Back to blog

Google Ads Attribution Models: Data-Driven vs Last Click

PPC Course

Google Ads Attribution Models: Data-Driven vs Last Click

By Daniel | PPC Performance Marketing Specialist

Which attribution models Google Ads offers, how data-driven and last click differ, how conversion windows work, and how attribution changes bidding and budgets.

Updated 31 October 2026 · Part of the PPC guide series

Short answer: An attribution model decides which ad interactions get credit for a conversion. Google Ads now offers two: data-driven attribution, which shares credit across the clicks and interactions that helped based on your account's data, and last click, which gives all credit to the final click. Google no longer supports first click, linear, time decay or position-based models. Your choice matters because Smart Bidding optimizes toward the conversions as they are credited, so attribution quietly shapes where your budget goes.

What attribution means in Google Ads

A customer might search a generic term, click an ad, leave, search again days later with your brand name and buy. Attribution decides how that purchase is credited across those clicks, the keywords behind them and the campaigns they belong to.

The models available today

Model How credit is given Good for
Data-driven Shared across ad interactions based on how they contribute, using your past conversion data for that action Most accounts; gives credit to earlier, non-brand touchpoints
Last click 100% to the last clicked ad and keyword Simple setups or when you need comparability with older reports

Google's help center states that first click, linear, time decay and position-based models are no longer supported, and conversion actions that used them were moved to other models. Some guides still list six models; only these two are available to set in Google Ads now. Data-driven attribution has been the default for new conversion actions since Google made it the default in 2021.

A worked example

A hypothetical path: generic search ad click → display remarketing click → brand search ad click → purchase worth $200.

Interaction Last click credit Data-driven credit (illustrative)
Generic search click $0 $80
Remarketing click $0 $40
Brand search click $200 $80

Under last click, the generic search campaign looks like it produced nothing and might be cut. Data-driven attribution recognizes that it started the journey. The exact split depends on your account's data; these numbers only illustrate the idea.

Why attribution changes bidding

Smart Bidding optimizes toward the conversions it is credited with. With last click, brand and bottom-of-funnel keywords receive almost all credit, so bidding favors them and starves the campaigns that create demand. With data-driven attribution, upper-funnel keywords and campaigns can earn partial credit, so they receive more budget when they genuinely help.

When you change a conversion action's attribution model, expect reported conversions per campaign to shift. Compare results over a full conversion cycle rather than reacting to the first days.

Conversion windows and lag

The conversion window is how long after an ad interaction a conversion can be counted. In Google Ads:

  • You can set it from 1 to 90 days, depending on the conversion source.
  • If you do not change it, the click-through window defaults to 30 days.
  • Engaged-view and view-through windows can be set separately, up to 30 days.

Set windows to match your real buying cycle. A short window on a product people research for weeks undercounts conversions from early clicks; a very long window on impulse purchases can over-credit old clicks. Look at how long conversions usually take after a click before choosing.

Example: Choosing a Window From the Data

A furniture store looks at its conversion path report and finds that most purchases happen 3 to 12 days after the first ad click, and a few take up to 3 weeks (illustrative). A 7-day window would miss part of the demand and make early campaigns look weaker than they are. The default 30-day click window fits the buying cycle better. An impulse-purchase store that sees nearly all conversions within 2 days could use a shorter window, so old clicks do not get credit they did not earn.

Brand versus non-brand credit

Brand searches often happen at the end of a journey that started elsewhere. Even with data-driven attribution, judge brand and non-brand campaigns separately:

  • Brand campaigns usually show low cost per acquisition because they capture existing demand.
  • Non-brand campaigns usually show higher cost per acquisition but create new demand.

Cutting non-brand campaigns because they look expensive under last click often reduces brand searches later.

Why Google Ads, GA4 and your CRM disagree

Source Why numbers differ
Google Ads Counts conversions on the date of the ad interaction, includes its own modeled conversions, uses your Google Ads attribution and window settings
GA4 Uses its own attribution settings across all channels, not just Google Ads
CRM or sales data Records the deal when it closes, with no ad credit unless you import it

Expect differences. Decide which source answers which question: Google Ads for optimizing within Google Ads, GA4 for channel comparison, CRM for business results. For tracking setup that makes these sources line up better, see conversion tracking for PPC.

Use attribution in budget decisions

  • Compare campaign performance under data-driven attribution, not last click, before cutting upper-funnel campaigns.
  • Look at the attribution reports in Google Ads to see common paths and how long conversions take.
  • For big budget moves, confirm with an incrementality test (for example, a geographic holdout) rather than relying on any attribution model alone.

For reporting these results to stakeholders, see PPC analytics and reporting.

Example: Testing Incrementality With a Geographic Holdout

Attribution assigns credit, but it does not prove that an ad caused a sale. A simple test is to pause a campaign in a few comparable regions while leaving it running elsewhere, then compare total sales (not just ad-attributed sales) in both groups over several weeks. If the paused regions lose much less than the ad-attributed sales suggested, the campaign was capturing buyers who would have bought anyway. Run it only when you can afford a few weeks of lower reach, and keep other changes steady during the test.

Check That Your Tags Are Installed

Attribution depends on tags firing correctly. A free SEO audit of your landing and thank-you pages detects common tracking pixels, such as the Facebook Pixel, and reports load speed, which helps you spot a missing or slow tag before you trust any model's numbers.

Common attribution mistakes

  • Judging non-brand campaigns only by last-click results.
  • Using a conversion window much shorter than the real buying cycle.
  • Comparing Google Ads and GA4 numbers as if they should match.
  • Changing attribution models and immediately judging the effect.
  • Treating attribution as proof of incrementality.

FAQs

Which attribution models are available in Google Ads?

Data-driven and last click. Google no longer supports first click, linear, time decay or position-based models.

Is data-driven attribution better than last click?

For most accounts, yes, because it gives credit to earlier interactions that helped and lets Smart Bidding value them. Last click is simpler but tends to overvalue brand and final-step clicks.

What is a conversion window in Google Ads?

It is the period after an ad interaction during which a conversion can be counted. It can be set from 1 to 90 days, and the click-through window defaults to 30 days if you do not change it.

Does Changing the Attribution Model Change My Past Data?

It changes how conversions are credited in reports and bidding for the affected conversion action, and Google describes how past data is handled in its help pages. Always note the date of the change in your reporting, so a shift in campaign numbers is not mistaken for a real performance change.

Do I Need an Incrementality Test?

Not for small decisions, but it is worth running before large budget changes, such as cutting a major campaign or doubling spend. No attribution model can tell you for certain what would have happened without the ads.

What does "7-day click, 1-day view" attribution mean?

It is a common way of describing conversion windows: conversions are counted if they happen within 7 days after a click or 1 day after an ad view. Platforms set these windows differently, so check each platform's settings before comparing results.