App Store Optimization: Get Found Before You Spend on Ads

The App Store and Google Play are the two biggest search engines most businesses never optimize for. Apple has published that roughly 65% of App Store downloads come directly from search, and around 70% of visitors use search to find apps in the first place. The app that ranks — and the page that converts — wins the install long before any ad budget enters the picture.

This guide is the ASO playbook we apply when we ship mobile products, from consumer apps to niche B2B tools. It covers keyword research, metadata, ratings velocity, page conversion, measurement and the growth loops that make rankings compound. Worked numbers come from a modelled consumer language-learning app and a niche field-service scheduling tool, and every modelled figure is labelled illustrative.

Here is the short version before the deep dive.

Key Highlights

• ~65% of App Store downloads come directly from search, per Apple's own Search Ads research — the store is a search engine first.
• 70% of App Store visitors use search to find apps (Apple), yet most app pages are never keyword-optimized at all.
• You control 60 indexed characters on iOS (a 30-character title plus a 30-character subtitle) and a 100-character keywords field — the highest-leverage text in your marketing stack.
• Google Play indexes an 80-character short description and up to 4,000 characters of long description; Apple does not keyword-index its long description.
• Page-testing vendors such as StoreMaven report double-digit conversion lifts from creative overhauls; our modelled example lifts search-traffic conversion from 24% to 31%.
• Apple allows only 3 rating prompts per 365 days per app — prompt timing decides whether you collect 3.3-star or 4.6-star feedback (illustrative).
• In our modelled language-app program, search impressions grew 2.6x in six months and search-driven installs tripled from 3.1K to 9.4K per month (illustrative).
• ASO is a compounding asset: keyword rankings you earn keep producing installs after the spending stops — the opposite of paid acquisition.

Where App Discovery Actually Happens

The scale of the opportunity is hard to overstate. Sensor Tower's State of Mobile reporting counts well over 250 billion app downloads a year across the two major stores, and almost every one starts the same way: a person types a need, scans the first handful of results, taps, and decides within seconds. Rank and page quality decide who wins that moment.

For a typical consumer app, the split is lopsided: store search dominates, browse and category traffic adds a slice, and referral links clean up the rest. In our illustrative model above, store search drives 62% of first-time installs — right in line with Apple's published share. Brand searches (people typing your app's name) add another 15%, which is why every piece of offline and social brand-building quietly feeds your store performance.

Niche B2B tools play a different game. A field-service scheduling app will never win "calendar app", but it can own long-tail terms such as "crew scheduling" or "job dispatch", plus its own brand name, where intent is high and competition is thin. Volume is smaller, but visitors convert at multiples of generic traffic — the same fundamentals behind our language-learning app case study.

Key takeaway: Treat the store as a search engine with a checkout attached. The same discipline we bring to web SEO — keyword research, on-page optimization, conversion testing — applies directly to ASO, as we cover in our SEO guide, and it compounds in exactly the same way.

Lever 1: Keyword Research for Stores, Not Google

Store queries are short, intent-heavy and more literal than web searches — people type "learn spanish offline", not "most effective methodology for acquiring a second language". Autocomplete, competitor metadata and review language are the three demand maps you can mine for free.

A structured keyword build takes one person about a day and sets up everything downstream:

• List 20–30 seed terms from your product's core jobs — what would a user type to solve the problem?
• Type each seed into both stores' search fields and harvest every autocomplete suggestion — those are real queries with real volume.
• Scan the top 5–10 competitors: steal nothing, but note which words appear in their titles and subtitles, because Apple indexes those fields heavily.
• Read one-star and five-star reviews of competitor apps for the vocabulary users actually use, which often differs from your internal product language.
• Score every term on relevance, volume and difficulty, and start with winnable long-tail terms — graduate to head terms as ratings velocity and conversion improve.

The compounding shows up fast once the list is right. In our modelled language-app program, tracked keyword impressions grew from an index of 100 in January to 261 by June — a 2.6x gain in six months (illustrative).

