The same non-brand Google search keyword cost B2B advertisers about 21% more between April and September 2026 than in the same months of 2025, after removing the effect of their own budget, ad position and Quality Score changes, according to an analysis of 14 advertisers by Grey Matter, a B2B marketing agency. The study, titled "Same Click, Higher Price" and dated October 2026, tracked 1,349 non-brand keywords over six quarters using the advertisers' own Google Ads data. Once the advertisers' own decisions were counted back in, the typical account's non-brand cost per click rose 36%. Every 10% added to campaign spend raised an advertiser's own click prices by about 1.4% on top of the market.

In Short

Grey Matter, an agency that runs marketing for B2B companies, followed the same 1,349 Google search keywords across 14 of its clients for a year and a half and found that an identical click cost about 21% more than a year earlier, after taking out the effects of the advertisers' own budget, ad position and ad quality changes. When those advertisers' own choices are added back, the typical one paid 36% more per non-brand click, with bigger budgets pushing prices up further and individual results ranging from a 30% fall to a 121% rise. The sample is small, drawn from one agency's clients, and carries a likely range of 1% to 44%, so the study points clearly upward for B2B search costs without fixing a precise rate or explaining what drove the increase.

Fourteen advertisers, 1,349 keywords

The report is written by Andy Hendricks, head of search at Grey Matter, and sets out a narrow question. Published CPC benchmarks, the agency argues, report a single average that blends new competitors, larger budgets, different keywords, and shifts in match type and bidding. "A raw average can't tell" whether the same click actually got more expensive, according to the report. Grey Matter wanted to know what happens when an advertiser buys the exact same click a year later.

The sample is 14 B2B advertisers drawn from Grey Matter's client base. Names, keywords and identifying details are anonymised, and every figure is reported either in aggregate or for unnamed accounts; the 11 accounts in the price index are labelled A to K. Only Google Ads Search campaigns are included. Performance Max, Shopping and Display are excluded.

The period runs for six quarters, from April 2025 through September 2026. Year-over-year figures compare April to September 2026 with the same months of 2025, a choice that, according to Grey Matter, controls for seasonality because B2B search costs can vary meaningfully from one quarter to the next.

Every keyword was tagged as brand, competitor or non-brand. Campaign names decided the label first; keywords elsewhere were classified by their text against each advertiser's brand names and a list of known competitors. Only non-brand keywords feed the main estimates, because brand and competitor CPCs, according to the report, "follow different dynamics."

Two further exclusions shape the result. Broad match keywords are left out because the searches they match drift over time: "a broad keyword's CPC can rise simply because Google starts matching it to pricier searches," the report states. Keywords in Maximize Clicks campaigns are also excluded, since their CPCs "mostly reflect how the bidding algorithm spreads a fixed budget, not what the auction charges."

The analysis controls for four things advertisers influence: campaign spend, impression share lost to budget, top-of-page position and Quality Score. Each advertiser carries equal weight regardless of spend, which keeps the largest budgets from dominating the result.

Two methods

Grey Matter uses two complementary approaches. A regression model follows each keyword across all six quarters, estimates the change in its CPC after the controls, and uses all 14 advertisers. It produces the headline number. A chain-linked keyword index, the method used to build consumer price indexes, matches keywords from one quarter to the next and multiplies the quarterly changes together. Keywords need at least five clicks in both quarters of each link, and an advertiser's matched keywords must cover at least 40% of its eligible spend. Eleven advertisers cleared that bar. Three had too little keyword continuity to be included.

The method summary counts 5,035 keyword-quarters across the 1,349 keywords. That works out at an average of about 3.7 quarters of data per keyword out of a possible six, and roughly 96 tracked keywords per advertiser, according to PPC Land's calculation from the report's figures.

The report is direct about what the design can and cannot do. "This is not a controlled experiment, and it can't prove what caused any single price change," it states. What it can do, according to Grey Matter, is measure how much the price of the same click changed across many advertisers and separate the change every advertiser faced from the change each one created for itself.

A quarter-by-quarter price index

The index sets the second quarter of 2025 at 100. The same keywords then moved to 110 in the third quarter of 2025, 111 in the fourth, dipped to 106 in the first quarter of 2026, and rose to 122 in the second quarter and 132 in the third.

Most of the increase came in 2026. Between the first and second quarters of 2026 the index rose about 15%, and between the second and third about 8%, according to PPC Land's calculation from the published index values. Compared with the same quarter a year earlier, spring prices were up 22% and summer prices up 20%, according to the report.

