Quality Score: What It Actually Measures (And What It Doesn't)

Every Google Ads guide tells you to "improve your Quality Score." Most of them then list 5 to 10 things to do. Almost none of them mention that Quality Score (QS) is not a lever you pull. It is a number Google calculates based on three structural inputs that live upstream of the Quality Score column itself.

Trying to improve Quality Score directly is like trying to lower your fever by sticking the thermometer in cold water. The fever is real. The thermometer reading is not where the cause lives. The patterns repeat across $150M+ in managed ad spend.

What Quality Score actually is (and what it isn't)

Quality Score is a 1-to-10 score Google gives every keyword in your account. You see it in the keyword view of the Google Ads dashboard. It updates on a batch schedule, not live.

Three component ratings feed it: Expected Click-Through Rate (CTR), Ad Relevance, and Landing Page Experience. Each one is rated Below Average, Average, or Above Average against other advertisers who showed for the same searches in the last 90 days. Google's own Quality Score documentation calls it "a diagnostic tool." That wording matters more than people realize.

Quality Score affects what you pay per click and where your ad shows. Google's auction documentation is direct about the mechanism: a high-quality ad can win a higher position at a lower price than a competitor with a higher bid but lower quality. That is the real lever. It is not a bonus you earn for being good. It is the auction rewarding the ad that is more likely to satisfy the searcher.

What Quality Score is not:

• A direct lever you can pull
• A reliable measure of campaign health on its own
• Calculated the same way in every account
• A real-time metric (it lags actual changes by 24 to 48 hours, often more)
• The cause of your CPC

That last one is where most operators get it wrong. CPC is set by the auction. Quality Score is one input, weighted against your bid, your competitor signals, and the searcher's context. The cost-per-click side of the picture goes deeper in how Google Ads cost depends on CPC and conversion rate. For the wider audit context, where Quality Score sits in the full Google audit method shows the 8-stage view.

The three component scores explained for operators

The 1-to-10 Quality Score is a roll-up of three component scores. Each one rewards a different decision. Each one punishes a different mistake.

Expected Click-Through Rate. Google's prediction of how often this keyword's ad will be clicked when it shows, normalized against the position it would appear in. The number is not your actual CTR. It is what Google expects, based on history.

What changes it: ad copy that matches the keyword's intent, headlines that include the search term word for word, and ad extensions that increase the visible surface area. What does not change it: higher bids, more keywords in the ad group, or fancier creative on the landing page.

Ad Relevance. A measure of how closely your ad copy maps to the keyword's intent. It rewards ad groups that are tight around one idea. It punishes ad groups that try to cover too many keywords with the same generic ad.

What changes it: splitting broad ad groups into thematically tight groups (5 to 15 keywords per group, one intent), and putting the keyword in at least one headline. Google's documentation on the three components confirms it is measured per keyword against the ad shown for it. What does not change it: the landing page, the offer, or the price.

Landing Page Experience. A rating of how relevant, useful, and fast your landing page is for the keyword and ad that drove the click.

What changes it: ad-to-page congruency (the page promises what the ad promised), page speed under 3 seconds on mobile, and a clean mobile layout. What does not change it: making the page longer, packing more keywords into the page, or adding generic trust badges in the footer.

Why most "improve your Quality Score" advice doesn't work

The internet is full of Quality Score boosters. Most of them target the wrong layer. Here are the four common ones and why each one fails.

1. "Add more keywords to your ad group." Expected CTR is calculated per keyword, not per ad group. Adding 50 broad keywords does not raise the score of the keywords already in there. It usually drops the score, because the new keywords dilute the ad group's thematic tightness. The ad you wrote for one intent now has to serve five intents poorly.

2. "Increase your bid to improve position." Expected CTR is normalized against the position the ad would appear in. A higher position with the same click rate does not raise Expected CTR. It just costs more. The auction rewards quality at every position, not bid alone.

3. "Add a privacy policy, trust badges, and a phone number." Landing Page Experience is measured against the relevance, speed, and mobile usability of the page for the specific keyword and ad. Adding generic trust signals does not change ad-to-page congruency. If the ad promised "30% off cleansers" and the page is a generic homepage, no badge fixes that gap.

4. "Use Dynamic Search Ads (DSA) to let Google match keywords." DSA auto-generates headlines from your site content. Those headlines are often loosely matched to the auto-targeted search query. The result is a low Expected CTR rating on the auto-generated keyword pile. Quality Score across DSA campaigns is usually one to two points below comparable manual campaigns.

The pattern across all four: they treat the symptom and never touch the cause. The structural fixes the audit method catches do more in one pass than every "QS booster" tip combined.

The order of operations to actually lift Quality Score

Order matters. Each step depends on the previous one being clean. Skip step 1 and step 2 will not stick. Skip step 2 and step 3 cannot earn the credit.

Step 1. Restructure ad groups for thematic tightness. One tight intent per ad group. Five to fifteen keywords per group, all variants of the same idea. If you cannot describe an ad group's intent in one short sentence, the group is too broad. Split it.

Step 2. Write ad copy that matches each ad group's intent. The keyword should appear in at least one headline. Sitelinks, callouts, and structured snippets should reinforce the same idea. Do not reuse one generic ad across 12 different ad groups. That is the single biggest lift on Ad Relevance.

Step 3. Build a landing page that matches the ad group. One page per ad group is the gold standard. One page per category is the realistic minimum. The page headline should echo the ad's promise. The offer should be visible in the first viewport. The page should load in under 3 seconds on a mid-range Android phone. That is what moves Landing Page Experience.

Step 4. Tighten match-type discipline. Exact match is the cleanest read on intent. Phrase match goes in a separate ad group so its data does not contaminate exact. Broad match is dangerous outside mature accounts with strong negative keyword hygiene. Match types matter here because Expected CTR is a per-keyword calculation, and broad match keywords often pick up off-intent queries that drag the average down. For the deeper post on this decision specifically, see the Match-Type Discipline.

Step 5. Wait. Quality Score updates on a delay. Most accounts see meaningful movement 2 to 4 weeks after structural changes. If you check it the next day and nothing has moved, that is normal. The auction needs new impression data before the components recalculate.

Do these in this order and Quality Score lifts on its own. Skip the order and you get to start over.

When Quality Score doesn't matter

The honest contrarian close. Four cases where chasing Quality Score is a waste of focus.

Quality Score rewards good structure. It punishes bad structure. But it does not cause good or bad performance. It reflects it. The accounts that move the number sustainably treat it as the readout, not the goal.