Marketing Attribution Models

While marketing attribution has evolved significantly over the past five years, many brands and marketers still struggle in their approach to developing models for success. In this four-part blog series, our aim is to provide important context to performance measurement of digital marketing campaigns and a simple guide for helping marketers build more impactful attribution models. In Part One of this series, we outline the basics of marketing attribution and highlight prominent models that are more widely used today.

———-

Marketing Attribution Models

Marketing attribution is the process of assigning credit to specific marketing channels and tactics for the conversions they generate. It helps businesses understand which marketing efforts are driving the most value and making the biggest impact on their bottom line. There are several different attribution models that businesses can use to understand the role that each marketing channel plays in the conversion process. Some of the most common models include:

First Interaction

In this model, all value generated with the conversion is attributed to the first touchpoint of the conversion. This model is extreme and favors specific types of campaigns, but it works well in campaigns that work to raise awareness.

Linear Attribution

In this model, the value generated by the conversion is evenly distributed over all campaigns that played a role in the conversion funnel of the customer.

Time Decay

In this model, all touch-points get a share of the conversion value. In addition, the closer a customer is to the conversion moment, the higher the percentage value it will be assigned.

U-Shaped

In this model, set percentages are assigned to the first and last touchpoint, while the rest are evenly distributed.

Last non-direct Click

In this model, attribution is used to redistribute the conversion value generated by campaigns that run in a specific period. Direct visits in this model do not have any campaign assigned to them and are instead seen as the baseline segment. However, if a direct visit was their last means of visiting the website before a campaign was a touchpoint, all value should go to the last campaign touchpoint.

Data Driven Conversion Attribution Models

Data-driven conversions work with the data that visitors generate and apply models. Unlike Attribution models, these models use past data and math to show probabilities to marketers.

Logistic Regression Model

This model calculates the probability of conversion to occur and gives the outcome of the probability of conversion for each channel. The Logistic Regression Model is accurate and intuitive, but actionability is low for this model, and the predictive value of results is limited.

The Hidden Markov Model

This model suggests a limited number of mental states for the audience you are trying to reach toward a conversion. These states include dormant (unaware), awareness, consideration, and purchase. The Hidden Markov Model calculates the visitor’s probability of falling into those states and the probability that the visitor will shift to another state. The outcome view of this model is the total conversion assigned to each channel and a visualization of the number of times the campaign was responsible for moving visitors into a new conversion stage. The major drawback of this model is that the specific interactions from the user must be of value for this model to succeed. As well as this, a certain amount of bias should be expected with this model.

In Summary

There are pros and cons to each of these attribution models, and the right model for a business will depend on its specific goals and needs. It’s important for businesses to experiment with different models and see which one provides the most accurate and actionable insights. Overall, marketing attribution is an important tool for businesses looking to understand the ROI of their marketing efforts and optimize their strategies for maximum impact. By understanding which channels and tactics are driving the most value, businesses can make more informed decisions about where to allocate their resources and how to improve their marketing efforts.

To learn how to begin building a framework for your own attribution model selection, read Part 2 of of our blog series here.

Stay In Touch.

Subscribe to our monthly email newsletter.