One figure in the report's summary panel needs care. The "+32%" it lists is the cumulative rise from the second quarter of 2025 to the third quarter of 2026, a span of five quarters rather than a year. The chart's 95% band also widens as each quarter is measured further from the base: by the third quarter of 2026 it stretches from just above 100 to nearly 170.

From 29% to 21%: what the controls remove

The 21% is what remains after three adjustments. Before any controls, keywords that advertisers bought in both years cost 29% more. Accounting for budget increases and cuts lowered that to 22%. Adjusting for ad position and Quality Score lowered it to 21%.

Budget, in other words, did most of the work. According to Grey Matter, advertisers' own spending decisions inflated the apparent market increase by about 7 points before any other adjustment. Position and Quality Score together moved the estimate by a single point.

Each step carries a wide interval. The uncontrolled figure has a likely range of 4% to 59%, the budget-adjusted figure 1% to 48%, and the main estimate 1% to 44%. The ranges are 95% intervals clustered by advertiser, according to the report's appendix.

Bigger budgets, dearer clicks

The study's second finding concerns the part of inflation advertisers create for themselves. Across the sample, every 10% increase in campaign spend raised an advertiser's own CPCs by about 1.4%, on top of the market. Doubling a budget, according to the report, adds roughly 11%. That corresponds to an elasticity of about 0.15, with a doubling adding about 10.6%, according to PPC Land's calculation.

The most extreme case nearly quadrupled its non-brand Search spend, an increase of 283%, and the same non-brand keywords cost it 132% more. In the third quarter, the advertiser's brand keywords, on a flat budget, rose only about 4%. "Same advertiser, same market, very different price changes, depending on how the campaigns were funded," the report states.

Grey Matter's explanation ties the effect to automated bidding. When a campaign on a strategy such as Maximize Conversions, one of Google's Smart Bidding strategies, gets more budget, the algorithm spends it by entering auctions it previously skipped: higher positions, more competitive searches and pricier times of day, according to the report. The keywords stay the same while the price paid for them climbs. The agency does not frame this as a mistake. More budget usually buys more clicks. "But each additional click costs more than the last, which matters for anyone scaling a budget," according to the report.

That account of the mechanism is an interpretation, not a measurement. The study does not report which bid strategy each advertiser ran, and Google Ads, as the report acknowledges, does not provide historical budget settings or bid targets, so spend and impression share lost to budget stand in for them.

A steeper line in the chart

The scatter chart that illustrates the finding plots each of the 11 indexed advertisers' change in campaign spend against the change in what the same keywords cost them, on log scales, with a fitted trend line. Read against those scales, the line rises by roughly 3.5% for every 10% of extra spend, or close to 30% for a doubling, according to PPC Land's reading of the chart. That is noticeably steeper than the 1.4% stated in the text. The line also crosses flat spend at around 30%, which matches the "spend trend, at flat spend" estimate of 31% that the report gives elsewhere, so the chart appears to show the advertiser-level spend trend rather than the coefficient from the keyword-level regression. The report does not reconcile the two.

The same chart shows how loosely spend and price are tied for individual accounts. One advertiser cut spend by roughly 40% and still saw the same keywords cost about a quarter more. Another raised spend by about 15% and saw the same keywords cost about a third less, according to PPC Land's reading of the chart.

Where this meets Google's August 17 change

The study's window overlaps with one of the most debated Google Ads changes of 2026. On August 17, 2026, budget-limited campaigns on Target CPA or Target ROAS began delivering toward their stated targets rather than beating them, across Search, Shopping, Performance Max, Demand Gen and Travel. Google had told advertisers about the change on July 2, and in early August Google's Ginny Marvin said the target becomes the efficiency lever in budget-capped campaigns.

Grey Matter's budget controls include impression share lost to budget, a measure of the same budget-limited condition that the August change keys on. Yet the report does not mention the change and does not split the third quarter before and after it. The timing in the index also cuts against a simple link: the steepest quarterly rise, about 15%, came between the first and second quarters of 2026, before August 17. The third quarter, which includes roughly six weeks after the change, rose about 8%, according to PPC Land's calculation.

The budget finding lands in a live argument. In October, two practitioners criticised Google Ads for turning budget into a performance lever, one of them calling budget the new Quality Score. Grey Matter's numbers put a size on part of that complaint for this sample: budget changes accounted for about 7 points of the raw increase, while position and Quality Score adjustments accounted for one.

