Moving your Quality Score from 5 to 8/10 cuts CPC by 22 to 34% depending on vertical (median on our audited accounts 2025-2026). It's the most underused Google Ads optimization lever in 2026: no additional media budget, visible effect in 14 to 28 days, and near-guaranteed ROI once you know how to diagnose which of the three components is pulling your score down.
This pillar guide answers a single question: how do you move from a median QS of 5-6 to a median QS of 8+ in 30 days? We break down the real weight of the 3 components (Expected CTR 42%, Ad Relevance 28%, Landing Experience 30% in 2026), lay out 10 proven tactics with their quantified impact, and provide a 30-minute checklist to audit your account. Audience: SMB and mid-market PPC managers running $5.5 to $220,000/month in Google Ads.
Is Quality Score still decisive in 2026?
Fundamental reminder: Quality Score is a 1-10 rating assigned by Google to each keyword in a Search campaign, recalculated at every auction in real time. The score shown in the UI is an average smoothed over 3 to 7 days — it doesn't reflect the instant score driving your position. Ad Rank is calculated at the exact moment of the query, keyword by keyword.
What changed in 2025-2026: three evolutions redistributed the weights. First, Landing Experience evaluation moved to Gemini at the end of 2024 — Google now analyzes semantic content, alignment with search intent, and complete UX. Then, Core Web Vitals thresholds became stricter: LCP ≤ 2.5s, CLS ≤ 0.1, INP ≤ 200ms (INP having replaced FID in March 2024). Finally, mobile-first became systematic: a landing that performs well on desktop but struggles on mobile caps at Landing Experience average. Official documentation on Google Ads support.
Business impact: QS is the most actionable variable on CPC in mature SEA. You don't control competitor bids or seasonality — but you fully control the three QS components. On our 2025-2026 panel, accounts that invested in a QS overhaul over 6 weeks show median CPC -22 to -34% by vertical, all other variables unchanged. For a wider view of combined levers, see our CPA reduction guide.
How does Google calculate Quality Score?
QS aggregates three sub-scores into a final score, each displayed in the UI as Above / Average / Below average. Google has never published the exact weighting, but our regression tests on more than 2,400 audited keywords 2024-2026 converge on: Expected CTR ~42%, Landing Experience ~30%, Ad Relevance ~28%. These weights have evolved — Landing Experience weighed ~22% in 2022 before the Core Web Vitals overhaul and rises under Gemini pressure.
Expected CTR (~42%) — the predicted click-through rate for this keyword, independent of position and extensions. Google uses the keyword's CTR history, the ad group's ad history, and contextual signals (time, device, query). Most sensitive component to the editorial quality of RSAs and keyword-headline match.
Landing Experience (~30%) — technical Core Web Vitals, semantic content evaluated by Gemini since 2024, query-landing alignment, mobile-first, absence of interstitials, HTTPS. The slowest to move but the most underused.
Ad Relevance (~28%) — semantic alignment keyword ↔ ad text, evaluated in NLP. An ad group triggering 50 heterogeneous keywords with a single RSA can't have all its ads be relevant. Ad group granularity is the main driver.
real weight varies slightly by vertical and account maturity. On B2B SaaS accounts where landings are well crafted and ad groups already granular, Expected CTR dominates (up to 48%). On high-competition e-commerce accounts, Landing Experience can rise to 34% (Core Web Vitals becomes discriminating). The priority order remains stable nonetheless: Expected CTR > Landing Experience > Ad Relevance.
How Google calculates Expected CTR
Expected CTR isn't your actual CTR over the last 30 days. It's a prediction Google makes of expected CTR for this keyword, at this position, with this ad — independent of extensions. Google isolates pure text contribution to prevent sitelinks from mechanically inflating the score.
Three levers move Expected CTR quickly:
1. Dynamic Keyword Insertion (DKI) with restraint. Insert {KeyWord:default} in headline 1 aligns the title with the matched query. Relevant use on thematic ad groups of 8-12 close keywords. Classic trap: DKI on an overly broad ad group produces absurd titles. On our audited accounts, well-used DKI produces +0.4 to +0.9 median QS point in 21 days.
2. Targeted ad group granularity. An ad group of 8 thematically tight keywords lets 3 RSAs cover all intents. An ad group of 50 heterogeneous keywords can't: expected CTR plateaus below average on 60 to 80% of keywords.
