Quality Score isn’t the whole reason your CPCs are rising in 2026
When CPCs rise, Quality Score is still one of the first places many PPC practitioners look.
The logic is not ridiculous because Google scores expected click through rate, ad relevance and landing page experience as markers for decreasing or increasing CPCs, so a low score appears to offer both a diagnosis and a plan. Leading marketers to tighten the ad groups, improve the ads and fix the landing page. And in an ideal world where best practices were legitimately correct, CPC should come back down.
Except the Quality Score you are trying to improve is not used in the auction.
That does not mean ad quality is irrelevant. Google evaluates quality in real time, and those signals can affect Ad Rank and CPC. But the score from 1 to 10 in your keyword report is a historical diagnostic. Your CPC is the outcome of a live auction shaped by your bid, competitors, search context, thresholds, assets and the traffic Google chose to match your queries with.
AI Max, broader query expansion and changes to target based bidding can alter the auctions behind your average CPC without leaving a neat explanation in the Quality Score column.
So before another account audit turns into a Quality Score clean up project, ask a more useful question: what actually changed in the auctions you were paying for?
Here is how to work it out without blaming Google for everything or pretending the answer always sits inside the keyword table.
What Google has been saying over the years
It is a historical diagnostic at keyword level, built from three components:
Expected click through rate
Ad relevance
Landing page experience
Each component is rated above average, average or below average by comparing your performance with other advertisers whose ads showed for the exact same search during the previous 90 days.
Google compares Quality Score to a warning light on a car’s engine. Yeah, okay - I understand the point, but I do not love the analogy.
The problem with this analogy is that a warning light suggests something is broken and that fixing it should switch the light off. Quality Score is closer to a blood pressure reading. It can flag that something deserves attention, but it does not diagnose the cause on its own.
You still need to examine the three components behind the number, then consider the wider account and auction context. Better stats does not automatically guarantee that every other outcome will improve. In the same way, a higher Quality Score does not automatically mean a lower CPC.
What actually enters the auction?
Google recalculates Ad Rank for every auction and for different ad locations. Its Ad Rank documentation names six broad factors:
Your bid
The quality of your ad and landing page at that moment
Ad Rank thresholds
The competitiveness of the auction
The context of the person’s search
The expected impact of your assets and other ad formats
Google also says higher quality ads can often lead to lower CPCs. Often matters. Better quality can help, but it is only one part of an auction in which bids, competition, thresholds, context and assets are also changing.
That is how a keyword can move from a Quality Score of 5 to 8 while its average CPC still rises.
Why the Quality Score myth is surviving extinction
“Your CPCs are up because your Quality Score is low” is a comfortable diagnosis. It’s concrete, fixable and contained neatly inside the account.
“Your CPCs are up because several factors within the auctions changed” is harder to explain, even when it is the more accurate answer.Clients and major stakeholders don’t always want the answer that’s technically accurate. They want an answer that is neatly packaged with a pretty bow on top, with a clear list of next steps afterwards. That sends the PPC practitioner (who may need billable hours anyway) straight down the optimisation rabbit hole, “fixing” whatever Google deems “improving”.
That is the trap. A diagnosis does not become correct just because it produces a familiar optimisation checklist for you to follow.
What can actually increase CPC’s?
Yes, I’m going to do the annoying thing and say, “It Depends”. There is no universal answer because accounts do not all move for the same reason. These are the areas worth investigating before prescribing a Quality Score project.
Competition and auction mix
More competitors, or existing competitors bidding more aggressively, can raise the price of auctions you already enter. If CPC increased, check whether competitor activity changed during the same time period.
This is why benchmark data needs context. WordStream reported an average CPC for a combination of industries of $5.26 in 2025 and $5.42 in its 2026 benchmark, based on more than 13,000 US search campaigns running between April 2025 and March 2026. The most useful element of these articles is to check if your particular industry is listed and if not, check your historical data, and if there isn’t any consult agencies that specialize in your industry that already have a CPC benchmark
Use benchmarks to understand the market, not to diagnose an account. Auction Insights, search terms and segmented account data are the evidence you need.
