Google Ads Quality Score, Explained: What It Actually Measures and Why Chasing the Number Is the Wrong Move

Type “what is a good quality score in google ads” into Google and you’ll get a dozen confident answers, most of them slightly wrong. Someone will tell you a Quality Score of 10 cuts your cost-per-click in half. Someone else will tell you to restructure your ad groups to boost it. A third person will tell you it doesn’t matter at all anymore because everything is automated now. None of these people are lying to you on purpose — Quality Score is one of the most misunderstood metrics in all of digital advertising, partly because Google has never been fully transparent about how it’s calculated, and partly because the metric itself has quietly changed meaning as the platform has automated around it.

This post is an attempt to explain what Quality Score actually is, what it actually does, and — more usefully — what to do about it in 2026, when a growing share of your account (Performance Max, broad match, AI Max) doesn’t show you a keyword-level score at all. If you run Google Ads for a living, or you’re the person at your company who’s ultimately responsible for the account, this is the version of the explanation that skips the recycled infographics and sticks to what’s verifiable.

Quality Score has been part of the platform since 2005, back when it was introduced under the AdWords name as a way to keep auction results honest — without some measure of relevance, the highest bidder would always win the top spot regardless of whether their ad or landing page actually served the person searching. Two decades later the core idea hasn’t changed: Google still wants the auction to reward ads that are genuinely useful to the person seeing them, not just ads backed by the biggest budget. What has changed is how much of the account that idea gets applied to visibly, and that’s the part most explanations of Quality Score haven’t caught up with yet.

What Quality Score Actually Is

Quality Score is a 1-to-10 rating Google assigns to your keywords in Search campaigns, built from three components. Google’s own documentation defines them as:

• Expected click-through rate (Expected CTR) — the likelihood that your ad will be clicked when it’s shown for that keyword, compared to other advertisers, normalized for ad position so a top-of-page ad isn’t unfairly rewarded just for being seen more.
• Ad relevance — how closely your ad’s messaging matches the intent behind a person’s search.
• Landing page experience — how relevant and useful the page someone lands on actually is, given what they searched for and what your ad promised.

Each component gets its own status — “below average,” “average,” or “above average” — visible at the keyword level in the Google Ads interface. The blended 1–10 number you see is a summary of those three signals. That much is well documented and not particularly controversial. Where most explanations go wrong is in what happens next: what that number is actually for.

The Misconception Everyone Repeats: Quality Score Isn’t a Bidding Input

This is the single most important sentence in Google’s own Quality Score documentation, quoted directly: “Quality Score is not an input in the ad auction.” It’s described explicitly as a diagnostic tool — something you look at to understand how your ads are affecting the experience of the people who see them, not a lever that gets pulled during each auction.

What actually happens in the auction is different, and it’s the source of most of the confusion. Every time your keyword is eligible to show, Google runs a real-time evaluation of the same three underlying components — expected CTR, ad relevance, landing page experience — at that specific moment, for that specific search. That real-time evaluation feeds into Ad Rank, alongside your bid, the competitiveness of the auction, the context of the search, and the expected impact of your ad formats and extensions. The 1–10 Quality Score you see in your account is a separate, aggregated, backward-looking summary of how those components have scored historically. It’s correlated with what happens in the auction, but it is not what happens in the auction.

That distinction sounds pedantic until you’ve sat across from a client asking why their “Quality Score 3” keyword is still getting clicks at a reasonable cost, or why fixing the visible number didn’t immediately drop their CPCs. The visible score is a rearview mirror. The auction runs on a live feed.

Why Your Quality Score Doesn’t Update the Way You Think It Does

Here’s the part almost no explainer states plainly: the number in your account is a trailing average, built mostly from your keyword’s historical exact-match-equivalent performance. It is not recalculated fresh for every query your keyword happens to match. That’s why two things that seem contradictory are both true at once — your account-level Quality Score can look mediocre while your actual auction performance, on a given day, for a given search, is fine. The visible number is lagging what’s really happening in real time.

This also explains why Quality Score changes don’t show up instantly. High-volume keywords can see the visible score update multiple times a day, because there’s enough traffic to recalculate the trailing average quickly. Low-volume keywords might not move for a week or more, simply because there isn’t enough fresh data to shift the average. If you fix a landing page or rewrite ad copy, the realistic expectation is one to two weeks before you see initial movement, and three to four weeks before the number stabilizes at its new level. If you check the day after a fix and see no change, that’s not evidence the fix didn’t work — it’s evidence the average hasn’t caught up yet.

Breaking Down the Three Components — and What Actually Moves Each One

Since restructuring campaigns doesn’t move Quality Score (more on that below), the only real lever is improving the three underlying components directly. Here’s what that looks like in practice for each one.

