The three numbers that tell you if your installs are worth anything

Why day 1 is a lie

Everyone counts installs on day one. That's the easy part. But an install is not a customer. An install is a promise that might be broken by 6pm the next day.

The real question isn't 'how many people installed my app?' It's 'how many of those people are still using it a week later?' That's retention. That's the metric that tells you whether your marketing actually worked or whether you just rented some attention.

Most indie developers and small studios don't have a way to answer that question without juggling spreadsheets and guessing. The attribution tools they might have heard of - Branch, AppsFlyer, Adjust - were built for teams with 50 people and enterprise budgets. They don't talk to you about retention in a way that matters for a team of three.

Cohorts are just buckets of people with a deadline

A cohort is a group of users who installed your app on the same day, tracked together over time. Simple idea. Powerful signal.

In Attribr, we track three moments: day 7, day 14, and day 30. Why those numbers? Because they tell different stories. Day 7 tells you if people found your app interesting enough to come back within a week. Day 14 tells you if they've made it past the novelty phase. Day 30 tells you if there's a real habit forming.

Here's the part that matters: each cohort is broken down by install source. So you can see that installs from Reddit had a 34% day-7 retention rate, but installs from TikTok had 18%. Suddenly your marketing math changes. The cheaper channel looked good until you saw the retention numbers. That's when you know where to double down and where to stop wasting money.

We built this directly into the dashboard. No exporting to a spreadsheet. No waiting for a support ticket to come back. You open Attribr, you see your cohorts, you see which sources are bringing you keepers and which are bringing you tourists.

Attribution has to come first, or the retention numbers lie

You can't measure retention by source if you don't know which source each install came from.

Attribr answers that question with both deterministic matching - we can see the actual install ID from your ad network - and probabilistic matching, which uses other signals when deterministic data isn't available. Neither approach is perfect on its own. Together, they work. And they work without requiring ATT permission, which means you get data on iOS 14.5+ even when users have opted out of tracking.

A lot of developers think this is complicated. It's not. The SDK is 50KB. It runs in less than 50ms on launch. Three lines of code in Swift or Kotlin and you're collecting data. No third-party dependencies. No bloat.

The SDK starts counting from day one. It phones home to tell us which source brought each user in. Then it quietly tracks whether they're still active on day 7, day 14, and day 30. You don't have to do anything except open the app and look at the numbers.

The Rippl angle: when attribution feeds back into marketing

There's a second layer to this that matters if you're using Rippl for performance marketing.

Rippl is community-driven CPI. Indie developers promote each other's apps. When someone in the Rippl network promotes your app and it converts, Attribr knows it was a Rippl install. More than that, Attribr knows which specific promoter drove it. That's unique. Most attribution SDKs stop at the network level - 'this came from Facebook' or 'this came from a mobile ad network.' But they don't tell you which person or creator drove the actual conversion.

With Attribr, you can see that Promoter A brings you a 40% day-7 retention rate while Promoter B brings 22%. You can measure whether a community-driven partnership is actually sustainable. You can pay commissions fairly because you have the data. That feedback loop makes performance marketing, well, actually performant.

What cohorts don't tell you, and why that matters

Retention cohorts are powerful but they're not magic. They tell you who came back and who didn't. They don't tell you why. A 25% day-7 retention rate could mean your app is fundamentally broken, or it could mean you're attracting the wrong audience for that particular campaign, or it could mean your onboarding is a wall of friction.

But once you know the number, you can ask better questions. You can A/B test onboarding. You can refine your targeting. You can look at which sources bring users who stick around and lean into those. The data becomes actionable.

We also show you fraud signals in the Pro plan. Not because we're a fraud detection company, but because if your cohorts look suspiciously perfect or suspiciously terrible, you should know. Sudden spikes in bot-like behaviour can skew your retention numbers and mask the truth about your real users.

The small-team advantage

One reason Attribr exists is because small teams need to be fast with their data. You can't wait three weeks for a support ticket. You can't afford per-event pricing that scales with your success. You need a tool that costs £29 a month and tells you whether your marketing is working by the end of the week.

The cohort dashboard does that. You see it in real time. Seven days after a campaign launch, you already know whether you've found a channel worth pursuing. Fourteen days in, you know if it's sustainable. By day 30, you either commit or you pivot. That speed is the whole difference between an indie team and a team with an analytics department.