You're spending money on marketing. Maybe you're running Instagram ads, sending emails, posting on TikTok, paying an influencer, or all of the above. Sales are coming in. The question you can't answer: which of those things is actually working?
That's what attribution solves. It connects a sale back to the marketing activity that caused it.
Why it matters more than you think
Without attribution, you're allocating budget based on feel. You spent £500 on Instagram ads last month and revenue went up, so you assume Instagram is working. But was it? Maybe those sales came from the email campaign you ran the same week. Maybe they came from a TikTok post that went mildly viral three weeks ago. Maybe someone found you on Google and the ads had nothing to do with it.
When you know which channels drive actual purchases, you can stop wasting money on the ones that don't. That sounds obvious, but most small businesses and indie developers don't have attribution set up at all. They're guessing.
How it actually works
Attribution boils down to two steps:
Step 1: Tag your marketing. Every link you share gets a unique identifier. Your Instagram bio link is different from your TikTok link, which is different from your email link. When someone clicks one of these, the system records which link they used.
Step 2: Match the click to a purchase. When that person buys something, the system looks back and says "this person clicked the Instagram link three days ago." The purchase is attributed to Instagram.
That's it at the core. The complexity comes from edge cases — what if someone clicked two different links? What if they clicked a link on their phone but bought on their laptop? What if they saw your ad but didn't click anything? These are real problems, but you don't need to solve all of them to get useful data.
The different models
Attribution gets complicated when you try to credit multiple touchpoints. If someone saw your Instagram ad, then clicked a link in your email, then bought — who gets credit?
Last-click attribution gives all credit to the last thing they clicked before buying. In this example, email gets 100% credit. This is the simplest model and what most small businesses should start with. It tells you which channel closed the deal.
First-click attribution gives all credit to the first touchpoint. Instagram gets 100% because that's where the person first discovered you. This is useful for understanding which channels bring in new audiences.
Linear attribution splits credit equally across all touchpoints. Instagram gets 50%, email gets 50%. This feels fair but can make every channel look mediocre.
Time-decay attribution gives more credit to touchpoints closer to the purchase. Email might get 70% and Instagram 30% because the email was more recent.
For most small brands and app developers, last-click is the right starting point. You can always get fancier later, but last-click gives you actionable data immediately: which link did they click before they bought?
What you actually need to set up
You don't need an enterprise platform. You need three things:
1. Tracked links. Each marketing channel gets its own link. When someone clicks it, you know where they came from. You can use UTM parameters on your own website, or a dedicated link tracking tool.
2. Purchase tracking. Your payment system (Stripe, RevenueCat, Shopify, whatever) needs to record purchases in a way that can be matched back to clicks.
3. Something connecting them. This is the attribution tool itself — the thing that says "click from Link A → purchase by User B → attributed to Instagram."
For website-based businesses, Google Analytics does a basic version of this with UTM parameters. For mobile apps, it's harder because the App Store sits between your marketing link and the actual purchase. That's why mobile attribution tools exist.
The cost problem
The mobile attribution industry was built for large companies. AppsFlyer, Adjust, and Branch are genuinely good tools, but they're priced for teams spending thousands on ads. If you're a solo developer or a small brand doing a few hundred in monthly revenue, paying £500/month for attribution doesn't make sense.
This is why a lot of small businesses skip attribution entirely. The tools that do it well cost too much, and the free alternatives (manual UTM tracking in spreadsheets) are too tedious to maintain.
There's a middle ground. Tools like LinkOwl charge per attributed sale (5p each) instead of a flat monthly fee. If you make 50 sales, you pay £2.50. If you make zero, you pay nothing. The cost scales with your revenue instead of eating into it.
Common mistakes
Tracking clicks but not purchases. Knowing that 200 people clicked your Instagram link is interesting. Knowing that 3 of those 200 actually bought something is useful. Click data without purchase data is vanity metrics.
Attributing everything to the last touchpoint. If you only look at last-click, you might undervalue channels that introduce people to your brand but don't directly close sales. Instagram might send someone to your site today, but they come back via Google three days later and buy. Google gets the last-click credit, but Instagram did the real work. Keep this in mind, even if you use last-click as your primary model.
Over-engineering it. You don't need a multi-touch attribution model with machine learning when you have 50 customers a month. Start simple. One link per channel. Last-click attribution. You can add complexity when your marketing spend justifies it.
Ignoring it because it's "too hard." Setting up basic attribution takes about 20 minutes. Create tracked links for each channel, connect your purchase system, and check the data once a week. That's it. The hardest part is starting.
Getting started today
Pick your two or three main marketing channels. Create a unique tracked link for each one. Share those links instead of your direct URL. Connect your payment system to record which link drove each purchase.
After one week, check the data. You'll immediately see which channel drives purchases and which drives clicks that go nowhere. That single insight is worth more than most marketing courses.
Attribution isn't about having perfect data. It's about having better data than guessing.