For a few years, the received wisdom on responsive search ads was relatively straightforward: pin your most important headlines, maintain message control, and accept a modest reduction in Google's ability to optimise combinations. A 2022 study backed this up, finding that pinned RSAs outperformed unpinned ones on conversion rate. A new analysis of 5,900 accounts suggests that picture has changed.
This matters because RSA pinning is one of the few genuine levers advertisers still have over ad copy delivery. As Google continues to expand automation - through Performance Max, AI Max and Smart Bidding - understanding where human control adds value versus where it gets in the way is a practical, ongoing question. This study helps answer it.
What Changed Between 2022 and Now
Google has made significant changes to how RSAs are served since 2022. The machine learning that drives headline and description combination selection has matured considerably. What worked in the earlier period - where pinning gave Google clearer signals and produced more predictable messaging - may no longer confer the same advantage now that Google's own optimisation has improved.
The Adalysis study, covering 5,900 accounts, represents a meaningful sample size. The 2022 version of the same research found pinned ads had higher conversion rates than unpinned. The updated findings reflect a platform that has genuinely moved on. That shift should prompt any advertiser who set their pinning strategy two or three years ago and left it alone to take another look.
The Case for Pinning Is Not Dead - It Is Narrower
Pinning still makes sense in specific situations. Compliance-sensitive sectors - financial services, regulated health, legal - often need exact messaging in position one. Brand-critical terms, unique value propositions, or messaging that must appear for regulatory or quality score reasons are all legitimate candidates for pinning. The issue is when pinning gets applied broadly as a default rather than a deliberate choice.
There is also a quality score consideration. RSAs with very high pin rates reduce the number of combinations Google can test. This limits performance data and can suppress the ad strength score - though ad strength itself is an imperfect proxy for actual conversion performance. The broader point is that pinning has a cost in terms of system flexibility, and that cost needs to outweigh the control benefit to justify it.
For lead generation advertisers in particular, where every conversion is a qualified enquiry rather than a low-commitment click, headline control can feel important. If your unique selling point or a specific offer needs to appear consistently, pinning position one is defensible. But pinning positions two and three on top of that starts restricting the variation Google needs to find optimal combinations.
How to Audit Your Current Pinning Setup
Start by pulling your RSA asset reports from Google Ads. Filter by ad strength and look at which ads have 'Poor' or 'Average' ratings - a high degree of pinning is frequently a contributing factor. Cross-reference with actual conversion data rather than relying on ad strength alone. An ad with poor strength but strong conversion performance is not broken; it may just reflect a tight, intentional setup.
Identify ads where three or more positions are pinned. These are the ones most likely to be underperforming relative to their potential, because Google has almost no room to test. For campaigns using Smart Bidding strategies like Target CPA or Target ROAS, restricted variation in ad combinations limits the signal quality the bidding algorithm can act on - tighter ad control compounds tighter keyword matching, and the system has fewer data points to work with.
Where you find over-pinned ads, consider a controlled test. Create a variant with pinning on position one only - protecting your key message - and allow positions two and three to rotate freely. Run both in an ad variation experiment with sufficient volume to reach statistical significance. Measure against conversion rate and, if you have offline import set up, against lead quality downstream.
Pinning and Performance Max: The Bigger Picture
RSA pinning does not exist in isolation. Most accounts running standard search campaigns are also running Performance Max, and many are either transitioning to AI Max or considering it. In PMax, asset control works differently - you cannot pin headlines in the same way, and Google assembles ad creative dynamically across channels. The lesson from the RSA data feeds into a broader strategic question: how much copy control do you actually need, and where should you apply it?
The honest answer for most lead gen accounts is that message discipline at the landing page level matters more than headline control at the ad level. If your value proposition, offer, and call to action are consistent and compelling on the page, the exact headline combination that drove the click becomes less critical. Conversion rate optimisation on the landing page has more leverage than precise pinning in most scenarios.
That is not an argument against all pinning. It is an argument for being deliberate about where you apply it. Use pinning to enforce must-have messaging, not as a defensive reflex against Google's automation. The data from 5,900 accounts suggests Google's RSA serving has improved to the point where broad pinning is, on balance, likely costing you conversions rather than protecting them.
What to Do With This Information
Review your RSA pinning as part of your next account audit. It is a low-effort, potentially high-impact check. Accounts that set their pinning strategy in 2021 or 2022 and have not revisited it are operating on assumptions that the platform has moved past. The ad serving environment is different now, and your copy strategy should reflect that.
If you run campaigns with Smart Bidding, prioritise giving the algorithm room to work - that means more headline combinations in rotation, not fewer. Reserve pinning for genuinely non-negotiable messaging. Test everything else. The accounts consistently performing at the lowest CPA are usually the ones treating their ad copy setup as an ongoing experiment rather than a fixed configuration.