Someone asks you for the marketing number. An investor on a board call, a co-founder building the model, an advisor who wants to see a plan. You go looking for a rule, and what comes back is a share of revenue. At pre-seed, you multiply that share by a revenue line that is mostly zero, get something close to nothing, and know it is wrong. The figure that holds at this stage is not a percentage at all. It is the number of months of runway you can lose without changing the hiring plan.
The percentage rules assume you already have revenue.
Budget as a share of revenue is a control, not a plan. It works at a company that already sells something: the machine runs, the return is roughly known, and the percentage keeps spend in proportion to what it produces.
Take the same rule to a company with no customers, and it returns either nothing, or a number derived from a revenue projection you wrote yourself. The second is worse than the first. It looks like arithmetic, and it is really a mood.
The rule starts to mean something later, once revenue is repeatable enough that a share of it describes a real machine. Borrowed early, it is borrowed authority.
The other common answer fails the same way. Copying the spend of a company one stage ahead sizes your budget to a motion that already works. Yours has to pay for finding out whether one exists.
Before revenue, marketing is paid for in months.
You do not have a marketing budget. You have cash, a burn rate, and a date. Money that leaves for marketing shortens the runway, so the honest unit is not a percentage of anything. It is time.
That changes the question from what is normal to something you can actually answer: how much of the runway this is allowed to cost. Almost no founder can defend a percentage. Every founder can tell you what happens if the raise moves a month later.
The test is the hiring plan.
A good marketing budget at pre-seed is the largest one that leaves the hiring plan untouched.
The hiring plan is the part of an early company that breaks loudest. An engineer you do not hire is a roadmap that slips; a roadmap that slips moves the raise, and a raise that moves is the only real emergency at this stage. Marketing that costs you the next hire is too expensive at any size. Marketing that costs you a few weeks at the far end of the runway is affordable at almost any size, as long as it is buying something.
So work the number out the way you would work out a hire. Take the cash. Subtract the burn already committed: payroll, infrastructure, the people the plan says you are hiring. Set the date you intend to raise and put the months of margin you want in front of it. What is left is the marketing budget. It is a ceiling rather than a target, and it is denominated in months.
Set it for a window, not a year. At this stage, the plan is rewritten at every raise, and a twelve-month budget written before a seed round is a number nobody follows past spring.
A budget has to buy an answer.
A number on its own is a wish. The budget becomes real once you can say what the money is meant to prove by the time it runs out.
The useful questions at this stage are small and awkward: whether you can reach these people at all, and whether anyone outside the founders’ network will pay for what you built. Spend that returns an answer is cheap even when the answer is no. Spend that returns activity is expensive even when it is small, because you have to buy the same answer again later with less runway behind it.
In ten years of this work, we learned it’s not the size of your budget—it’s how you spend it. Large budgets don’t get cut because they cost too much; they get cut for lack of ROI and tangible results, because nobody can articulate the outcomes they paid for.
Decide the number you are accountable to before the money moves, not after it has gone.
Most of a first budget is not media.
Founders assume the budget is ad spend. Early on, it usually is not. The money goes to the things the company cannot yet say clearly: who it is for, what it is called, and what the site says when a stranger arrives. Media buys attention for a message. It does not produce one.
That order matters, because ads are the fastest way to find out you have nothing to say, and the most expensive. Positioning and the site come first, then channels in the order that suits the company.
The same logic sets the ceiling on people. A budget that would cover one full-time marketing hire and nothing else is usually the wrong shape at pre-seed, and when to hire and what kind is a separate decision that sits downstream of this one.
Runway sets the ceiling. The targets set what the money has to return. We would rather size an initial marketing budget against actual pipeline and revenue targets than an arbitrary benchmark. Then, strategically invest in Brand, Positioning and Website before any of it goes to media.
Tell us what you’re building.