Responsive Search Ads and Ad Strength in Google Ads

1. How an RSA is built today

An ad is a set of assets:

• up to 15 headlines of 30 characters each;
• up to 4 descriptions of 90 characters each;
• a final URL and a display path.

Not all of it shows. One headline and one description always appear. The second and third headline and the second description show up when the system predicts a better result, or when there is room on screen.

Three changes that often go unnoticed:

• Single-headline ads. Since February 2024 Google may show only one headline when it predicts that will perform better. There is no way to switch this off;
• Campaign-level headlines and descriptions. Up to three headlines and two descriptions can be set once for a whole campaign and fed into every RSA inside it;
• Automatically created assets. The system adds its own headlines and descriptions based on your site and existing assets, and they compete with yours.

The practical consequence of the first point: every headline has to make sense on its own. If the meaning only appears in the pair "Headline 1 + Headline 2", there is a fair chance the user sees half a sentence.

2. Ad Strength: what it is and what it does not do

Ad Strength evaluates the relationship between the assets you supply and the keywords the ad group targets. It looks at how many headlines there are, how different they are from one another, how relevant they are to the keywords, and whether assets are in place.

The most important thing about this indicator is stated in Google’s own documentation: Ad Strength is a feedback tool for asset diversity and combination testing, and it is not used to calculate Ad Rank, Quality Score or auction wins. The rating also does not decide whether an ad is eligible to serve at all.

So moving the label from "Average" to "Excellent" is not the same as improving the advertising. It is the system telling you that you have given it more material to test with.

3. Is "Excellent" worth chasing

What Google says. Advertisers who raise Ad Strength for responsive search ads and sitelinks from "Poor" to "Excellent" see 15% more conversions on average. The figure is official, but it describes a move from the lowest rating — not from "Good" to "Excellent".

What practice says. In May 2025 Matt Bowen of Logical Position pulled data from 3,660 non-brand search campaigns over 30 days and added two individual account audits. The picture: ads rated "Excellent" showed higher CPCs and weaker results than ads rated "Average" and "Good".

The reason is simple. To earn "Excellent" you have to unpin everything and give the system complete freedom over combinations. The user then sees headline sets that nobody assembled deliberately, and the specifics of the offer get lost in them.

The same analysis notes something worth keeping in mind: large advertisers such as Best Buy and Nike pin their headlines, despite Google recommending against it.

A working benchmark. "Poor" is a signal that the ad genuinely is unfinished and needs fixing. "Average" and "Good" are a normal state for most accounts. "Excellent" is not a goal in itself: if you reached it by removing every pin and padding the ad with generic headlines for the sake of variety, you traded quality for a label.

4. Pinning headlines: when to and when not to

Pinning locks a headline or description to a specific position. Google does not recommend it for most advertisers because it cuts the number of possible combinations. There are situations where it is justified.

The working balance is one or two pinned headlines with the rest free. That keeps the core message in place and still leaves the system room to test.

5. What actually drives RSA performance

If Ad Strength plays no part in the auction, the work belongs where the auction actually happens.

• Headlines that differ in meaning, not in wording. Fifteen rephrasings of one idea give the system nothing. What helps is headlines covering different motives: price, turnaround, terms of service, range, geography, how you work together;
• Alignment with the landing page. The promise in the headline should be visible on the page without scrolling. That is what genuinely feeds Quality Score, through ad relevance and landing page experience;
• Ad group structure. The same assets produce different Ad Strength in different groups precisely because the keywords differ. If a group mixes intents, no set of headlines will be relevant to all of them;
• The number of ads in a group. Two or three RSAs let you compare approaches rather than individual phrasings;
• The asset performance report. It shows which headlines the system rated best and which it barely uses. That is direct material for replacement — unlike the overall Ad Strength label;
• Bidding strategy and budget. The best ad will not rescue a group whose bids keep it out of the auction. We covered how to choose one in the article on Google Ads bidding strategies.

If clicks are coming in and orders are not, the ad is rarely the whole story — we unpacked that case separately in clicks with no orders.

6. Ad Strength, Quality Score and Optimisation Score

Three different metrics that get confused regularly.

A word on Optimisation Score: it rises when you accept the system’s recommendations, and some of those recommendations widen reach or add automated assets that do not suit every account. Treating it as a measure of how well the account is run is a mistake.