Quality Score and Ad Rank in Google Ads 2026: How Your Click Price Is Set
Two advertisers enter the same auction on the same query. One pays $0.90 a click and sits first; the other pays $1.60 and sits third. The second one bids more. The difference is Ad Rank — and the part of Ad Rank you can actually move shows up in the interface as Quality Score.
By 2026 this topic has accumulated real confusion: automated bidding hides the bid, AI Max reshapes which queries you appear on, Ad Strength gets mistaken for Quality Score, and half your keywords show a dash in the Quality Score column. This guide covers how position and click price are actually determined, what is diagnostic, what is optimizable, and what is noise you should stop chasing.
What Ad Rank is made of
Ad Rank is a number recalculated on every single impression. It decides three things: whether you appear, in what position, and which assets show with you. Its inputs:
- Your bid. The maximum you are willing to pay. Under automated strategies the system sets it, but it does not disappear.
- Auction-time quality. Expected click-through rate, ad relevance, and landing page experience.
- Ad Rank thresholds. Minimum quality bars below which the ad does not show at any price.
- Auction context. The query itself, device, location, time of day, the other ads competing, and the content of the results page.
- Expected impact of assets and formats. Sitelinks, callouts, images — how much Google predicts they will improve the outcome of this impression.
Two consequences change how you work:
- Ad Rank is recalculated every time. There is no stored position. The same keyword at 9am on mobile and 11pm on desktop competes in two different auctions with different thresholds.
- Thresholds are a filter, not a multiplier. Below the bar, raising the bid changes nothing — the ad simply does not appear. This is the familiar “we tripled the bid and impressions did not move.”
Why a higher bid does not mean a higher position
Actual cost per click in Search is derived not from your bid but from what it takes to beat the advertiser below you:
Actual CPC ≈ (Ad Rank of the advertiser below / your quality) + $0.01
The economic consequence is direct: better quality lowers your click price at the same position. An advertiser with above-average quality pays less for the identical slot — that is arithmetic from the formula, not a Google talking point. Turning that into campaign-level numbers is covered in unit economics and how much a click is worth.
Quality is not a “good advertiser” badge. It is a coefficient that simultaneously admits you to the auction and sets what you pay for the same slot.
What Quality Score is — and what it is not
Quality Score is a diagnostic estimate from 1 to 10 at the keyword level in Search campaigns. It aggregates impression history on exact matches of the query to your keyword and shows how you look relative to other advertisers competing in the same auctions.
What Quality Score is not:
- Not the number plugged into the auction. The auction uses a quality estimate computed at impression time with full context. Quality Score is a lagging, averaged reflection of it in the UI.
- Not an assessment of your business. It is relative: the same landing page can score 8 on a narrow keyword and 4 on a broad one, because competitors and user expectation change.
- Not a client-report metric. Average account Quality Score is close to meaningless because it is not weighted by traffic.
- Not applicable to Performance Max or Demand Gen. No Quality Score exists there — diagnosis runs through Ad Strength and asset-level reporting.
A dash in the column means insufficient data: too few impressions on exact matches. A dash is not “bad”, it is “nothing to measure yet.”
The three components: how they work and what moves them
Each component is reported as “below average”, “average” or “above average”, benchmarked against other advertisers who served on the same keyword over the last 90 days.
| Component | What it measures | Main lever | Response time |
|---|---|---|---|
| Expected CTR | Likelihood of a click on this query, normalized for position and assets | Headline–query match, specificity of the offer | 1–2 weeks |
| Ad relevance | How well ad copy matches the intent behind the query | Ad group structure, keyword precision in headlines | Days |
| Landing page experience | Relevance, transparency, usability and speed of the page | Query-specific content, speed, mobile layout | 2–4 weeks |
Expected CTR
This is a forecast, not the CTR in your report. Google normalizes for position: an ad in slot one at 8% CTR and an ad in slot four at 4% can carry the same expected CTR. So “my CTR is great, why is this below average” is a normal situation — you are being compared with advertisers who do even better from the same slot.
What moves it: a headline containing what the person actually searched for, and specificity instead of abstraction (“Delivered in 2 hours” beats “Wide selection”). What does not move it: stacking assets for the sake of count, and exclamation marks. Testing methodology is in A/B testing responsive search ads.