3. Complete extensions. Sitelinks (4-6), callouts (8-10), structured snippets, images, call, lead form. Even though Expected CTR is evaluated excluding extensions, their presence improves observed real CTR which enters history. An account with 100% complete extensions shows above-average Expected CTR on 62 to 74% of its keywords, vs 32 to 44% without.
On our 2025-2026 sample, placing the exact keyword in pinned headline 1 produces +0.6 to +1 median QS point in 14 to 21 days. Impact is more pronounced on commercial keywords than informational. See also our RSA writing method.
Ad Relevance: aligning keyword and ad
Ad Relevance measures semantic alignment between the triggered keyword and the visible ad text. Evaluated in NLP by Google, it isn't enough to mechanically repeat the keyword. Example: on "CRM software for small business," an ad talking about "team project management" will be below average even if "software" appears 3 times. Intent must be addressed explicitly in at least two headlines.
The four rules of above-average Ad Relevance:
- 15 unique headlines per RSA, not 5 variants of the same message. Google needs material to test combinations. Accounts limiting to 5-8 headlines mechanically cap at Ad Relevance average.
- 4 descriptions with different angles: main benefit, social proof, guarantee, urgent CTA. Google picks the one best matching detected intent.
- Dosed pinning: max 1 pin per ad. Pin the keyword in headline 1, let Google freely combine the other 14. Over-pinning (3+ pins) throttles the algorithm and degrades Ad Relevance.
- Coherent ad group theme. If you can't write one description that speaks to all the ad group's keywords, you need to split it.
for each ad group, ask yourself: "If I had to summarize the common intent in one sentence, what would it be?" If the resulting sentence exceeds 15 words or contains "or / as well as / and also" conjunctions, the ad group is too broad. Splitting into 2-3 thematic ad groups typically produces +0.8 to +1.2 median QS points in 21 days, variable depending on initial keyword dispersion.
For match type choice which also impacts Ad Relevance (a poorly managed Broad dilutes relevance), see our 2026 match types guide. And for the bidding strategy that orchestrates this setup, our Smart Bidding comparison.
Landing Experience: Core Web Vitals 2026
Landing Experience moved from secondary component to near-parity with Expected CTR in 2026. Two reasons: the late-2024 Gemini overhaul of semantic evaluation, and strict application of mobile Core Web Vitals thresholds since March 2024. In our 2025-2026 audits, it's below average in about 38% of accounts — second driver of mediocre QS after ad group granularity.
The four 2026 pillars to hit:
1. Core Web Vitals at the 75th percentile. LCP ≤ 2.5s, CLS ≤ 0.1, INP ≤ 200ms, measured at the 75th percentile of real mobile traffic (CrUX), not on an isolated desktop Lighthouse. Official documentation on web.dev/vitals and Google Search Central.
2. Message match headline ↔ H1. The ad headline must be found nearly word for word in the landing's H1. If the ad promises "CRM Software for Small Business — Free Trial" and the landing shows "Unified commercial management," Gemini penalizes the mismatch.
3. Strict mobile-first. 68 to 74% of Google Ads traffic is mobile in 2026. A landing at 1.8s LCP desktop but 4.2s mobile sees its mobile Landing Experience flip below average. Systematically test on Mobile 3G-Fast in PageSpeed Insights.
4. Content semantically aligned with intent. Gemini evaluates whether the page actually answers the intent. A generic homepage used for 50 different keywords caps at Landing Experience average. Solution: dedicated landing pages per intent. See our Google Ads landing pages guide.
Measured impact: on accounts moved above the 75th percentile CWV, we measure +14 to +22% median CVR, variable by initial quality. An LCP moving from 3.8s to 2.1s alone produces +0.8 to +1.4 median QS points in 21 days.
10 tactics to move from QS 5 to QS 8+/10
Here are the 10 tactics with measured impact. Each tactic targets a dominant component, with an effect interval observed on our 2025-2026 panel. Applying them in sequence (1 to 5 first, then 6 to 10) produces a cumulative gain of +2 to +3 median QS points in 4 to 6 weeks.
1. Exact keyword in pinned headline 1. Guarantees a visible match and a strong signal for Google. Avoid pinning more than one headline.
2. 3 RSAs per ad group with 15 unique headlines. Three distinct RSAs, 15 truly different headlines (not variants), 4 descriptions with varied angles. Google has the material to personalize by query.
