The typical B2B advertiser spends 25% of budget creating demand, and 16% buying clicks.
A starting point, not a target.
The advertiser in the middle of the B2B market spent 25% of paid budget creating demand, 7% chasing it with retargeting and 16% buying clicks, and buying clicks is the line that swallowed $12.7M across this dataset and handed back 66 leads. Three percentages, on 109 advertisers with more than $10K of 2025 spend, that let you say out loud where your own plan sits. One thing this page will not do, on purpose: tell you what your split should be.
Pull three percentages out of last quarter’s spend before the next planning meeting, what you put into creating demand, into retargeting, and into traffic and click campaigns, and walk in with the market beside them. Across 109 advertisers, the quietest quarter put 4% of budget into creating demand, the middle 25%, the boldest quarter 51%; retargeting sat at 7% in the middle, and buying clicks at 16%. That click line is the market’s weakest-defended money, and the easiest of the three to move.
2025 · 109 advertisers above $10K spend for how each one split its own budget · 153 advertisers for the share of all the dollars · Methodology · Suppression rules · Typical vs best-case: how to read this
Read this before you copy these numbers
Copying the middle of the market is how you reproduce the 0.56x.
This is where the market sat, not what worked. The middle is partly a middle of mistakes: it describes what a mixed set of advertisers did in 2025, and the same dataset shows that all that spending returned 0.56x for every ad dollar in revenue we can trace back to an ad. Use it to place yourself, “we are in the quietest quarter on creating demand”, then argue about what each line brings back, not about matching a number the market happened to land on.
Executive summary
- The benchmark: how 109 advertisers each split their own budget. Creating demand: the quietest quarter 4%, the middle 25%, the boldest quarter 51%. Retargeting: 0%, 7%, 19%. Buying clicks: 4%, 16%, 40%.
- The one line to defend: traffic and click campaigns. Across the dataset they took $12.7M and produced 66 leads, $191,686 apiece, across 112 advertisers. If your click line is near the 40% the heaviest quarter of advertisers spends, that is the first conversation.
- What this is not: a recommended split. Copying the middle of the market is how you reproduce the 0.56x the same dataset reports. A healthy split is a judgement call that depends on your sales cycle, how mature your category is and how much demand already exists for you. None of which this dataset can see. We publish the range; the judgement stays yours.
Playbook
- The play: compute your own three percentages, creation, retargeting, traffic, from last quarter’s spend before anyone asks for them, and bring them to the planning meeting with the market bands beside them.
- The likely finding: the click line is bigger than anyone intends. 16% in the middle of the market, 40% in the heaviest quarter, and across the dataset it bought leads at $191,686 each.
- The move that follows: re-point that click budget at a lead-generation goal and spend it on the format that priced best in this dataset, document creative on cold audiences made a lead for $148 against $358 for video. Cold and retargeting themselves came in at $187 and $196, close enough to call a tie, so the format is where the argument has evidence. See finding 01.
How does the market actually split a paid budget?
Wider than any “recommended mix” suggests. Across the 109 advertisers here with more than $10K of 2025 spend, the one in the middle put 25% of paid budget into creating demand, cold prospecting audiences of people who have never touched the brand, 7% into retargeting, and 16% into traffic and click campaigns. The spread matters more than the middle: on creating demand, the quietest quarter of advertisers spent 4% and the boldest quarter 51%. Half the B2B market is somewhere between barely creating demand and spending half its budget on it.
Count every dollar instead of every advertiser and the picture tilts: 52.5% of all 2025 spend went to creating demand, 22.9% to retargeting and 21.9% to traffic and click campaigns. The dollar share for creation runs above the middle advertiser because the largest spenders lean further into it. Which is itself a benchmark: the biggest budgets in this dataset create more demand than the typical one, not less.
What did the click line buy?
Across the dataset, almost nothing you can count in leads. Traffic and click campaigns took $12.7M of 2025 spend across 112 advertisers and produced 66 leads, $191,686 per lead. 61 advertisers put more than 30% of their budget there. Some of that money is doing real brand work a lead count cannot see, which is exactly why it needs a stated job and its own scoreboard: a click campaign bought to build the brand is a decision, and a click campaign sitting inside a cost-per-lead report is an accident. The full argument is finding 03.
