Google has now eliminated all but two attribution models in both Google Analytics and Google Ads: last-click and data-driven. This move has a significant impact on ecommerce companies using Google’s reporting tools. Such attribution models as first-click, linear, time decay, and position-based will no longer be available in these Google reporting platforms. This reflects a broader shift in marketing attribution as machine learning-based models replace the older “one size fits all” heuristic attribution models.
Google’s Data Driven Attribution is a machine learning model. Machine learning attribution offers the advantage of being much more reflective of the user journey to purchase that is occurring on ecommerce brands’ websites. Most rules-based models overvalue certain touch points while undervaluing others — for example, first-click disregards other touch points on the customer journey, while position-based overvalues first and last clicks while undervaluing intermediate touch points. Machine learning-based models on the other hand calculate a weighting for the touchpoints in user journeys to purchase based on the behavior of the site’s users. Because no two sites are the same ML models are reflective of what is actually transpiring on a site, as opposed to the older “rules-based” heuristic attribution models.
But this change is not without its challenges for ecommerce stores that have a wealth of experience and knowledge in the older models. The change also abandons models that had a useful purpose. First-click is useful in gaining an understanding of the ability of each channel or campaign to recruit first-time users and, if subsequent behavior is tracked in understanding each channel or campaign’s long-term value in contributing to subsequent purchases. Relying on Google means giving up that option.
This topic was explored in depth in an article in Retail TouchPoints by Phil Dubois, AdAmplify’s CEO.
Read “How Google’s Approach to Attribution is Changing and Why It Matters for Ecommerce Companies“