Quality Score is a diagnostic, not a goal

Quality Score is reported as 1–10 per keyword. Treated as a score to maximise it is a distraction. Treated as three separate diagnoses it is one of the more useful things in the interface.

The three components

Each keyword carries three sub-ratings, each Below Average, Average, or Above Average:

Expected click-through rate — will people click this ad for this query, relative to competitors at the same position.

Ad relevance — does the ad text correspond to the query.

Landing page experience — is the page useful, relevant, fast and navigable for someone arriving from this query.

The composite hides which one is failing. The components tell you which team has the work.

Reading the combinations

Ad relevance below average, everything else fine. The keyword is in the wrong ad group. It is being served an ad written for adjacent intent. This is the cheapest of all fixes: move it, write the ad for it.

Expected CTR below average, ad relevance fine. The ad is on-topic but unappealing. Usually a weak offer, or no differentiation from three competitors saying the same thing.

Landing page experience below average. Frequently the query is a mismatch for the destination — the ad group sends everything to one page regardless of what was asked. Sometimes it is genuinely speed or mobile usability.

All three below average. The keyword probably does not belong in the account.

What it is worth financially

A better Quality Score reduces the cost of a given ad position. That is real money, but the effect is often overstated in ways that encourage optimising the number for its own sake.

The correct framing: Quality Score components tell you where the mismatch is between query, ad, and page. Fixing the mismatch improves conversion rate, which is worth considerably more than the CPC discount.

What not to do

Do not pause low Quality Score keywords reflexively. A keyword with a 4 that converts profitably is a keyword that converts profitably. The score is Google's estimate of experience, not your estimate of value.

Do not build hundreds of single-keyword ad groups chasing relevance. That was a rational strategy once. With modern matching and automated bidding it fragments your conversion data across too many small buckets to learn from, which costs more than the relevance gains.