What should you change?
Put the three percentages on the plan itself, creating demand, retargeting, buying clicks, and make each one be defended against the market range, not against last year. A budget that cannot state its own split cannot be compared to anything, and “the middle of the market spends 25% creating demand” is a far better planning input than a vendor’s recommended mix.
If you spend more than the market’s middle 16% buying clicks, move one slice of it into creating demand for a quarter and hold everything else steady. Judge it on cost per lead and on closed-won revenue per dollar, and keep whatever click spend has a stated brand job, on its own scoreboard, out of the cost-per-lead report.
“The benchmark says we should be at 25/7/16.” It says the advertiser in the middle was there. Copying that is how you reproduce the 0.56x. The ranges are wide on purpose, 4% to 51% on creating demand alone, and this dataset cannot see which split built the best business, only what the market chose.
Work out your own split this week: last quarter’s paid spend by campaign goal and audience type, as three percentages. Most teams have never produced this number, and it takes an afternoon in the ad platforms. Bring it to planning with the market range next to it rather than waiting to be asked.
Re-point one click campaign at a cold prospecting audience with a lead-generation goal, same offer and same creative. Click campaigns recorded leads at $191,686 while cold prospecting bought them at $187; the point of the test is to see that gap on your own account, and to give the click spend you keep a job the lead report was never able to read.
“Traffic campaigns are always waste.” A click campaign is bought to produce clicks, so a lead count is the wrong scoreboard for it, part of that 12.7M is deliberate brand spend. The defensible position is that it needs a stated job and its own reporting, not that it should be zero.
What are the exact numbers?
Two views of the same money. The first three columns describe how one advertiser splits its own budget, with every advertiser counting once. A $50K advertiser counts the same as a $2M one. The last column describes where all the dollars went. Both cover calendar-2025 spend.
| Share of paid budget | Lowest quarter | The middle | Highest quarter | Share of all the dollars |
|---|---|---|---|---|
| Creating demand (cold prospecting) | 4% | 25% | 51% | 52.5% |
| Chasing demand (retargeting) | 0% | 7% | 19% | 22.9% |
| Traffic and click campaigns | 4% | 16% | 40% | 21.9% |
The quarters and the middle: 109 advertisers with more than $10K of 2025 spend, each counted once. The last column: all 153 advertisers with every dollar counted, so the two answer different questions and will not match. Neither adds up to 100%. Awareness, engagement, video-view and other goals hold the rest, and each figure is the middle of a different line-up of advertisers. What the click row bought, $12.7M for 66 leads, is proved in finding 03.
How to find your own split in this
In the bottom quarter
You put less into this bucket than three quarters of the market does. That is a position, not an error, but it should be a stated one. Down here on creating demand usually means your pipeline is living off demand somebody else made.
In the middle half
You are where half of B2B advertisers sit. The benchmark has nothing to say about you here; the argument moves to what each line brings back, not what share of the budget it takes.
In the top quarter
You are more committed to this bucket than three quarters of the market. On creating demand that is a growth posture; on buying clicks at 40% or more, it is the single largest unexamined line in most plans.
The dissent we are publishing against ourselves
One reviewer called this a vanity cut, and the objection is worth printing: the three middle numbers do not add up to 100%. They come to roughly 48%, 25% creating demand, 7% retargeting, 16% buying clicks. Which looks broken if you read the row as one budget being sliced up.
It is not one budget. Each number is the middle of a different line-up of the 109 advertisers: the advertiser in the middle on creating demand is usually not the advertiser in the middle on buying clicks, and the rest of every individual budget sits in awareness, engagement, video-view and other goals outside these three buckets. So these are three separate markers you can hold yourself against, not three slices of a pie. Which is exactly why they can place you and cannot be added up, stacked in a chart, or handed to a planner as a mix.
The reviewer’s conclusion, do not use it to set a target. Is one we agree with and have built the page around. Where we differ: knowing that half the market sits between 4% and 51% on creating demand is a useful thing to know about your own position, provided nobody mistakes the middle of that range for a recommendation.
