Pause the losing ad: comparing ads in Google Ads

What it flagsThe weaker of two ads in the same group

AdBuilder sets two responsive search ads in an ad group by default, so Google can pit them against each other. This signal looked at such a pair and flagged the ad that clearly underperformed the other in the same group. That way you could see at a glance which ad performed better and which one lagged.

It deliberately only compared ads within the same ad group. Those show on the same keywords and compete for the same impressions, making a fair comparison possible. Comparing ads across different groups or campaigns would be comparing apples to oranges, since different keywords, audiences and bids come into play there.

It was a purely informational recommendation: AdBuilder changed nothing in your account. You got the insight and decided yourself whether to pause the losing ad.

Why it mattersWhy a weak ad costs you budget

When two ads compete for the same impressions, every impression that goes to the weak variant comes at the expense of the strong one. The weaker ad siphons off traffic that the better ad would probably have converted into more clicks or conversions. Over time that quietly drags down your return without you noticing right away.

Pausing a poorly performing ad concentrates traffic on the winner. That's the core idea of an ad A/B test: run two variants, measure which performs better, then continue with the best one. The tricky part is picking the right moment, since a difference that looks big on a handful of impressions can be pure chance.

How it worksHow AdBuilder compared the ads

The comparison only started once both ads had enough data. The threshold was a minimum of 150 impressions per ad, so there was something meaningful to compare and a random blip wouldn't immediately count as a real difference.

The ads were compared on conversion rate first. If there were too few conversions to judge reliably, the comparison fell back to click-through rate (CTR). The clearly weaker of the two was flagged as a candidate to pause, giving the winner more room.

Important: this standalone signal has since been retired. It overlapped too much with two other flows: the weak-ads recommendation, which improves or replaces weak responsive search ads, and A/B test wrap-up, where you cut over to the best ad once there's a clear winner. AdBuilder no longer emits this alert and automatically clears out any old, still-open alerts. The threshold and comparison logic described above explain how the signal used to work, not a recommendation you'll still see appear on its own.

• At least 150 impressions per ad before any comparison happens.
• Compared on conversion rate; falls back to CTR if there are too few conversions.
• Only ads within the same ad group.
• Informational: no automatic change in Google Ads.
• Now folded into the weak-ads flow and A/B test wrap-up.

What you can doWhere to find this now

If you still want to pick the best ad and clean up the weak variant, you can do that through the recommendations that took over this task. They offer more than just a signal: they actually help you improve or replace the weak ad.

• Improve weak ads: tackles responsive search ads with low CTR or Ad Strength, with an improved or replacement variant.
• A/B test winner found: wrap up a running test once there's a clear winner and cut over to the best ad.
• Pause manually: you can of course always pause the weaker ad yourself in Google Ads, for example after checking impressions, CTR and conversion rate per ad.