How often keywords carry no Quality Score at all

You meet it the first time you add the Quality score columns on the Keywords page and export the result. Instead of a number from 1 to 10, a large share of the rows show a dash in the Quality Score column.

The tempting reading is that this is a new keyword problem which will clear itself once traffic builds. Google's explanation does not promise that. It names a condition, and a keyword can sit below that condition indefinitely.

Google's explanation is narrow and worth quoting closely: a dash in the Quality Score column means there are not enough searches that exactly match your keywords to determine a keyword's Quality Score. Quality Score rests on historical impressions for exact searches of the keyword, so the quantity that decides whether a score exists is exact match search volume, not clicks, not spend, and not impressions won through broader matching.

Google does not publish the threshold. There is no number to aim at and no way to tell from the interface how close a keyword sits to it, and this page does not guess at one.

Two documented details follow from the same rule. Broadening a match type does not change how Quality Score is calculated, but it does let a keyword enter more auctions, which raises the chance of accumulating enough exact match data for a score to appear. And a broad match keyword can report a score with no impressions of its own, provided a corresponding exact match keyword in the same ad group has impressions in the last 90 days, in which case both report the same score.

The cost lands in reporting. Almost every account level average Quality Score is computed over the keywords that have one, because those are the only rows with a number to average. On these readings that is under half the account on the first measurement and just over a third on the second.

So the headline figure describes a self selected subset, and it moves for reasons that have nothing to do with quality: which keywords happened to clear the exact match volume bar that week. A dashboard tile showing an average Quality Score, and a client reading it as the state of the account, are describing different things.

It matters because the blank triggers work. A missing score gets read as a bad score, and it produces landing page rebuilds, ad rewrites and bid changes aimed at a problem the data does not show. The documented constraint is exact match search volume, which is mostly outside the advertiser's control and is not fixed by editing the ad.

Knowing the unscored share is normal changes what you promise. Coverage is not a target, and a keyword without a score is not a keyword to pause.

A null score is not a zero and not a one out of ten. It is an absence: Google reported nothing for that keyword on that day.

It is not a penalty and it is not evidence that the ad or the landing page is poor. Google describes Quality Score as a diagnostic tool, says it is not an input in the ad auction, and says it should not be optimised or aggregated with the rest of your data.

It is not a stable quantity either. These two readings are nine days apart and differ by roughly twelve percentage points, so treat any single figure, including both of these, as a snapshot.