Expected CTR. This is a relative measure — how your ad is expected to perform against competitors bidding on the same keyword, normalized for position. The most direct way to influence it is writing ad copy that matches the actual intent behind the search rather than generic category copy. If someone searches “emergency plumber same day,” an ad that leads with availability and speed will outperform one that leads with your company’s 20-year history — not because the second ad is worse writing, but because it’s answering a different question than the one being asked. Pinning your most relevant Responsive Search Ads headlines to the top position for high-intent queries, and using ad extensions/assets that Google’s own systems flag as increasing expected engagement, both feed this component.

Ad relevance. This is about the tightness of the match between the keyword, the ad, and the query — not the CTR you’re getting, but whether the ad’s content is topically aligned. Loose ad groups with a wide spread of unrelated keywords sharing the same handful of ads are the most common cause of a weak ad relevance signal. Tightening ad groups around a single theme, and writing RSA headlines and descriptions that use the actual language of that theme rather than generic brand messaging, is the fix. This is also where a periodic look at your search term report pays off twice over — it tells you which queries are triggering your ads, and misaligned queries are usually the first sign that an ad group’s relevance is drifting.

Landing page experience. This is judged on relevance to the ad and the search, plus more practical usability factors — page speed, mobile-friendliness, transparency, ease of navigation, and whether the page actually delivers what the ad promised. A landing page that makes someone hunt for the thing the ad advertised, or that loads slowly on mobile, or that feels like a generic homepage rather than a page built for that specific intent, drags this component down regardless of how good the ad copy above it is. This is usually the most neglected of the three components, because it lives outside the Google Ads interface — nobody’s staring at a landing page’s load time inside their campaign dashboard, so it’s easy for it to quietly degrade after a site redesign or a plugin update nobody flagged as ads-relevant.

A concrete version of this: a service business runs an ad for “same-day AC repair” that sends traffic to the homepage instead of a page built around emergency repair. The homepage covers installation, maintenance plans, financing, and repair all in one scroll — technically related, but it makes the visitor do the work of finding the thing the ad promised. Expected CTR might still look fine, because the ad copy itself is well written and closely matched to the search. Ad relevance might also hold up, since the keyword and ad text are aligned. But landing page experience quietly drags the blended score down, because the page isn’t built around the specific intent behind that specific search. The fix usually isn’t a full site rebuild — it’s a dedicated page, or even just a page section, that speaks directly to “same-day” and “AC repair” in the first screen someone sees, with the fastest possible path to booking or calling.

The Restructuring Myth: There Is No Account-Level Quality Score

One of the most persistent bad habits in PPC management is reorganizing ad groups and campaigns in the hope of nudging Quality Score upward — duplicating a keyword into a fresh, isolated ad group, or splitting a campaign into smaller pieces, without actually changing the ad copy or the landing page. Google has addressed this directly in its own materials: there is no ad-group-level, campaign-level, or account-level Quality Score. Moving a keyword into a new structural container, without changing what the ad says or where it sends people, does not change its quality signals. Google’s own analogy for the metric, from a whitepaper written by its Chief Economist, is a car’s engine warning light — a diagnostic signal, not something you’re meant to spend your afternoon staring at and fine-tuning directly. If restructuring alone moved the number, everyone would already be doing it, and Quality Score would be a solved, gameable metric instead of the thing it actually is: a byproduct of genuinely relevant ads and pages.

Quality Score in an Automated Account: Performance Max, Broad Match, and AI Max

Here’s where most Quality Score content — including a lot of what still ranks on page one — falls behind. An increasing share of a typical Google Ads account no longer runs on the kind of keyword-level structure Quality Score was designed to score.

Performance Max doesn’t use traditional keyword targeting, so there’s no keyword-level Quality Score displayed for it at all. That doesn’t mean the underlying quality signals — expected engagement, relevance, landing page experience — stop mattering. They still feed into how PMax’s automated systems select and prioritize creative and targeting; you just lose the direct visibility into them that a Search campaign gives you. Google has also tightened automated creative-quality checks inside Performance Max, flagging low-quality assets — blurry stock imagery, stretched logos, AI-generation artifacts — at the asset-group level, and suppressing “Asset Strength” until they’re replaced. That’s effectively the automation-era descendant of the old ad relevance and landing page checks, just surfaced through a different label.

Broad match campaigns present a related but distinct problem: their visible Quality Score often looks artificially low, because the score’s trailing comparison window is still weighted toward exact-match-equivalent performance, and it doesn’t fully capture the breadth of queries broad match expansion actually serves. A broad match keyword that’s converting well across a wide range of related queries can still show a mediocre visible score, simply because the reporting hasn’t caught up to how the keyword is actually being used.

