WATCH A DEMOGET STARTED
Platform overview
Documentation
Updates
Predictive Bidding AlgorithmsAudience SegmentationBudget PlannerMarketing Mix ModelingMulti-Touch AttributionIncrementality TestingBusiness Reporting
CustomersPricing
BlogPodcast
AboutNewsroom
Log in
← Back to Bonsai University

What is multi-touch attribution?

Who closed that deal — your LinkedIn ad, or the email that came after? If you're crediting just one, you're lying to yourself.

The problem with single-touch

Most marketers use single-touch attribution. First-touch says: "The billboard gets all the glory." Last-touch says: "The sales call seals it — nothing else matters." Both are like giving the Oscar to only the lead actor and ignoring the director, the writer, and the gaffer who kept the lights on.

Multi-touch attribution fixes this. It assigns credit to every touchpoint in the customer journey — the ad that sparked curiosity, the webinar that built trust, the case study that handled objections. Everyone gets their share of the trophy.

The models

There's no one-size-fits-all. Linear attribution splits credit equally — democratic, but naive. Time-decay gives more weight to touchpoints closer to conversion — recency bias, but often realistic.

U-shaped models favor the first and last touch while acknowledging the middle. And data-driven models? They let machine learning do the math.

The right model depends on your sales cycle. Selling enterprise software with a six-month journey? That's different than selling sneakers with a three-day impulse window.

Example

Picture this: A CFO sees your display ad on Monday. Ignores it. Googles you Wednesday after a colleague mentions your product. Downloads your ROI calculator Thursday. Gets retargeted Friday. Finally books a demo the following Tuesday after opening your nurture email.

Last-touch attribution? The email gets 100% credit. Multi-touch? You see the actual assist chain — and realize that display ad started everything.

Takeaway

Multi-touch attribution shows you which marketing actually moves the needle — so you stop guessing and start optimizing.