Quality Score

Quality Score is a rating from one to ten that Google Ads shows against each keyword in a Search campaign, summarising how well your keyword, your advert and the page behind it fit together. It is a report card. Understanding what it does not do turns out to be more useful than understanding what it does.

In short

• A one to ten rating of how relevant your keyword, advert and page are.
• Google says plainly that it is not an input in the auction.
• Built from three parts, and the parts matter more than the total.
• It measures relevance, not whether the traffic is profitable.

What it does not do

Almost every explanation of this number says the same thing: raise your Quality Score and you will pay less and rank higher. Google’s own documentation says something different, in as many words. The score is a diagnostic tool. It is not consulted when the auction runs.

What runs the auction is Ad Rank, calculated fresh for every single search. Google names six inputs: your bid, the quality of the advert and landing page assessed at that moment, the thresholds it requires before showing an advert at all, how competitive the auction is, the context of the search including device and location, and the expected effect of any extra formats attached to the advert.

So the causal story is the other way round from how it is usually told. Relevance influences the auction directly. Quality Score describes that same relevance to you afterwards, using data averaged over past impressions. Accounts with good scores do pay less, which is why the myth survives, but improving the number is not the mechanism. Improving what the number measures is.

The three parts

The total is the least informative thing on the screen. Each component is rated below average, average or above average, and each points somewhere different.

• Expected click-through rate. Whether people are likely to choose your advert when they see it for this keyword. Below average here is an advert problem. The copy is not answering the search.
• Ad relevance. Whether the advert reads as though it was written for that keyword. Below average usually means one ad group is covering too many different searches at once.
• Landing page experience. Whether the page delivers what the advert implied, quickly and on a phone. Below average here is a website problem that no amount of ad rewriting will fix.

Add the three columns to your keyword view before you do anything else. A total of four tells you something is wrong; the components tell you which department it belongs to.

The fix for each

Ad relevance is usually the easiest and the most rewarding, because the cause is almost always structural. An ad group containing twenty loosely related keywords cannot have adverts that speak to any of them precisely, and it usually means nobody has separated the search intent behind them. Splitting it into groups of a few closely related terms, each with adverts written for those terms, fixes relevance and expected click-through rate together.

Landing page experience is the one people avoid, since it means changing the website rather than the advert. Sending every campaign to the homepage is the usual cause. The page needs to be about the thing that was searched for, load quickly on mobile data, and make the next step obvious. That is a landing page problem and it is worth the work, because it is the only one of the three that also improves what happens after the click.

Expected click-through rate responds to the advert saying plainly what the searcher asked about. It is measured against other advertisers for the same keyword, so it moves when your advert is a better match than theirs, not when it is more enthusiastic.

Where chasing the number goes wrong

Treating Quality Score as the objective produces some genuinely counterproductive behaviour, and it is common enough to name. Money usually leaks through missing negative keywords rather than through a low score, and yet businesses pause keywords with low scores that were quietly producing customers. They split ad groups past the point of usefulness, ending up with dozens of groups too small to gather data. They rewrite adverts that were converting well in order to chase a component rating. In each case a diagnostic has been promoted into a target, and the account gets worse while the dashboard gets better.

The number to manage is what an inquiry costs you. Quality Score is one of several explanations for why that number is where it is.

The words on the report

• Ad Rank. What actually decides whether and where your advert shows. Calculated per search.
• Expected click-through rate. Google’s estimate, before the advert runs, of how likely it is to be clicked.
• Ad relevance. How closely the advert matches the keyword.
• Landing page experience. Google’s assessment of the page behind the click.
• Ad group. The container holding a set of keywords and the adverts written for them. The unit most Quality Score problems are solved in.
• Below average. The component rating that names your problem. Compared against other advertisers bidding on the same keyword.

What to do this week

Open the keywords view, add the three component columns, and sort by cost rather than by score. Look only at the expensive keywords with a component rated below average. That short list is where relevance work pays for itself, and it is usually a handful of rows rather than the whole account.

Ignore everything with a dash instead of a number. Those keywords have not been seen enough for Google to have an opinion, and neither should you.