App Store Optimization changes as your product grows. What you measure in month one won't matter in month twelve. A stage-based framework keeps your team focused on the right questions, cuts false alarms, and speeds up decisions. This guide breaks down what to track in Presale, Launch, and Scale, which formulas matter, and how to know when you're ready to move forward.
Why Stage-Based KPIs Work
Each stage brings different risks. Presale risks building the wrong message for the wrong audience. Launch risks slow learning and weak product-market fit. Scale risks decay as rankings slip, competitors copy your wins, and costs drift upward.
Your KPIs should evolve from visibility readiness to conversion quality to durability. Think of them as a truth funnel: Can people find you, trust you, and stick around?
Key Terms and Formulas
Impressions show how often your listing appeared in store results from search, browse, or other sources.
Taps count clicks from results to your product page.
Tap-Through Rate divides taps by impressions. It reveals if your title and icon match what people expect.
Product Page Views track visits to your store page.
Conversion Rate divides installs by page views. It shows how well your page persuades visitors.
Retention measures the percentage of installers returning on days one, seven, or thirty. Always track by cohort to avoid mixing different user types.
Average Rating is your rolling weighted star rating.
Rating Velocity counts new ratings per thousand installs in a set window.
Indexation confirms the store recognizes your keyword and ranks you in the top two hundred.
Keyword Coverage shows your share of tracked keywords where you rank in the top ten.
Ranking Distribution groups your keywords into bands, such as positions one through three, four through ten, and so on.
Incrementality measures installs caused by paid campaigns versus your organic Baseline.
Organic Uplift shows the percent change in organic installs when you turn paid campaigns on versus off.
Experiment Win Rate divides statistically significant wins by total completed tests.
Treat estimates like organic uplift as directional trends. Use consistent methods to spot real patterns over time.
Presale: Build Readiness and Set Baselines
Your goal is simple. Be discoverable on day one, ship creatives that match real search intent, and set a baseline you can beat.
What to Track
Keyword Universe Size: Build a list of phrases split into branded, competitor, and generic terms. Aim for one hundred to three hundred high-intent generic keywords for your category plus top competitor terms. The exact count depends on your niche, but breadth prevents tunnel vision.
Indexation Readiness: Check that core phrases appear in your title, subtitle, and keyword fields. Track a readiness score by dividing the must-have phrases placed by the total must-have phrases.
Coverage by Intent: For each intent bucket, forecast your top-ten coverage using competitor baselines and your on-page signals. Since you won't have factual ranks, you'll revisit this weekly after launch.
Creative Readiness Ratio: Count required assets like icons, screenshots, preview videos, and feature graphics across locales and platforms. Divide the produced and quality-checked assets by the total required assets. Target one hundred percent before launch.
Hypothesis Bank Coverage: Build a backlog of testable ideas linking audience, value proposition, and creative execution. Map each to a metric like tap-through rate or conversion rate. Track prioritized hypotheses with testable assets divided by top hypotheses.
Localization Readiness: Prioritize locales by size, revenue potential, and competitive intensity. Measure the percent of localized assets and the string quality pass rate. Ship at least your top three to five locales with native copy and culturally adapted screenshots, not just translated text.
Technical Readiness: Create custom product pages, deep links, campaign taxonomies, and experiment infrastructure. Score your readiness by confirming that naming conventions, tracking systems, and experiment slots are assigned to hypotheses.
Baseline Benchmarks: Collect competitor tap-through and conversion rate proxies from public tools and early tests. Document category rank ranges and rating distributions. Baselines anchor your launch goals.
Launch Forecast: Combine conservative assumptions around expected impressions, tap-through rate, and conversion rate to forecast day-seven installs by traffic source. Document this as a range, not a single number.
Ready to Launch When
You hit ninety percent creative readiness, eighty percent hypothesis coverage for your first two test sprints, complete localization in priority locales, finish technical setup, and sign off on your launch forecast.
Launch: Learn Fast and Lock Your Story
Prove your store assets, persuade the right traffic and create sustainable ranking momentum. Two loops, visibility and persuasion, need to work together.
Visibility Metrics
Indexation Velocity: Track how fast you enter the top two hundred for tracked terms after release. Measure days to the first index for each must-have keyword, then days to the top fifty. If velocity stalls, revisit your metadata and semantic variety.
Ranking Distribution Shift: Watch weekly movement across rank bands. Track net flow upward, like fifteen keywords moving into the top ten while three fall out. Plot branded, competitor, and generic terms separately. Early wins in branded terms are common, but generic gains prove category fit.