3. Thematic ad groups ≤10 keywords. One intent per ad group. Splitting a 40-keyword ad group into 4 10-keyword ad groups produces a QS gain greater than any other individual tactic.
4. Match types in 30/50/20 mix. 30% Exact (precise control), 50% Phrase (coverage), 20% Broad (exploration with Smart Bidding). Broad without Smart Bidding: guaranteed QS drift.
5. Complete extensions. Sitelinks (4-6), callouts (8-10), structured snippets, images, call, lead form. Gains in visible surface and real historical CTR.
6. Shared negatives ≥ 200 entries. Account-level list, fed monthly from the Search Terms report. Eliminates irrelevant clicks that dilute Expected CTR.
7. Landing Core Web Vitals 75th percentile. Audit via PageSpeed Insights or CrUX. Direct QS gain and secondary CVR gain.
8. Message match headline ↔ H1. The landing H1 semantically echoes the main headline. Coherence evaluated by Gemini since 2024.
9. Mobile-first LCP < 2.5s. Prioritize mobile loading. A landing that performs mobile necessarily performs desktop, the reverse isn't true.
10. Continuous A/B test top 10 keywords. 2 RSA variants permanently on each top 10 ad group. Eliminate the loser at 28 days, launch a new one. Significant cumulative annual effect.
Measured impact: -22 to -34% CPC depending on vertical
QS affects CPC via a simplified formula: real_CPC = (Ad_Rank_competitor_below / your_QS) + $0.01. The higher your QS, the larger the number Google divides the next competitor's Ad Rank by, so the lower your CPC. The effect is non-linear: moving from 5 to 7 produces a more pronounced gain than moving from 7 to 9. Here are the relative multipliers observed on our 2025-2026 panel, normalized to QS 5 = 1.0× baseline.
Main stat to remember: on accounts audited between 2024 and 2026, moving median QS from 5 to 8 produces -22 to -34% median CPC, variable by vertical (e-commerce higher, B2B SaaS more moderate, local lead gen the most sensitive). On an $11,000/month budget, that represents $2,420 to $3,740 of recurring savings, or $29-45k/year at constant traffic. Or, put another way, +28 to +52% clicks at identical budget.
QS is the only variable with immediate and recurring impact: Smart Bidding takes 2-4 weeks to readjust, a creative overhaul takes 6-8 weeks, an audience strategy takes 90 days. QS moves in 14 to 28 days, the gain stabilizes, and it doesn't disappear as long as you maintain the tactics. It's the only Google Ads optimization whose ROI is near-guaranteed given method, without market or seasonality dependence.
To frame this trade-off alongside other account optimization levers, see our Google Ads audit checklist.
What mistakes crash Quality Score?
Eight mistakes concentrate 80% of the mediocre QS we observe in audits. Recognizing them lets you avoid months of trial and error — and for most, the fix only takes a few hours of work.
- 1. An ad group with 50-100 heterogeneous keywords. Most frequent case. A single RSA can't be relevant for 50 intents. Ad Relevance plateaus below average on 60 to 80% of keywords. Fix: split into thematic ad groups of ≤10 keywords.
- 2. A single RSA with 3-5 headlines instead of 15. Google lacks material to personalize. Expected CTR plateaus at average. Fix: move to 15 unique headlines per RSA, 3 RSAs per ad group.
- 3. Generic homepage landing for all campaigns. No message match, systematic Landing Experience below average. Fix: dedicated landings per intent or per ad group.
- 4. No shared negatives = ignored Search Terms report. Irrelevant clicks dilute real CTR and degrade Expected CTR by ricochet. Fix: shared negatives list ≥ 200 entries, monthly review.
- 5. Mobile landing with LCP > 3.5s. 68-74% of traffic is mobile; Landing Experience plateaus below average. Fix: Core Web Vitals audit, optimize images, reduce blocking JS.
- 6. Keyword completely missing from RSA headlines. Non-relevance signal for Google, Expected CTR and Ad Relevance degraded. Fix: exact keyword must appear in at least 2 headlines, ideally pinned in headline 1.
- 7. Broad match without Smart Bidding or solid negatives. Massive traffic drift toward irrelevant queries, collapsed CTR. Fix: Broad only with active tCPA/tROAS, negatives list ≥ 200 entries, tight ad groups.
