Blog - Growth, Engineering & AI

You can't beat a competitor you're not measuring. The good news: the mobile ecosystem leaks an enormous amount of competitive data—category rankings, download estimates, revenue trends, keyword positions, update cadence, pricing changes, even review sentiment. None of it requires anything shady; it's all modelled from public store data. The teams that grow fastest treat this as core infrastructure with a weekly cadence, not an occasional curiosity when a board meeting looms.

This guide covers the full stack—what you can get for free, what each paid tier actually buys you, how accurate the estimates really are, and the dashboard format that turns raw tracking into decisions.

Level 0: the stores themselves (free, underrated)

Before paying for anything, exhaust what Apple and Google publish directly:

Category rank charts. Both stores show Top Free and Top Grossing by category and country. A competitor's Top Grossing position is the closest free proxy for revenue momentum that exists—check it weekly and write it down, because the stores don't show history.
Update history. The "Version History" on the App Store shows every release with notes and dates. Release cadence tells you how fast their team ships; the notes tell you what they're betting on. Three consecutive releases about onboarding means they have a conversion problem—or just fixed one.
The review stream. Sort a competitor's reviews by most recent and most critical. Recurring complaints are your positioning shortcut: whatever their users hate is a headline your ads can make. Watch their developer replies too—response speed is a proxy for how well-resourced the team is.
Screenshot and metadata changes. Screenshot redesigns and title/subtitle changes are ASO tests happening in the open. If a competitor's new screenshots survive three months, they converted better than the old ones.

The limitation of level 0 is obvious: no numbers, no history, and it doesn't scale past two or three competitors. That's what the paid tiers fix.

Level 1: download & revenue estimates

Sensor Tower (which absorbed data.ai), AppMagic, and Apptica model downloads and in-app revenue for any app from store rank data, using apps whose real numbers they know as calibration. Two things to understand about these estimates:

Accuracy is good enough for trends, not accounting. For established apps with steady rank, estimates typically land within 20–30% of reality. For small apps (under ~1,000 downloads/day) or apps with heavy web/offline monetisation, they can be off badly. Never quote the absolute number in a deck without a range; always trust the direction of the curve.
The derivative is the signal. What matters is never "they did 400k downloads last month"—it's "their download curve inflected upward two weeks after the new onboarding shipped" or "revenue jumped 40% when they moved the paywall." You're reverse-engineering cause and effect from public timing.

On pricing: Sensor Tower is the enterprise standard and priced like it (five figures a year). AppMagic delivers most of the practical value at a fraction of the cost and is where we'd start; Apptica is a solid mid-priced alternative with strong ad-creative coverage.

Level 2: keyword rankings — the ASO battleground

Tools like AppTweak, ASOMobile, and Checkaso track which keywords any app ranks for, in which position, in which country (we compared the whole category in our ASO tools guide). This is where competitor tracking turns directly into growth, through a workflow we call gap mining:

Pull the full keyword set your top three competitors rank for (top 10 positions).
Subtract the keywords you also rank for. What's left is the gap list—proven demand you're invisible to.
Score the gap list by volume and difficulty, pick the top ten, and work them into your title, subtitle, and keyword field over your next two releases.
Track weekly. Store algorithms move fast, and a keyword you win can be lost in a month.

Gap mining is the cheapest organic growth roadmap that exists, because your competitors already validated the demand—you're just showing up where they proved users search.

Level 3: ad spend signals — paid or organic?

Pair your ranking data with the ad libraries (Meta Ad Library, TikTok Creative Center, Google Transparency Center—we wrote a full guide on spying on competitor ads) and, if budget allows, Sensor Tower or Apptica's ad intelligence, which shows which networks a competitor buys and their share of voice. The point of pairing the datasets is diagnosis. When a competitor's rank spikes, there are only three explanations, and each demands a different response from you:

Rank spike + burst of new ad creatives → they found a winning ad. Decompose it, pattern it, respond in the auction within two weeks.
Rank spike + no new ads → product change, press, or a viral moment. Check their update notes and TikTok organic before you touch your ad budget.
Rank spike + featured placement (check the store editorial pages) → store featuring. Temporary; don't overreact.

Teams that skip the diagnosis step chronically misread paid pushes as product wins and copy the wrong thing.

The dashboard: one sheet, fifteen minutes a week

Resist the data swamp. Everything above compresses into one spreadsheet with five competitors as rows and seven columns: category rank, top-grossing rank, estimated weekly downloads, new keywords entered (top 10), new ads spotted, store listing changes, notable reviews. Update it every Monday—fifteen minutes once the sources are bookmarked. Add a one-line "so what" note whenever something moved.

After a quarter, this sheet does something no purchased report can: it predicts. You'll see the pattern of how your market responds to paywall changes, which creative angles precede download spikes, and which keywords are winnable. Every competitor experiment becomes free R&D for your roadmap.

Speed of response is the whole point

Competitive data is only worth what you do with it, and its value decays in weeks. When a competitor's new onboarding drives a rank jump, you want to know within seven days—while there's still time to test the same mechanic before the market normalises it. Discovering it in a quarterly review is trivia, not intelligence.