Multi-touch attribution (MTA)
Also known as: MTA
In short
An attribution approach that distributes conversion or revenue credit across multiple observed marketing interactions in a customer journey.
Share credit across the observed journey
Multi-touch attribution recognizes more than one interaction before an outcome. A model decides how much credit each eligible touchpoint receives within the selected scope and lookback period.
A linear model can divide credit equally. Other models use different rules or data-driven estimates. These are general model concepts, not a claim that every analytics platform offers every option. Google Analytics, for example, has retired several older rules-based models.
The result depends on the interactions the system can observe and connect. Missing devices, untracked channels or inconsistent identifiers change the journey available to the model.
Illustrative allocation
Imagine a fictional customer journey containing an email click, a search ad click and a later purchase worth $90. An illustrative linear allocation across those two touchpoints assigns $45 to email and $45 to paid search.
The store still received one $90 order. The model changes its allocation in the report; it does not create extra revenue or establish the customer’s reason for buying.
Separate allocation from causal impact
Attribution distributes credit for recorded outcomes. It does not directly measure what would have happened without advertising. Use controlled incrementality testing for that question. Compare models only when the conversion definition, reporting window and channel coverage are understood, and avoid adding separately attributed platform totals as if they were unique orders.