Google Ads更偏向高广告强度RSA，但低评分广告转化更好

If you run responsive search ads, Google is now steering more impressions toward whichever ad in an ad group carries the higher ad strength rating. That ad has the better conversion rate only 45% of the time.

A new study by Brad Geddes, co-founder of Adalysis, analyzed more than 5,900 Google Ads accounts to answer a question PPC managers have argued about since RSAs became the default ad format. Does higher ad strength lead to better performance?

His answer is yes for some ads and no for others. The difference comes down to one setting, which is pinning.

I’ve known Brad since May 2011, when he became the PPC Faculty Chair at Market Motive while I was the faculty chair of Video and Content Marketing. And as one of the co-authors of Digital Marketing Fundamentals, which Wiley published in 2023, I asked Brad to contribute the chapter on Digital Advertising. So, I can vouch for his Experience, Expertise, Authoritativeness, and Trustworthiness (E-E-A-T) in PPC and paid media.

When Brad previewed the findings in a LinkedIn post last week, calling ad strength “one of Google Ads’ most misunderstood metrics,” I went back to read his full study, which follows up on the company’s 2022 ad strength study. That gives us a four-year view of how Google’s ad serving has drifted.

Here is what the data shows, where it collides with what Google says, and what I’d change in your accounts before holiday traffic peaks.

What Google Says Ad Strength Does

Start with the official position, because Brad’s numbers only become interesting against it.

Google’s Help Center describes ad strength as a guide for the ad creation stage. It says the rating “doesn’t directly influence your ad’s serving eligibility” and isn’t used to calculate Ad Rank, Quality Score, or auction wins.

Ginny Marvin, Google’s Ads Product Liaison, has been just as direct. In a March 2024 post on X, she wrote that ad strength “isn’t used directly in the auction & shouldn’t limit your testing.”

I take Google at its word on the mechanics. But “not an auction input” and “doesn’t affect which of your ads gets shown” are two different claims, and Brad’s data speaks to the second one.

What Happens When 2 Ads Go Head To Head

The most useful part of Brad’s study isolates ad groups with two active RSAs that carry different ad strength ratings. Both ads share the same keywords, bids, and budget, which strips out a lot of noise.

In those ad groups, the higher-rated ad received more impressions 63% of the time. An Excellent ad beat a Poor ad on impressions in 78% of ad groups. Ads two ratings apart, such as Excellent versus Average or Good versus Poor, won 66% of the time. Even the closest matchup, Good versus Average, went to the higher-rated ad 59% of the time.

Performance told a different story.

The higher-rated ad had the higher CTR in 52.5% of ad groups, barely better than a coin flip. It had the higher conversion rate in only 45%, which means the lower-rated ad converted better in 55% of ad groups. On conversions per 1,000 impressions, a metric Adalysis calls CPI, the lower-rated ad won 54.3% of the time.

Then there’s cost. The lower-rated ad had the lower CPA in 51.5% of ad groups, and across all ad groups the higher-rated ads ran an 11.8% higher CPA. In Brad’s example, a lower-rated ad with a $100 CPA would face a higher-rated rival at $111.80.

Brad’s warning for anyone running ad tests is Google “will serve the higher-rated ad more often, even if it has a worse conversion rate or CPI.”

4 Years Of Drift

Because Adalysis ran a similar study in 2022, Brad can show which way Google’s serving has moved.

How Often The Higher-Rated Ad Came Out Ahead	2022	2026
More impressions	56.8% of ad groups	63% of ad groups
Higher CTR	48.5%	52.5%
Higher conversion rate	45.8%	45.1%

Google is leaning harder on the higher-rated ad when it picks which ad to show, and that pick lines up with clicks a little better than it did four years ago. It does not line up with conversions any better. Brad concludes that Google “has increased its reliance on ad strength to determine what ad to show in an auction over the years.”

That’s his reading of observational data, and Google would likely dispute the word “reliance.”

Why The Lower-Rated Ad Often Wins

Before anyone decides ad strength is useless, look at who the lower-rated ads are.

Brad attributes most of their conversion edge to pinning. In most of the mixed-strength ad groups, one ad was pinned and the other wasn’t. Pinned ads tend to earn lower ratings, and they also tend to post higher conversion rates.

