Nederlands 
English Multi-touch attribution explained
Multi-touch attribution distributes the credit for a conversion across all the touchpoints that preceded it. That is fairer than last-click, but the distribution remains an assumption.
Multi-touch attribution is a method that distributes the value of a conversion across multiple digital touchpoints in the customer journey, according to a chosen allocation rule.
The common multi-touch models
Multi-touch attribution has several allocation rules. Linear gives each touchpoint equal value. Time-decay gives more weight to touchpoints closer to the conversion. Position-based places emphasis on the first and last contact.
There is also data-driven attribution, where an algorithm bases the distribution on patterns in conversion paths. That sounds objective, but the algorithm only sees what it measures, and that is almost always digital.
Each model gives a different outcome on the same data. The choice of model partly determines the conclusion.
Where multi-touch attribution runs into problems
Multi-touch is an improvement over last-click, but three limitations remain.
Digital touchpoints only
TV, radio and print leave no click and therefore fall outside the model.
The allocation rule is an assumption
Linear, time-decay or position-based: every rule is a choice, and every rule gives a different outcome.
Increasingly difficult in a cookieless world
Multi-touch relies on tracking users across channels. That becomes increasingly difficult without third-party cookies.
No incrementality
A touchpoint receives value because it was in the path, not because it caused the conversion.
When to look at MMM instead
Multi-touch attribution is useful for comparing digital campaigns against each other. For the question of how to allocate budget across online and offline combined, it falls short.
If you want all channels in one model, including TV and radio, and based on incrementality, then Marketing Mix Modeling is the appropriate method. See also the MMM versus multi-touch comparison.
Frequently asked questions
Is multi-touch attribution better than last-click?
For digital channels, yes. Last-click gives all the credit to the final contact and overvalues bottom-of-funnel channels. Multi-touch distributes value, though that distribution remains a chosen assumption.
Which multi-touch model is best?
There is no objectively best model. Linear, time-decay and position-based each give different outcomes. The choice depends on your funnel, and that dependency is precisely a weakness of the method.
Does multi-touch attribution measure TV and radio?
No. Multi-touch works on digital touchpoints with a measurable click or impression. Offline channels fall outside the model. Marketing Mix Modeling is needed for those.
Does multi-touch attribution work without cookies?
With difficulty. Multi-touch relies on tracking users across channels. As third-party cookies disappear, that becomes increasingly unreliable.
A model that includes all channels.
Book a demo and see how Datafy looks beyond the digital customer journey.