Why Your Google Ads Quality Score Is Low (And the Fix)
Written from AfM's working notes and the public sources listed at the end. Examples marked illustrative are hypothetical, not client results.
Updated 27 September 2026
3 sources cited.
First published 17 February 2026
AfM guide
Search, brand voice, publishing systems, and topical authority.
On This Page
The short answer
Quality Score is a rating from 1 to 10, shown per keyword, that estimates how relevant and useful your ads and landing pages are for people searching that keyword, compared with other advertisers (Google Ads Help). Google describes it as a diagnostic tool. It is not an input in the ad auction itself. The auction uses its own real-time assessment of quality at the moment of each search, and Quality Score is a summary that helps you see where that quality is weak.
That distinction matters. Chasing a higher number is pointless. Fixing what the number points at (ads that do not match the search, or pages that disappoint) improves how your ads compete and what happens after the click.
The three components
Quality Score combines three components, each rated Above average, Average or Below average, based on comparison with other advertisers whose ads showed for the same search over the last 90 days:
- • Expected click-through rate: how likely your ad is to be clicked when shown for that keyword.
- • Ad relevance: how closely the ad matches the intent behind the search.
- • Landing page experience: how relevant and useful the page is to the person who clicked.
Add these three columns to your keyword report. The overall score tells you there is a problem; the components tell you which one.
Diagnose by component
Below-average Quality Score components: usual causes, fixes and where AI copy helps
| Component below average | Usual causes | Fix | Does AI copy help? |
|---|---|---|---|
| Ad relevance | Ad group contains keywords with different intents; headlines never mention the searched theme | Split ad groups by intent; include the keyword theme in two or three headlines | Yes. Drafting theme-specific headlines per ad group is fast with AI, provided you check accuracy |
| Expected CTR | Generic headlines; no reason to choose you over the ads beside you; missing sitelinks and callouts; showing for loosely related searches | Write distinct reasons to click (price, proof, speed, specific use case); add assets; tighten matching and negatives | Partly. AI can draft angles, but the reason to click must be real, and matching problems are fixed with keywords and negatives, not copy |
| Landing page experience | Page does not continue the ad's promise; slow on mobile; hard to find the next step; generic homepage used for every keyword | Send each theme to the page that answers it; match the headline; speed up mobile; one clear action | Rarely. The problem is the page, its speed and its structure. AI can draft page copy, but someone must build and test the page |
Fix in this order
1. Sort by money, not by score. A keyword with a score of 3 and almost no spend is not your problem. Filter to keywords with meaningful spend and at least one Below average component.
2. Fix structure before copy. If an ad group mixes "accountant for landlords", "limited company accounts" and "tax return help", no single set of headlines can be relevant to all three. Split them. This often lifts ad relevance and expected CTR together.
3. Rewrite the ads for each new group. Each responsive search ad takes up to 15 headlines of up to 30 characters and 4 descriptions of up to 90 characters (Google Ads Help). Use two or three headlines to echo the search theme and the rest for distinct reasons to click. This is where AI drafting saves time.
4. Clean the matching. With broad match and AI Max, ads can show for searches you would never have chosen. Low expected CTR is often a matching problem. Review the search terms report and add negatives.
5. Fix the landing page last, but do fix it. It is the slowest to change and often the biggest lever for conversions. The page should repeat the promise of the ad in its headline, load quickly on a phone, and make the next step obvious.
6. Wait before judging. Components are based on historical comparison. Give changes several weeks of traffic before reading the new statuses.
An illustrative example
This is an illustrative example with a hypothetical firm.
An accountancy firm has one ad group, "accountant", containing twelve keywords from "landlord accountant" to "self assessment help". The keyword "landlord accountant" shows Quality Score 4: ad relevance Below average, expected CTR Average, landing page Below average. The ads talk about "Expert Accountants for Your Business" and click through to the homepage.
The fix: a new ad group for landlord keywords; headlines such as "Accountants for Landlords" and "Rental Income Tax Returns"; a landing page about landlord accounting with the same headline; the homepage no longer used. Ad relevance and landing page are the components expected to move. Whether conversions rise is the real test.
Where AI copy helps and where it misleads
AI is well suited to producing theme-specific headline sets for many ad groups quickly, which is the most common fix for poor ad relevance. It is poorly suited to deciding what is true about your business, and it cannot fix a slow or off-topic page. Give it the keyword theme, your substantiated facts and a character limit; review everything it writes.
Do not confuse Quality Score with Ad Strength. Ad Strength rates how an individual responsive ad's assets are set up and is not an auction factor either; see why Ad Strength shows Poor. For a full Google Ads build process, see how to create Google Ads with AI.
FAQ
Does a higher Quality Score lower my cost per click?
Quality Score itself is not used in the auction, but the quality factors it summarises are part of how the auction ranks ads. Improving real relevance and landing page experience tends to improve position or cost; raising the reported number without changing either does nothing.
Why do some keywords show no Quality Score?
Quality Score needs enough impressions for that exact keyword to compare with other advertisers. New or low-volume keywords often show a blank until they have enough history.
My landing page is fine, so why is it rated below average?
The rating is relative to other advertisers and to the specific search. A good general page can still be below average for a specific keyword if competitors send that search to a page built for it. Check whether the page answers that keyword's question in its first screen.
Should I pause keywords with low Quality Score?
Only if they also convert poorly. Some low-scoring keywords are profitable because the buyers they bring are valuable. Decide on cost per conversion and conversion value, using Quality Score to diagnose rather than to judge.
Sources
- About Quality Score for Search campaigns, Google Ads Help. Cited 27 September 2026 from a published summary; the primary page has not been re-read. 1 to 10 keyword-level score; three components with status ratings from a 90-day comparison; diagnostic tool, not an input in the ad auction.
- 5 ways to use Quality Score to improve your performance, Google Ads Help. Cited 27 September 2026 from a published summary; the primary page has not been re-read. Using Quality Score as a diagnostic in pursuit of performance goals.
- About responsive search ads, Google Ads Help. Cited 27 September 2026 from a published summary; the primary page has not been re-read. Headline and description counts and character limits.
Want to know where your search ads lose quality?
Request a free written breakdown of your public ads and landing pages, plus any figures you share. No logins needed.
Request a breakdownRelated Articles

AI Content Creation: Scale Without Losing Quality
Where AI content loses quality, what Google actually penalises, the input pack that prevents generic drafts, a scored editorial gate you can use on any draft, and why review hours set your real volume.
Read more →
Scaling Content Production with AI on a Fixed Team
AI can draft an article in minutes, but your content output is set by how much a person can properly check. Work out your real capacity, move the bottleneck deliberately, and scale without new hires or lower standards.
Read more →
AI Blog Writing: From Brief to Published Post
The nine steps between an idea and a published post when AI does some of the writing: the brief template, the prompts, the checks a human must own, and what to measure afterwards.
Read more →