Data-Driven Attribution

Data-driven attribution is GA4's default way of deciding which channels get credit for a conversion. Instead of a fixed rule like "give it all to the last click," it uses machine learning on your own conversion data to estimate how much each touchpoint actually contributed. Channels that genuinely move people toward converting earn more credit; ones that don't earn less.

Why it matters

Fixed-rule models guess. Data-driven attribution measures — at least within the limits of your own history. It can hand a channel one-and-a-half conversions or a third of one, because credit is shared proportionally rather than dumped on a single touch. For most small sites it's the most honest default, since it adapts to how your specific customers actually behave.

A concrete example

You compare paid search under last-click and under data-driven attribution. Last-click gives it 12 conversions. Data-driven gives it 18.5 — because the model sees paid search reliably starting journeys that other channels finish. The extra 6.5 conversions are credit the last-click view was quietly handing to whichever channel happened to close.

The common misreading

People expect data-driven numbers to be whole and to match last-click. They won't. Fractional conversions are normal, and the totals will look different from any single-touch model. A channel's count dropping after a switch to data-driven doesn't mean it got worse — it means credit is being shared more accurately.