What's Your Experience with the Data Driven Attribution Model?

Data Driven Attribution Model in Google AdWords
Data Driven Attribution Model in Google AdWords

Hey Roy, have you had any experience working with the Data Driven Attribution Model in Google AdWords? 

Do you have any recommendations on when you'd use it?


Nathan, happy to share my perspective on the Data Driven Attribution Model.  Google defines the Data Driven Attribution Model as follows:

Data-driven attribution gives credit for conversions based on how people search for your business and decide to become your customers. It uses data from your account to determine which ads, keywords, and campaigns have the greatest impact on your business goals. You can use data-driven attribution for website and Google Analytics conversions from Search Network campaigns.

You can learn more about the Data Driven Attribution Model by visiting Google's support site.

The Data-Driven is currently a black-box, and I don't trust it.  Attribution model and performance are chicken and egg, as the model becomes a self-fulfilling prophecy.  If you picked First Click, then you'd invest more into high-funnel campaigns, and then high-funnel campaigns would perform "better", for example.  If they're claiming that they can find the combination of model variables that increases the total at the bottom of the funnel, then that'd be cool... but the Conversion Optimizer, eCPC, Ad Optimization, and all of the other "magic black box" features in AdWords generally underperform (or outright fail) for mid-size retail accounts, as far as I've found.

I find that most of the time when Google launches some automated data analysis system that the first 6 months to 1 year are not very good and the early adopters don’t do well unless they have so much data the system can learn very quickly.  It seems that it really takes 1-2 years to work the bugs out to where it generally works.

So, let's put it in the "wait and see" category, but I'm very skeptical.  Till I see evidence suggesting I'm wrong (which happens from time to time), I'm going to stick with Time Decay.

Thanks for the question Nathan.  I love to talk shop so don't hesitate to reach and drop me a line.

Up and to the right,