In one sentence
Marketing attribution is how you decide which channels and touchpoints get credit for a sale, like the ad, email or AI answer that brought a shopper in, so you know where your next marketing dollar should go.
Common marketing attribution models
- Last click: all credit to the final touch before the order
- First click: all credit to the touch that introduced the shopper
- Linear: credit split evenly across every touch
- Data-driven: credit assigned by a model based on your own conversion paths
No model is true. Each is a way of sharing credit. Pick one, use it consistently, and check it against totals like MER.
Why AI answers are hard to attribute
A shopper can read an AI answer, remember your brand, and type your URL a day later. That visit looks like direct traffic, and the AI answer gets no credit.
Even clicked links don't always show where they came from. Some AI apps don't pass a referrer, so visits can land in direct traffic too. AI referral traffic only captures the visits that do.
How to attribute sales from AI search
- Track AI referrers as their own channel in your analytics.
- Ask "How did you hear about us?" at checkout and include AI assistants as an option.
- Watch branded search and direct traffic alongside changes in your AI visibility.
- Tie orders back to the questions and pages that sent the shopper, where you can.
Marketing attribution FAQs
Which attribution model should a small store use?
Start with what your analytics tool gives you, often last click or data-driven, and add a post-purchase survey. The survey catches the channels clicks miss.
Can I use UTM parameters for AI traffic?
Only on links you control, like ones you share in posts or emails. AI assistants write their own links, so UTM parameters won't tag most AI visits.
Why do my ad platforms claim more sales than I made?
Each platform counts sales its ads touched, so one order can be claimed several times. Compare their totals to your real revenue.
Related terms