Ask the internet how much to spend on marketing and you'll get a confident answer: 7–10% of revenue, or 12–20% if you're growth-stage. These benchmarks aren't useless, but they're the wrong starting point, because they describe what companies spend, not what your growth goal costs. Two companies with identical revenue can have wildly different correct budgets depending on what they're trying to do next year.
Start from the goal, not the percentage
The defensible budget is built backwards from the revenue target. The chain looks like this: how much new revenue do you need next year? At your average deal size, how many new customers is that? At your close rate, how many qualified opportunities? At your qualification rate, how many leads? Now you have a number that marketing must produce, and you can price it, channel by channel, using your actual historical cost per lead where you have it and honest estimates where you don't.
When a budget is built this way, the CFO conversation transforms. You're not asking for "10% of revenue because that's the benchmark." You're saying: "the growth target costs this much lead flow; here's the cost per lead evidence; fund the target or adjust the target." Both are legitimate answers, but now it's one decision, made with clear eyes, instead of a negotiation about a percentage.
The 70/20/10 allocation
However large the number is, how you divide it matters as much. The split we recommend to most growing companies: 70% to proven channels: whatever has already demonstrated it produces leads at an acceptable cost; 20% to promising experiments: one or two new channels with a defined test budget and a kill criterion; 10% to brand infrastructure: the unglamorous compounding assets like your site, your case studies, your email list. The ratio protects you from the two classic failure modes: pouring everything into one channel until it saturates, and chasing every shiny new platform with money that proven channels would have converted.
Where the benchmarks actually help
After you've built the bottom-up number, then check it against the benchmarks. If your goal-derived budget lands at 2% of revenue, your growth target is probably too timid, or you're assuming conversion rates you haven't earned. If it lands at 35%, your model likely has an expensive assumption hiding in it (usually cost per lead or close rate). The benchmark is a sanity check on your math, not a substitute for it.
What to do this quarter
Build the chain: revenue target → customers → opportunities → leads → cost. If you're missing the conversion numbers to complete it, that's finding #1, start measuring them now, because a company that doesn't know its close rate and cost per lead isn't ready to spend efficiently at any budget level. Then take the completed chain to your next leadership meeting and watch the budget conversation get dramatically shorter.