Quality Score

The three components, and why they matter more than the number

Google reports Quality Score per keyword on a 1-10 scale, alongside three component ratings, each shown as Below average, Average or Above average. The single number is a summary; the diagnosis lives in the components.

• Expected click-through rate — how likely a click is when your keyword triggers, measured against competitors with position effects stripped out. It is mostly a verdict on whether your ad's promise matches what that query wanted.
• Ad relevance — how closely the ad text matches the intent behind the keyword. A generic ad stretched across a broad ad group is the usual cause of a Below average rating here, and it is the cheapest of the three to fix.
• Landing page experience — whether the page delivers what the ad promised, quickly, on a phone, without a hunt. It is the one component that is not about advertising at all, which is exactly why it is the one most advertisers leave alone.

What it controls, and what it does not

Quality Score matters because ad rank is not just a bid. Google combines your bid with quality signals and the expected impact of your ad assets, so a more relevant advertiser can outrank a higher bidder and frequently pays less per click for the same position. That is the real mechanism behind the folk wisdom, and it is why "just raise the bid" stops working sooner than most founders expect.

What it is not is the number in the auction. The auction recalculates quality signals for every individual query in real time; the 1-10 in your keyword table is a historical, aggregated diagnostic built from exact-match history. Treat it as a smoke alarm — worth acting on when it goes off, useless as a weekly KPI. You also cannot optimise it directly: you tighten the ad group, rewrite the ad, or fix the page, and the score follows weeks later or it does not, which is itself information. For the page half of that work, see conversion rate optimization.

What a small account actually sees

Here is the part the optimisation guides skip: at a small budget, most of your keywords will have no Quality Score at all. In our own Google Ads account for AgentCeres — the AI Growth Officer at agentceres.com — every daily snapshot we have taken since late July 2026 shows roughly three in four enabled keywords reporting no Quality Score, because Google needs impression volume before it will estimate one. A dash in that column is not a bad score; it means Google has not seen enough to guess. Every guide assumes you are reading a full column. A young account is reading a mostly empty one, and the temptation is to over-interpret the handful of rows that do have data.

The rows that did have data taught us more as a pattern than as scores. On our 2 August 2026 snapshot, of the keywords carrying a Quality Score, landing page experience was rated Below average on all but two — while ad relevance was split roughly evenly between Above and Below average. That is a coherent story rather than a mystery: the ads had been rewritten per ad group several times, and the traffic still landed on pages written for a general visitor. When your components disagree, the disagreement is the instruction — it names which of the three jobs you have actually been doing. Deciding whether to keep spending at all is a separate question, covered in should I use Performance Max for my startup.