That makes sense to anyone who has pinned a headline on purpose. You pin a price, a required disclaimer, or a qualifier because you want to filter the click rather than chase it. Ad strength rewards variety and keyword coverage, and a pinned ad written to pre-qualify buyers works against both.

The unpinned results look very different. Among unpinned ads, CTR rose with every step up in ad strength, and so did CPI. Conversion rates rose slightly at each step except Poor to Average. For those ads, Brad’s conclusion is that chasing a higher rating generally pays off.

So, the honest summary isn’t that ad strength doesn’t matter. It’s that ad strength predicts performance well when you leave Google’s system alone and poorly when you’ve deliberately overridden it.

How Many Headlines Excellent Actually Takes

Only a third of unpinned ads in the dataset reached Excellent, and Brad’s asset counts show why stuffing in headlines won’t get you there.

Without pinning, an ad can reach Good with as few as seven assets. For Excellent, Brad recommends at least 11 different headlines. Volume alone doesn’t do it, though. Only 44% of ads with 15 or more headlines reached Excellent, and in the most extreme case, an ad with 38 assets was still rated Poor.

Once you pin, the minimum asset count to reach Good jumps to 10. Pinning two positions now caps an ad at Good and often leaves it at Poor or Average, which is stricter than what Adalysis saw in 2022. If you pin all three headline positions, Brad suggests aiming for about 13 assets.

There is a workaround. Pinning three or more different assets to the same position still leaves room for Excellent, while pinning only one or two caps the ad at Good. Position matters too, since headline one carries more weight than headline two, which carries more than headline three.

Where I Come Down

I’ve been in paid search since 2003, and Google has asked advertisers to trust a diagnostic label for most of that time. Quality Score came first. Google’s Ads Product Liaison has said it is also a diagnostic rather than an auction input, yet the expected CTR, ad relevance, and landing page signals behind it clearly shape what we pay. Ad strength looks like the same kind of label.

My read is that Google and Brad can both be right. The rating itself may never enter the auction. But the things that raise it, more distinct headlines, better keyword coverage, and fewer pins, give Google’s system more combinations to work with when it predicts which of your ads will earn the click. In your account, the outcome is the same either way. The higher-rated ad gets the traffic.

That’s where I think practitioners should push back. Google’s ad selection appears tuned to clicks, and your business generally gets paid on conversions. An 11.8% CPA gap across ad groups is not a rounding error at holiday budgets. Treat ad strength as a traffic allocator you have to manage, and stop treating it as a score you have to win.

What To Change In Your Account This Week

Sort your RSAs by pinning status before you read a single rating. Tag each ad as unpinned, partly pinned, or fully pinned in Google Ads Editor or a report export. The rating means something different in each group, so judge each group on its own.

Push unpinned ads to Excellent with variety, not volume. Brad recommends at least 11 headlines that mix keywords, benefits, features, calls to action, and authority statements. If an unpinned ad sits at Average with 15 headlines, the likely culprit is repetition. Rewrite the near-duplicates before you add more.

Rebuild pinned ads instead of abandoning them. Brad’s process is to remove the pins, build the ad to Excellent, then re-pin and keep adding assets until the rating reaches at least Average, or Good if you pin a single headline. Where you must pin, pin three or more variations to that position. If you pin all three headlines, plan on about 13 assets.

Judge tests on CPI and CPA, and give mismatched tests more time. Brad says it’s fine to run ads with different ratings in the same ad group. Expect the lower-rated ad to collect far fewer impressions, though, since an Excellent ad took the impression lead in nearly four of five matchups against a poor one. That stretches the time it takes to reach statistical significance. If the lower-rated, pinned ad wins on CPA, keep it running. Don’t pause a winner to fix a label.

The Machine Still Decides Who Sees Your Best Ad

Brad closes his study with a line every PPC manager should keep handy. He writes that “often, the human ability to create compelling ads will outperform the machine’s ability to serve the ads properly.”

I agree with one caveat. The machine still decides how often your best human-written ad gets seen. So, ad strength is something to check against your own conversion data, ad group by ad group, rather than a grade to celebrate.

Google tells us what the rating isn’t. Brad’s 5,900 accounts tell us what it does.

More Resources:

Featured Image: mohammadhridoy1/Shutterstock