Bucketing the list keeps priorities honest as volume grows:

The niche B2B tool's winning list looked completely different: "crew scheduling", "job dispatch app", "field technician scheduling" — modest volume, but conversion rates several times higher than generic terms. Expect any keyword list to go stale within weeks, so re-run the autocomplete and competitor scan monthly. Seasonality matters too: a metadata refresh timed to "taxes" in April or "fitness" in January is free visibility.

Key takeaway: Keyword research for stores is demand archaeology — autocomplete, competitor metadata and review language tell you exactly what to rank for. Rebuild the list monthly and start where you can win.

Lever 2: Title, Subtitle and Description That Rank and Sell

Metadata is where ranking and selling meet, and the two stores treat it very differently. Apple indexes the title, subtitle and hidden keywords field; Google Play indexes the title plus short and long descriptions. Getting the mechanics right starts with the limits:

On iOS, the highest-volume keyword belongs in the title and the second in the subtitle, with the 100-character keywords field reserved for variants your title does not already cover. Apple's own guidance is to skip spaces after commas, avoid duplicating words already in your title, and use singular terms (the index matches plurals). Do not burn 30 characters on a tagline — "MindLoop: Daily Calm" tells the algorithm nothing, while "MindLoop: Meditation & Sleep" tells it everything.

On Play, the 80-character short description is simultaneously a billboard and an index, so it should lead with the benefit and carry the primary keyword naturally. The long description is crawled for relevance, but keyword-stuffing it risks metadata-policy flags, from listing rejection to suspension. Write for humans first and repeat the two or three priority terms naturally; Apple's long description, by contrast, is pure sales copy because iOS does not keyword-index it. Refresh the promo text monthly — it is the only iOS field you can change without a new build.

Key takeaway: Metadata is 160 indexed characters on iOS and a 4,080-character canvas on Play — the cheapest ranking leverage you will ever get. Put the keyword where the algorithm looks and the benefit where the human looks.

Lever 3: Ratings Velocity — The Ranking Flywheel

Both store algorithms weigh your rating average, and industry analysis consistently finds that velocity and recency matter as much as the raw number: a 4.6 collected this month beats a 4.7 frozen two years ago. Ratings also sit above the fold, so they convert as well as rank. The catch is that the stores ration your ability to ask.

Apple's rating API allows at most 3 prompt opportunities per 365 days per app, and Google Play's in-app review API works on a similar quota logic. That means every prompt is an asset you spend, and timing decides whether you bank a five-star or a complaint. The modelled language app makes the point: prompting after a completed lesson produced a 4.6 average, prompting on a random app open produced 3.8, and prompting right after signup produced 3.3 (illustrative).

Same app, same users — the only variable was when we asked. The operating rules that follow from this:

• Prompt after a visible win — a completed lesson, a booked session, a saved invoice — never during onboarding and never after an error or a crash.
• Never gate features on ratings or offer rewards for reviews; both Apple and Google explicitly prohibit it, and getting caught costs far more than the stars.
• Reply to reviews, especially one- to three-star ones, on Google Play where replies are public and read by every future installer.
• Watch velocity weekly in App Store Connect and Play vitals, not just the average — a slipping rate of new five-star ratings is an early-warning indicator.

Key takeaway: You get three chances a year to ask, so ask at the moment of maximum delight. Ratings velocity is a product decision, not a marketing one.

Lever 4: From Page Views to Installs — Conversion Is Half the Game

Installs equal impressions times click-through times page conversion, which means conversion multiplies everything upstream — including paid. Double your page conversion and every dollar of ad spend effectively halves its cost per install. That is why creative work is growth work, not decoration.

The gains from a structured creative overhaul are real and repeatable. Page-testing vendors such as StoreMaven report double-digit lifts from creative revamps as a norm, and our modelled program follows the same shape: search-traffic conversion improved from 24% to 31%, browse traffic from 19% to 26%, and web-to-store traffic from 12% to 18% (illustrative).

Work the page top-down, because attention runs out fast:

The tooling is free and built in. Apple's Product Page Optimization supports up to three simultaneous treatments, and up to 70 custom product pages match page to keyword — a "learn spanish" search can land on a page whose first screenshot is a conversation, not a word list. Google Play offers native listing experiments plus up to 50 custom store listings for the same play; our fitness training app case study shows how intent-matched pages compound.

