Is Ad Strength worth chasing in Google Ads?

Mostly no. Getting a responsive search ad off Poor is worth an afternoon of copywriting. Pushing it from Good to Excellent often costs you relevance, because the last step usually means unpinning headlines you pinned for a reason. Ad Strength has no effect on your Ad Rank, your position or what you pay per click. What it changes is how many ad combinations Google has available to test, which matters up to a point and then stops mattering.

So the question worth asking concerns the final step, and whether reaching Excellent is worth what it costs you.

What Ad Strength measures, and what it ignores

Start with what the rating does not do. Google Ads Liaison Ginny Marvin, replying to advertisers on X in April 2024, said that “Ad Strength is not used in Ad Rank” and is not a factor in stopping a keyword from going to auction. She added: “Ad Strength is a distinct diagnostic tool. And to reiterate, it’s not a factor in the auction.”

So an ad rated Poor can outrank an ad rated Excellent, on the same keyword, in the same auction, whenever its relevance to the search is better. The rating is Google’s opinion of your asset set. The auction never reads it. (For what the auction does read, see how Ad Rank works.)

Google’s Ad Strength help page uses five ratings: Incomplete, Poor, Average, Good and Excellent. Incomplete means you have not supplied the minimum assets, and the other four describe how much variety you have given the system. The page also says Ad Strength does not determine whether your ad is eligible to serve, so a Poor ad still runs. The rating is built from three inputs:

• Asset and keyword relevance. Whether your headlines and descriptions contain the keywords you are targeting.
• Enough headlines and descriptions. You can supply up to 15 headlines and 4 descriptions, and more assets mean more possible arrangements.
• Sitelink sufficiency. Whether you have six or more sitelinks available at account, campaign or ad group level.

Every item on that list is an input. None of them is an outcome. An Excellent ad can convert nobody and the rating will not move, because nothing in the calculation looks at what happened after the click.

The third input is telling. Sitelinks can sit at account or campaign level, so adding six of them once can lift the rating on ads right across the account. That shows how loosely the label tracks the quality of any individual ad.

Ad Strength also has no bearing on Quality Score. The two are separate diagnostics: Quality Score rates a keyword on expected clickthrough rate, ad relevance and landing page experience, while Ad Strength rates the variety of assets in one ad.

Google’s case for chasing it

Google publishes a specific claim on the same help page: advertisers who improve Ad Strength for their responsive search ads and sitelinks from Poor to Excellent “see 15% more conversions on average”, based on Google internal data from August 2025.

That figure deserves to be taken seriously, and read precisely. The comparison is Poor against Excellent.

It does not tell you what the final step from Good to Excellent is worth on its own. An ad rated Poor is usually missing assets entirely, running three headlines where fifteen were available. Fixing that is a real improvement, and it is reasonable to expect much of the gain sits in escaping Poor rather than in reaching Excellent. Google does not publish the breakdown, so nobody outside Google can say for sure.

The two jobs are nothing alike. Going from Poor to Good means writing copy you should have written anyway. Going from Good to Excellent usually means giving up control you chose to keep.

The evidence that complicates it

Matt Bowen of the agency Logical Position looked at 3,660 generic (non-brand) search campaigns over 30 days, writing in Search Engine Land in May 2025. Each campaign was grouped by the predominant Ad Strength across its ads, then performance was compared.

Poor-rated ads behaved as expected, with higher costs per click and lower click-through rates. The surprise came at the other end. Excellent-rated ads also showed higher costs per click and weaker performance, while campaigns rated Average and Good held up better. It is one agency’s data and a single 30-day window, so treat it as a strong signal rather than a law, but it cuts directly against the idea that higher is always better.

Pinning is the likely hinge, and the mechanism is unglamorous. Reaching Excellent usually requires unpinning headlines so Google can assemble more combinations; Google’s own page notes that pinning “reduces the overall number of headlines or descriptions that can be shown”.

Unpinned headlines produce arrangements you never wrote. Some drop the keyword entirely, some read as three unrelated fragments stacked together, and both weaken relevance to the search. Weaker relevance raises what you pay, and relevance is a signal the auction does read.

Large advertisers seem to have noticed. The same article points to live search ads from Best Buy, Sweetwater and Nike with keyword-matched headlines pinned first, even though Google advises against pinning.

You can check the effect in your own account without any study. Open the combinations report on a responsive search ad and read what Google has been serving. If your keyword is missing from some of the combinations, unpinning further will produce more of them.

How both findings can be true

Google’s figure and the agency data are not in conflict, and seeing why gives you the rule.

Ad Strength is a decent proxy for whether you have done the basics, and it stops being a proxy once you have. Below a certain point, a low rating means your ad is thin, and fixing it helps. Above that point the rating starts rewarding variety for its own sake, and variety and relevance are different things.

That produces a curve rather than a line. Performance climbs as you move off Poor, flattens through Average and Good, and can bend back down if you strip out every pin to reach Excellent.

There is a reason the rating leans this way. Google benefits when you supply more assets, because more combinations give its systems more room to find what works across every auction. Your interests and Google’s overlap for most of that journey, then separate slightly at the end.

What to do with your rating

Get off Poor, and treat that as urgent

A Poor rating usually means missing assets. Write 10 to 15 distinct headlines and all 4 descriptions, using different angles rather than reworded versions of one claim.

This is copywriting work rather than account work. Headlines that describe the same benefit five ways give Google nothing useful to test, which is where writing as a discipline earns its place in a paid search account.

A workable spread for a service business: two headlines carrying your main keyword, two naming your location, two on price or guarantee, two on speed or availability, two on proof, and the rest on your specific offer. Ten different claims beat fifteen variations of one.

Stop at Good

Good is the point where you have supplied enough variety for the system to work with. Pushing to Excellent from there usually means unpinning something you pinned on purpose.

If your rating sits at Good with two or three pinned headlines holding your keyword in position one, leave it alone. You have traded a label for relevance, and relevance is what sets your bill.

The one exception is a brand new ad group with no history. There, let Google run unpinned for a few weeks, read the combinations report, and pin only what the data tells you to protect.

Judge by the columns that cost you money

Cost per conversion, conversion rate and click-through rate decide whether an ad works. Ad Strength is not on that list.

When the two disagree, believe the money. An Average ad delivering leads at AUD 60 beats an Excellent ad delivering them at AUD 95, and no rating in the interface changes that arithmetic.

Fix the ad group before the ad

A weak rating sometimes reflects a weak ad group rather than weak copy. If an ad group carries three different intents, no single ad can be relevant to all of them, and the rating is telling you about your structure. That fix usually starts with how many ads belong in one ad group rather than with the headlines themselves.

Know where the rating does not apply

Shopping campaigns have no Ad Strength rating, because there is no ad copy to rate. Google assembles the listing from your product data instead, which is why Shopping optimisation happens in the feed. Performance Max shows an Ad Strength on each asset group, and the same reasoning applies: more assets give the system more to test, and past a reasonable level the rating stops telling you anything about performance.

If your account is being managed toward ratings rather than results, that is usually visible within minutes of opening it, and it is one of the first things we check in a Google Ads review.

The rule of thumb: treat Ad Strength as a checklist you complete once, not a score you keep working on. Get every ad to Good with a full set of distinct headlines and descriptions, keep the pins that hold your keyword in place, and from then on let cost per conversion decide what changes, even if that means an ad sits at Good for the rest of its life.