Google has also announced that Dynamic Search Ads, Automatically Created Assets, and campaign-level broad match are being folded into “AI Max,” continuing a multi-year shift where more of the account runs on automated matching and keyword-level visibility keeps shrinking. Practically, this means the honest, current answer to “how do I check my Quality Score” is no longer just “look at the column in your keyword report.” For the growing share of an account that doesn’t expose that column, you have to lean on proxy signals instead: impression share lost to rank (a sign your ad quality or bid isn’t competitive enough to show), Asset Strength ratings in PMax and RSAs, and search term reports that reveal whether the queries actually triggering your ads still match what you’re targeting. If you’re running Performance Max, it’s also worth treating a full audit of that campaign as the closest equivalent to a Quality Score check-up you’re going to get, since PMax simply won’t hand you the keyword-level number a Search campaign would.

The Weighting Myth: There Is No Published Formula

Search around long enough and you’ll find articles confidently stating that expected CTR is “39% of the score,” or presenting a named formula with specific point allocations for each component. Treat these with real skepticism. Google has never published exact weights for how the three components combine into the visible 1–10 number, and the specific percentages circulating online don’t trace back to any primary source — they get invented once, then copied and re-copied across enough blog posts that they start to look authoritative through repetition alone.

The more useful, if less satisfying, answer is that Google deliberately doesn’t disclose the exact weighting, and that’s arguably the right call: a published formula becomes something to game rather than a genuine signal of ad and page quality. The absence of a formula is not a content gap you need to fill with a fabricated one — it’s the actual state of the documentation, and any post that gives you exact percentages is telling you something Google itself hasn’t said.

Does Quality Score Still Move Cost Per Click the Way People Say?

You’ll frequently see a table claiming a Quality Score of 10 costs roughly half what a Quality Score of 5 costs, with costs climbing sharply as the score drops — a Quality Score of 1 sometimes quoted at nearly 400% more than the baseline. These figures get repeated across a huge number of current PPC blogs, almost always uncredited, and they trace back to older AdWords-era analysis rather than any current, dated, primary source from Google. That doesn’t make the underlying idea wrong — better quality signals genuinely do tend to correlate with more competitive Ad Rank and therefore lower costs at a given position — but the specific percentages you’ll see quoted should be treated as a commonly repeated industry estimate, not a confirmed, current Google statistic. Worth knowing before you cite a number in a client report as if it came from Google itself.

How Often It Updates, and How Long to Wait Before You Judge a Fix

Because the visible score is a trailing average, its update speed depends heavily on traffic volume. High-volume keywords can show movement within the same day. Low-volume keywords can sit unchanged for a week or more simply because there isn’t enough fresh data to shift the average meaningfully. After a genuine fix — a rewritten ad, a faster or more relevant landing page — a reasonable expectation is the first signs of movement within one to two weeks, with the number fully settling into its new level over three to four weeks. Checking daily and concluding a fix “didn’t work” after 48 hours is one of the more common ways people convince themselves Quality Score is unpredictable or unresponsive, when it’s actually just slower than the rest of the account.

A Practical Way to Use Quality Score Without Obsessing Over It

Given everything above, here’s a framework that treats the metric as the diagnostic tool it’s meant to be, rather than a target to chase for its own sake.

• Use it to triage, not to grade. Scan for keywords with a genuinely low score (industry convention treats below 5 as worth a look, 7+ as solid — this is a practitioner heuristic, not an official Google threshold) and check which specific component is dragging it down before touching anything.
• Fix the component, not the number. If expected CTR is the weak link, work on ad copy and format relevance. If it’s ad relevance, tighten the ad group’s theme. If it’s landing page experience, that’s a job for whoever owns the site, not just the ads account.
• Don’t restructure to chase it. New ad groups and campaigns don’t reset or improve quality signals on their own — only genuinely better ads and pages do.
• Give fixes time before judging them. One to two weeks for initial movement, three to four for stabilization, longer on low-volume keywords.
• Where the number doesn’t exist — Performance Max, broad match, AI Max — use the proxy signals instead: impression share lost to rank, Asset Strength, and whether your search terms still match your intended targeting.
• Treat a profitable, low-scoring keyword as fine. If a Quality Score 3 keyword is converting at a cost you’re happy with, there’s no rule that says you have to fix it. The score is a diagnostic, not a report card you’re required to improve everywhere.

None of this is a one-time project. Ad copy that’s relevant today drifts as competitors update their own ads and as the queries triggering your keywords shift over time; a landing page that scored well six months ago can quietly degrade after an unrelated site change nobody flagged as ads-related. Quality Score, impression share lost to rank, and Asset Strength are all the kind of signals that are far more useful caught early, on a normal Tuesday, than discovered three weeks later during a monthly review when the underlying spend has already happened. That’s a big part of what continuous, day-to-day account monitoring is actually for — not a one-off audit, but something that’s watching the account daily and flags the kind of quiet drift these signals represent before it compounds. That’s the specific problem Growera’s Google Ads management is built around: keeping an eye on the account every day rather than waiting for a scheduled check-in to notice something’s changed.

