Most B2B Google Ads accounts have a column called Quality Score that nobody really agrees on. Half the team treats it as a top-line KPI, the other half ignores it because Smart Bidding and broad match are doing the work. Both views miss the point. Quality Score still moves what you pay per click, but in 2026 it works as a diagnostic of how Google sees your ad-keyword-landing-page fit, not as the throttle on your campaign.
For the broader playbook on running Google Ads campaigns that actually generate B2B pipeline, see our complete Google Ads for B2B SaaS guide. This post focuses on one specific lever inside that playbook: how Quality Score is calculated, when it is worth fixing, and which components actually matter for B2B advertisers running long-cycle, high-CPC campaigns.
What Quality Score Actually Measures in 2026
Quality Score is a 1 to 10 metric on each keyword in your account. According to Google's official documentation, "Quality Score is calculated based on the combined performance of 3 components: Expected clickthrough rate (CTR), Ad relevance, and Landing page experience". Each component gets one of three statuses: Above average, Average, or Below average, evaluated against other advertisers whose ads showed for the exact same search over the last 90 days.
The crucial bit most B2B teams miss: the visible Quality Score is a diagnostic, not the number Google uses in the live auction. Frederick Vallaeys, who helped build Quality Score during his time at Google and now runs Optmyzr, put it bluntly in March 2026: "Quality Score was never designed to tell you whether a campaign is profitable. It's not a KPI. It's a diagnostic that reflects how the system views relevance and expected usefulness in the auction". Ad Rank uses the same three underlying components, but on a per-query basis, which is why a keyword with a visible Quality Score of 6 can still convert beautifully and a keyword sitting at 8 can still cost too much.
How Quality Score Affects Your CPC
The mechanism is simple even if the auction maths are not. Ad Rank is bid times quality. Your actual CPC is the minimum needed to beat the next ad below you, plus one cent. If your quality is higher, you need a smaller bid to clear the same threshold, so your CPC drops. The often-quoted discount table from Store Growers (updated 2026) shows the relative effect against the average score of 5: a Quality Score of 8 gives a 37% CPC discount versus average, a 9 gives 44%, and a 10 gives 50%. A Quality Score of 3 inverts the same maths into a 67% CPC penalty.
For B2B advertisers paying 5 to 15 EUR per click on category keywords like "ERP software" or "B2B CRM platform", that swing is the difference between a sustainable campaign and one that bleeds budget. The catch is that you cannot bid your way out of low quality. Bidding higher pushes the ad up the page but does not change the underlying Quality Score. The only fix is the three components themselves.
The discount maths is symmetrical, not the impact. Going from 7 to 9 saves 15 percentage points on CPC. Going from 4 to 6 saves 25 points. Fixing low scores is more profitable than chasing 10s.
The Three Components: Where B2B Accounts Actually Lose Points
Expected CTR is relative, not absolute. As Optmyzr explains, "Your CTR could be 50% for a term but if Google's expecting 70%, well, that's going to be below average". For B2B, this hits hardest on broad category keywords. "CRM software" sits next to consumer ads with screaming offers and free trials. Your B2B ad with a "Book a Demo" CTA looks dull by comparison and gets flagged as Below average even when the actual CTR is solid for the audience.
Ad relevance is the easiest to fix. If most keywords in an ad group share Below-average ad relevance, the ad copy probably does not mention the keyword theme. Tightening ad groups so each one covers a single keyword theme, then mirroring that theme in headlines, lifts ad relevance fast. This is also where Responsive Search Ads earn their keep: feed in 10 to 15 headlines covering different angles of the same theme and let Google find the strongest combinations.
Landing page experience is the slowest to fix and the highest leverage. Page speed, mobile usability, and on-page topic match all feed into this. If you send "ERP software for manufacturing" traffic to a generic homepage, expect Below average. Send it to a page whose H1 says "ERP software for manufacturing" and which loads in under 2.5 seconds, and the score climbs.
When to Optimise Quality Score and When to Ignore It
Smart Bidding and broad match have changed the calculus. If your account is hitting CPA and ROAS targets, a visible Quality Score of 5 or 6 on broad-match keywords is usually noise, not signal. The auction is using the per-query Ad Rank components anyway, and chasing the visible number can lead to dismantling structures that work.
Optimise Quality Score actively when impression share is dropping because of rank, when landing page experience is consistently Below average across an ad group, or when red "Rarely shown (low Quality Score)" warnings are spreading in tightly themed campaigns. Those are the cases where the diagnostic is pointing to real friction in the auction. If CPA is healthy and revenue is steady, leave the column alone and focus on the components Ad Rank actually consumes.
For B2B specifically, the highest-ROI Quality Score work is almost always on landing pages. Most B2B sites have one or two over-loaded landing pages handling traffic for ten different ad groups. Splitting them into themed pages, each matching a single keyword cluster, lifts landing page experience and conversion rate at the same time. Pair that work with proper conversion tracking - see our guide on Google Ads conversion tracking under Consent Mode v2 - and the score moves within weeks.
Conclusion
Quality Score in 2026 is a diagnostic with real money attached, not the KPI it once was. The number on the keyword line tells you how Google sees the relevance triangle of search query, ad copy, and landing page. The same three components feed Ad Rank in the live auction, which is where the actual CPC is set. For B2B advertisers, the practical move is to ignore the visible score when CPA is healthy, dig in when impression-share-lost-to-rank climbs or red warnings spread, and treat landing page experience as the highest-leverage fix. Tighten the keyword-ad-page triangle, let Smart Bidding take care of the auction, and the discount in the table above shows up where it matters: in your blended CPC. For the wider campaign structure that benefits from these fixes, return to our Google Ads for B2B SaaS guide.
Frequently Asked Questions
Does Quality Score still matter with Smart Bidding and broad match?
Yes, but indirectly. The visible Quality Score column is increasingly a diagnostic, not a real-time auction input. Ad Rank in every auction still uses Expected CTR, Ad Relevance, and Landing Page Experience on a per-query basis. So the components matter even when the aggregated number looks misleading. The practical rule: if Smart Bidding is hitting CPA targets, ignore the visible score and only intervene when impression share lost to rank rises or low-quality warnings spread.
How long does it take for Quality Score to update after I make changes?
Days to weeks, depending on impression volume. Quality Score is calculated from historical performance for exact searches, so a keyword with high volume picks up signal fast. Long-tail B2B keywords with 50 impressions a month can take a full quarter to reflect ad copy or landing page changes. Do not panic if the score does not move in week one. Track Expected CTR, Ad Relevance, and Landing Page Experience separately to see which component is moving.
Should I pause keywords with a Quality Score of 1 or 2?
Not automatically. First check whether they convert. A keyword with a Quality Score of 2 that converts at twice the account average is making money, just expensively. The right move is to fix the underlying component (usually landing page experience or ad relevance) rather than pause the keyword. Pause only when the keyword is both unprofitable and untunable, for example when it is a broad-match symptom of someone else's poor structure rather than an intent worth bidding on.