Impressions by Surface: Split impressions across search, browse, and referrals. In competitive categories, search should dominate early. If browse carries you, you may index for broad terms but fail to persuade searchers.
Persuasion Metrics
Search Result Tap-Through Rate: This measures clicks from search results to your page. It diagnoses icon, title, and subtitle effectiveness for each query. Low tap-through with high impressions means mismatched copy or a weak icon.
Product Page Conversion Rate: Track installs are divided by page views and broken down by traffic source and locale. Tie conversion to creative sets and value proposition hierarchy. If tap-through is healthy but conversion is weak, your page storytelling or proof needs work.
Custom Product Page Conversion: Map ad intent to tailored pages. Custom pages should outperform defaults on matched intents. Track custom page share of installs and win rate relative to the default.
New Ratings Velocity and Average: Launch volatility is normal. Track new ratings per thousand installs and the rolling average star rating. Set guardrails, like keeping the average above four points, five stars. Respond quickly to each locale's first two hundred to five hundred reviews.
Quality and Stability Metrics
Retention Day One and Day Seven
Retention proves you attract the right users. If installs spike but day-one retention crashes, you have persuasion without value alignment. Read quality by cohorts split by source and variant to avoid mixing groups.
Crash Rate
Store algorithms penalize poor stability. Treat this as non-negotiable. If stability dips, pause aggressive tests and fix the issue.
Experimentation Metrics
Experiment Win Rate and Time to Decision
Track the share of completed experiments with significant lift and median time to decision. A low win rate teaches you what doesn't work. A slow win rate wastes time.
Ready to Scale When
You hit sustainable top-ten coverage on priority generic keywords, stabilize search tap-through and conversion at or above competitor medians, maintain ratings above four point five with healthy velocity and fast responses, keep retention within expected ranges, and validate at least one winning creative direction across locales.
Scale: Compound Growth and Defend Your Position
Add countries, deepen coverage, and protect gains while improving unit economics.
Portfolio Metrics
Keyword Portfolio Health
Track your share of top five, top ten, and top twenty positions across your keyword universe by intent. Add ranking decay rate, which measures the percentage of keywords dropping a band week over week. Healthy portfolios show slow decay and steady gains in the top five.
Category Rank Sustain
Day-of-week and seasonal swings are normal. The goal is to sustain target rank bands without excessive spending. To cut noise, use seven-day medians.
Organic Uplift Versus Paid Cannibalization
Measure change in organic installs when paid campaigns run versus when they're off in the exact locations and queries. Use holdouts or time-series models. Target positive uplift with shrinking cannibalization as your ranks strengthen.
Incrementality of Paid
For Apple Search Ads and other channels, track incremental installs per dollar and blended return on ad spend or lifetime value to cost per install. App Store Optimization and paid search should work together. Custom pages and ranking wins can raise both organic conversion and paid efficiency.
International Metrics
Locale Parity
Conversion, tap-through, retention, and rating velocity should converge toward your primary locale benchmarks. Gaps reveal cultural or competitive issues. Track each metric's percentage of locales within ten to fifteen percent of the primary locale performance.
Localization Depth
Complete localization goes beyond translated text. Track percent of locales with fully localized screenshots, video captions, pricing cues, and social proof. Deep localization lifts both tap-through and conversion.
Reputation and Cadence Metrics
Rating Quality Distribution
Average ratings hide the mix. Track the share of one- and two-star ratings, the share of five-star ratings, and the median time to respond to critical reviews. Aim to reduce low-star share over time for large user bases while keeping velocity high.
Release Cadence Impact
Tie release notes and version updates to tap-through, conversion, and retention shifts. A steady cadence with meaningful improvements correlates with better browser exposure and user trust.
Growth Efficiency Metrics
Creative Throughput
Track the number of high-quality experiments per month and how many get localized. Measure experiment throughput times and median lift. Plateaus signals it's time to refresh your hypothesis bank.
In-App Events and Promo Surfaces
Measure lift in impressions, tap-through, and conversion during events like seasonal promos or feature launches. Good programs raise browser contribution and reactivation without hurting search position.
Measurement Tips That Prevent Bad Decisions
Split everything by cohort. Read retention and revenue by install week, version, locale, and traffic source. Mixing sources hides the truth.
Map intent carefully. Custom pages or tailored listings should mirror the query's language and promise. Don't route broad terms to narrow creatives.
Use consistent windows. Seven- and fourteen-day windows work well for launch. For scale, read twenty-eight days to capture slower effects.
