Data Driven Attribution (DDA)

As per my previous posts e.g here on multi click attribution; we should all be keen to understand the value that marketing channels have to in turn make better business decisions. Data Driven Attribution (DDA) can help enable this goal.

For some context in GA4 (google analytics 4) the other 4 key attribution models (first click, linear, time decay, and position-based) have in Q2 2023 been removed (reference) leaving only 2 models remaining (last click and DDA). As a result I believe you will see many more companies adopting DDA over the coming months (particularly those who rely on Google Ads/SA360 and GA4).

First up; what is data driven attribution?

Data-driven attribution is where customer behaviour is analysed > attributed to identify patterns among those users who convert, compared to those who don’t.

Data-driven attribution looks at engagements and interactions across your channels and in Google ads accounts will build a a paid history to better understand how your ad campaigns work (DDA is specific to ‘you’) to attribute > optimise > drive conversions + sales.

Sounds perfect? No!…

While data-driven attribution in Google Ads and Analytics is less biased and helps distribute value between all touchpoints more fairly, it’s not perfect.

There are limitations around requiring a minimum number of conversions for DDA to run (at time of writing for G-Ads = 3,000 ad interactions and 300 conversions over 30 days), it relies on online data (i.e not omnichannel) and it’s a tad “black box” as to how credit (each touchpoint or channel) is exactly assigned.

In summary, and despite the limitations, I believe DDA will be the main attribution model for many businesses moving forward so my recommendation is to understand > utilise > test & learn.

Leave a Comment