Ad relevance
The fastest component to fix and the most reliable symptom of a structural problem. “Below average” nearly always means one thing: the ad group mixes keywords with different intent, and no single ad can answer all of them.
The fix is splitting the group, not rewriting copy. Check match types at the same time — broad match without disciplined negatives drags in queries the ad was never written for and pulls relevance down. See keyword match types, and its mandatory companion, routine cleanup via search terms analysis and negative keywords.
Landing page experience
The slowest component: site changes take weeks to register. Google looks at content relevance to the query, originality and usefulness, transparency (who you are, how to reach you, what happens to data), and technical behaviour — speed and mobile usability.
The two most expensive mistakes here are one generic landing page for the entire account, and a slow mobile build. The first kills relevance on every narrow keyword at once; the second damages every keyword simultaneously. Technical thresholds are in landing page speed and Core Web Vitals, platform-level requirements in Google Ads destination requirements.
Ad Strength is not Quality Score
This confusion costs real money. Ad Strength rates asset completeness and variety: how many headlines you have, how different they are, whether keywords appear. It is a packaging hint, not an auction quality estimate.
| Quality Score | Ad Strength | |
|---|---|---|
| Level | Keyword | Ad / asset group |
| Based on | Impression history and user behaviour | Asset composition and variety |
| Affects click price | Yes, via Ad Rank | Not directly |
| Scale | 1–10 plus three component ratings | Poor / average / good / excellent |
“Excellent” Ad Strength alongside below-average ad relevance is a perfectly common combination, and it means you have plenty of assets that are about the wrong query. Reading that report correctly is covered in Ad Strength and the RSA asset report.
How quality behaves under Smart Bidding and AI Max
With Smart Bidding the system sets the bid, but quality works exactly as before: it determines which position and which price you receive at a given target. So “Quality Score does not matter on automation” is wrong — the bid lever moved to the system, the quality lever stayed with you.
AI Max and broader query matching add a new effect: the set of queries you appear on becomes wider and more varied. Average ad relevance typically dips — not because the ad got worse, but because the query set expanded. The right reaction is not panic over a falling Quality Score but a read on money: if CPA holds, the expansion is working. Mechanics in AI Max for Search campaigns.
Diagnosis: what to look at, in what order
- Add the columns. In the keyword report: Quality Score, Expected CTR, Ad Relevance, Landing Page Experience, plus their historical versions so you can see movement.
- Sort by spend, not by score. A keyword scoring 3 that spends $5 a month is noise. A keyword scoring 6 that spends $2,000 is the job.
- Find the common denominator. If landing page experience is below average across every keyword in a group, the page is the problem. If relevance is below average in one group out of fifteen, that group’s structure is the problem.
- Cross-check the auction. Falling impression share with stable quality means rising competition, not your decline. Tooling: impression share and auction insights.
- Change one thing at a time. Copy and landing page simultaneously means you learn nothing from the result.
What not to do
- Chase 10/10. The score is relative; in competitive niches 7 may be the market ceiling. The objective is money, not a column value.
- Delete low-score keywords that convert. An expensive click that converts beats a cheap one that does not.
- Report an account-wide average. Unweighted by impressions, that number means nothing.
- Expect instant results from site changes. Landing page experience is inert — check again after two to four weeks.
- Confuse Quality Score with Optimization Score. The latter is a recommendations index, not an auction input; see optimization score and recommendations.
A 30-day checklist
| Window | Action | What to expect |
|---|---|---|
| Days 1–3 | Export keywords above a spend threshold, tag them by the three components | A problem map — usually 2–3 clusters, not a hundred individual cases |
| Days 4–7 | Split mixed-intent ad groups, rewrite headlines to match the query | Ad relevance responds first |
| Days 8–14 | Search terms cleanup, negatives, match type adjustments | CTR rises, irrelevant impression share falls |
| Days 15–21 | Landing pages: dedicated pages per cluster, speed, mobile layout | Landing page effect lands near the end of the window |
| Days 22–30 | Reconcile CPC, impression share, CPA. Decide on bids | Lower CPC at the same position is the key success signal |
Why your assets do not always show
Ad Rank decides not only your position but the composition of the ad. Sitelinks, callouts, images and prices appear when Ad Rank clears separate thresholds calculated for those formats. So “all assets filled in, none of them showing” is not an interface bug — it is a direct consequence of quality and position.
