Attribution & Conversion Counting

Data-Driven Attribution vs Last-Click Attribution

In short: Both are live, selectable attribution settings inside Google Ads today, which makes this one of the few comparisons in this family where you are actually choosing between two active options rather than reconstructing a retired one. Last-click is a fixed rule: the final click always gets 100 percent of the credit. Data-driven attribution compares converting and non-converting paths and distributes fractional credit based on what it observes each touchpoint actually contributing, and it needs enough conversion volume to have something to learn from. Data-driven attribution is now the default for conversion actions in Google Ads and the default reporting model in GA4; last-click remains available in Google Ads as the simple fallback when that volume is not there. Use data-driven attribution whenever your account has the conversion history to support it, and drop back to last-click only when volume is too thin for a model to learn anything.

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-Click Attribution

Gives all conversion credit to the final ad click before the conversion, ignoring every earlier touchpoint in the path.

All credit lands on the final click before the conversion, which makes reporting simple and reproducible but flattens upper-funnel work. Advertisers use it when paths are short or when they need an auditable baseline everyone can recompute. Google Ads still offers it while defaulting to data-driven attribution. The usual mistake is reading last click as what the ad caused rather than simply where the path ended.

Full definition

Side by side.

The differences that actually change what happens in your account.

	Data-Driven Attribution	Last-Click Attribution
What gets credited	A fractional, path-specific share of credit assigned by comparing your converting and non-converting paths.	The single final click before the conversion, 100 percent every time.
Minimum data needed to function	Enough conversion volume for the model to detect patterns; too little and it cannot distribute credit meaningfully.	None - works from day one on a single conversion.
Where the credit split comes from	Modeled from your own account's actual path data.	A fixed rule, not modeled from anything.
Default status today	Default model in both Google Ads and GA4.	Still selectable in Google Ads, but no longer the default there.
Visibility across channels	Sees only what happens inside that platform's own tracked data - it cannot credit a channel it never observed.	Same limitation - only sees the final click it recorded.
Reproducibility	Can shift over time as the underlying model retrains on new path data, so the same historical conversion can get re-weighted.	Fully static - a conversion's credit never changes after the fact.
Failure mode	Low-volume accounts get a model with too little signal, producing credit splits that look precise but rest on thin data.	Gets read as what the ad caused when it is really just where the path happened to end.

What actually separates them.

Data-driven attribution is a statistical model that gets re-estimated as new conversion data comes in, so the same past conversion's credit split can move over time; last-click is a static rule that never changes once recorded.

Data-driven attribution requires a minimum amount of conversion history before it can distribute credit meaningfully, while last-click works identically whether you have one conversion or one million.

Both models are boxed in by the same walled-garden limit - neither can see or credit a touchpoint the platform never tracked - but data-driven attribution compounds that limit by needing volume within that same walled garden to model anything.

Switching a campaign's read from last-click-style reporting to data-driven attribution can visibly reshuffle which keywords or audiences look efficient, since data-driven attribution may credit assist touches last-click was silently ignoring.

Data-driven attribution's output looks precise because it hands out fractional credit down to a decimal point, but that precision is a property of the model's confidence, not proof of causation, in exactly the same way last-click's 100 percent number is not proof either.

Which one should you use?

Use Data-Driven Attribution when

Your account has enough conversion volume for the platform to say it has sufficient data to model attribution.
You are managing Smart Bidding strategies that are already designed to work with fractional conversion credit.
You want credit for upper- and mid-funnel touches without manually picking a rule-based model.
You are comfortable with an attribution answer that can shift slightly as more data accumulates, rather than needing it to be permanently fixed.

Use Last-Click Attribution when

Your account does not have the conversion volume for data-driven attribution to model anything reliably.
You need a static number that will not change on you later when the model retrains.
You want the simplest possible cross-check against data-driven attribution's fractional output.
You are auditing a low-volume campaign where a modeled answer would be less trustworthy than a plain rule.

Common questions.

Do I have to choose between data-driven and last-click attribution?

Not really - data-driven attribution is the default in both Google Ads and GA4 today, so unless you have deliberately switched a conversion action's setting, you are already on it. Last-click is there as a manual fallback, mainly useful when conversion volume is too low for the modeled version to have learned anything reliable.

Why did my conversion credit change even though nothing happened last month?

That is a property of data-driven attribution, not a bug - it re-estimates its model as new path data accumulates, so historical credit splits can shift slightly as the platform sees more converting and non-converting paths. Last-click does not have this behavior since it is a fixed rule with nothing to retrain.

How much conversion volume do I need for data-driven attribution to work well?

There is no single published universal threshold, and it varies by account, but the general principle is that the model needs enough recent conversions across enough different paths to detect a pattern; very low-volume conversion actions get much less reliable splits. If your volume is thin, treat the modeled output cautiously and keep last-click as a sanity check.

Does data-driven attribution see conversions that happened outside Google Ads or GA4?

No. Both last-click and data-driven attribution can only credit touchpoints the platform itself tracked, so a channel that never fired a trackable click or event - offline referrals, some organic social, word of mouth - is invisible to either model regardless of how sophisticated the math is.

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