Google Ad Strength is the most visible optimisation target in a search account, and the largest independent analyses find it does not predict performance. Across roughly twenty thousand accounts, adverts rated Excellent produced the worst cost per acquisition in the dataset.
Adverts rated merely Average produced the best.
This is for you if somebody on your team is working towards Excellent ratings and you want to know what that effort buys.
What the twenty thousand account analysis found
Optmyzr analysed responsive search ads across roughly 20,000 active Google Ads accounts in April 2026.
Ads rated Excellent: cost per acquisition of $28.68. Ads rated Average: $12.43. Roughly half, in favour of the lower rating.
The same analysis found headlines under twenty characters at $9.35 against $18.27 for longer ones, and sentence case beating title case by a factor of 3.7, at $7.46 against $27.47.
This was not a one-off. The same publisher had already found, across more than a million ads and over 22,000 accounts, no correlation between ad strength and conversion performance.
Two analyses, different years, different sample sizes, same conclusion about the rating.
What Google says, and why both belong in the room
Google’s guidance states that improving from Poor to Excellent is worth 15% more clicks and conversions.
Set that beside the independent findings and they disagree. Neither party is lying and both have interests.
Google benefits when advertisers supply many unpinned assets, because that gives its matching system more freedom. The rating encourages exactly that behaviour, so the guidance and the commercial interest align.
Optmyzr sells optimisation software, benefits from having something to correct, and draws its dataset from customers who are already relatively sophisticated. That skew is worth stating.
So read both, notice the interests, and weigh sample size. Twenty thousand accounts and a million ads is a substantial body of evidence, and Google has not published a comparably detailed counter-analysis.
Why Excellent Google Ad Strength might perform worse
The studies report the effect rather than the cause, so this is reasoning.
Reaching Excellent rewards asset quantity and variety. The straightforward way to satisfy it is to add more headlines and descriptions until the indicator moves.
Under time pressure, those additions tend towards near-duplicates. Same message, slightly rephrased. The rating rises because the boxes are full, and the system gains nothing genuinely new to combine.
Worse, diluting a sharp message across fifteen similar assets can produce combinations that say less than your original three did. So the process of reaching Excellent may actively work against the specificity that makes an advert land.
That would explain the direction of the finding without requiring the rating to be badly designed. It is measuring completeness and being used as though it measures quality.
Where Google Ad Strength is genuinely useful
At the bottom of the scale rather than the top, and that distinction matters.
A Poor rating usually means something real. Too few assets, excessive repetition, or nothing for the system to work with. Fixing that is worth doing.
Between Average and Excellent is where the evidence says the signal breaks down. Moving from Average to Good to Excellent involves adding more of what you already have, and the data suggests that stops helping and may hurt.
So use it as a floor check. If it flags Poor, act. If it says Average, that is not a problem to solve, and the accounts producing the best cost per acquisition in that dataset were sitting exactly there.
What to optimise instead
Four changes, all from the same dataset, all cheaper than chasing a rating.
Sentence case rather than title case. The largest single effect measured, free to implement, reversible in minutes.
Shorter headlines on at least some assets, under twenty characters where the message allows.
Partial pinning rather than full pinning. The same analysis found partial performed best and full worst, which follows from the mechanism you are relying on.
And distinct messages rather than variations. Supply assets that say genuinely different things, so the combinations are meaningfully different rather than rephrasings.
Every one of those improves the advert without reference to the indicator, which is the point.
The general lesson about platform indicators
This is a clean example of something worth carrying beyond search.
Platforms provide diagnostic indicators. Those indicators serve the platform’s model of good practice, which is shaped by what the platform’s systems work best with.
That is not manipulation. It is what an indicator built by an interested party looks like, and it means the indicator measures conformity to a recommended approach rather than outcomes.
So treat any platform quality score the same way. Useful for catching omissions, unreliable as a performance target, and never a substitute for the number attached to money.
The equivalent trap exists on Meta with quality rankings, and the discipline transfers directly.
What Google Ad Strength cannot see
Worth naming, because it explains the disagreement neatly.
The rating cannot see your landing page, your offer, your margin, your close rate or who your customer is. It evaluates the assets in front of it against a pattern.
Cost per acquisition sees all of those, because it is the outcome of all of them.
So a rating and a result are not measuring the same thing and were never going to agree reliably. Expecting an asset completeness score to predict business outcomes was always optimistic, and the data has now established how optimistic.
What to tell a client about Google Ad Strength
This comes up often enough to be worth scripting, because the rating is visible to anybody who logs in.
A client seeing Average next to their adverts will reasonably ask why it is not Excellent. The honest answer takes two sentences.
The rating measures how complete your asset set is, not how well the adverts perform. Across the largest independent analysis available, the adverts rated Average produced roughly half the cost per acquisition of those rated Excellent.
Then show them the cost per acquisition. That moves the conversation from a label to the number the business actually cares about, and it usually ends there.
Having that explanation ready matters, because without it the pressure to chase the rating comes from outside the account and is hard to resist in the moment.
What to change this week
Three steps.
Check whether Google Ad Strength appears as a target in any report or objective in your account, and remove it if so.
Then apply the four changes above, starting with sentence case, since it is free and has the largest measured effect.
Then compare cost per acquisition across your ad groups against their strength ratings. If your best performers are rated Average, you have replicated the finding in your own account.
Responsive search ads covers the wider disagreement between the two sources, ad headlines covers the brevity and casing evidence, fund the ad set and read what the algorithm chose is the position underneath, and creative benchmarks are mostly invented covers how to check any figure you are given. What I do across paid, SEO and the website treats platform indicators as diagnostics rather than goals. To review what your account is currently optimising towards, book a teardown.