The complete guide to ad attribution models (and which one to use)

Attribution is the most misunderstood concept in digital advertising. It's also the one that determines whether you think your ads are working or failing. Here's the plain-English guide.

What is an attribution model?

An attribution model is the rule that decides which ad interaction gets credit for a conversion. If a customer sees your TikTok ad, clicks your Google ad three days later, then buys — who gets the credit? The model you choose answers that question, and it changes everything about how you read your data.

Last-click attribution (the default, and the most misleading)

Last-click gives 100% of the credit to the final interaction before purchase. It's the default on most platforms — including Google Ads and Meta Ads. The problem: it systematically undervalues the ads that introduced the customer to you. Your TikTok and Meta awareness campaigns will always look 'unprofitable' under last-click, even though they're generating the demand that Google later converts.

First-click attribution

Gives all credit to the first interaction. The opposite bias — it overvalues awareness campaigns and undervalues the closing ads. Useful if you're specifically trying to measure which channel generates new interest, but misleading for budget allocation.

Linear attribution

Splits credit equally across every touchpoint. A customer who saw 4 ads before buying gives each one 25% of the credit. Fairer than last-click, but it treats every interaction as equally important — a brand search at the bottom of the funnel gets the same credit as a random display ad impression.

Data-driven attribution (the gold standard)

Uses machine learning to weigh each touchpoint based on its actual contribution to the conversion. Google offers this in Google Ads (if you have enough conversion volume — 300+ per month). It's the most accurate, but it's also a black box — you can't always see why it assigned credit the way it did.

Which model should you use?

Use last-click for quick budget decisions — it's what every platform reports by default. But always pair it with a platform-agnostic view that shows every channel's spend and ROAS together. Without that, you'll keep over-funding the platforms that happen to be last in the funnel and starving the ones that are actually generating demand.

AdPulse shows every platform's spend, conversions and ROAS in one view — so you can see the full picture regardless of attribution model. Start free.