Canyon's data-driven attribution model in Google Analytics was giving too much credit to direct traffic. So racing bike company Canyon sought to create a custom attribution model that was better suited to the Canyon customer journey.
With Canyon products generally being high-value and high-consideration purchases, we introduced a 90-day lookback window that would be in line with the long conversion period. The model also downweighted direct traffic when it was the final step, based on the assumption that the customer had already made the decision before this last step.
The model reduced credit given to direct traffic by 38% when compared to the last non-direct click model, and distributed it across the other channels, giving Canyon a much more accurate ROI for all marketing activity, particularly display, and allowing the retailer to make better-informed marketing decisions.