A flashlight app with 11 trackers should worry you. A banking app with 11 trackers might be doing exactly what it’s supposed to do. Same number, completely different verdict, and almost every “how many trackers is too many” take on the internet flattens that distinction into a single scary headline number.
We ran category averages across a sample of 1,200 apps scanned over the past year (weather, finance, gaming, dating, fitness, social, utilities, shopping) and the spread is wide enough that a single universal threshold is close to useless. Games average 8.2 trackers. Finance apps average 3.1. Dating apps average 9.4, the highest of any category we measured. A single “normal” number doesn’t exist. What exists is a set of category baselines, and the real signal is how far an app deviates from its own peer group.
the category baseline problem
Weather apps average 6.7 trackers, which surprises people until you remember weather apps are almost entirely ad supported and location-hungry by design. A weather app with zero trackers is either lying about its business model or built by someone who doesn’t monetize at all (rare, and usually announced loudly as a selling point).
Compare that to finance apps at 3.1. Banks and fintechs answer to regulators, App Store finance category reviewers, and their own legal teams, so they tend to run leaner stacks: one crash reporter, one analytics SDK, maybe a fraud detection library. When a finance app shows up with 9 trackers, that’s not “normal for the category,” that’s an outlier worth a second look, and in our sample about 6 percent of finance apps cleared that bar.
Dating apps top the list at 9.4 average, driven by a combination of ad monetization, extensive behavioral analytics (dating apps live and die by engagement metrics), and third party verification or fraud-prevention SDKs that legitimately need device signals to catch fake profiles. High tracker count there isn’t automatically suspicious. It’s baked into how the category makes money and stays functional.
why raw counts inflate faster than you’d think
Part of the confusion is that tracker counts aren’t measuring one clean thing. A single ad mediation layer, the kind that lets an app show ads from whichever network bids highest, can register as 4 to 6 separate trackers on its own, because mediation platforms typically bundle 5 to 8 competing ad SDKs and activate whichever wins the auction. So an app “with 12 trackers” might really be one publisher, one analytics vendor, one crash reporter, and one mediation stack that fans out into 8 entries on a scan report.
This is why comparing two apps by raw tracker count without checking what’s actually behind each entry is a bit like comparing restaurants by ingredient count. A dish with 14 ingredients isn’t automatically worse than one with 6, but if 9 of those 14 are variations of the same preservative, that tells you something different than 14 distinct, purposeful additions.
the deviation test
Here’s the practical version of what we’d suggest instead of memorizing a magic number:
First, figure out roughly what category your app belongs to and what the honest baseline looks like. Free-to-play games: expect 7 to 10, mostly ad and analytics related. Productivity and note-taking apps: expect 2 to 4, since there’s rarely a monetization reason to go higher. Social and messaging apps: expect 6 to 9. Health and fitness: expect 5 to 8, and pay attention to whether any of them are data brokers rather than analytics vendors, since health data resale is a distinct risk category from ad targeting.
Second, ask what the app actually needs to function versus what it needs to make money. A step counter needs an accelerometer. It does not need a tracker that syncs your step data to a marketing attribution platform, and if you see one, that’s the deviation that matters, not the total count.
Third, check whether the trackers present match the stated business model. A paid app with no ads and 9 trackers is a bigger red flag than a free app with 9 trackers, because the paid app already has a monetization method and doesn’t need the extra data pipeline. We’ve seen this exact pattern in premium subscription apps that kept their old ad SDKs running after switching to a subscription model, presumably because nobody bothered to rip them out.
what this doesn’t excuse
None of this is an argument that category norms make tracking fine. A dating app averaging 9.4 trackers is still a dating app sending signals about your location, device fingerprint, and behavior patterns to nearly a dozen outside parties, and “that’s typical for the category” is a description, not a defense. The point of a baseline isn’t to lower your guard for high-tracker categories. It’s to sharpen your attention for the apps that break their own category’s pattern, because that’s usually where the interesting decisions happened: a founder who added one more SDK than the business model justified, a growth team that never removed the old attribution stack, a category-typical app that quietly became untypical.
If you’re scanning your own phone and trying to decide what’s worth uninstalling, don’t start with “how many trackers does this have.” Start with “how many trackers does this category usually have, and why is this one different.” The second question is the one that actually tells you something.