What Is Attribution in Google Ads (and Why Last-Click Is Lying to You)

What attribution actually means

Most customers don’t buy on the first ad they click. They see a Search ad, don’t buy, get retargeted by a Display or YouTube ad a few days later, search your brand name directly, and finally convert. Attribution is simply the rule Google Ads uses to decide which of those clicks — or how much of each — gets the credit for that conversion.

Why does the rule matter? Because whichever touchpoint gets credited looks like it’s “working,” and whichever doesn’t looks wasteful. Get the rule wrong and you’ll cut a campaign that was actually doing the hard work of starting the journey, while over-funding the one that just happened to be there at the end.

This isn’t an abstract reporting quirk. It feeds straight into automated bidding. Smart Bidding strategies like Target CPA and Target ROAS optimise towards whichever conversions your attribution model tells them count — so if the model is crediting the wrong touchpoints, the algorithm is quietly optimising towards the wrong signal too, not just misreporting after the fact.

Last-click: simple, and systematically wrong in one direction

Last-click attribution gives 100% of the conversion credit to the final ad click before someone converted. It’s easy to understand and, for a long time, it was the Google Ads default — which is exactly why so many accounts are still structured around it without anyone deciding that on purpose.

The problem is what last-click ignores: everything that happened before that final click. If a customer saw a Display remarketing ad on Monday, ignored it, then searched your brand name on Friday and clicked a branded Search ad to buy, last-click hands 100% of the credit to that branded search — the ad that was arguably the easiest, cheapest click of the entire journey. The remarketing ad that brought them back gets nothing, even though it may be the reason they searched your brand at all.

The bias runs in one predictable direction: last-click always favours whatever sits closest to the point of conversion, and always undersells whatever opens the journey or nudges someone back into it. That’s not a random measurement error you can average out — it’s a structural blind spot, which is why simply switching models, rather than tweaking budgets under the old one, is usually the actual fix.

Where last-click typically under-credits

• Remarketing / Display — rarely the last click before a purchase, because it’s designed to bring people back, not close the sale itself
• YouTube ads — usually an early-stage touchpoint; someone watches, doesn’t click, converts days later through a different channel
• Broad, top-of-funnel Search terms — the generic query that starts the research undersells itself next to the branded query that ends it
• Branded Search — the opposite problem: it tends to be over-credited under last-click, since it’s disproportionately the final click even when another channel did the work of creating that brand awareness

A worked example

Say a customer runs a broad search for “double glazing quotes,” clicks a Generic Search ad, and lands on your site without converting. Three days later they see a Display remarketing ad and click through, but still don’t convert. A week after that, they search your brand name directly, click a Branded Search ad, and request a quote.

Under last-click attribution, the Branded Search ad gets 100% of the credit for that conversion. Generic Search and Display get nothing. Look at the account through that lens and Generic Search and Display both look like they’re burning budget for no return — even though, in this journey, they’re the reason the customer knew who you were by the time they searched your brand.

Under data-driven attribution, credit is split across all three touchpoints based on how much each one actually influenced the outcome, learned from your account’s own conversion data rather than a fixed rule. Generic Search and Display both get a meaningful share of the credit, because the model can see the pattern: conversions that pass through those touchpoints happen more often than conversions that don’t.

How to actually see this in your own account

1. Open the Attribution reportsInside Google Ads, the Attribution section (under Insights and reports) shows conversion paths, path length, and time-lag between first click and conversion — this is where the multi-touch journey becomes visible instead of theoretical.
2. Check assisted conversions, not just last-click conversionsAssisted conversions show how often a campaign appeared earlier in a path that ended elsewhere. A campaign with low direct conversions but high assisted conversions is doing upper-funnel work that a last-click-only view would miss entirely.
3. Look at path length and time lagIf most of your conversion paths involve two or more clicks over several days, that’s a strong signal your account rewards patience — cutting a campaign for looking underwhelming after one day is premature.
4. Judge upper-funnel campaigns on their role, not their finish lineBefore pausing Display, YouTube or broad Search on the grounds that they “don’t convert,” check what happens to your branded Search and direct conversions if you turn them off for a test period — the real cost often shows up somewhere else in the account.

What data-driven attribution changes in practice

Google made data-driven attribution (DDA) the default model for new conversion actions, and first-click, linear, time-decay and position-based models have all been retired as manual options — last-click is the only other model still selectable, and DDA is what most accounts run on by default. In practice, that means the credit-splitting is already happening in most accounts whether or not anyone set it up deliberately.

What it changes for you day to day is how you read performance. Campaigns and keywords that look expensive or low-converting under an old last-click mindset — broad Search terms, Display remarketing, YouTube — often earn more credit under DDA once the model can see their real role in the journey. The practical shift isn’t a setting you flip; it’s remembering, when you look at a report, that a conversion showing against one campaign was very possibly influenced by two or three others before it.