Grey Matter Found B2B Clicks Up 21% on Identical Keywords (Plan 2027 for 20%)
B2B agency Grey Matter tracked 1,349 non-brand Google Search keywords across 14 B2B advertisers from April 2025 to September 2026 and found the same click cost about 21% more year over year, with budget, top-of-page position and Quality Score held constant. Its data also suggests each 10% budget increase lifts your own CPC by roughly 1.4%. Plan 2027 search budgets with about 20% CPC inflation baked in.
The report is called "Same Click, Higher Price," and PPC Land walked through the methodology this morning in a lot more detail than most agency studies get. Andy Hendricks, Grey Matter's head of search, built two estimates: a regression that follows each keyword across six quarters, and a chain-linked price index that only matches keywords with at least five clicks in both quarters. The regression says +21%. The index lands at +22 points. Every sanity check they ran falls between +17% and +32%.
So no, this isn't your account manager being sloppy. Something in the market moved.
Why the 21% matters more than the bigger numbers
Most CPC benchmarks you've seen compare average CPC this year to average CPC last year. That number mixes everything together: new keywords, match type changes, budget increases, a new bid strategy someone switched on in March. You can't tell if the click got more expensive or if you just started buying more expensive clicks.
Grey Matter tried to strip that out. The uncontrolled number was +29%. After adjusting for budget, it dropped to +22%. After position and Quality Score, +21%. The remaining 21% is, roughly, the market charging more for the identical thing.
For context, Dreamdata's B2B benchmark found non-branded search CPC rose about 29%, from $4.13 to $5.34, between August 2024 and July 2025, while CTR fell from 5.47% to 4.04%. Interesting that Dreamdata's raw number and Grey Matter's uncontrolled number are both 29%. I wouldn't read too much into that coincidence, but it does suggest the "about 20% after controls, about 30% before" shape isn't a one-agency fluke.
Compare that with the blended picture. Alphabet's Q3 2026 10-Q shows average cost-per-click up just 4% year over year across Google Search and other properties, with paid clicks up 8%. And the WordStream/LocaliQ benchmark put the 2026 average CPC at $5.42, roughly 3% up on the prior year, with cost per lead actually falling for the first time since 2020.
On paper, those numbers contradict each other. They probably don't. Google's 4% is a global average across every vertical, every country and a huge pile of cheap consumer clicks. B2B non-brand search is the narrow, crowded end of that pool, where a handful of SaaS competitors bid on the same 90 keywords. Averages hide the expensive corners.
The part where your budget bids against you
This is the finding I'd actually forward to a finance team. Grey Matter estimates that every 10% increase in an advertiser's own spend raised its CPCs by about 1.4% (doubling spend adds roughly 11%). In its most extreme account, non-brand spend went up 283% and non-brand CPC went up 132%, while that same advertiser's brand CPCs rose about 4% on a flat budget.
Same advertiser, same market, very different price changes, depending on how the campaigns were funded.
That line is from the report, and it's the whole thing really.
One caveat that PPC Land flagged and I think matters: eyeballing the report's own scatter chart suggests closer to 3.5% per 10% of spend, or about 30% per doubling. The report doesn't reconcile the two. So the true budget penalty could be somewhere between "annoying" and "painful." I'd plan around the higher end until your own data says otherwise.
And to be fair, this isn't entirely new. Smart Bidding has always reached for marginal clicks when you hand it more money. It just seems a lot less forgiving now that the base price is also climbing. Two forces stacking on the same click.
What's actually driving it (nobody knows, honestly)
The report is upfront that it's "not a controlled experiment" and can't name a cause. A few candidates are floating around.
The obvious one is Google's August 17 change to budget-limited tCPA and tROAS campaigns, which now deliver toward the target instead of banking the surplus. We covered how that ate a 20% ROAS cushion in three weeks for one advertiser. But here's the awkward bit: Grey Matter's steepest jump (about +15% from Q1 to Q2 2026) happened before August 17. So it can't be the whole story.
Shrinking click supply is the other suspect. Grey Matter's earlier "Great Declick" study found informational-query CTR fell 38% across 18 B2B accounts. Fewer clicks available on the same queries, same number of competitors, and prices go one direction. It's a bit like a parking garage closing two floors and keeping the same number of monthly passes. Nobody raised the official rate, but finding a spot got expensive.
The same squeeze shows up outside B2B. Channable's European ecommerce data shows CPC up 15% year over year on Shopping and PMax, with PMax ROAS down 46% between June 2025 and June 2026. Different sample, different campaign types, same direction.
How much to trust a 14-advertiser study
Some, not blindly. The sample is 14 clients from one agency, countries and currencies aren't disclosed, and the report lives on Grey Matter's site next to an invitation to commission similar work. Excluding the biggest spender, the 95% interval runs from -2% to +40%, which technically includes zero. Two of the eleven indexed accounts have fewer than 12 comparable keywords per quarter, and those two posted +121% and +77%.
Also, the headline excludes broad match. Including broad match pushes the estimate to +31%. Given how hard Google pushes broad match with Smart Bidding, I suspect a lot of real accounts sit closer to that number than to 21%.
Still, from what I've seen, this is the most careful attempt anyone has published to separate market price from account behavior. I'd rather work from a careful small study than a big sloppy average.
Re-pricing the 2027 plan before Q1 kickoff
If your 2027 paid search forecast assumes flat CPCs, it's already off. A $40,000/month non-brand budget that bought about 7,500 clicks at a $5.34 CPC buys closer to 6,200 clicks at +21%. That's roughly 1,300 fewer clicks a month, before any budget increase penalty.
Here's what I'd do this week, roughly in this order:
- Build your own same-keyword index. Pull non-brand keyword data for Q3 2025 and Q3 2026. Filter to exact and phrase keywords with at least 5 clicks in both periods (Grey Matter's threshold). Compare CPC on that matched set only. Takes maybe 20 minutes in a spreadsheet. If your matched-set CPC rose less than 10%, you're doing better than the study's market. Above 30%, something account-specific is also going on.
- Split the variance before anyone blames the agency. Grey Matter's typical account saw non-brand CPC up 36%: 22 points market, 3 points budget, minus 3 for keyword and bid changes, and 14 points "other." Do the same breakdown. It changes the conversation with a CFO from "PPC got worse" to "the price went up, here's what we controlled."
- Stage budget increases in 15-20% steps. Then judge each step on the marginal cost of the extra clicks, not the blended average. If the extra 20% of spend buys clicks at 1.5x your current CPC, that's where to stop.
- Price the conversion before raising Maximize Conversions budgets. Decide what one more qualified lead is worth. Then check the next step against it.
Personally, I'd put a 20% CPC inflation line in the 2027 plan and a separate line for whatever you expect budget growth to add on top. Finance teams handle "the market rate went up" much better than they handle surprise overspend in March.
A guess, with stakes
My prediction: at least two more agencies publish matched-keyword CPC studies before the end of Q1 2027, and both land somewhere between 15% and 30%. Once that happens, "CPC inflation" stops being a vibe in client calls and becomes a standard line in every B2B search forecast, the way agencies already pad for Q4 CPMs.
I don't think the people who come out of 2027 fine are the ones who found a clever bidding trick. They're probably just the ones who stopped assuming last year's click price was a fixed cost.