Quality Score is Google's 1-to-10 rating of the overall experience your ad and landing page provide to someone searching for a given keyword. It is calculated from three components: expected click-through rate, ad relevance, and landing page experience. Think of it as Google's shorthand for how well your ad answers the question someone just typed in. It is not a single metric stored somewhere in a vault; it is recalculated at auction time, and the number you see in your account is a rounded snapshot, not a live feed.
Quality Score directly affects how much you pay per click and whether your ad shows at all. A higher score means Google charges you less for the same position, because it trusts you to deliver a good experience. An account full of 4s and 5s is bleeding money on every click. We have seen accounts cut cost-per-acquisition by 30% or more simply by fixing the underlying Quality Score issues, without touching bids or budgets.
Google evaluates three sub-components independently. Expected CTR predicts whether your ad will get clicked, based on historical performance for that keyword and match type. Ad relevance measures how closely your ad copy matches the searcher's intent. Landing page experience assesses load speed, mobile-friendliness, content relevance, and ease of navigation. Each sub-component is rated "Below Average," "Average," or "Above Average," and these combine into the 1-to-10 score. The practical lever is this: fix whichever sub-component is dragging, and the score moves.
The biggest mistake is chasing the number itself. Agencies will restructure entire accounts to push a 6 to a 7, spending hours on something that barely moves CPA. The score is a diagnostic tool, not a KPI. Another common miss: ignoring landing page experience entirely and trying to fix Quality Score with ad copy alone. If your page loads in six seconds on mobile or dumps the visitor on a generic homepage, no amount of keyword insertion will save you. We also see marketers panic over low scores on brand terms, which is almost always a data lag issue, not a real problem.
Straight answers about how Quality Score actually works and where most people get it wrong.
For non-brand keywords, 7 or above is solid. Scores of 8-10 are excellent and typically mean you are paying below the average CPC for that auction. For brand terms, expect 8-10 as a baseline; anything lower usually signals a structural issue like a competitor bidding on your name or a mismatch between your ad and landing page. Do not obsess over getting every keyword to 10. A 7 that converts profitably is worth more than a 10 that drives irrelevant traffic.
Yes, and that is usually the right move. The three levers are tighter ad copy relevance, better landing page experience, and improved expected CTR through more specific ad groups. Rewriting ads to match search intent more closely and sending traffic to dedicated landing pages instead of your homepage are the two changes that move the needle most. Bids influence auction dynamics but they are not a direct input to Quality Score.
The 1-to-10 Quality Score you see in your account is specific to Search campaigns. Display, Video, and Performance Max use their own relevance signals behind the scenes, but Google does not surface a visible score for those formats. The principles are the same, though: relevance, experience, and expected engagement all affect what you pay and where your ads appear.
The visible score in your account updates periodically, not in real time. Google recalculates the actual quality assessment at every single auction, but the number in the interface can lag by days or even weeks. This is why you should never make reactive decisions based on a score that dropped yesterday. Track the sub-components over time and look for sustained trends, not daily fluctuations.
We treat Quality Score as a diagnostic, not a target. During an audit, we look at sub-component breakdowns across the account to spot patterns: are landing pages consistently rated below average? Is ad relevance low across a whole campaign? Those patterns tell us where the structural problems are. From there, the fix is specific, whether that is landing page rebuilds, ad group restructuring, or copy overhauls. We have done this across 250+ client accounts and the pattern recognition makes the difference between generic advice and advice that actually moves performance.