- 8. Ignoring below-average for 3+ months. QS doesn't fix itself. A below-average left untreated settles in and ends up affecting the entire account via Smart Bidding. Fix: monthly QS audit, top 20 below-average prioritized.
mistake #1 (overly broad ad group) often hides behind seemingly well-maintained accounts. Quick check: export keywords from your 5 largest ad groups. If any contains more than 15 keywords, there's an 80% chance its Ad Relevance is below average on at least 40% of keywords. Splitting is the priority before any other optimization.
30-min checklist to audit your QS
Sequential procedure to identify below-average keywords and prioritize actions. 30 minutes of work, executable today, basis of the 30-day optimization plan.
Step 1 — Export the Keywords report (5 min). Google Ads > Campaigns > Search Keywords. Columns > Modify columns, check in Attributes: Quality Score, Expected CTR, Ad Relevance, Landing Page Experience. Sort by descending Impressions. Export as CSV.
Step 2 — Filter below-average of top 20 (5 min). In Sheets, filter on Expected CTR OR Ad Relevance OR Landing Page Experience = "Below average." Limit to top 20 by impressions. These 20 keywords typically absorb 40 to 60% of budget — to fix first.
Step 3 — Diagnose the limiting component (10 min). Map each keyword to its action: Expected CTR below → "RSA refresh + extensions"; Ad Relevance below → "Split ad group"; Landing Experience below → "Optimize landing." In most cases, 60% involve Ad Relevance, 25% Landing Experience, 15% Expected CTR — but your distribution may differ.
Step 4 — Apply tactics (variable duration). Each component has its playbook (sections 3 to 5). Start with the 5 highest-volume keywords: they produce 60 to 70% of total gain. The next 15 in weeks 2-3.
Step 5 — Measure impact at 14 days (10 min). Re-export the report, compare QS and CPC before/after. Typical gain: +1.5 to +2.5 median QS points, -18 to -28% median CPC. If < +0.5 pt after 21 days, revisit the initial diagnostic.
Audit CTA: our free SteerAds audit fully automates the checklist above: 2-minute import, automatic detection of below-average top-20 keywords, component-by-component diagnostic, prioritized tactic recommendation. To industrialize continuous monitoring and detect QS drifts in real time, our auto-optimization module watches the 3 components daily. For broader industry context, coverage by Search Engine Land remains a reference to follow.
Sources
Official sources consulted for this guide:
FAQ
Is Quality Score still important in 2026?
Yes, more than ever on mature accounts. QS remains the most actionable variable of Ad Rank in 2026: Google has simply added Gemini-powered signals on landing evaluation and stricter Core Web Vitals weighting. In our sector panel, moving an account from median QS 5 to median QS 8 produces -22 to -34% CPC depending on vertical, or a gain of 20 to 30% clicks at identical budget. It's the only optimization lever with near-immediate effect (7 to 21 days for Google recalibration) and zero media cost.
How long to improve your QS?
Google recalculates Quality Score at every auction, but the score shown in the UI only stabilizes after 3 to 7 days of fresh history. For a complete overhaul (restructured ad groups, new ads, optimized landing), expect 14 to 28 days before seeing average QS rise by 2 points. Low-volume keywords take more time (up to 6 weeks) because Google needs impressions to recalibrate its Expected CTR. In most cases, 80% of the gain arrives in the first 4 weeks, the remaining 20% over 60 to 90 days.
QS 6/10: is that bad?
QS 6 is the neutral zone: neither penalized nor rewarded. Compared to a QS 5 baseline, a QS 6 pays about -15% CPC. But compared to a QS 8 reachable with the 10 tactics in this guide, you leave -22 to -34% CPC on the table. QS 6 is mostly a signal that one of the three components is below average. The diagnostic in section 9 lets you identify which: in 60% of cases it's Ad Relevance (ad group too broad), in 25% Landing Experience (Core Web Vitals), in 15% Expected CTR (weak RSA).
Can you have a good QS without a good landing?
No, not in 2026. Since the late-2024 Gemini-powered landing evaluation update, Landing Experience weight has risen to around 30% of total QS, nearly on par with Expected CTR. A slow landing (LCP > 3s), not aligned with the keyword, or without an H1 matching the headline caps QS at 6/10 even with perfect RSAs. On our audited accounts, moving Core Web Vitals to the 75th percentile produces +0.8 to +1.4 median QS points in 21 days. It's the most neglected component and the most profitable to fix.