The question sounds like it should have a number for an answer. It doesn’t. It has a method for an answer, and the method matters more than any figure a benchmark chart will hand you.
Percent-of-revenue benchmarks are astrology for budgets. They answer “what do companies like mine spend?” when the only question that matters is “what does my growth goal cost?” Those are different questions, and only one of them survives contact with a CFO. A marketing budget built from your own unit economics — your revenue target, your pipeline math, your demonstrated cost of pipeline — defends itself in the board meeting. A percentage cribbed from someone else’s spreadsheet cannot.
This post walks through the honest method. It’s arithmetic, not alchemy. And it starts by explaining why the number you’ve probably been quoted is the wrong starting point.
Why Are Percent-of-Revenue Benchmarks the Wrong Starting Point?
Because they import other companies’ goals, margins, business models, and marketing maturity into your budget. A benchmark average blends a bootstrapped services firm growing at a comfortable pace with a venture-backed software company burning cash to take share. Neither of those companies is yours, and their blended spend tells you nothing about what your specific growth goal costs.
There’s a second problem: the ranges are wide enough to justify anything. When the low end of a benchmark and the high end differ by a multiple, the “data” becomes a Rorschach test. The CFO sees the low end. The marketing vendor sees the high end. Both cite the same chart. That’s not evidence; that’s a negotiating position wearing a lab coat.
And consider the sources. Most percent-of-revenue benchmarks are published by parties selling marketing — agencies, software platforms, media companies. Every vendor’s benchmark conveniently supports a bigger number. That doesn’t make them liars. It makes them poorly positioned to tell you what you should spend, because their answer to “how much to spend on marketing” was decided before you asked.
Here’s the fair use of a benchmark: sanity-checking a finished budget. If you build your budget from the ground up and it lands wildly outside what any company your size spends, that’s worth a hard look at your assumptions. A benchmark can catch an error. It cannot produce a budget. Using it as the starting point is outsourcing the most consequential resourcing decision of your year to an average of strangers.
What Is the Goals-Backward Method?
Start with the revenue you want and work backward through your own conversion math until you arrive at a cost. Five steps, in order, each one an equation your finance team can audit.
Step one: set the revenue goal for the year. Not an aspiration — the number the board expects, the number your plan is built on. Everything downstream derives from this, so if the goal is soft, the budget will be too.
Step two: isolate the new revenue marketing must produce. Your revenue goal isn’t all new business. Subtract what retention and expansion will deliver from the existing base. What remains is the gap that new pipeline has to fill. This step alone kills most benchmark logic, because two companies with identical revenue can have wildly different gaps depending on churn and expansion — and therefore need wildly different marketing budgets.
Step three: convert the new-revenue gap into pipeline. Divide the new revenue needed by your close rate on qualified pipeline. If you close one in four qualified opportunities, you need four dollars of pipeline for every dollar of new revenue. Your close rate — not an industry average — is the divisor. This is where “how much pipeline do we actually need?” stops being a debate and becomes division.
Step four: convert pipeline into opportunities and leads. Divide required pipeline by your average deal size to get the number of qualified opportunities you need. Then walk backward through your funnel conversion rates — opportunities from qualified leads, qualified leads from raw inquiries — to size the top of the funnel. Each stage is a ratio you either know or need to establish.
Step five: price it. Multiply the opportunities you need by your demonstrated cost-per-opportunity — what it has actually cost your business, historically, to put a qualified opportunity in front of sales. Not what a vendor says it should cost. What it has cost you. That product is the core of your marketing budget.
Notice what this method does. It replaces “what do companies like mine spend?” with a chain of your own numbers: goal, gap, close rate, deal size, conversion rates, cost-per-opportunity. Every link is auditable. Every link is yours. When the board asks why the budget is the size it is, you don’t point at a chart. You point at the arithmetic.
One honest caveat: the method assumes your demonstrated costs hold as you scale spend. They rarely hold perfectly. Cheap channels saturate; new channels start expensive and improve. Build the budget on demonstrated costs, then treat the assumption as something the first quarter will test, not something you’ve proven. More on that below, because that honesty is exactly what makes the budget defensible.
What If You Don’t Know Your Own Numbers?
Then that is the first budget line. Fix the measurement before scaling the spend.
A company that can’t compute its cost-per-opportunity has no business writing a large check for growth, because it can’t tell working spend from wasted spend. The goals-backward method requires close rate, deal size, funnel conversion rates, and cost-per-opportunity. If two or more of those are guesses, your first quarter’s budget should fund instrumentation and a baseline, not scale.
What that looks like in practice: CRM (customer relationship management) hygiene so opportunities have sources and stages that mean something — and if your CRM isn’t capturing every lead, everything downstream of it is already unreliable. Attribution that’s honest about its limits rather than precise about fiction. A defined funnel with stage definitions sales and marketing both accept. Then a baseline quarter — spend at a modest, deliberate level across your most plausible channels for the express purpose of producing the numbers the real budget will be built on.
