Google Analytics Attribution for the Assist

If you have read any of my past blog posts, you might notice that I can’t help but use a sports reference every once in a while.  My previous posts Google for the Assist and Turning Your Paid Search Campaigns into a Slam Dunk both look at the world of online marketing through the eyes of a sports fanatic. So, last week when I watched NBA Player Steve Nash take the spot for the third most assists in NBA history, I couldn’t help but relate it to this post about attribution I had planned.

Steve Nash is currently 40 years old, and in the NBA you won’t find many players that age still holding their own on the court. Although Nash’s points per season have been on the decline since the 2006-07 basketball season, his contracts keep getting renewed, and he’s still seen as a competitive threat on the court. The number one reason for this is his ability to assist.

The importance of the word “assist” is often overlooked in the world of digital marketing. Many people think that if a tactic is not how the user converted then it’s alright to give 100% of the value to the source with the last interaction, but that couldn’t be further from the truth.

Starting the attribution conversation is the first step in accurately measuring your ROI [Tweet this] and getting better insights into which channels actually work for you. Getting some initial insights into the value of attribution is as easy as looking at your Google Analytics data and comparing how the pre-set attribution models stack up for your site.

Find the Best Attribution Model

In the following chart we compare the Last Interaction, First Interaction, and Position Based attribution models. While many companies still use Last Interaction and Last Click conversion models, it’s extremely important to set up models that best fit your needs.  As you can see, Paid Search has nearly 10,000 less conversions in the Last Interaction model compared to the First Interaction model, while Direct is the opposite.

analytics_attribution

So what insights can we gain by looking at this data?  Paid Search is capturing a significant amount of “upper funnel” traffic that comes back later and converts through another channel. And even though Paid Search was a major part of capturing that traffic, the last click models give Paid Search no credit for thousands of users.

Set Goals Carefully

The type of model you use could significantly change the way you set your goals and/or look at performance. If a client’s cost-per-lead goal was $30, then in the example below, we are not hitting that goal using a last interaction or 40-20-40 Position-based model. Based on the results you see using whichever model you are using, it is important that you truly analyze and understand the data so you can make the most informed budget allocation and planning decisions.

analytics_attribution_2

In addition to the three models compared today there are several more preset models in Google Analytics and customizations that fit your needs and goals the best.  However, before you rush and choose an attribution model, take time to talk your team, department, and other departments to create a model that best fits everyone’s goals – for now and the future.

Noah J. Brooks

About Noah J. Brooks

Noah graduated with a marketing degree from Bowling Green State University in 2009. Currently serving as a Digital Strategist on Fathom’s EDU team, he specializes in the education industry. With clients ranging from small, local schools to national higher education providers he has gained invaluable knowledge to help clients increase leads and enrollments.With a sometimes unhealthy love for Excel, Tableau and data, Noah is constantly striving to come up with creative solutions to exceed his clients' goals. When he is not busy exploring the world of Digital Marketing Noah enjoys spending time with his wife, playing basketball, the occasional poker game, or his newest hobby of cycling.

Leave a Reply