Each keyword in a Google Search ad group has its own Quality Score. And I knew the rough formula was ad relevance × expected CTR × landing page experience. So how do I structure my ad groups to push that score up — and bring CPC down?

I cared because ad group structure is one of the few things you fully control in Search ads. Get it right and you lower CPC without touching bids.

The wrong assumption I had

I was treating Quality Score as the lever Google plugs into the auction. It’s not. Quality Score is a diagnostic report card at the keyword level. The actual auction uses something different: auction-time ad quality.

The factors are similar — expected CTR, ad relevance, landing page experience — but the live auction also pulls in real-time signals like search term, device, location, time of day, and competition. The 1–10 number in your dashboard is Google’s diagnostic that correlates with auction-time quality. It’s not what gets plugged in.

What I learned

Two things changed how I think about this.

First, the right question isn’t “how do I maximize Quality Score?” It’s: how do I group keywords so each ad group has one clear search intent, one relevant ad, and one relevant landing page? When that’s true, your auction-time quality goes up — and lower CPC tends to follow.

The decision rule I’m using now: keep keywords in the same ad group only if all three are true.

• They can use almost the same ad headline
• They deserve the same landing page
• They represent the same search intent

If any of those is no, split them.

A “running shoes” searcher and a “shoe repair” searcher want completely different things. One ad and one landing page can’t serve both well.

Second — and this is the part that mattered most for how I think about Search ads — don’t over-optimize for Quality Score itself. A keyword with a lower Quality Score can still be valuable if it converts well. The goal is profitable conversions, not a perfect score. QS is a useful health check, not a target.