Fourteen B2B companies that raised their Google Ads budgets between 2025 and 2026 ended up paying more for the exact same keyword, not simply more keywords or more clicks. It is the kind of pattern this site has tracked before, in how AI defaults have quietly taken over bidding decisions in Google Ads and Meta Ads accounts without anyone auditing what changed. The data behind this particular pattern comes from Grey Matter, a paid-search agency, which published it in an October 2026 report titled Same Click, Higher Price.
Grey Matter tracked 1,349 non-brand keywords across six quarters, from April 2025 to September 2026, and found that the average cost per click rose 21% year over year on an identical keyword, after accounting for the usual explanations: higher spend, lost impression share, better position, improved Quality Score. For a business weighing how much to add to a 2027 Google Ads budget, that 21% matters more than the account-wide average CPC a dashboard usually shows, because the average blends two different things: the mix of keywords you bid on, and how much each one now costs.
What Grey Matter measured
The sample is modest by design: 14 B2B advertisers, all existing Grey Matter clients, tracked individually rather than averaged into a blended industry figure. The agency excluded Performance Max, Shopping and Display campaigns, along with broad-match keywords and Maximize Clicks strategies, to isolate Search campaigns running on conversion-focused automated bidding. A regression model then separated out how much of the CPC change came from the advertiser’s own spend, lost impression share, average position and Quality Score, leaving a residual attributable to market-wide price pressure and anything the model could not explain.
Grey Matter is explicit about what that residual does and does not prove: the report itself states this is not a controlled experiment and cannot establish what caused any single price change — which Search Engine Journal’s coverage summarizes by calling the report “a reference point for B2B non-brand Search, not a forecast for any single account.” The 21% figure carries a wide 95% confidence interval, from 1% to 44%, which is what a 14-advertiser sample produces. Those limits do not erase the pattern, but they do set its scope: directional evidence from one agency’s book of business, not an industry constant to plug into a forecast.
The same keyword, a higher price
Before any statistical control, the 14 advertisers’ CPC on the same keyword rose 29% year over year. Controlling for the extra spend itself brought that down; adding position and Quality Score narrowed it further, to the reported 21%. Spend still has its own, separate effect: each 10% increase in an advertiser’s budget came with roughly a 1.4% rise in CPC on that same keyword, a relationship that holds even before counting the extra clicks that spend buys.
Looking at the 11 advertisers with enough continuous data to build a chained price index, the average non-brand CPC rose 36% year over year. Grey Matter’s model attributes about 22 of those 36 points to market-wide price inflation that hit every advertiser regardless of their own choices, and roughly 3 points to the advertiser’s own budget increases; changes in which keywords and match types they bid on actually pulled the number down by about 3 points. That leaves roughly 14 points the model cannot explain, which Andy Hendricks, Grey Matter’s head of search, attributes in part to category-level competition the regression does not capture.
One advertiser in the sample nearly quadrupled its budget and saw its CPC on the same keyword rise 132%, the most extreme case in the group. It is not representative on its own, but it shows the range a budget increase can produce once automated bidding starts reaching into pricier territory.
Why a bigger budget buys pricier clicks
Hendricks ties the pattern to how Google Ads’ automated bidding decides where to spend a larger budget. Google Ads describes Maximize Conversions, the bidding strategy most of these accounts run, as a system that sets a bid in real time for every auction, based on historical account data and live signals such as device, location and time of day, aiming to win as many conversions as the budget allows. When that budget goes up, the strategy does not just buy more of the same auctions: it starts bidding into auctions it previously skipped, which tend to be higher positions, more competitive searches and pricier time slots.
That is a different mechanism from the one a federal court documented in Google’s own internal search ads pricing systems, where price increases were reportedly built into Google’s own auction mechanics rather than triggered by an advertiser’s budget choice. Grey Matter’s report makes no such claim, and conflating the two would overstate what this data shows. What the agency’s numbers describe is narrower: an automated strategy spending a bigger budget by reaching for more expensive clicks, not a hidden lever Google turns independently of advertiser behavior.
Manual CPC bidding does not carry this particular blind spot: the advertiser sets a ceiling for every keyword, auction by auction. That is also why many B2B accounts moved away from it, since reviewing bids one keyword at a time stops scaling once an account runs into the hundreds of them. Automated bidding trades that manual ceiling for scale, and Grey Matter’s data is really describing the price of that trade once a budget grows large enough to reach further into the auction.
What this means before you raise a 2027 budget
Grey Matter’s own recommendations, built from this data, are worth taking at face value rather than as generic advice: raise 2027 budgets in stages and check the cost of each additional click at every step, rather than approving one large increase and reading the result months later. For campaigns running Maximize Conversions or a similar automated strategy, set a conversion value target and a performance floor before the budget goes up, not after, so the system has a boundary to respect once it starts reaching for pricier auctions.
The more useful diagnostic, though, is simpler: track CPC on a fixed set of keywords over time, separate from the account-wide average. An account average mixes two things that move independently, how much of your budget goes to each keyword and how much that keyword now costs, and it can look flat even while the keywords you actually depend on are getting more expensive. That same-keyword view is also where clean conversion data earns its keep: a budget increase that quietly buys pricier clicks is only a problem if those clicks are not converting at a rate that justifies the price, which is the same discipline this site has described for reading PPC return on investment past the first click.
None of this turns a 21% figure from 14 advertisers into a rule for any specific account. It is a flag to check your own numbers against, not a forecast to build a 2027 plan on.
The number worth checking before the dashboard average
The account-level CPC average on a Google Ads dashboard was never built to catch this. It blends keyword mix and keyword price into one number, which is exactly why a correlation like Grey Matter’s can sit undetected inside an account for quarters. Before committing more budget to an automated bidding strategy in 2027, the data worth pulling is narrower and less comfortable than the dashboard default: what the same keyword cost a year ago, and what it costs now.
If you want to run this same-keyword check against your own account data, we are glad to talk through what it shows.
Photo credit: Pixabay.







