Quality Score is a 1–10 diagnostic estimating how relevant your ad, keyword and landing page are to a searcher. It is not the number used in the live auction — that is a real-time ad rank calculation — but it is a reasonable proxy and the only view you get.
The three components
Each is reported as below average, average or above average, relative to other advertisers on the same keyword over the past 90 days.
Expected click-through rate. How likely your ad is to be clicked, adjusted for position. The heaviest-weighted component in most cases.
Ad relevance. How closely the ad text matches the intent behind the keyword.
Landing page experience. Whether the page delivers what the ad promised, loads quickly, works on mobile and is transparent about who you are.
The reported Quality Score is keyword-level and historical. A new keyword starts with an estimate based on account history, which is why established accounts have an advantage on day one.
What it actually affects
Two things: whether your ad shows, and what you pay per click.
A higher quality signal means you can win the same position at a lower bid. The effect is real and frequently overstated — improving from average to above average on all three components typically reduces cost per click meaningfully, not by an order of magnitude.
It also affects eligibility for ad extensions and for certain positions, which matters more than the cost difference in competitive auctions.
Improving expected click-through rate
This is mostly an ad copy and targeting problem.
Include the keyword in the headline, because matched terms are bolded and that visibly increases clicks. Write headlines about the outcome rather than the product. Use all available headline and description slots, since more real estate means a larger, more clickable ad.
Extensions matter more than most advertisers assume: sitelinks, callouts and structured snippets increase the ad’s size and give the searcher more reasons to click.
And remove keywords that consistently underperform rather than trying to rescue them. A low-relevance keyword drags the account and rarely recovers.
Improving ad relevance
This is an account structure problem more than a copywriting one.
Ad groups containing fifty loosely related keywords cannot have ad copy relevant to all of them. Tight themes — a handful of closely related keywords per ad group, with ad copy written for that specific theme — is the structural fix.
Modern automated bidding and broad match have made very granular structures less necessary than they once were, but the principle holds: the ad should read as though it were written for the search.
Negative keywords matter here too. Irrelevant traffic matched by broad keywords produces poor engagement, which feeds back into the signal.
Improving landing page experience
The component most often ignored, and the one where improvement also benefits everything else.
Message match. The page should continue the ad’s promise. An ad about a specific feature should land on a page about that feature.
Speed. Slow pages are penalised here and lose conversions independently. See Core Web Vitals.
Mobile. A large share of paid traffic is mobile, and pages that technically work but are awkward on a phone underperform.
Transparency. Clear information about who you are, how to contact you, and what you do with data. This component includes trust signals explicitly.
Our notes on landing page conversion cover the structure that works.
What you cannot control
Three things worth accepting rather than fighting.
Competitor behaviour. The components are relative. If a competitor improves, your relative score can fall with nothing changed on your side.
Keyword inherent difficulty. Broad, high-volume keywords with mixed intent will rarely score well, because no single ad can be relevant to everyone searching them.
The lag. The score reflects 90 days of history. Changes take weeks to show, and the data accumulates slowly on low-volume keywords.
Using it correctly
Treat Quality Score as a diagnostic that points at problems, not as a metric to optimise for its own sake.
The practical workflow: sort keywords by spend, look at the component breakdown for the highest-spending keywords with below-average components, and fix those specifically. A keyword with a score of 4 and no spend is not worth an afternoon.
And keep the real objective in view. A score of 10 on a keyword that produces no conversions is worthless; a score of 6 on a keyword producing profitable sales is fine. Cost per acquisition and return on ad spend are the numbers that matter.
Account structure in the era of automation
Automated bidding has changed what good structure looks like. Very granular single-keyword ad groups, once standard practice, now often starve the algorithm of the data it needs to optimise.
The current consensus is fewer, larger campaigns with enough conversion volume for automated bidding to learn, combined with tight thematic ad groups so relevance stays high. Exactly where the balance sits depends on your conversion volume — below a few dozen conversions a month, automated strategies struggle regardless of structure.
Search terms, not keywords
The single most productive recurring task in a search account is reading the search terms report — the actual queries that triggered your ads, as opposed to the keywords you bid on.
With broad match and automated bidding, the gap between the two is wide. Reading the report weekly produces three kinds of action: negative keywords for irrelevant queries that are consuming budget, new keywords for queries converting well that you were not targeting deliberately, and ad copy ideas from the language people actually use.
Build a shared negative keyword list at account level for the categories that are never relevant — jobs, free, tutorials, and competitor names where you do not want to compete — and a campaign-level list for the finer distinctions.
Reading the account by spend
Account optimisation attracts attention to the wrong places because the interface sorts by convenience rather than by importance.
Start every review by sorting by cost, descending. The top twenty line items by spend usually account for most of the budget, and a problem there is worth more than a dozen fixes in the tail.
Then check three things on each: is it converting at an acceptable cost, is the search term report clean, and does the landing page match. Most accounts have one or two expensive keywords quietly failing all three, and fixing those moves the account more than any structural reorganisation.
A short weekly routine
Most search accounts need about an hour a week rather than a quarterly overhaul, and the routine is consistent.
Read the search terms report and add negatives. Check spend and conversions against the prior week for anything that moved sharply. Look at the top five keywords by cost and confirm each is still earning its place. Check that nothing is disapproved or limited by budget unexpectedly.
Leave structural changes for a monthly session. Accounts that are restructured weekly never accumulate enough stable data for automated bidding to learn, which makes everything worse.
Frequently asked questions
What is a good Quality Score?
Seven or above is healthy for commercial keywords. Branded keywords often reach nine or ten; broad generic keywords rarely exceed six.
Does Quality Score affect organic rankings?
No. They are entirely separate systems, although the work that improves landing page experience often benefits both.
How quickly does it change?
It reflects roughly 90 days of history, so expect several weeks before improvements are visible, and longer on low-volume keywords.
Should I pause low Quality Score keywords?
Only if they also perform poorly commercially. A keyword with a score of 5 that converts profitably should stay.