This feels slow to a founder under growth pressure. It isn’t. The alternative is spending a full year’s budget at unknown efficiency, which is slower — you just don’t find out until the year is gone. A baseline quarter costs you one quarter. Flying blind costs you the year and, often, the board’s confidence in marketing altogether. If you’ve already cycled through hires or agencies that couldn’t show their math, you know what that erosion feels like. The measurement line item is how you stop the cycle.
What Belongs in the Budget Beyond Media?
More than most budgets include, which is why most budgets stall. Media-only budgets fail predictably: the ads run, and nothing around the ads was funded, so nothing compounds.
Build the full budget holistically. The commonly forgotten lines:
- Content production. Media buys attention; content converts it. Landing pages, case studies, sales enablement, the assets that make a click become an opportunity. Unfunded content is the most common reason paid media underperforms its potential.
- Tooling. The CRM, the automation platform, the analytics stack. Not glamorous, but the pipeline math above is impossible without it.
- People and agency capacity. Someone has to deploy the media, write the content, read the data, and adjust. Budgets routinely fund the spend and forget the operators. Capacity to absorb the budget is part of the budget.
- Measurement infrastructure. Attribution, dashboards, reporting cadence. This is what lets you present the chain to the board next year with better numbers than this year.
- A testing reserve. A deliberate allocation for channels and messages you haven’t proven yet. Demonstrated cost-per-opportunity tells you what worked historically; the testing reserve is how you find what works next, before your current channels saturate.
The proportion between media and everything else varies by business, and I won’t pretend there’s a universal split. The principle doesn’t vary: a dollar of media without the surrounding system is a dollar working at partial strength. When you compose the budget, fund the whole engine, not just the fuel.
How Do You Defend the Budget to a Board or CFO?
Present the chain, in order: goal, pipeline math, demonstrated costs, budget. Every line traces to a number the CFO can audit, which is precisely what a percent-of-revenue figure can never offer.
The presentation writes itself once the method is done. Here is the revenue goal we all agreed to. Here is the new revenue marketing must produce after retention and expansion. Here is the pipeline that requires at our close rate, the opportunities that requires at our deal size, and the cost at our demonstrated cost-per-opportunity. Here is the budget, including the content, tooling, capacity, measurement, and testing lines that make the media work. The budget stops being a request and becomes a derivation. A CFO can argue with an assumption in the chain — good, that’s a productive argument. A CFO cannot argue with the structure, because the structure is their own discipline applied to marketing. For a deeper walkthrough of structuring that conversation, see how to build a board-ready marketing report.
Then do the thing most budget presentations skip: include the honest sensitivity analysis. Say plainly which assumptions are strong and which are thin. Your close rate over the last two years is probably strong. Your cost-per-opportunity in a channel you’ve barely tested is thin. Your assumption that costs hold as spend scales is thin by definition. Name each one, and name what the first quarter will verify. “By the end of the first quarter we’ll know whether cost-per-opportunity holds at the new spend level, and here’s the decision we’ll make if it doesn’t.”
This is counterintuitive to founders who feel they need to project certainty to win the budget. The opposite is true with a good CFO. Certainty about the unknowable reads as either naivety or salesmanship, and boards have seen plenty of both. A budget that says “these numbers are demonstrated, these are estimates, and here is how we’ll convert estimates into demonstrated numbers” reads as decision ownership. It also builds the feedback loop that makes next year’s budget conversation shorter: you’ll walk in with a year of verified assumptions instead of a fresh set of guesses.
One more benefit worth naming. A budget built this way survives a bad quarter. When a percent-of-revenue budget underperforms, there’s nothing to diagnose — the number was arbitrary, so the miss is unexplainable, and the usual response is to cut. When a goals-backward budget underperforms, you can locate the broken link. Close rate held but cost-per-opportunity spiked in one channel; here’s the adjustment. The budget becomes a living model of the business rather than a line item to defend.
The Budget Is a Byproduct
Work through the method and something becomes clear: the budget was never really the question. The question is whether you understand your own revenue engine well enough to price a goal. Companies that do treat the marketing budget as an output of arithmetic. Companies that don’t reach for benchmarks, because a benchmark is what you cite when you can’t show your work.
Getting to that arithmetic is the work — mapping the funnel, establishing the baselines, pressure-testing the assumptions, and composing a budget where every line answers to pipeline. It benefits from an outside marketing operator who can build the model and help you defend it to the people who fund it. If next year’s budget conversation is coming and you want to walk in with a derivation instead of a percentage, that’s a conversation worth having before the planning cycle closes.