Smart Bidding may be buying more expensive clicks on purpose
Smart Bidding does not optimise for the cheapest click. It bids according to the likelihood of a conversion or the conversion value it predicts.
Google says Target CPA can set higher CPC bids for clicks it considers more valuable, then set lower bids for clicks it considers less valuable.
That means a CPC increase is not automatically evidence that Smart Bidding failed. Ask what came with it:
Did conversion rate improve?
Did lead quality or revenue improve?
Did traffic shift towards searches with higher commercial value?
Did CPC rise while CPA, ROAS or profit remained acceptable?
The useful diagnosis is not “automation made clicks expensive”. It is “the bidding system paid more for this traffic, and here is whether that trade was worth it”.
Your query targeting changed
Starting on 1 September 2026, Google began converting Search campaigns using automatically created assets or campaign level broad match into AI Max settings. Dynamic Search Ads followed a separate timeline and should not be bundled into that September change. PPC Land reported the migration details.
AI Max can expand matching beyond the searches your keywords would have reached on their own. Google describes it as using broad match and keywordless technology to find new queries. That changes the population of auctions behind your average CPC.
It does not guarantee that CPCs will rise. Expansion may find cheaper traffic in one account and more competitive searches in another. Either way, comparing the new average with the old one as if the underlying traffic stayed constant is a bad comparison.
If CPC moved around the September migration, check:
Change history and AI Max settings
Search terms before and after the change
Brand, competitor and generic query mix
Match type and keywordless expansion
Search Partners compared with Google Search
Conversion quality, not just click cost
For a deeper look at the trade offs, read Google’s AI Max for Search: What the Data Actually Shows in 2026.
The August bidding update changed how some campaigns behave
On 17 August 2026, Google started rolling out a change to target based bidding for campaigns that were limited by budget. The rollout was completed on 27 August.
The scope matters. According to Google’s guidance, the change applies to campaigns using Target CPA or Target ROAS across most eligible campaign types. Target CPC is included for Demand Gen only. Campaigns that were not limited by budget were not affected.
Affected campaigns now optimise more consistently towards the target entered. If a campaign had been beating that target, its previous performance was no longer a safe baseline.
That can change traffic and cost patterns, but it does not guarantee a higher CPC. First confirm that the campaign was limited by budget, used an affected strategy and had been outperforming its target. Then check what changed after the rollout.
PPC Live covered how practitioners prepared for the update, what they did afterwards and whether ROAS targets are becoming performance ceilings.
The context of the search changed
Google considers the search term, location, device, time, other ads, other search results and additional user signals when calculating Ad Rank.
Two searches connected to the same keyword can therefore produce different auctions and different prices.
Before blaming Quality Score, segment CPC by:
Device
Location
Day and hour
Network
Top compared with other positions
Brand compared with non brand intent
New compared with returning users, where your measurement setup allows it
A shift in any of these can move the account average even when nothing is wrong with the keyword.
Do not turn AI Overviews into a convenient explanation
AI Overviews are changing search behaviour. A Pew Research Center analysis found that users clicked a traditional search result in 8% of visits when an AI summary appeared, compared with 15% when one did not.
That is evidence of changing organic click behaviour. It is not evidence that every lost organic click becomes a new paid bidder, or that AI Overviews caused the CPC increase in a particular account.
Treat that connection as a market hypothesis. If you want to use it as an account diagnosis, you still need account evidence.
Do not confuse expensive traffic with invalid traffic
Invalid traffic is a traffic quality and billing issue. It is not automatically an explanation for a higher average CPC.
