What is paid search attribution?

Paid search attribution is how credit for a conversion is split across the ad clicks and interactions that came before it. In Google Ads the choices are now data-driven attribution, the default, and last click.

Is data-driven attribution the default in Google Ads?

Yes. Google made data-driven attribution the default for new conversion actions in 2021, and its help page says it is the default for most conversion actions. You can still switch a conversion action to last click.

Were rule-based attribution models removed from Google Ads?

Yes. Google removed first click, linear, time decay, and position-based models in 2023 and moved conversion actions still using them to data-driven attribution. GA4 removed the same four models in November 2023.

Does data-driven attribution need a minimum number of conversions?

No hard minimum since 2021. Google recommends at least 200 conversions and 2,000 ad interactions in supported networks within 30 days for best results, but lower-volume conversion actions can still use it.

What does data-driven attribution in Google Ads miss?

It only sees Google ad interactions. Touches on Meta, LinkedIn, email, organic search, or offline before the Google click are invisible, so use it for allocation within Google Ads, not for comparing channels.

Paid Search Attribution: Last Click vs Data-Driven

Q: What attribution model gives all credit to the last click?

Last click. It assigns 100% of the conversion credit to the final ad click before the conversion and nothing to earlier interactions, which overvalues brand and retargeting campaigns and undervalues upper-funnel ads.

Google Ads now offers two attribution choices: data-driven, the default, and last click. Here is why last click misleads paid search budgets, how data-driven attribution works, what it cannot see, and what to do about the gap.

Paid search attribution is how conversion credit gets split across the ad clicks and interactions before a sale or lead. In Google Ads today there are two real choices: data-driven attribution, the default, which spreads credit based on your own conversion paths, and last click, which gives all the credit to the final click.

The older rule-based models are gone. Google removed first click, linear, time decay, and position-based models from Google Ads in 2023 and switched conversion actions still using them to data-driven attribution, per the Google Ads Help announcement. GA4 dropped the same four models in November 2023. So the practical question is no longer which of six models to pick. It is whether you trust data-driven attribution, when last click still has a place, and what neither of them can see.

What Attribution Model Gives All Credit to the Last Click?

Last click. It assigns 100% of the credit to the final ad click before a conversion. It was never designed as an accurate model of marketing contribution. It became a standard because it was easy to implement and easy to report. Calling it a model makes it sound like a deliberate choice; really it means "stop counting touches before the final one."

When people search for non-last-click attribution, they usually mean any model that gives credit to earlier touches. In Google Ads, that now means data-driven attribution.

Why Last-Click Attribution Misleads Paid Search

The bias is not random. Last click systematically overvalues brand keywords and retargeting, and undervalues non-brand search, YouTube, display, and anything that happens days before the conversion.

Consider a typical B2B buyer journey:

A user searches a non-brand term ("project management software for agencies"), clicks a YouTube ad, and leaves.
Three days later they search a category term ("best project management tool"), click a Search ad to a blog post, and leave.
A week later they search the brand name, click a branded Search ad, and request a demo.

Last click gives 100% of the credit to the brand Search campaign. The YouTube ad and the non-brand Search ad get nothing. Brand Search looks cheap per conversion, non-brand looks expensive, and the obvious recommendation is to move budget to brand.

Follow that advice for long enough and fewer people enter the funnel, because the touches that introduced them were defunded. Brand search volume eventually softens because fewer people know the brand. The cause is a measurement model that could not see the top of the funnel. I see this pattern often in paid search accounts optimized on last click in competitive categories.

How Data-Driven Attribution Works in Google Ads

Google's data-driven attribution help page describes it this way: it uses your conversion data to calculate the actual contribution of each ad interaction by comparing the paths of customers who convert to the paths of customers who do not. It covers Search (including Shopping), YouTube, Display, and Demand Gen.

Three facts that older articles, including an earlier version of this one, get wrong:

It is the default. Google made data-driven attribution the default for new conversion actions in 2021 (Google Ads and Commerce blog), and the help page says it is the default for most conversion actions.
There is no hard minimum. Google removed the old data requirements in 2021. It now recommends at least 200 conversions and 2,000 ad interactions in supported networks within 30 days for best results, but lower-volume conversion actions can still use it.
Last click still exists. You can switch a conversion action to last click. Linear, time decay, and position-based are no longer options.

What it gets right: it gives credit to the upper-funnel and mid-funnel interactions that last click ignores, and because Smart Bidding uses those credits, bids follow contribution instead of proximity to the sale.

