BlogOur Latest News

Is Last Touch Attribution TRULY Dead?



A quick search on Google will show you hundreds of articles about the failings and death of Last Touch (Last Click) Attribution. Despite these claims, Last Touch Attribution is still the industry standard, so how can it be dead?

The discrepancy comes from two competing forces in the Marketing world - the top of the funnel vs the bottom of the funnel. Last Click Attribution favors only the bottom of the funnel, limiting the accuracy of the collected data. The use of Fractional Attribution models spreads credit across all the touch points in the funnel.

Last Click Attribution is the most favorable option for any advertising company such as Google Adwords or Facebook Ads because it allows them to artificially inflate their ROAS (Return On Ad Spend). Online purchasers behave in a seemingly crazy manner - in an effort to get the best deal, a customer can run multiple product searches, open multiple tabs, search for coupons, visit multiple sites, etc. This history can be difficult to sift through, so for channels in the bottom of the funnel it's easiest to simply take credit for the entire event.

Marketers focusing on SEO, remarketing, offline campaigns and other indirect channels need Fractional Attribution models to understand how their efforts have contributed to brand growth. For e-commerce, Positional Attribution is also effective because it relies on a u-curve model - weighting credit more heavily on the first and last touch, while giving proportional credit to the remaining channels in the middle. For lead generation, the Time Decay Attribution model can be used to capture the credit due for each touch that builds customer trust with your business and ultimately leads to a conversion.

The goal at oobyMetrics is to remove the inherent biases from attribution modeling by helping businesses find the perfect model(s) to assist in building brand growth and channel equity. Helping each of our clients make their Marketing dollars go farther is our business.