the conversion rate of your product page, and the volume and freshness of your ratings. Pick one high-intent keyword you can realistically own, put it in your app name, and build everything else around proving to Apple that people who search that term install and keep your app. Everything else — ads, press, social — is an accelerant, not the engine. And the single most underrated input is time: rankings compound over months, not weeks.

Pick one keyword and go all in

Most first-time developers approach keyword research like a shopping trip. They open a research tool, sort by search volume, and try to grab twenty terms at once. The result is an app that ranks 60th for everything and first for nothing, which on the App Store is functionally the same as ranking nowhere. Roughly 70% of App Store downloads still start with a search, and the tap-through beyond the first three results falls off a cliff.

The better strategy is narrower and more uncomfortable: pick the single most important keyword in your category and commit to it, even if the difficulty score looks brutal. Sebastian Röhl, the indie developer behind HabitKit, did exactly this. He couldn't find a keyword with a friendly volume-to-competition ratio, so he stopped looking for one and went all in on "habit tracker" — the hardest, most obvious term in his niche. He renamed the app on the store to "Habit Tracker – HabitKit," reserved the subtitle for his second and third priorities ("Streaks & Accountability"), and waited. Within a month he was top ten in the UK, and within ten months he was top five across multiple countries.

That story is instructive not because the keyword was clever but because the decision was singular. Apple's algorithm rewards relevance density. When your name, subtitle, screenshots, reviews, and the actual behaviour of your users all point at the same term, you're giving the ranking system an unambiguous signal. When you hedge across ten terms, you're giving it noise.

The practical test for choosing your term: would a stranger who wants exactly what you built type this phrase? Not "best productivity app" — nobody searches that with intent to install a specific thing. Something like "habit tracker," "book summaries," "expense splitter," "sleep sounds." High intent beats high volume every time, because a searcher with intent converts, and conversion is what compounds.

If you're building in a category where the obvious term is genuinely locked up by an incumbent with ten million ratings, the move isn't to give up on specificity — it's to find the adjacent phrase that the incumbent didn't bother to own. "Book summaries" is competitive; "nonfiction book summaries," "book summary app," and "audio book summaries" are all separate search behaviours with their own rankings.

Where to get the data

You can't eyeball this. Apple publishes a popularity score for search terms, and you need a tool to see it. Röhl's current recommendation is Astro, the ASO tool from Appfigures, which handles keyword popularity, difficulty scoring, and competitor teardowns in one place. The established alternatives are Sensor Tower, AppTweak, Mobile Action, and App Radar; ASOMobile and Appfigures' own free tier are the cheaper entry points if you're pre-revenue.

The free option most people skip: run a low-budget Apple Search Ads campaign and read the search terms report. That's Apple's own data on what real users typed before tapping your listing — not a third-party estimate of it. Cross-check whatever your tool tells you against that report before you commit a name change.

What Apple actually indexes

Apple's search index pulls from a smaller set of fields than most people assume, and knowing the list changes how you write them.

The app name (30 characters) carries the most weight by a wide margin. The subtitle (30 characters) is second. The keyword field (100 characters, invisible to users) is third and still meaningful. Apple also indexes your developer name, your in-app purchase names and descriptions, and — this is the one most people miss — the text of your user reviews, which is part of why reviews mentioning your core use case are worth more than generic five-star ratings.

Your long description is not indexed on the App Store. It matters for conversion, and it is indexed on Google Play, but stuffing keywords into it does nothing for Apple search. Write it for humans.

A few mechanics worth internalising. Don't repeat words across fields — if "tracker" is in your name, it's already indexed, and using the slot again in the keyword field wastes characters. Skip plurals; Apple handles singular and plural variants. Skip spaces in the keyword field and separate with commas only, which buys you extra characters. And don't include your own app name or category name in the keyword field — Apple indexes those automatically.

Apple also builds combinations. If your name contains "habit" and your keyword field contains "streak," you're eligible to rank for "habit streak" without ever writing that phrase. This is why a tight set of strong root words outperforms a long list of full phrases.

Conversion rate is the hidden ranking factor

Here is what separates apps that climb from apps that plateau: Apple doesn't just measure whether you're relevant to a query, it measures what happens after someone sees you. If your listing appears for "habit tracker" a thousand times and gets fifteen installs, and a competitor gets ninety, the algorithm draws the obvious conclusion.

That makes your product page a ranking asset, not just a marketing asset. Three elements do almost all the work.

The icon is judged in about half a second at thumbnail size. Test it at 60 pixels, not on your 27-inch monitor. Detailed illustrations turn to mush; a single bold shape with high contrast survives.

The first two screenshots are the only ones most people see, because they're what's visible in search results without scrolling. They shouldn't be raw device captures. They should be captioned — a short benefit line above a cropped screen — and the first caption should restate the promise of the keyword the person just searched. Someone who searched "book summaries" should see the words "book summaries" in the image.

