Google Ads Quality Score: what actually moves the number

Quality Score is an estimate, on a one-to-ten scale, of how well your ad answers a search. It isn't a performance grade or a judgment of the business — it's a prediction, and understanding that changes what you do with it.

The three components

Expected click-through rate. How likely Google thinks people are to click your ad for that keyword, compared with other advertisers. It carries the most weight.

Ad relevance. How well the ad copy corresponds to what the person searched for.

Landing page experience. Whether the page delivers what the ad promised, loads fast, and works on a phone.

Each appears as "above average", "average" or "below average". That classification is more actionable than the final number, because it points at where the problem is.

Why it affects price

Ad Rank combines your bid with estimated quality. With better quality you get the same position at a lower bid, or a better position at the same bid.

That's the whole mechanism, and it explains why two advertisers bidding the same amount pay differently. It also explains why working on the ad and the page usually pays more than working on the bid: one improves the price structurally, the other only changes the ceiling.

What you can move and what you can't

Of the three components, two are under direct control and one only partly.

Ad relevance is the fastest to fix: it's almost always corrected by splitting ad groups that mix topics and writing specific ads for each. An afternoon of work with an effect in weeks.

Landing page experience is also actionable, though slower because it involves more people. What weighs most in practice: that the page addresses the same topic as the search, loads fast on mobile, and doesn't make anyone hunt for the promised information.

Expected click-through rate is the hardest to move directly, because it depends on how people react. It improves as a consequence of the other two more than through direct action.

The mistake of treating it as a goal

There are accounts where someone decided the score had to reach eight, and every decision started serving that number.

That's a mistake, because the score is a means. A keyword with a score of five and excellent cost per conversion is a better deal than one with a ten and bad cost per conversion.

The right read: use the score as a diagnostic when cost is high, not as a management metric. If cost per conversion is where it needs to be, a mediocre score isn't a problem that needs solving.

When a low score is expected

Some situations produce a low score with nothing done wrong.

Generic top-of-funnel keywords. Broad searches have structurally lower click-through rate because intent is diffuse.

New terms with no history. The score needs data; early on it works from conservative estimates.

Markets where you compete against very well-known brands. Expected click-through rate is measured against everyone else, and a recognized name wins clicks through recognition, not through a better ad.

In all three cases, insisting on raising the score is fighting the structure of the market.

Diagnosing in practice

Add the three component columns to your report, not just the overall score. Then look at the pattern.

If the problem is in relevance, the fix is structural: split groups, write specific ads.

If it's in the page, the fix lives outside Google Ads: speed, message match, clarity of the offer.

If it's in expected click-through rate and the other two are fine, it's probably competition or diffuse intent — and the question becomes whether that keyword deserves to stay in the account at all.

The component you fix off-platform

Worth dwelling on landing page experience, because it's the component most people write off and the one with the most room.

Three concrete things move it, and none requires redesigning the site.

Message match. If the ad says "export consulting" and the landing page is the agency's homepage, the person has to hunt for what they were promised. A specific page per campaign topic solves that.

Mobile speed. Most traffic arrives from a phone, and a page that takes several seconds loses people before showing anything. It's usually heavy images and third-party scripts, not architecture.

Clarity of the action. What the person is expected to do has to be visible without scrolling.

All three improve conversion as well as the score, so the return on the work doesn't depend on Google recognizing it.

What changed with automation

With automated bidding and more automated campaign types, Quality Score lost some of the prominence it once had. Google sets the bid based on conversion probability, and manual position tuning matters less.

What hasn't changed is the underlying point: an ad that matches the search and a page that delivers the promise perform better and cost less. The score is just how Google makes that principle visible.

That's why it's still worth reading, even though it stopped being the central lever of optimization.

The ad group size question

Most relevance problems trace back to one decision: how many keywords live in an ad group.

An ad group with twenty keywords spanning different topics can't have an ad that's very relevant to any of them. The copy has to be generic enough to cover everything, and generic copy scores poorly against every specific search.

The opposite extreme — one keyword per ad group — was popular for years and has aged badly. It fragments the data so much that neither Google nor you accumulate enough signal per group, and it creates accounts nobody can maintain.

The useful middle: group keywords that would be answered by the same ad and the same page. If you can't write one headline that fits every keyword in the group, the group needs splitting.

What to review now and then

Once a month, filter the highest-spending keywords and look at their three components. No need for more frequency: the score moves slowly, and reacting to weekly variation creates noise.

What matters in that review isn't the number — it's finding the expensive keyword whose problem has a cheap fix, which is almost always a badly built ad group.

One last reading caveat: the score shown in the interface is the current value, not the one applied in past auctions. Comparing today's score with last week's cost can confuse more than it clarifies.

If the symptom is the price of the click, see high CPC on Google Ads. For the most frequent cause of wasted spend, see Google Ads search terms. For what a click actually costs and why, see how much Google Ads costs; for a local business deciding where to start, Google Ads or Business Profile. And for the weekly read without exporting, see the paid traffic report.