Skip to main content

Attribution Models


Digital advertising terms don't have to be overwhelming. Check out our digital glossary to quickly look-up important definitions.

Need Help?

For more information, resources, or support, please feel free to reach out to our team at

In advertising, there are often numerous sources and impressions that contribute to a desired action. Here's how marketers assign the importance of each touch point, leading up to a conversion.  

What is Attribution and How Do We Measure It?

Attribution is the process of crediting the appropriate marketing channels for the role they played in a subsequent conversion. There are several models marketers use for attribution:

Attribution Models Digital Marketing Guide

How to Choose the Right Attribution Model

To choose the right attribution model, you should consider factors like time of year, product category, amount spent, customer loyalty, initial and final touch-point channels, and length of conversion path.

The most recent method for qualifying attribution is called Fractional Attribution though a specialized vendor is required to measure this way. Consumers interact with duplicate ads throughout the conversion cycle, and fractional attribution removes repetitive impressions from the attribution model to allocate value as accurately as possible to each media partner and ad impression.

For example, if the conversion was around the holiday season for a consumer electronic item less than $200, which was promoted for a very limited time, a linear attribution model might make sense. However, for a major life decision like the purchasing of a new car, it might make more sense to use fractional attribution, which addresses a longer conversion cycle.

What does the industry think?

Many marketers believe the most holistic view of attribution is the most accurate, giving a precise weight to every touchpoint before conversion. This model  in concept is called the "Full Funnel Attribution", but to date there is no way to precisely quantify the weight of each impression. Fractional attribution seems to be the closest model available today, but also requires resources and analysis well beyond that required for traditional attribution methods.