The video preview is optional and autoplay-muted. It helps most for apps where the value is motion-dependent (games, editing tools) and helps least for utilities where a static shot explains everything.

Then measure. Apple's Product Page Optimization lets you run true A/B tests on icon, screenshots, and preview video against live traffic, splitting up to four treatments. It's free, it's built into App Store Connect, and a surprising number of indie developers have never opened it. A conversion lift from 3% to 4.5% is a 50% install increase from the same ranking — and because installs feed ranking, it usually becomes more than 50%.

Reviews: the flywheel most developers under-use

Ratings do three jobs at once. They influence ranking directly, they influence conversion (an app at 4.7 stars converts far better than the same app at 4.1), and their text feeds the search index.

The mechanics of asking matter more than the asking. SKStoreReviewController gives you three prompts per user per year, so the timing of each one is expensive. Prompting on app launch, or right after onboarding, catches people before they've experienced anything worth rating. The better trigger is the first moment of delivered value — Röhl's HabitKit asks after a user logs their first habit completion. For a book summary app, that's finishing a first summary; for a fitness app, the first completed workout.

The second source of reviews is email, and it's badly underused. When a user writes to you and you reply personally, you've already created goodwill that most apps never generate. Röhl closes those replies with a single line: "PS: If you're enjoying the app, I'd be grateful if you could leave a review on the App Store. It helps others discover the app and gives it a boost!" People say yes at a rate that would embarrass any in-app prompt.

Reply to reviews, including bad ones — publicly, specifically, and without defensiveness. Apple lets users update a review after a developer response, and a meaningful share of one-star reviews get revised upward when someone actually fixes the problem. The responses also read as evidence of an active developer to everyone else scrolling past.

One caution: your rating resets are in your control. You can reset your rating with each new version, which is almost never a good idea unless you've had a genuinely catastrophic release. Ratings accumulate slowly and are the hardest asset to rebuild.

Apple Search Ads: useful, but not magic

There's a persistent belief that running Apple Search Ads lifts your organic ranking. Apple has never confirmed it, and Röhl is admirably honest about his own experiment: $100 a month, roughly 40 installs at €2.50 each, no measurable conclusion, and a note in the middle of his own article that writing it out made him realise he should probably cut it.

The defensible use of Search Ads isn't ranking manipulation — it's research. A small campaign shows you the exact search terms driving taps to your app, which is data you cannot get any other way. Run it for a few weeks with broad match to harvest terms, feed the winners into your keyword field, and then decide whether the paid spend pays for itself on its own merits. If your cost per install exceeds what a user is worth to you, turn it off. Ads are a measurement tool that sometimes happens to be profitable.

Off-store signals and building in public

Apple has said publicly that it takes web traffic into account, and Apple Search Ads reporting now surfaces web referrer data. Links to your App Store page from real sites carry weight, and more importantly they carry people.

This is where building in public earns its keep. Röhl shared the HabitKit journey on X in detail — revenue numbers, mistakes, feature decisions — and the community that formed around it generated organic mentions and links that no press push would have bought. The mechanism isn't mystical: an indie developer telling an honest story is more shareable than a product announcement, and shares become links, and links become traffic, and traffic becomes installs.

The window that matters most is the week after launch, when Apple gives new apps a temporary visibility boost. What you do with that week determines whether the boost decays into obscurity or into a stable baseline. Line up your Product Hunt launch, your newsletter, your subreddit posts, and any press for that window rather than spreading them across three months.

Time, iteration, and the part you can't control

Ranking is a lagging indicator. HabitKit's climb to the top five took ten months of consistent shipping and review accumulation, not a clever launch trick. Apple's algorithm appears to weight sustained performance heavily, which is a polite way of saying it distrusts spikes.

The compounding loop is simple: ship improvements, which generate better reviews, which improve conversion, which improves ranking, which increases impressions, which generates more reviews. Every turn of that loop is small. Twenty turns is a business.

And some of it is luck. A term you can't crack, a competitor who launches a free tier, a feature by Apple that never comes. Röhl's first app didn't work; HabitKit did. The correct response to a strategy that isn't moving after six honest months is not to try harder at the same thing — it's to change the keyword, change the positioning, or change the app.

Start with the name field

If you take one action after reading this, make it the boring one: open App Store Connect, look at your app name, and ask whether it contains the exact phrase your ideal user would type. Most apps fail this test, and fixing it costs a version submission and nothing else.

Then set up a Product Page Optimization test, move your review prompt to your first real value moment, and give it two quarters. ASO isn't a growth hack; it's a maintenance habit that pays out on a delay.

For 15-minute non-fiction book summaries of best-selling books, check out sumizeit.com.