What it leaves out: it only sees Google ad interactions. Meta, LinkedIn, email, organic search, and offline touches before the Google click are invisible to it. GA4's data-driven model sees more channels, including organic, but still only what GA4 can track.

Where to Find Attribution Reports in Google Ads and GA4

In Google Ads, the attribution reports live under Goals > Attribution: conversion paths, path metrics, assisted conversions, and model comparison (attribution reports help). Model comparison is the fastest way to see how much credit non-brand and upper-funnel campaigns gain under data-driven attribution versus last click.

In GA4, reporting attribution can be set to data-driven, paid and organic last click, or Google paid channels last click. The default lookback window is 90 days for most key events and 30 days for acquisition events like first visits (GA4 attribution settings). For channel-level journeys, the Advertising section's attribution paths report splits touchpoints into early, mid, and late (attribution paths report).

The Cross-Channel Attribution Gap

Data-driven attribution inside Google Ads is a real improvement over last click inside Google Ads. It does not solve cross-channel attribution.

A user who sees a Meta video ad, then a LinkedIn post, then clicks your branded Google Search ad converts in a journey that Google Ads credits entirely to Google Ads, because Google can only see its own interactions. Meta's report will claim its share of the same conversion. Add the platform reports together and you usually get more conversions than you actually had.

Privacy changes widen the gap. Browser restrictions on cookies and app-level tracking consent mean every user-level tracking system sees fewer touches than actually happened, and none of them can tell you exactly how many they miss. Any single-platform attribution number should be read as that platform's view, not the truth.

What You Can Actually Do About It

1. Run data-driven attribution in Google Ads, with clear eyes. Use it to allocate budget within Google Ads. Do not use it to decide Google Ads versus other channels.

2. Fix what you are counting. Data-driven attribution is only as good as the conversions it learns from. If the conversion fires on every form fill, including spam and unqualified leads, the model learns to produce more of those. Import real outcomes through offline conversion imports and enable enhanced conversions, which send hashed first-party conversion data to Google. Google's offline import guidance notes that offline conversions uploaded more than 90 days after the last click are not imported, so upload regularly. Then set conversion values that reflect what each outcome is actually worth.

3. Use a cross-channel tool when budget decisions span platforms. Platforms such as Northbeam, Triple Whale, and Rockerbox use first-party and server-side data plus modeling to rebuild journeys across channels. They are more complete than any single-platform report, not perfect. I compare these and Google's native options in more depth in paid search attribution tools.

4. Use media mix modeling at scale. MMM works on aggregate spend and outcome data instead of user-level tracking, so cookie loss does not break it. Google publishes Meridian, an open-source MMM that supports calibration with experiment results and budget optimization. MMM is slow and needs statistical skill, and it gives directional channel answers rather than keyword-level ones.

5. Test incrementality directly. Turn a channel off in a set of regions for several weeks and compare against similar regions that kept it on. That gives you a direct read on what the channel adds at that spend level. Google also offers Conversion Lift studies for some accounts, which measure the causal impact of ads; Google notes it is not available for every account.

Operating Under Attribution Fog

Most accounts cannot afford MMM or clean holdout tests and still have to make budget decisions. The framework I use:

Keep data-driven attribution on in Google Ads and accept its limits.
Check assisted conversions and the model comparison report for every campaign. Campaigns that show up often in paths but rarely as the last click are the ones last click undervalues.
Watch for the death spiral: a channel looks expensive on last click, you cut it, volume into the channels that depended on it falls, total conversions fall, and the pressure to cut grows. Brand efficiency declining after cuts to non-brand or upper-funnel spend is the usual signal.
Keep channel diversity even when the numbers favor concentration. A portfolio built only on brand keywords and retargeting has no way to grow the pool of people who could search the brand.
Ask customers. A short survey asking new customers how they first heard about you captures offline and cross-device journeys that no tracking tool sees. It has recall bias, but it is first-party data you own.

Recommended Tools and Resources

Google Ads data-driven attribution: native, no extra cost, the default for most conversion actions
Google Ads attribution reports (Goals > Attribution): model comparison and assisted conversions
Northbeam: cross-channel attribution for e-commerce
Triple Whale: attribution built around Shopify stores
Rockerbox: multi-touch attribution for longer, multi-channel journeys
Google Meridian: open-source MMM; free, but needs data science resources

For teams building custom attribution models, wiring offline conversion data, or running incrementality tests, The Voice of Cash (thevoiceofcash.com), the team behind this site, works with performance marketing teams on measurement infrastructure.

Paid Search Attribution: Why Last Click Misleads and How Data-Driven Attribution Works Now