A retail study from invalid traffic detection company Lunio found 72% more invalid traffic in retail Search campaigns using AI Max than in standard Search campaigns by invalid traffic detection company Lunio. The finding is worth knowing, but so are its limits. It was a third party retail study, and the wider dataset of more than 414 million clicks covered several advertising platforms. It was not 414 million AI Max clicks.
Google’s Invalid Activity Credit Report shows credited clicks, interactions and amounts by campaign and network for Search and Performance Max.
Use it to separate auction pricing from traffic quality and billing. Do not use it as proof that invalid traffic raised CPC unless your account data supports that conclusion.
What Quality Score is still useful for
Quality Score is useful when it gives you a place to investigate. It becomes a problem when the number itself becomes the end all be all strategy.
Use it to:
Spot a sudden change in expected CTR, ad relevance or landing page experience
Decide which part of the ad, query and landing page journey needs attention
Prioritise high value keywords where better relevance could improve performance
Identify weak landing page experiences that may also be hurting conversion rate
The component ratings are more useful than the headline score. “Landing page experience: Below average” gives you somewhere to start. “Quality Score: 6” does not.
It is also a Search diagnostic at keyword level. It does not provide an equivalent view across Performance Max or every auction reached through newer automated targeting.
The underlying work still matters. Relevant copy, sensible structure and a landing page that answers the query can improve auction time quality, conversion rate and the value of the traffic you buy.
Just do not promise that moving a historical diagnostic from 5 to 8 will reverse a change in competition, bidding or traffic mix.
How to diagnose a CPC increase properly
1. Confirm that the increase is real
Compare equivalent date ranges and account for seasonality, promotions, brand activity and conversion lag. Do not compare a partial week with a full one.
2. Break apart the account average
Segment CPC by campaign, network, device, location, time, brand status, query category and match source. The average may have risen because the traffic mix changed.
3. Check the auctions
Review Auction Insights, impression share, top impression rate and outranking changes. Look for new competitors and changes in how often existing competitors appear.
4. Check reach, bidding and platform changes
Review Change history, AI Max settings, search terms, negative keywords, URL expansion, Search Partners, bid strategy, targets and budget status. This is where you connect the timing of the CPC change to something that altered eligibility or bidding.
5. Check auction time quality
Use the Quality Score components as clues, then inspect the actual relationship between the search, ad and landing page. Do not stop at the score from 1 to 10.
6. Check what the higher CPC bought
Compare conversion rate, qualified lead rate, average order value, revenue, profit and customer value. Then review the Invalid Activity Credit Report separately if traffic validity is a concern.
The point of this process is not to prove Quality Score innocent. It is to identify the cause before choosing the fix.
How to explain rising CPCs without losing the room
Lead with what happened in the account
Do not hide behind an industry benchmark. Start with the observable change:
“CPC increased by 14% after the query mix shifted towards non brand searches. Competitor overlap also increased, but conversion rate improved enough that CPA stayed within target.”
That is an explanation. “CPCs are up everywhere” is a defence.
Separate evidence from hypotheses
Say what the data proves, what it suggests and what you are testing next. A plausible theory is not the same as a confirmed cause.
Separate what you control from what you can influence
You can improve ads, landing pages, conversion data, targeting controls and bidding inputs. You can respond to competition and market demand, but you cannot optimise them out of existence.
Move the conversation from cost per click to value per click
If CPC rises, the durable responses are better conversion rates, stronger lead quality signals, more accurate values and a channel mix that does not depend on one auction.
The question is not only, “How do we make the click cheaper?”
It is also, “How much value did the more expensive click create?”
Conclusion: At the end of the day stop chasing the Quality Score number
Quality Score can tell you where to look. It cannot tell you the full story of why CPC changed.
That story may involve auction time quality, but it may also involve competition, query mix, bidding targets, platform changes or the context of the searches you entered. The job is to separate those forces, show what changed and decide whether the more expensive traffic created enough value.
Improving ad relevance and landing page experience is still good PPC. Treating the score from 1 to 10 as the diagnosis is not.
Comments