What advertisers actually paid

The typical advertiser's average non-brand CPC rose 36% year over year. Using the chain-linked index for the 11 advertisers with enough keyword continuity, Grey Matter split each advertiser's change into four parts:

  • Market: +22 points, the increase every advertiser faced.
  • Budget: +3 points, the extra cost of advertisers spending more.
  • Keywords: -3 points, changes in keywords, match types and bid strategies.
  • Other: +14 points, increases the first three parts do not explain.

Keyword and match type changes nearly cancel out on average, because some advertisers shifted toward pricier keywords and others toward cheaper ones. The "other" bucket is the larger surprise. At 14 points, it is bigger than the budget and keyword effects combined.

According to the report, the "other" bucket comes mostly from a few advertisers whose CPCs roughly doubled, "likely" because of competition specific to their categories that a market-wide estimate cannot capture. That is offered as an explanation, not tested. The report's per-account chart adds a detail. The three largest increases belong to accounts J, A and K, and two of them, J and K, are among the three accounts with fewer than 12 comparable keywords per quarter. Much of the largest movement therefore sits where the keyword evidence is thinnest.

Two smaller points about the arithmetic. The 36% matches the geometric mean of the 11 per-account changes shown in the report; a simple average of the same figures would be 43%, and the median is 28%, according to PPC Land's calculation. And the report uses three slightly different market figures: 21% from the regression on 14 advertisers, 22 points from the index on 11, and "about 20%" in its conclusions.

Eleven accounts, eleven different outcomes

Averages hide most of the story. Ten of the 11 indexed advertisers paid more for the same keywords than a year earlier, and nine paid more overall. Actual CPC changes ranged from a 30% decrease to a 121% increase.

The per-account figures for actual CPC change were: J +121%, A +116%, K +77%, I +48%, B +40%, E +28%, D +28%, C +28%, G +21%, F -4% and H -30%. Account H was the only one that paid less for the same keywords, and it paid about 30% less overall.

The gap between what the same keywords cost and what an advertiser actually paid comes from its own account changes. Those changes moved individual advertisers' CPCs by as much as 17% in either direction. One advertiser that kept its budget flat saw the same keywords rise 20%, then paid 40% more overall after moving spend into pricier phrase match keywords. The two figures are consistent: 1.40 divided by 1.20 is about 1.17. "For any single advertiser, its own account decisions can matter almost as much as the market," according to the report.

How solid is 21%?

With 14 advertisers, the report concedes that any single estimate carries real uncertainty. The main estimate of +21% has a likely range of +1% to +44%, so Grey Matter tested other ways of measuring the market.

An advertiser with a flat budget, the closest thing in the data to a pure market reading, saw the same keywords cost 20% more. The spend trend read at flat spend gives 31% across all 11 indexed advertisers, or 20% when the three with the fewest comparable keywords are removed. Leaving out the biggest spender brings the main estimate down to +17%. Including broad match keywords raises it to +31%, which the report reads as consistent with broad match drifting into pricier searches. The spend trend without the biggest spender gives +32%.

Every method and check falls between +17% and +32%. "The true figure could be meaningfully higher or lower, but every way we measured it points to a substantial increase," according to the report.

Those checks have limits of their own. They re-run the analysis on the same 14 or 11 advertisers, so they test modelling choices rather than whether the sample represents B2B advertisers at large. One specification also matters more than the summary suggests. Without the biggest spender, the appendix gives a likely range of -2% to +40%, an interval that includes zero. In that version, at the 95% level, the study cannot rule out no market increase at all.

How the number compares with other CPC data

Grey Matter's estimate sits well above other published CPC measures for 2026, though it measures something different. WordStream's 2025 benchmarks had average cost per click rising 12.88% to $5.26 across more than 16,000 search campaigns for April 2024 to March 2025. Its 2026 edition put the figure at $5.42, while cost per lead fell for the first time since 2020, a CPC rise of about 3%, according to PPC Land's calculation. Channable data showed CPC up 15% year over year on Shopping and Performance Max, with return on ad spend down 46% on Performance Max, comparing June 2025 with June 2026.

Alphabet's own disclosure is lower still. Its quarterly filing showed paid clicks up 8% and average cost-per-click up 4% in the third quarter of 2026. That is an average across every advertiser, market and Google property, the kind of blended figure Grey Matter argues cannot answer the same-click question. A fixed basket of B2B non-brand Search keywords in a single agency's accounts and a company-wide average measure different things. They do not contradict each other. They do suggest that this particular basket rose several times faster than Google's average.