Common Quality Score Questions, Answered Directly

A handful of questions come up constantly enough — in forums, in client calls, in agency onboarding docs — that they’re worth answering plainly rather than folding into the sections above.

Does Smart Bidding make Quality Score irrelevant? No, though it changes how visible the connection feels. Smart Bidding strategies like Target CPA, Target ROAS, and Maximize Conversions are optimizing toward a conversion outcome, and they do that by adjusting bids in real time based on signals that include the same quality and relevance factors Quality Score is built from. Automating the bid doesn’t remove ad and landing page quality from the equation — it just means you’re no longer the one manually reacting to the number, which makes it even more important that the underlying ad and page quality is solid, since you’ve handed the moment-to-moment reaction to the algorithm. If you’re weighing which bidding strategy fits your account, that’s a related but separate decision — see our breakdown of Target CPA vs. Target ROAS vs. Maximize Conversions for how to think about that choice on its own terms.

Why is my Quality Score low on a keyword that converts well? Because Quality Score measures relevance and expected engagement, not profitability. A keyword can have a mediocre expected CTR or ad relevance score and still convert at a cost you’re happy with, especially in a low-competition niche where you’re not fighting for position against aggressive bidders. There’s no rule requiring you to “fix” every low-scoring keyword. If it’s hitting your numbers, the practical move is to leave it alone and spend your attention on keywords that are both underperforming and low-scoring — that combination is where the diagnostic actually earns its keep.

Should I pause a keyword just because it has a Quality Score of 1 or 2? Not automatically. Check what it’s actually costing you and what it’s returning first. A very low score paired with a high cost and no conversions is a real signal to pause or rework. A very low score paired with occasional cheap conversions from a long-tail, low-volume query is often not worth the time it would take to fix.

Does adding negative keywords improve Quality Score? Not directly, but it can help indirectly. Negative keywords stop your ad from showing on irrelevant searches, which means your ad’s performance data stops being diluted by clicks and impressions on queries it was never a good match for. Over time, that can make your expected CTR and relevance signals more accurately reflect how well the ad matches the searches it should be showing for. It’s a supporting move, not a lever you pull directly — the same discipline that shows up when you’re working through a search term report to find negatives in the first place.

Is a Quality Score of 7 good? By the informal industry convention most practitioners use, yes — 7 and above is generally treated as solid, and below 5 is treated as worth investigating. This isn’t a number Google has published as an official threshold; it’s a heuristic the PPC community has settled on because it’s roughly where the “above average” status on the individual components tends to cluster. Treat it as a useful rule of thumb, not a pass/fail line from Google itself.

Does Quality Score differ by device? Yes — Google’s own guidance confirms device (mobile versus desktop) is factored into how ad quality is calculated, since expected CTR, relevance, and landing page experience can genuinely differ between a mobile search and a desktop one for the same keyword. A landing page that renders beautifully on desktop but loads slowly or displays awkwardly on mobile can drag down mobile-specific quality signals even while the desktop experience scores fine, which is one more reason a landing page audit needs to happen on an actual phone, not just a widescreen monitor.

The Bottom Line

Quality Score is real, it’s grounded in three specific, well-documented components, and it’s worth understanding — but it’s not an auction input, there’s no account-level version of it, restructuring alone won’t move it, and nobody outside Google actually knows the exact weighting formula, whatever confident-sounding percentages you find online. The visible 1–10 number is a trailing, aggregated summary of quality signals that are evaluated fresh, in real time, for every auction your keywords enter — which is why the number in your dashboard and the reality of your auction performance don’t always seem to agree.

The more the account shifts toward Performance Max, broad match, and AI Max, the less that visible number is going to be there for you at all — which makes the underlying discipline (relevant ads, relevant pages, genuinely matching what people are searching for) more important than the score itself, not less. If you’re building or reviewing a Google Ads account in 2026, the useful habit isn’t checking a single number periodically — it’s watching the components that number was always trying to summarize in the first place, whether or not Google still shows you the summary. And if you’re evaluating how much of that watching you want to do by hand versus with help, that’s worth factoring into how you think about what account management should actually cost you in time as well as money.

None of the individual pieces here are complicated on their own — write ads that match intent, keep ad groups tight, make landing pages fast and relevant, don’t restructure for the sake of a number. What’s hard is doing all of it consistently, across every campaign, every week, as queries shift and competitors update their own ads around you. That’s less a one-time skill problem than an attention problem, and it’s the reason “check Quality Score” tends to slide down the priority list the moment an account gets busy — right up until a quietly degraded landing page or a drifted ad group shows up in the cost-per-conversion column a month later, well after the spend already happened.