There is no recommended split in this report, and that is deliberate. Copying the middle of the market is how you reproduce the 0.56x. The right split depends on your sales cycle, how mature your category is, how much demand already exists for your brand and whether paid is buying new customers or just keeping you visible. None of which this dataset can see. We publish the range so a split can be placed and defended, not a prescription that would be a guess dressed up as a benchmark. The one line the evidence does argue with is buying clicks, and even there the recommendation is a stated job and its own reporting rather than a ban.
How do we know?
Every number above had to clear the same rules before it reached this page. In brief:
- Who is counted: 109 advertisers with more than $10K of calendar-2025 spend. The floor exists so that a $600 test account cannot show up as an advertiser putting 100% into creating demand; below it, one campaign decides the whole percentage.
- How a budget is bucketed: by what the campaign was bought to do and who it was aimed at. “Creating demand” is spend on cold prospecting audiences. Firmographic, technographic, intent and account-list targeting of people who have never touched the brand. “Chasing demand” is retargeting. “Traffic and click campaigns” is spend on campaigns bought for traffic or clicks rather than leads.
- Two ways of counting, on purpose: the quarters and the middle count every advertiser once; the share of all the dollars counts every dollar once, across all 153 advertisers. Where they disagree, 52.5% of dollars against 25% for the advertiser in the middle. The disagreement is the finding: the largest spenders split their budgets differently from the typical one.
- What has to be true before we print a number: at least 5 advertisers in a group, $50,000 of combined spend and no advertiser above 50% of that spend. These budget splits rest on 109 advertisers, well above the floor.
- We watched, we did not experiment: this is what the market did in 2025, not what worked. Nothing here says one split caused a better result, and this dataset cannot rank splits by business outcome. The revenue numbers that would settle it do not survive our own publication rules at this cut, because single advertisers hold too much of the result. That is the whole case against treating the middle as a target: the behaviour it describes is the same behaviour that returned 0.56x in revenue we can trace back to an ad.
- What else could explain this: advertisers label their own campaigns, so some “traffic” spend is genuinely brand work and some is a goal set by mistake; from platform data the two look identical. Company stage matters too. A category leader with strong inbound rationally creates less demand with paid than a challenger does. Read the range as context for a conversation, not as a scorecard.
Full rules in the Closed-Won Protocol, including the claims we refused to publish, with the exact reason each one was killed and the number it cost us. New here? Start with typical vs best-case: how to read any benchmark.
Cite as: Metadata 2026 B2B Ad Spend Benchmark (n=153 advertisers, $57.6M, 2025), metadata.io/benchmark-report-2026
What do you do different tomorrow morning?
One named change per seat. The columns reorder to match your reading lens; both are always on the page.
Make the split a line item
The one change: require every paid plan to state three percentages, creating demand, retargeting, buying clicks, and to defend each against the published range.
- Ask for last quarter’s three numbers first, before the next plan is written. If nobody can produce them in a day, that is the finding. The split is currently something that happened to you, not something you decided.
- Compare each to the market: creating demand 4% / 25% / 51%, retargeting 0% / 7% / 19%, buying clicks 4% / 16% / 40%. Each trio reads lowest quarter, middle, highest quarter. Sitting outside the market is fine; sitting outside it without a reason is not.
- Give the click line a stated brand job and its own reporting, or move it. It is the biggest unexamined position in the typical plan and the cheapest one to fix.
- Then use the Reality Gap calculator for the other half of the self-benchmark: where your cost per lead sits in the real per-advertiser range, not against the quoted average.
Produce the split before anyone asks
The one change: pull last quarter’s spend by objective and audience type, and turn it into three percentages.
- Export spend by campaign for the quarter, tag each campaign as creating demand, retargeting, clicks or other, and total the four. An afternoon of work, and it survives every planning conversation for the next year.
- Put the market range in the same table so the numbers arrive with context instead of an opinion.
- Flag anything in the top quarter of the market with a one-line reason. The click bucket is where that sentence is usually hardest to write, which is the useful part.
- Take the reallocation argument to finding 01: cold and retargeting priced at $187 and $196, close enough to call a tie rather than a ranking. The gap worth acting on is format, where document ads made a cold lead for $148 against $358 for video.