Mind seasonality. Compare like weeks year over year and normalize around holidays or store-wide promotions.
Track review operations. Set response time targets and tag themes like pricing, onboarding, or bugs so product teams can act fast.
Build Stage-Specific Dashboards
Skip the mega-dashboard. Build scorecards for each stage.
Presale Scorecard tracks readiness across creative, localization, technical setup, hypothesis bank coverage, universe size, and forecast. Add a single gate-to-launch indicator tied to exit criteria.
Launch Scorecard tracks indexation velocity, ranking distribution shift, tap-through rate, conversion rate, custom page performance, ratings velocity and average, day-one and day-seven retention, crash rate, and experiment win rate. Add a weekly learning summary and subsequent tests.
Scale Scorecard tracks portfolio health across top positions, decay rate, category rank sustain, organic uplift, incrementality, locale parity, rating quality, release cadence impact, and creative throughput.
Set alerts for sharp tap-through drops on top keywords, conversion dips after releases, rating shocks in prominent locales, unexpected ranking decay, and high cannibalization.
Diagnose Problems and Fix Them
Low tap-through with decent impressions means your icon, title, or subtitle is off, or query matching is weak. Fix the first promise with clarity, category signals, and differentiation. Test icon motifs that cue the job to be done. Put the strongest descriptive term first in your title.
Good tap-through with low conversion means your page story or proof is weak. Strengthen social proof, reorder screenshots to show outcomes first, and align primary visuals with top queries landing on that page.
Good conversion with poor retention means you persuade the wrong users. Revisit targeting and claims. Avoid over-promising features you don't deliver in the first session.
Ranking stalls despite good conversion means you need to expand your keyword graph. Add semantically adjacent phrases, refresh metadata to capture new variants, and deepen backlinks or mentions that help store understanding.
Ratings slide means you must identify top complaint themes, escalate fixes, and run in-app review prompts only after a proven moment of delight. Never prompt after friction points.
High cannibalization means paid campaigns absorb installs you'd get organically. Refine match types and bids. Let organic win where you hold consistent top-five positions.
Keep Momentum with Governance
Hold a weekly stand-up during launch and bi-weekly on a larger scale. Review the scorecard, top three learnings, following three tests, and any guardrail breaches. Maintain a living experiment log with hypotheses, assets, metrics, and results. Close the loop with product and support teams so store promises match app reality.
Quick Reference: Formulas and Guardrails
Tap-Through Rate = Taps divided by Impressions
Conversion Rate = Installs divided by Product Page Views
Top Ten Coverage = Keywords ranked ten or better, divided by Tracked Keywords
Ranking Decay Rate = Keywords that dropped a band divided by Tracked Keywords
Rating Velocity = New ratings divided by Installs times one thousand
Organic Uplift roughly equals Organic installs with paid on minus Organic installs with paid off, divided by Organic installs with paid off
Incremental Installs per Dollar = Total installs with paid on minus Baseline installs, divided by Spend
Stage Guardrails
Tune these to your category, but use them as starting points.
Presale needs ninety percent creative, localization, and technical readiness, plus a hypothesis bank ready for two test sprints.
Launch needs tap-through at or above category median, conversion at or above category median, ratings above four points five with substantial velocity, retention within expected bands, and validation of at least one winning creative direction.
Scale needs top-ten coverage on generics growing or holding steady, low ranking decay, positive organic uplift, healthy paid incrementality, and performance parity across top locales within ten to fifteen percent.
FAQ
Are These KPIs “Official Store Metrics” Or A Framework?
They are both. The growth stage model is a framework, while impressions, page views, installs, and conversion rate are official store metrics in App Store Connect and Google Play Console.
What Exactly Is An “Impression” In App Store Connect?
An impression is counted when someone sees your app in App Store surfaces like Search, Today, Games, Apps, or on the product page, depending on the report view.
Is “Taps” An App Store Connect Metric?
No. Taps is an Apple Ads metric for ad interactions. In App Store Connect, the closest equivalent step is product page views.
How Does Google Play Define Conversion Rate?
In Play Console, conversion rate is shown in Conversion analysis as acquisitions divided by store listing visitors, and you can segment it by traffic source, country, and other dimensions.
Conclusion
App Store Optimization works best as continuous learning, not a one-time makeover. Stage-based metrics keep that learning honest. In presale, you earn the right to launch by proving readiness. In launch, you earn the right to scale by proving persuasion and early quality. You protect and compound in scale by watching portfolio health, international parity, and unit economics. Measure what matters for your current stage, and your team will know exactly what to do next.