I've watched this play out dozens of times. Someone builds an app, gets a few hundred downloads, starts experimenting with marketing. TikTok posts, Instagram stories, maybe a Reddit thread or two. Sales trickle in. Ten a week, fifteen on a good week.
Then comes the question: which of those channels is actually working?
The obvious answer is to use an attribution tool. The less obvious answer is that most attribution tools cost more per month than the app is earning.
The maths don't work
AppsFlyer's cheapest plan runs around $500/month. Adjust is similar. Branch has a free tier, but the features you actually need (the ones that tell you which link led to a purchase) sit behind enterprise pricing.
If your app makes £300 a month and you're spending £200 tracking where that revenue comes from, you've lost before you've started. The tool that's supposed to help you grow is eating your margin.
This is the core problem with subscription-based attribution for small apps. The pricing is designed for companies spending five or six figures on advertising. It assumes you've already figured out product-market fit and you're optimising at scale. It assumes wrong.
What bootstrapped developers actually need
Most indie devs don't need multi-touch attribution models or incrementality testing. They need to answer one question: did that TikTok post I made on Tuesday drive any purchases?
That's it. A link, a click count, and a purchase count. Maybe broken down by source. Nothing more complicated than that.
The problem is that "simple attribution" doesn't really exist in the traditional tool market. You either get Google Analytics (which can't track app purchases properly) or you get a full mobile measurement partner that's overkill for your situation.
There's a gap between free-but-useless and enterprise-but-expensive. That's where most small app developers sit, staring at their App Store Connect dashboard trying to guess whether their revenue came from organic search or that Instagram reel they posted last week.
Pay when you earn, not before
The pay-per-sale model flips the economics. Instead of a flat monthly fee, you pay a few pence per attributed purchase. No purchases, no charge.
Here's why that matters when you're bootstrapped:
Month one. You're testing. You create a few tracked links, share them on different platforms. Maybe you get 8 attributed purchases. Your attribution cost is 40p. You spent 40p to learn that Reddit drove 5 of those sales and TikTok drove 3. That's useful information for less than the price of a stamp.
Month six. You've figured out what works. Reddit and TikTok are your channels. You're doing 200 purchases a month. Attribution costs you £10. You're making enough that £10 is noise, and the data you're getting back is actively helping you double down on what works.
Month twelve. You've scaled to 2,000 purchases. Attribution costs £100. But you're making thousands in revenue, and you know exactly which channels and which creators are driving it.
The cost grows with the revenue. It never outpaces it. That's the whole point.
What this looks like in practice
With LinkOwl, setup takes about ten minutes. You create tracked links for each marketing channel or campaign. When someone clicks one and later makes a purchase, that sale gets attributed back to the link.
Say you're promoting your app on three channels:
- A TikTok account where you post weekly
- A newsletter where you mention the app occasionally
- An influencer who agreed to try it
You create three links. Each one redirects to your App Store listing but carries attribution data. When purchases come through (via RevenueCat or Superwall webhooks), LinkOwl matches them to the original click.
At the end of the month, you can see that the influencer drove 45 purchases, TikTok drove 12, and the newsletter drove 3. Next month you negotiate a longer deal with the influencer and stop worrying about the newsletter.
No SDK integration required if you're already using RevenueCat. The webhook handles everything.
Compared to guessing
The alternative to attribution isn't "nothing." The alternative is guessing, and guessing has its own cost.
I've seen developers run influencer deals for months based on follower count alone. They pay a creator £500 for a post, see a vague uptick in downloads, and assume it worked. Maybe it did. Maybe the uptick was from an App Store feature that happened the same day. Without attribution, you can't tell.
That uncertainty compounds. You keep spending on channels that might not work. You ignore channels that might be carrying the weight. You make decisions based on correlation rather than causation, and your marketing budget (however small) leaks.
Even if attribution only saves you from one bad £200 influencer deal, it's paid for itself many times over at 5p per sale.
When free tools fall short
Google Analytics tells you about website traffic, not app purchases. App Store Connect shows downloads but not which marketing channel drove them. RevenueCat tracks subscriptions and purchases beautifully, but it has no idea whether that subscriber came from TikTok or a Google search.
Each tool sees one piece of the picture. None of them connect the marketing activity to the purchase. That connection is what attribution does, and it's the one thing bootstrapped developers are usually missing.
The bottom line
If you're making enough from your app that you'd like to make more, attribution is worth having. The question is whether you can afford it right now.
With pay-per-sale pricing, the answer is always yes. Five pence per attributed purchase means you start paying when you start earning. The tool costs nothing when it's not useful, and costs very little when it is.
LinkOwl charges £0.05 per attributed purchase with no monthly fee. Your first charge doesn't even hit until you've accumulated £5 in attribution fees (100 purchases). Before that, it's free.
If your app is making money, you should know where that money comes from. If it's not making money yet, you should know that too, before you spend more on the channels that aren't converting.