What Is Multi-Touch Attribution?
Multi-touch attribution splits credit for a conversion across the touchpoints that led to it. How it works, the common models, and why privacy changes have limited what it can see.
Multi-touch attribution (MTA) assigns credit for a conversion to the touchpoints that came before it: the ads someone saw or clicked, the emails they opened, the pages they visited. Instead of handing all the credit to one interaction, it spreads it across the journey.
For years, MTA was the main alternative to last-click reporting. Understanding what it does well, and what it can't do, helps explain why measurement has moved on.
From single touch to multi-touch
Single-touch attribution gives 100 percent of the credit to one interaction:
- Last click credits the final touchpoint before the conversion, often a search ad.
- First click credits the touchpoint that started the journey.
Both are simple and both are misleading, because almost nobody buys after a single interaction. Last click in particular rewards the channels that close sales and ignores the ones that create interest.
Multi-touch attribution fixes the most obvious part of that problem by recognizing that several touchpoints played a role.
How multi-touch attribution models divide credit
The common models differ in how they split the credit:
- Linear: every touchpoint gets equal credit.
- Time decay: touchpoints closer to the conversion get more credit.
- Position-based (U-shaped): the first and last touchpoints get most of the credit, and the rest is shared among those in between.
- Data-driven or algorithmic: a statistical model compares journeys that converted with ones that didn't and assigns credit accordingly.
Whichever model you choose, the output is the same kind of answer: a share of credit for each channel or campaign, based on the journeys the system could see.
What multi-touch attribution can't see
That last part is the catch. MTA can only credit what it can track, and there are three big gaps.
Channels people don't click. Connected TV, linear TV, radio, podcasts and out-of-home rarely leave a trackable click. MTA largely misses them, even when they're driving demand.
Journeys it can no longer follow. MTA depends on recognizing the same person across sites and devices. Apple's App Tracking Transparency, browser tracking protections and the decline of third-party cookies have cut deeply into that. Journeys that used to be connected now look like separate, anonymous visits.
Cause and effect. Even with perfect tracking, attribution answers "which touchpoints were present?" not "which touchpoints caused the sale?" Someone who was always going to buy may click a brand search ad on the way. Attribution gives that ad credit. It didn't create the sale.
Where attribution fits today
MTA is still useful for tactical questions within channels you can track well, such as comparing creative or keywords inside one platform. But it is no longer a reliable basis for deciding how to split a budget across channels.
For that, marketers are turning to approaches built on aggregate data: marketing mix modeling, which estimates each channel's contribution to total sales, and incrementality measurement, which estimates the sales each channel actually added. Both include channels people don't click, and neither depends on following individuals around the web.
The question has shifted from "who touched this customer?" to "what actually made a difference?" Attribution answers the first. Budgets depend on the second.
Frequently asked questions
What is the difference between single-touch and multi-touch attribution?
Single-touch attribution gives all the credit for a conversion to one touchpoint, usually the first or the last. Multi-touch attribution splits the credit across several touchpoints along the way.
What are the common multi-touch attribution models?
Linear (equal credit to every touchpoint), time decay (more credit to recent touchpoints), position-based or U-shaped (most credit to the first and last touchpoints), and data-driven or algorithmic models that assign credit statistically.
Does multi-touch attribution still work without third-party cookies?
Only partly. It still works inside platforms and on your own site where you can identify users, but it loses sight of many cross-site and cross-device journeys, so its picture of the customer journey is less complete than it used to be.
Is multi-touch attribution the same as incrementality?
No. Attribution divides credit between touchpoints that were present before a conversion. Incrementality estimates how many conversions would not have happened without a channel. A channel can collect plenty of attributed credit while adding few incremental sales.
What should I use instead of, or alongside, multi-touch attribution?
Many marketers now pair or replace it with marketing mix modeling and incrementality measurement, which work from aggregate data, include channels people don't click, and estimate what actually caused sales.