B2B marketing budget allocation: what the split actually looks like at $2M, $10M, and $50M ARR

Most board decks put the marketing budget on one line with the same footnote: spend as a percentage of revenue, benchmarked against a peer group. It is the single worst way to set a demand gen budget, and it survives because it is easy to defend in a room where nobody wants to argue about pipeline math.

The percentage rule assumes the budget is doing the same job at every stage. It is not. At $2M ARR the budget is buying a company's first real pipeline from close to nothing. At $10M it is buying efficient capture of demand that now actually exists. At $50M it is buying share of a market that competitors are starting to commoditize. Three different jobs, three different splits, and a fourth number sitting on top pretending they are the same problem.

Why percentage of revenue breaks first at the extremes

A five to ten percent of revenue rule sounds sensible until you run it at both ends. Five percent of $2M is $100,000. That does not cover a single senior demand gen hire, a paid media budget with any statistical significance, and a content function, in the same year. Companies that follow the percentage rule at this stage are underfunded by definition, then blame the channel mix when nothing works.

At the other end, five percent of $50M is $2.5M, which sounds generous until you notice it is funding brand, multi-channel demand creation, a much larger paid capture footprint, and a stack of tooling that a $10M company does not carry, all from one number that has not moved proportionally to the job it now has to do. The rule does not fail because the percentage is wrong. It fails because one percentage cannot describe three different jobs.

The four buckets every demand gen budget actually has

Underneath the label 'marketing budget' are four distinct jobs, and the mistake most finance models make is tracking spend by channel instead of by which of these four jobs the spend is doing.

• Demand capture: paid search, retargeting, comparison pages, review site placements. Converts buyers who are already looking. Fast, measurable, and finite, because it can only capture demand that exists.
• Demand creation: outbound, LinkedIn posts from the executive team, original research, category education content, founder-led content. Builds the pipeline that will show up as capture demand in two to four quarters.
• Brand: sponsorships, events at scale, unmeasured awareness spend, anything whose payoff is a lower cost of capture next year rather than a pipeline number this quarter.
• Tooling: the CRM, the ad platforms, the attribution layer, the enrichment data. Invisible in a pipeline review and impossible to run the other three buckets without.

The $2M ARR budget

What the money is actually buying

At $2M ARR, the company usually does not have enough existing search demand to make capture spend efficient on its own, and it rarely has the story or the case studies to make brand spend worth anything yet. Nearly everything has to go toward creation, because creation is what builds the first wave of pipeline from a market that has not heard of the company.

The $10M ARR budget

The pivot point

Somewhere between $5M and $10M ARR, enough category awareness exists that capture spend starts converting efficiently, and this is the point where the split should move, not stay fixed. Companies that keep the $2M weighting past this point are leaving cheap, high-intent pipeline on the table. Companies that flip too early, before enough demand exists to capture, starve the creation motion that got them here.

The $50M ARR budget

Where brand stops being optional

At this stage, capture channels are usually close to saturated. The company is bidding against itself in paid search, LinkedIn frequency has hit a ceiling in the ICP, and the cost per opportunity in capture channels has been climbing for several quarters. The lever that still has room is brand, because a company with real market share and no distinct position is vulnerable to a smaller competitor with a sharper story and a fraction of the budget.

The split across all three stages

What to cut first in a downturn

The instinct in a downturn is to cut evenly across every line, five or ten percent off everything, because it feels fair and avoids a hard conversation about priorities. It is the worst way to cut a budget, because it protects the weakest spend at the same rate as the strongest.

1. 1

   Cut unmeasured brand spend first

   Anything without a stated proxy metric and a review date is the easiest cut to make and the hardest to argue against, because nobody can point to a number it is currently producing.

2. 2

   Cut the channel with no stated expectation, wherever it sits

   Not every channel with a soft story is brand spend. Some are capture or creation lines that have quietly run for years without ever being asked to justify themselves. Find them by the same test: no number, no date, no defense.

3. 3

   Protect anything with a proven payback under six months

   This spend is close to self-funding. Cutting it does not save the company money in any real sense, it just defers the pipeline it would have produced into a quarter where the target has not gotten any smaller.

4. 4

   Hold tooling spend that the whole function depends on

   Cutting the CRM or the attribution layer to hit a budget number looks like a saving on a spreadsheet and costs the team the ability to make any of the other three decisions correctly for the next year.

The failure mode: funding channels instead of outcomes

A bloated budget almost never gets that way through one bad decision. It got that way because the budget is built by channel line item, carried forward from last year with a growth percentage bolted on, and nobody re-litigates a line that already exists. A channel budget survives by default. An outcome budget has to be re-earned.

The fix is not a dramatic rebuild every quarter. It is a standing rule: every line item states the pipeline outcome it is expected to produce and the date by which that outcome should be visible, before the money goes out the door. A budget built this way naturally reallocates, because a line with no stated outcome has nowhere to hide when the review happens.

A channel keeps its budget by existing. An outcome has to earn its budget every quarter. Build the plan around the second one.

The strongest case for the percentage rule

What I would do Monday

Re-sort the current budget into the four buckets before touching a single number. Most teams cannot answer, without doing this exercise, what percentage of spend is actually creation versus capture, because the finance chart of accounts is organized by vendor and channel, not by job. Once the split is visible, compare it honestly to the stage the company is actually at, not the stage the last plan assumed, and fix the biggest gap first.