The day we stopped asking for permission

What actually changed on iOS 14.5

Let me be clear about what happened. Apple didn't remove attribution. They removed a single identifier that made attribution trivial. Before 14.5, you could identify a device, match it across apps and ad networks, and have high confidence about where an install came from. It was deterministic. One signal, one answer.

The ATT prompt gave developers the option to ask users for permission to track them. Some users said yes. Most said no. Those who said yes kept that deterministic path. Those who didn't? Developers were left with probabilistic methods, inference, and a lot of guesswork. The problem was that most attribution providers made that guesswork expensive and only marketed it to enterprise teams.

For indie developers and small studios, the choice looked like: spend thousands a month on AppsFlyer or Adjust and still get probabilistic data, or give up on attribution entirely and hope your gut feelings about marketing were right.

Why works without ATT permission actually matters

Here's the thing that took us a while to understand. When people said 'we need attribution that works without ATT', they didn't mean they wanted a magic solution that magically knew everything. They meant they wanted a practical answer to three concrete questions: where did each install come from, are those users still around at day 7, 14, and 30, and if they came through a community promoter, who was it?

Those three questions are answerable without IDFA. Not with perfect confidence, but with enough accuracy to actually make decisions.

We built Attribr around that constraint. The SDK is 50KB. It does two things well: deterministic matching for the users who give permission (via ATT), and probabilistic matching for everyone else. The probabilistic side uses time of install, app version, device characteristics, and network information to build a fingerprint. It's not foolproof. But it's good enough for indie developers to understand where installs are coming from without paying enterprise pricing.

The real unlock was that this approach meant we could keep the SDK lightweight. No heavy tracking libraries. No third-party dependencies. Just the essentials.

A moment with a developer who needed this

About six months in, a developer reached out. They'd built a puzzle game. Small team, tight budget. They were running marketing campaigns across TikTok, Facebook, and referral partnerships. They'd tried setting up one of the big platforms, spent two grand a month, and still couldn't reliably tell which channel was actually driving installs that lasted past day 7.

They integrated Attribr in three lines of code. Within a week, they had their answer: their referral partnership was bringing users with 40 percent day 30 retention, but their Facebook spend was pulling users with 18 percent. Their gut had been wrong. They rebalanced their budget, cut Facebook spend, doubled down on referrals, and their LTV per dollar spent went up.

They were never going to show up in Adjust's case studies. They spent maybe 200 quid a month on Attribr. But attribution fundamentally changed how they allocated marketing. That's what works without ATT permission meant to them. Not a replacement for enterprise platforms. A practical, affordable answer to the questions that actually matter to their business.

The retention piece nobody talks about

Knowing where an install came from is useful. Knowing whether that install is still opening your app on day 30 is everything.

We built day 7, 14, and 30 retention tracking into Attribr because attribution without retention tracking is just vanity. You can know that Facebook brought you a thousand installs, but if those users are gone by day 7, that number is meaningless.

The combination matters. Attribr tells you not just which channel brought users in, but which channel brought users who actually stick around. That's the piece that indie developers needed and that most attribution platforms buried behind premium tiers.

We also added a bridge to Rippl, our performance marketing community platform. If your install came through a Rippl promoter, Attribr connects those dots. You see the actual promoter behind the CPI install. For developers working with community-driven marketing, that's the final piece of the puzzle.

What this actually costs

I'm going to be direct here because I remember the frustration of indie budgets.

Attribr is free for your first thousand installs per month. That alone changes the math for small teams testing attribution for the first time. If you're doing two thousand to twenty-five thousand installs a month, it's £29. If you're at the high end for indie development, up to one hundred thousand installs, it's £99 and you also get fraud signals and the ability to plug in your own ad network data.

It's not a free product masquerading as one. It's genuinely affordable for the audience it was built for. And if you're larger than that, we talk to you directly about what you actually need.

The question that matters more than features

Everything we've built into Attribr sits on top of a simple principle: attribution shouldn't require you to either give up your data or your budget.

iOS 14.5 forced a choice on the industry. Big platforms chose complexity and premium pricing. We chose simplicity and transparency. You don't need ATT permission to know where your installs come from. You need the right approach and the right tool.