Google to sunset 4 attribution models in Ads and Analytics

Google will soon retire four rules-based attribution models in Google Ads and Google Analytics:

• First-click.
• Linear.
• Time decay.
• Position-based.

Why we care. This will impact anyone who still uses these models in Google Ads. Anything that isn’t last-click will be much harder to track because the data-driven attribution formula is different for each advertiser – and not visible.

• “In the past, you could use linear and give each touch the same credit. You will no longer be able to do so. There will be no impact on last touch, as that is still available. But finding some of the first-touch information will be much muddier because there will no longer be a way to see the formulas that compute the attribution scores,” said Greg Finn, director of marketing for Cypress North.

Why the change. These attribution models will be sunset due to “increasingly low adoption rates, with fewer than 3% of conversions in Google Ads using these models” according to a Google spokesperson.

• Switching to this model typically results in a 6% increase in conversions for advertisers, Google said.

Sunset timeline. The first-click, linear, time decay and position-based models will all sunset in Google Ads and Google Analytics beginning in September. Before that, for newly created conversion actions, these attribution models will no longer be available in:

• GA4 starting in May.
• Google Ads starting in June.

Data-driven attribution as the default attribution model. Google announced its shift to data-driven attribution in September 2021. For advertisers who didn’t want to use data-driven attribution, Google Ads made five rule-based attributions available – last click, first click, linear, time decay and position-based.

Dig deeper. Data-driven attribution: How to think about Google’s default attribution model

The post Google to sunset 4 attribution models in Ads and Analytics appeared first on Search Engine Land.

https://ift.tt/KDfa4HL
from Search Engine Land https://ift.tt/5bvzoIp
via IFTTT