Book a call

What is data-driven attribution (DDA)?

Analytics & measurement

By Paweł Skibiński, Growth Strategist, Paid SocialUpdated 3 November 2026

How does data-driven attribution work?

Attribution means giving credit for a sale to the ads and clicks on the path before it. A simple rule gives all credit to one touchpoint. DDA instead looks at many paths, including ones that ended without a purchase, and estimates how much each interaction mattered.

Two details are worth knowing.

  • It is built from your data. The model is fitted per advertiser and per key event, so two accounts can get different splits.
  • It sees Google's view of the path. It works on the touchpoints Google can observe.

In Google Ads, the models on offer are last click and data-driven, and data-driven is the default for most conversion actions2. First click, linear, time decay and position-based models are no longer supported2. See last-click attribution for the simpler rule.

Is data-driven attribution proof that ads work?

No. This is where people slip. DDA is a better rule for dividing credit among the touchpoints that were recorded. It does not tell you what would have happened without the ads.

Customers who were going to buy anyway still leave paths. DDA will give some touchpoint a share. Asking whether the ads caused the sale is a different question. It is answered by incrementality, usually through a test.

Compare it to multi-touch attribution: faster and more granular than a test, but the same limit applies.

For ZFD: We read DDA as a credit report from Google. For where it sits next to other methods, see MMM vs MTA vs incrementality tests.

Easily confused with

  • Data-driven attribution vs Incrementality

    The extra sales caused by marketing, over and above what would have happened anyway. DDA does not measure this.

  • Data-driven attribution vs Last-click attribution

    All credit to the last clicked channel. DDA splits it.

  • Data-driven attribution vs Multi-touch attribution

    Any rule that splits credit across touchpoints. DDA is one Google version of it.

  • Data-driven attribution vs "AI attribution"

    A marketing label. DDA is a statistical model with a name, not a verdict.

FAQ

A rider in a red cape and bicorne hat on a rearing yellow plush horse, after David's Napoleon Crossing the Alps

Is data-driven attribution better than last click?

It splits credit using your account's past data instead of a fixed rule2. Neither measures cause. Which suits a decision depends on the decision.

Does DDA show incremental sales?

No. It divides credit among touchpoints that were recorded. Incrementality needs a comparison with what would have happened without the ads.

Can I trust DDA numbers in Google Ads?

Trust them as Google's credit split. Check them against shop revenue and, where you can, a holdout test.

Sources

  1. Get started with attribution - Analytics HelpPrimary source
  2. About attribution models - Google Ads HelpPrimary source

Let’s find what’s actually working.

Next step. Check what your ROAS looks like on explicit assumptions. Use the Is your ROAS real? calculator.

Is your ROAS real? calculator