The diagnostic order when assets do not show:
- Check the position. Lower slots simply do not support some formats, no matter how good your assets are.
- Check asset status. A disapproved or limited asset will not serve and will not announce itself prominently.
- Check the level. Campaign-level assets override account-level ones; a stale campaign set will suppress a good account set.
- Check the count. Consistent serving needs noticeably more assets than the minimum — the system assembles a combination per query and needs room to choose.
- Only then, quality. If your position is persistently mid-page, improving quality raises both position and the odds of an extended format at once.
Three common scenarios and the first move
| Symptom | Usual cause | First action |
|---|---|---|
| Few impressions, high bid, impression share lost to rank | Quality below the Ad Rank threshold — usually relevance plus landing page | Split the group by intent, build a dedicated landing page |
| Impressions fine, CPC climbing month over month at the same position | Competition rose; your quality did not change, the background did | Check auction insights, look for quality gains rather than bid increases |
| Quality Score dropped after moving to automation | The query set widened and average relevance diluted | Watch CPA and conversions; tighten with negative keywords |
The same principle holds in all three: Quality Score is a dashboard indicator, not a lever. You move it indirectly — through group structure, headline-to-query match and the state of the landing page. Treating “raise Quality Score” as a standalone task almost always ends in cosmetic copy edits that change nothing.
Account quality and ad quality are different things
Worth separating two ideas that get conflated in conversation. Quality Score describes the keyword–ad–page triangle. Account reputation is your policy compliance history, advertiser verification status and account age. The second is not a direct Ad Rank input, but it governs access to formats and betas, review speed, and resilience under scrutiny. That is why a brand-new account with a flawless structure still starts slower than a mature one.
If the goal is to operate at normal review speed with format access rather than wait half a year for history to accrue, it is more practical to start from agency Google Ads accounts that already carry it. Baseline settings and checklists for doing it yourself are in our Google Ads guide.
Two companion pieces published today: collecting enquiries without a landing page in lead form assets in Google Ads, and how feeds now shape the search results page in Search campaigns for Travel.
FAQ
What is Ad Rank in plain terms?
A number Google recalculates on every impression to decide whether your ad shows, in what position, and with which assets. Inputs are your bid, auction-time quality, Ad Rank thresholds, query context, and the expected impact of assets.
Does Quality Score affect cost per click?
Yes. Actual CPC derives from the Ad Rank of the advertiser below you divided by your quality. All else equal, higher quality means a lower price for the same position.
Why does a keyword show a dash instead of a score?
Insufficient data — too few impressions on exact matches of the query to that keyword. It is an absence of measurement, not a bad grade.
I raised my bid and impressions did not increase. Why?
Most likely your quality sits below the Ad Rank threshold. Thresholds act as a filter: under them, bidding does not help.
Is there a Quality Score in Performance Max?
No. PMax and Demand Gen have no Quality Score — diagnosis runs through Ad Strength and asset-level reporting.
How is Ad Strength different from Quality Score?
Ad Strength rates asset completeness and variety in the ad; Quality Score reflects interaction history on a specific keyword. Only the latter feeds click price through Ad Rank.
Which component should I fix first?
Ad relevance. It responds fastest and most often points at a structural issue — mixed intent inside one ad group.
How long does landing page experience take to change?
Usually two to four weeks after genuine page changes. It is the most inert of the three components.
Should I delete keywords scoring 3–4?
Not because of the score. If the keyword converts at an acceptable cost, keep it. Pause on economics, not on a column value.
Can I improve quality without touching the site?
Partly: group structure, headline precision, negatives and match types address two of the three components. The third does not move without site work.
Why is expected CTR “below average” when my CTR is high?
The metric is normalized for position and benchmarked against competitors on the same keyword. A high absolute CTR in a top slot can still be below expectation for that slot.
Does quality drop when moving to AI Max?
Average relevance often dips because the query set widens. Judge the migration on CPA and conversion volume, not on the Quality Score movement.
One sentence version: your bid buys a place in the queue, your quality buys the price and the right to stand in it. Automation took the first lever; the second is still entirely yours.
