RSA headline pinning: lock the brand line in position 1 or let the ads rotate?

Pinning in Responsive Search Ads is one of those settings people either swear by or never touch. Pinning your brand name to position 1 is usually fine. Pinning everything else quietly kills the test the format is built for.

Take a dental clinic that pins "Book Online" to headline 1 and "Free Consultation" to headline 2. Google can only rotate the remaining headlines, so the combinations it can test shrink fast and the ad ends up behaving like a manual ad with extra steps. The counterpoint is real, though. If legal or the brand team needs one exact phrase in one exact slot, pinning is the only tool you have, and that outweighs the learning loss.

My rule would be to pin only when a phrase is mandatory, and otherwise leave everything free and judge the asset report after a few weeks of volume. Does anyone see pinning help on low-volume accounts, where rotation never gets enough data to learn anyway?

On low-volume accounts I'd pin without much guilt. Rotation rarely gets enough data to learn anything, so you aren't giving up much optimisation. What you get back is predictability: the offer line always lands in a slot you picked, and you can read results without guessing which of the dozens of combinations served.

Where I'd push back is on how you judge it. The asset labels ("Low", "Good", "Best") need a lot of impressions before they mean anything, so on a small account they're mostly noise. A cleaner test is to pin the one line you need, run the ad for a fixed window, then launch a second ad with a different headline set and compare cost per conversion between the two.

Also, pin one slot, not three. Once you pin several, Google has almost nothing left to rotate, and you're back to a manual ad with a different name.

Agreed on the single pin, but the test setup needs tightening. Two RSAs in the same ad group don't get an even split, because Google pushes impressions toward whichever one it thinks is winning, so the early cost per conversion numbers are muddy. Better to put each headline set in its own ad group with the same keywords and a similar budget. If the ads have to share a group, set ad rotation to "rotate indefinitely". My rule of thumb: pin only what legal or the offer truly requires, one slot at most, and compare ad groups instead of ads.

Agreed on separate ad groups. One thing that trips people up is that "rotate indefinitely" only stops Google from optimizing the ad itself. The ad groups still compete in the same auction, so keep bids and budgets matched too. Otherwise a bigger budget in one group skews the comparison.

Fair on the budget, but I think the keyword overlap leaks more. If both ad groups bid on the same terms inside one campaign, they compete in the same auction, and ad rank leans heavily on expected CTR. So the headline set that already looks stronger wins most of those auctions. What you end up measuring is which ad wins the auction, not which headlines convert, and the cost per conversion gap mostly reflects who got the impressions.

The cleaner route is Google Ads campaign experiments. They split traffic between the original campaign and a trial version at a percentage you set, so the two headline sets never fight over the same auctions. Change only the headline set in the trial and leave everything else alone. On a low-volume account that can mean several weeks, so decide the stopping point and the conversion threshold before you start.