The comparison also shows why the WordStream result matters for reading this one. Grey Matter measures the price of a click, not the cost of a lead or a sale. Conversion rates, lead quality and pipeline value are outside the study, so a rising CPC on its own does not show whether these advertisers' acquisition costs rose by the same amount.

What the study does not establish

The limits are worth setting out in one place.

  • Representativeness. The 14 advertisers are clients of one agency. Their industries, countries, currencies and spend levels are not disclosed. Grey Matter's site lists manufacturing, private equity, professional services and software and technology among the industries it serves, but the report does not say which of these the sample covers.
  • Cause. The market component is what remains after the controls. The study does not separate competition from Google's own auction mechanics, product changes or anything else, and it says so: it measures correlation, "not proven causation."
  • Coverage. Performance Max, Shopping, Display, broad match and Maximize Clicks are all excluded. That matters because Search campaigns are a shrinking slice of many accounts: in Brainlabs data, Search fell to 41% of Google Ads clicks in August 2026 from 62% in January 2024, and its share of spend dropped from 86% to 71%.
  • Bid settings. Historical budgets and bid targets are not available from Google Ads, so the controls use spend and impression share lost to budget as proxies.
  • Thin baskets. Three of the 11 indexed advertisers average fewer than 12 comparable keywords per quarter.

The exclusion of broad match deserves a note of its own. Google has long promoted AI keyword matching through broad match paired with Smart Bidding. Grey Matter left broad match out to keep the basket stable, and including it lifted the estimate from 21% to 31%. The headline figure therefore covers phrase and exact match keywords only, while Google's own product direction has favoured broad match.

The report also appears on Grey Matter's website beside its service pages, pricing links and an invitation to commission similar work, and the author's biography notes availability for podcast interviews and featured articles. None of that changes the arithmetic. It is context for a study that does not report any independent review.

Both sides of the results page

The report frames itself as a sequel. Grey Matter's earlier study, "The Great Declick," dated July 2026, found that click-through rates on informational queries fell 38% year over year, from 0.90% to 0.56%, across 18 B2B accounts with ranking position held constant. Other measurements have pointed the same way: Seer Interactive found that organic click-through rates fell 61% and paid click-through rates 68% on informational queries showing AI Overviews. The growth of zero-click search is the backdrop to both.

Grey Matter's argument is that B2B advertisers are squeezed from both directions at once. Organic search sends fewer clicks, and paid search charges more for each. The report closes with five planning recommendations of its own: build roughly 20% annual CPC inflation into forecasts before setting growth targets; scale budgets in stages and judge each step by the cost of the extra clicks it buys; separate market movements from account changes before attributing a cost jump; decide what additional conversions are worth before raising budgets on Maximize Conversions; and plan SEO and paid search together.

For the paid search industry, the study adds a measured number to a debate that has mostly run on averages and anecdote. PPC Land has tracked the tension between Google's revenue optimisation and advertisers' own through 2026, and the question of what Google charges has also reached the legal system: law firm Keller Postman has brought mass arbitration claims alleging Google's conduct inflated ad prices since August 2016. Grey Matter's study does not test those allegations and says nothing about Google's intent. What it does offer is a method that others with access to their own accounts could repeat: hold the click constant, strip out the budget, and see what is left.

Timeline

Summary

Who: Grey Matter, a B2B marketing agency, and its head of search Andy Hendricks, who led the study. The data come from 14 anonymised B2B advertisers among the agency's clients. The findings bear on B2B advertisers buying Google Search ads and the agencies that plan their budgets.

What: An analysis of first-party Google Ads data covering 1,349 non-brand keywords and 5,035 keyword-quarters. After controlling for campaign spend, impression share lost to budget, top-of-page position and Quality Score, the same keyword cost about 21% more year over year (likely range 1% to 44%). The typical advertiser's actual non-brand CPC rose 36%, split into market (+22 points), budget (+3), keyword changes (-3) and unexplained (+14). Every 10% increase in spend added about 1.4% to an advertiser's own CPCs. Individual accounts ranged from -30% to +121%. Broad match, Maximize Clicks, Performance Max, Shopping and Display were excluded.

When: The study is dated October 2026. Its data cover April 2025 to September 2026, with year-over-year figures comparing April to September 2026 against the same months of 2025.

Where: Google Ads Search campaigns. The report does not state the countries or currencies involved.

Why: Published benchmarks blend market prices with advertisers' own changes in budget, keywords and bidding, so they cannot say whether the same click became more expensive. The study attempts to separate the two. It cannot establish what caused the market increase, does not measure conversions or lead costs, and rests on a small sample from one agency's clients.