Quality Score is three ratings, not one number
Google reports Quality Score as a 1 to 10 number, per keyword, in Search campaigns only. That number summarizes three component ratings, each shown as Above average, Average, or Below average:
The three Quality Score components
- Expected CTR: how likely your ad is to be clicked when it shows.
- Ad relevance: how closely your ad matches the intent behind the search.
- Landing page experience: how relevant and useful your page is to the people who click.
Every rating is relative. Google compares you against the other advertisers whose ads appeared on the same searches over the previous 90 days. "Average" means you're in the middle of that pack, not that your page is fine.
Use it as a diagnostic
Quality Score is a diagnostic, not a metric to optimize. Google says so in its own docs. The live auction uses a real-time quality estimate, not the 1 to 10 number sitting in your report, so moving that number from 5 to 8 isn't the goal. Finding which component is dragging and fixing the cause is.
What goes into landing page experience
Google doesn't publish a formula. It does publish the inputs it cares about, and they're consistent across its documentation.
What Google looks for
- Relevance to the query and the ad: someone searched for a specific thing, clicked an ad promising that thing, and should arrive on a page about that thing. Pointing "SOC 2 compliance software" traffic at your homepage is the most common cause of a below-average rating, and the hardest to fix with copy tweaks.
- Useful, original content: real information about the product, not thin copy assembled to catch traffic. Google's crawlers read the page, and its broader page-quality signals feed the rating.
- Ease of navigation: important information near the top, and nothing in front of it. Google names the number of links on a page as a factor, which is one reason a full site nav plus footer sitemap scores worse than a focused page. Interstitials and popups count against you too.
- Speed and mobile behavior: most paid search clicks are mobile. Slow pages and layouts that break on a phone both show up here.
- Transparency: clear about who you are, what you collect, and what happens after someone fills the form.
Every item on that list is a proxy for one question: did the person who clicked get what the ad promised? Google's own summary is "give your users what they're looking for and good performance should follow."
Why below-average ratings persist
Most teams sitting at below average already know all of this. The constraint is structural.
A typical Series A or B SaaS team runs 15 or 20 ad groups against three landing pages, because three pages is what one or two marketers could get through design and web review this quarter. Every ad group past the first points at a page that's approximately right. Google notices, charges more per click, and the team responds by rewriting ad copy, because ad copy is the part they can change this week.
The page is the expensive part of that loop. That's what Flint changes.
What Flint does about it
Flint generates a page per ad group instead of one per campaign. It reads the keywords and search terms from your connected Google Ads account, groups them by intent, and writes each page in the language of the query rather than your internal product vocabulary. That's the change that moves landing page experience, because it addresses relevance directly and the other fixes don't.
The rest is maintenance. Flint compares your live ad headlines and descriptions against the page they point to and flags where they've drifted apart, which happens on its own: ads get rewritten and pages don't. It reads the per-keyword component columns from your account and tells you which keywords are rated below average, so you hear it from Flint instead of from a CPC that crept up over a quarter. And when a page change ships, it tracks the historical component columns for those keywords alongside CPC and CPA, so you can see whether the change did anything.
Pages published through Flint are fast and responsive by default, with no popups or interstitials. That's table stakes rather than a differentiator, but it means load time and mobile layout stop being things you have to remember.
How to read the results yourself
In the Google Ads keyword report, enable Quality Score, Landing Page Exper., Ad Relevance, and Exp. CTR, along with the historical versions of each. The historical columns are the useful ones: they let you compare the rating before and after a page change instead of only seeing today's state.
Give it time. Because the rating is computed against a rolling 90-day comparison window, expect two to four weeks after publishing a new page before it settles, and expect it to move in steps rather than smoothly. Keywords pointing at the same URL usually move together.
Then judge the change on cost per click and cost per acquisition, not on the rating. A page that lifts landing page experience from Below average to Average and does nothing to CPA hasn't earned its place.
A caveat if you're running AI Max
With final URL expansion turned on, Google may replace the final URL you set with a different page on your domain when it thinks that page is a better match for the query. Under that setting, the quality of every page Google can send traffic to matters, not just the one you configured, and a new page you built may not be getting the traffic you think it is.
You can control which URLs are eligible for expansion in campaign settings. Check that before concluding a page change had no effect.
Sources
- About Quality Score for Search campaigns (Google Ads Help)
- 5 ways to use Quality Score to improve your performance (Google Ads Help)
- Optimize your ads and landing pages (Google Ads Help)
- Landing page experience (Search Ads 360 Help)
- About Final URL expansion in Search (Google Ads Help)
