The fraud problem nobody talks about (until it's too late)

The blind spot in indie attribution

When you're building an app at indie or small-studio scale, attribution tools feel like a luxury. You pick one of the big names, wire it up, and assume that once you've got install source tracked, you're done. But here's what happens in reality: you start spending real money with ad networks. Maybe it's Facebook, maybe it's AppLovin, maybe it's smaller regional networks. You run for a few weeks. The numbers look good. Then you check retention and something feels off.

The problem is that attribution alone doesn't tell you whether an install is fraudulent. It tells you which network sent it. But if that network is serving bot traffic, click spam, or incentivised installs from users who have no intention of using your app, you've got a problem on your hands. You've paid for an install. Attributed it correctly. And then watched it churn out by day 14.

Most indie developers don't have the budget or engineering time to build fraud detection themselves. The enterprise tools bundle it in, but they're also bundled with six-figure contracts and features you'll never use. So you're left guessing.

Why retention curves lie (and how to read them)

Here's what we learned from that developer's email, and from a dozen conversations that followed: you can spot fraud signals in your own data if you know what to look for. A healthy install cohort has a predictable shape. Users drop off, but the curve is smooth. If you're seeing sudden kinks, flat spots, or retention that makes no sense relative to your product quality, that's a signal worth investigating.

When we built the fraud signals feature in Pro, we didn't want to be another black-box fraud vendor. We wanted to give indie developers visibility into what Attribr was seeing in their own install streams. So we built roll-up data across your ad networks. You can see cohort retention broken down by source. You can spot which networks are sending you users who disappear by day 7. You can cross-reference that against your spend. And then you can make a decision: pause that network, adjust your CPI cap, or dig deeper.

The feature isn't a kill switch. We're not saying we'll block traffic for you. We're saying: here's what the data shows, and here's what normal looks like in your category. Make the call.

Ad networks and the transparency problem

If you've run campaigns with multiple networks, you know the frustration. Each platform has its own dashboard. Each one reports slightly different numbers. Some report post-install events, some don't. Some have a lag. Most importantly, none of them will tell you whether the users they sent you are real.

The ad-network roll-up in Attribr Pro was built to solve this. It's not magic; it's just consolidation. You see installs by network, day 7 / 14 / 30 retention by network, and the cohort charts that matter. It sits in one place. You don't have to log into five dashboards and squint at conflicting data.

One studio told us they'd been running two networks in parallel for eight weeks before they realised one was sending them 40% more installs but 60% lower day-7 retention. They found it by accident when comparing screenshots. With roll-up visibility, that would've surfaced in week two. They would've re-allocated budget instantly.

That's not flashy. It's not a silver bullet. But it's the difference between guessing and knowing.

The three-question test

Every time a developer tells us about their fraud concerns, it boils down to three questions. One: where did this install come from? Two: is the user still here at day 7, 14, 30? Three: am I wasting money? Attribr answers the first two out of the box. The fraud signals feature helps with the third.

But here's what's important: you still have to decide. We're giving you the data. The decision to pause a network, reduce spend, or investigate further is yours. We're not running an approval process. We're not gatekeeping your campaigns. We're just showing you what we see.

For teams with tight budgets, that distinction matters. You might tolerate a network with 22% day-7 retention if the CPI is low enough. Another studio might not. We can't make that call for you. We can only make sure you have the information to make it yourself.

Why this matters now

iOS 14.5 changed the game in a lot of ways, most of them painful for indie developers. Fingerprinting became harder. Deterministic matching became more valuable. The ad networks had to adapt, and some did better than others. What nobody talks about is that as tracking became harder, the incentive to ship bot traffic increased. If a network can't be proven out, they might as well send whatever moves metrics.

That's cynical, but it's the conversation we've had with studios that have been hit. They're not paranoid. They're just reading their retention charts and seeing a story that doesn't add up.

We built fraud signals and ad-network roll-up because we wanted Attribr to be useful at the moment when it matters most: when you're spending real money and trying to know whether it's working. It's not a replacement for trust. But it's a lot better than guessing.