Day Seven, Not Day One: What Really Matters After Launch
A developer messaged me last month with a problem I've heard a hundred times: 'My user acquisition looks great. I'm getting installs cheap. But my game is dead by week two.' She'd been chasing the wrong metric since launch day.
The install-count trap
When you launch an app, the first metric everyone watches is installs. It's visible, it's immediate, it's easy to celebrate. Your marketing is working. You got 500 installs on day one. Then 300 on day two. The curve flattens. You feel good about it.
But here's what you're not seeing: how many of those 500 are still opening your app on day seven.
I learned this the hard way with our own apps. We'd crack a new acquisition channel, launch a campaign, and watch the install numbers climb. Then, two weeks later, we'd notice retention was catastrophic. We'd spent weeks optimising for the wrong thing.
The real question isn't where the install came from. It's whether the person who installed is still there. A thousand dead installs from a cheap source is worse than fifty installs from a source where users actually stick around.
Why day seven changes the conversation
Day one tells you if someone clicked. Day seven tells you if they cared.
Most users who delete an app do it in the first week. They open it, decide it's not for them, and move on. If your users make it past day seven, retention curves typically stabilise. Not always up, but the trajectory becomes predictable. You can see which sources are sending engaged users versus which are just sending lookalike audiences with no real interest.
I remember working with a studio last year who'd been running ads through multiple channels. Their day-one install cost looked identical across all of them: roughly the same CPI. But when we looked at day-seven retention, one channel had 40% of users still active, and another had 8%. The cheap channel wasn't cheap at all. The expensive one was a bargain.
That's the shift. You're not measuring acquisition cost anymore. You're measuring actual customer quality.
How to know where your best users come from
The technical piece is simpler than most people think. You need three things: which source sent the install, whether that user is still there seven days later, and ideally, fourteen and thirty days too. You need to tie that back to your marketing sources so you can see which channels drive engaged users.
For indie devs who don't have the budget for enterprise attribution platforms, this used to mean either building it yourself or accepting that you'd never know. Building it yourself is expensive in time. Accept ignorance and you're flying blind.
Attribr answers those three questions directly. It shows you exactly where each install came from, using deterministic matching when possible and probabilistic matching when the signal isn't there. More importantly, it tracks cohorts at day seven, fourteen, and thirty. You can see retention curves by source. You can see which Rippl promoters are driving installs that actually stick around, which is unique if you're using performance marketing through the Rippl community.
The technical overhead is minimal. Fifty kilobytes of SDK code, no external dependencies, sub-50ms launch impact. Three lines to integrate into Swift or Kotlin. You're not slowing down your app to measure it.
What changes when you measure the right thing
Here's what happens once you have day-seven retention data: you stop guessing.
You see which marketing sources are actually valuable. You see which app store placement experiments work. You see whether a feature launch actually kept people engaged or just looked good on a metrics dashboard. You make decisions based on real user behaviour, not on optimised funnels that leak from the bottom.
A couple of months back, one of our users disabled a campaign they thought was working well. Their install volume dropped by 30%. But their retention improved by 8 percentage points. The campaign was cheap because it was bringing in the wrong audience. By turning it off, they freed up budget to focus on the sources that actually worked. Their revenue per user went up even though they had fewer installs.
That decision would have been impossible without measuring day seven. They'd still be celebrating volume instead of value.
The indie developer's advantage
There's an irony in app development: the bigger you get, the harder it becomes to move fast on insights. Enterprise platforms lock you into contracts and reporting cadences. You're stuck waiting for quarterly reviews while the market moves.
Indie developers don't have that problem. You can see your data, understand what's working, and change direction in days. The only thing stopping that is access to the right metrics.
If you're running user acquisition and you're not measuring day-seven retention by source, you're not running acquisition at all. You're running volume. Volume feels good, but it doesn't pay bills. Retention pays bills.
The infrastructure for this used to be expensive. Enterprise pricing, long sales cycles, complex implementations. It shouldn't be. Indie developers need attribution too. That's why we built Attribr specifically for this scale, specifically for this audience. Not for venture-backed studios that can afford anything. For people who are actually building sustainable apps with real users.
When was the last time you looked at retention by marketing source? What would change if you actually knew which of your users were still around on day seven?
Ready to try Attribr by MRVL?
One tap to download. No sign-up wall.