Key takeaway: Conversion is the multiplier on every channel you run. Test the first screenshot before anything else, and match pages to keywords the stores let you customize.

Lever 5: Measure What the Stores Won't Tell You

Both consoles show impressions, page views and conversion, but the splits that matter — which keyword drove the view, which page variant converted, what those installs did after launch — are partial or hidden. Rank-tracking tools fill the keyword gap, mobile measurement partners connect installs to behavior, and the consoles remain the source of truth for store-side numbers. The trick is a single weekly scorecard that stitches them together:

The payoff of measurement is knowing what compounds. Our modelled language app grew search-driven installs from 3.1K to 9.4K per month in six months, and the dashed tail shows the modelled trajectory to roughly 14.3K by September if rankings hold — no incremental spend, purely the compounding of the levers above.

Two honesty rules keep the scorecard useful. Never report rankings without the conversion attached — a number-one rank on a term nobody types is a vanity metric. And attribute conservatively: when a rank change and a creative test land the same week, hold one variable steady the following week.

Key takeaway: ASO becomes a program, not a project, the week it has a one-page scorecard and one owner. Track rankings with conversion attached, and let the trend line — not opinion — decide the next test.

Lever 6: Free Growth Loops That Compound ASO

Single levers add; loops compound. Each costs engineering or design time rather than media budget, and each feeds the ranking signals the stores reward:

• The ratings flywheel — better rank brings more installs, more installs create more prompt opportunities, better velocity pushes rank higher.
• Share moments — build one thing worth sending (a results card, a streak screenshot, a shared itinerary); every referral install also feeds the ranking flywheel.
• Localization — metadata in German, Spanish or Japanese opens new keyword markets where competition is a fraction of English; the listing text is cheap, the traffic is not.
• In-app events and offers — these surface in search results and editorial rails, so a seasonal event is effectively a new keyword landing page.
• Web-to-app SEO — pages that rank for "best app for X" queries hand warm traffic straight to your listing, as our hotel booking app case study illustrates for localized, travel-intent markets.

Localization deserves the emphasis because it is the most under-used lever in the list. Translating metadata into the top two non-English markets by install share costs a few hundred dollars and a release, and multiplies the keyword surface you compete on.

Discovery, though, is only half the business — installs that never convert to revenue are an expensive hobby. Paywalls, trials and pricing are their own discipline, which we cover in our app monetization guide. Run the two together and every ranking gain becomes a revenue gain.

Key takeaway: ASO stops being a checklist and becomes a growth loop when product, localization and discovery feed each other. Build one loop per quarter and let them compound.

What Your Business Should Do

Here is the 90-day plan we run with clients to move discovery before any ad spend.

Days 1–30: Audit and research.

• Pull your current metadata, ratings trend and top-20 keyword positions into a one-page scorecard.
• Build the keyword sheet: seed terms, autocomplete mining, competitor scans and intent notes.
• Fix hygiene basics: correct category, every metadata field filled, screenshots matching your top keyword's intent.

Days 31–60: Rewrite, test and prompt.

• Ship the new title and subtitle on iOS plus the short description on Play, and update the keywords field.
• Launch one page test — new first screenshot and captions — and create two custom product pages for your two biggest keywords.
• Move the rating prompt to your strongest success moment and track velocity weekly.

Days 61–90: Localize and systematize.

• Localize metadata into your top two non-English markets by install share.
• Reply to every recent one- to three-star review on Play and fix the top recurring complaint.
• Lock the weekly scorecard (rankings with conversion, impressions, velocity, retention by source) and assign one standing owner.

The apps that rank without ads are not lucky — they ran these loops on a calendar. Ninety days of unglamorous execution beats any one-off audit you can buy.

The Bottom Line

ASO is the cheapest durable acquisition channel an app business owns. The levers are unglamorous — sixty indexed characters, a well-timed rating prompt, a better first screenshot — but they compound while ad costs keep climbing.

If you want a second pair of eyes on your store presence, that is a normal Tuesday for us. Book a free strategy call and we will audit your keywords, page and ratings setup together — or send us a message and we will reply within one business day.