Attribution & Conversion Counting

Data-Driven Attribution vs Last Non-Direct Click

In short: Both are live inside GA4 today, but they answer different questions for different reports. Data-driven attribution (DDA) models fractional credit across a full conversion path using your account's converting and non-converting data, and powers GA4's dedicated Attribution reporting plus Google Ads conversion counting. Last non-direct click hands 100 percent of the credit to a single touch, skipping direct sessions, and powers GA4's Traffic acquisition report and default channel grouping. They can disagree on the same conversion because they are not trying to do the same job, one spreads partial credit across a modeled path, the other picks one winner by a fixed rule. If you are optimizing bids or budget allocation, trust DDA; if you are reading a quick channel-mix report, you are already looking at last non-direct click.

By the AdFlint research team · Fact-checked against current Google and Meta platform behavior · Last reviewed July 2026

Data-Driven Attribution

Uses a platform's own modeling of converting and non-converting paths to distribute fractional credit based on observed contribution rather than fixed rules.

The platform compares paths that converted with paths that did not, then assigns each touchpoint a fractional share reflecting its measured contribution. It is the default in Google Ads and GA4 and needs enough conversion volume before the model has data to learn from. Two things get misread: the credit is modeled rather than observed, and it remains in-platform, so it cannot see channels outside that platform's data.

Full definition

Last Non-Direct Click

Ignores direct sessions when assigning credit, handing the conversion to the last identifiable marketing source that preceded the visit.

When someone types your URL or arrives with no referrer, that session is skipped and credit passes back to the previous campaign, email, or organic visit. Analytics tools default to this because direct traffic is usually a returning user rather than a discovery channel. It shows up in GA4's cross-channel models. The catch: if the entire path is direct, direct keeps the conversion, and untagged campaigns get absorbed into it.

Full definition

Side by side.

The differences that actually change what happens in your account.

	Data-Driven Attribution	Last Non-Direct Click
How many touches get credit	Potentially every touch in the path, weighted by modeled contribution.	One, the most recent identifiable non-direct touch.
Where each one is used inside GA4	GA4's dedicated Attribution reporting section, and the basis for Google Ads conversion counting.	GA4's Traffic acquisition report, and default channel grouping.
Basis for the credit	Statistical comparison of converting versus non-converting paths.	A fixed rule: most recent non-direct touch wins, direct sessions are skipped.
Sensitivity to conversion volume	Needs sufficient volume to model reliably; thin data produces a less trustworthy result.	None, the rule applies identically regardless of volume.
Treatment of direct traffic	Direct sessions are just another data point the model can weigh, no special skip rule.	Explicitly skipped, credit rolls back to the last identifiable marketing touch before it.
Stability of results	Can shift as the model retrains on new data.	Never changes, it is a fixed rule applied consistently.
What it costs you when wrong	Low-volume accounts get a model built on thin evidence that can look more precise than it is.	A fully direct path leaves credit stuck on direct traffic, and untagged campaigns get folded into it.

What actually separates them.

DDA distributes fractional credit across a modeled path; last non-direct click assigns all credit to a single touch chosen by a fixed skip-direct rule.

DDA needs a conversion volume threshold to produce a reliable model; last non-direct click has no such requirement and runs the same regardless of scale.

DDA has no special handling for direct sessions, they are just another data point; last non-direct click explicitly excludes direct sessions from being credited.

The two live in different parts of GA4, DDA in Attribution reporting and Google Ads conversion counting, last non-direct click in the Traffic acquisition report, so the same conversion can be read two different ways depending on which report you open.

DDA's output can shift over time as the model retrains on new data; last non-direct click's output for a given path never changes because it is a fixed rule, not a trained model.

Which one should you use?

Use Data-Driven Attribution when

You are setting or reviewing Smart Bidding targets and want the conversion credit feeding the bidder to reflect modeled contribution, not a single-touch rule.
You have enough conversion volume for the model to be trustworthy.
You are trying to understand which upstream touchpoints actually correlate with conversions, not just which one happened last.
You are reading GA4's Attribution reporting section specifically, where this is the model in use.

Use Last Non-Direct Click when

You are reading GA4's Traffic acquisition report and need to understand what is already being counted there.
You want a fast, single-touch answer without setting up or trusting a modeled result.
You specifically care about excluding direct sessions from getting credit that belongs to an earlier marketing touch.
Your conversion volume is too low for DDA to model reliably, and you need a rule-based fallback.

Common questions.

Why does the same conversion show a different channel in GA4's Traffic acquisition report versus the Attribution report?

They are built on different models by design. The Traffic acquisition report uses last non-direct click, crediting one touch and skipping direct sessions, while GA4's Attribution reporting section uses data-driven attribution, spreading modeled credit across the full path. Neither report is wrong, they are answering different questions about the same conversion.

Which model does Google Ads actually use to count my conversions?

Google Ads conversion counting and optimization run on data-driven attribution, not last non-direct click. Last non-direct click is specifically a GA4 reporting convention for its Traffic acquisition report, not something Google Ads applies to its own conversion actions.

My account has low conversion volume. Should I trust data-driven attribution or fall back to last non-direct click?

If your monthly conversion volume is too thin for DDA to have meaningful data to learn from, its output can look confident while resting on very little evidence, so a transparent rule like last non-direct click is a reasonable, honest fallback until volume grows. Once volume is sufficient, DDA's modeled credit is generally the more useful signal for optimization decisions.

Or stop choosing between them.

AdFlint picks the setting, writes the ads, and keeps optimizing inside the Google and Meta accounts you already own.

Try AdFlint free See how it works

Related comparisons

All comparisons