I paid an influencer £150 for a post about my app last year. They had 45,000 followers. The post got 3,200 likes and about 80 comments. My download numbers that day? Exactly the same as every other day.
I had no way to prove the post did nothing. I also had no way to prove it did something. I just had a feeling and a lighter wallet.
That's the problem with influencer marketing when you don't track it properly. You're left guessing, and guessing is expensive when you're bootstrapped.
The metrics that don't matter (as much as you think)
Follower count tells you reach potential, not results. A creator with 8,000 engaged followers in your niche will almost always outperform one with 200,000 followers who covers everything.
Likes are even less useful. People like posts for all sorts of reasons. I've liked posts about products I'd never buy. You probably have too.
Comments are slightly better — at least someone cared enough to type something — but most comments on sponsored posts are other creators tagging friends or saying "need this!" without ever actually clicking through.
Impressions and views tell you the content was shown to people. That's it. Being shown something and buying it are very different actions.
None of these metrics are worthless. They give you a rough sense of whether the content landed. But they can't answer the question that actually matters: did this person sell anything?
What actually matters: attributed purchases
The only metric that tells you whether an influencer campaign made money is purchases you can trace back to that specific campaign.
Not downloads. Not sign-ups. Not "engagement." Purchases.
Downloads are a step in the right direction, but a download that never converts to a paid user is worth £0. I've seen campaigns that drove 200 downloads and zero purchases. Without purchase tracking, those 200 downloads look like a success.
Here's the basic maths. Say you pay an influencer £200 and they drive 15 purchases at £4.99 each. After Apple's 15% cut (if you're on the small business programme), that's about £63.60 in revenue. Your ROI is negative — you lost about £136.
Now say another influencer charges £50 and drives 8 purchases. That's £33.90 in revenue. Still negative, but you lost £16 instead of £136. And if those 8 customers have a lifetime value higher than a single purchase — if they renew or upgrade later — that second deal might actually be profitable over time.
You can't make these calculations without knowing which influencer drove which purchases.
How to actually track it
The setup is simpler than most people expect.
Step 1: Give each influencer a unique link. Not your regular App Store link. A tracked link that routes through an attribution tool first. When someone clicks that link, the tool records which link they used before sending them to the App Store.
Step 2: Connect your purchase data. When that person later makes a purchase in your app, the purchase needs to be matched back to the click. Tools like LinkOwl do this through RevenueCat or Superwall webhooks — when a purchase happens, the webhook fires and the attribution tool checks if that user came from a tracked link.
Step 3: Check the numbers. Your dashboard now shows: Influencer A drove 15 clicks, 8 installs, 3 purchases. Influencer B drove 42 clicks, 20 installs, 0 purchases. Decision made.
The whole setup takes about 20 minutes if your app already uses RevenueCat or Superwall. If it doesn't, add maybe an hour.
Calculating ROI properly
The formula is straightforward:
ROI = (Revenue from attributed purchases - Cost of influencer) / Cost of influencer × 100
So if you paid £100 and made £60 in attributed revenue, your ROI is -40%. You lost money.
But there are complications worth thinking about.
Lifetime value matters more than first purchase. If your app charges £4.99/month and the average subscriber stays for 6 months, each converted user is worth about £25 after Apple's cut. A campaign that looks unprofitable on day one might be profitable by month three.
Attribution windows affect your numbers. Someone might click a link on Monday, install the app on Wednesday, and buy on Friday. If your attribution window is 24 hours, that purchase gets counted as organic. Most tools default to 7 or 14 days. Longer windows catch more purchases but might over-attribute — someone who clicked a link two weeks ago and then found your app independently through search probably wasn't really influenced by that link.
Not every purchase gets attributed. Some people will see an influencer's post, search for your app directly in the App Store, and buy without ever clicking your tracked link. You'll never attribute that sale to the influencer, but they did contribute to it. This is the blind spot in all last-click attribution. Accept that your numbers undercount slightly.
What good ROI looks like
There's no universal benchmark because it depends entirely on your margins and customer lifetime value.
For a one-time purchase app at £4.99 (keeping about £4.24 after Apple's cut), paying more than £4 per acquired customer means you're losing money forever. That's tight. Most influencer campaigns won't be profitable here unless the influencer is very cheap or very effective.
For subscription apps, the maths changes. If your average subscriber pays for 4 months at £4.99 (about £17 net), you can afford to pay up to £17 to acquire them and still break even. Suddenly that £150 influencer who drives 10 subscribers is profitable.
A rough guide: if your influencer cost per attributed purchase is less than your average customer lifetime value, you're making money. If it's more, you're not.
The comparison spreadsheet
After a few campaigns, you'll want to compare influencers side by side. The columns that matter:
- Influencer name
- Amount paid
- Clicks on their link
- Installs attributed
- Purchases attributed
- Revenue from those purchases
- Cost per purchase (amount paid ÷ purchases)
- ROI percentage
Sort by cost per purchase. The influencers at the top of that list are the ones you rebook. The ones at the bottom are the ones you don't.
I keep mine in a simple spreadsheet. Nothing fancy. The attribution tool provides the purchase data; I just add what I paid.
Common mistakes
Judging too early. Give a campaign at least 7-14 days before drawing conclusions. Some audiences take time to convert.
Ignoring small creators. A nano-influencer with 2,000 followers who drives 5 purchases at £20 total cost has better unit economics than a big creator with 100,000 followers who drives 10 purchases at £500. Run the maths, not the vibes.
Not having tracking in place before the campaign starts. I've done this. Paid an influencer, they posted, I scrambled to set up tracking, and by the time it was live I'd already missed the initial wave of clicks. Set everything up before they post.
Treating all influencer costs the same. A £50 gifted product has different economics than a £500 flat fee. Factor in everything you spend — product costs, shipping, agency fees, your time setting up the campaign.
Getting started
If you're running your first influencer campaign, the minimum viable tracking setup is:
- Create a tracked link for the influencer (through LinkOwl or whatever attribution tool you use)
- Make sure your app sends purchase events to RevenueCat or a similar tool
- Connect the webhook so purchases get attributed
- Give the influencer their unique link
- Wait, then check the numbers
You don't need a complicated dashboard or a degree in analytics. You need one number: how many purchases did this person drive? Everything else follows from that.
The influencers who actually sell things will become obvious very quickly. And the ones who don't — well, at least you'll know, instead of guessing.