A SaaS startup closes a seed round and someone on the board asks what the marketing budget is. The founder searches for a benchmark, finds a percentage of revenue built for companies that have revenue, and either copies it or picks a round number that feels responsible. Neither tells the team what the money is for.
A seed budget is mostly time, not media.
At seed, the largest marketing cost is rarely ad spend. It is the time of whoever does the work: the founder's hours, a first hire, or an outside team on a monthly plan. Media, tools and production sit on top of that and are usually the smaller lines.
That changes what a budget conversation should cover. The question is less how much to spend on ads and more who is doing marketing, how much of their week it is and what they are expected to produce. A budget that funds a channel nobody has time to run is money set aside for a disappointment.
How much a startup should spend on marketing covers the general case. SaaS adds a few specifics worth planning around.
Name the experiment before the number.
A seed SaaS company does not yet know which channel will bring buyers at a cost it can carry. The budget's job in the first months is to find out. So start with the experiment, not the amount.
An experiment has a channel, a buyer, a message, a time box and a result that would count as an answer. Running search ads on a few high-intent queries for eight weeks, to see whether the trials they bring convert at a rate the business can live with, is an experiment. Doing some paid is not. Once the experiment is written down, the budget is whatever it costs to run it properly, plus the time to read the results.
Write the experiment on one page and share it with whoever approves the budget. When the time box ends, the same page says what was learned, and the next experiment is written beneath it. After a few quarters, that page is the most useful marketing document the company has.
Fund one or two channels, not five.
SaaS founders are offered a long menu: paid search, paid social, content, SEO, partnerships, product-led loops, outbound. A seed budget spread across all of them buys a thin test of each and a clear answer from none. Marketing channels a startup should try, in order sets out the sequence, and the budget should follow it.
One or two channels, funded long enough to produce a readable result, is what a seed budget can support. The others wait. Waiting is not a failure of ambition. It is how the team ends the quarter knowing something.
Budget for the trial, not just the traffic.
In SaaS, marketing does not end at the sign-up. A trial or free plan that loses people in the first session wastes the spend that brought them there. Part of the budget, usually time rather than cash, belongs to onboarding emails, the first-session experience and the follow-up when a trial ends. How a startup turns free trials into customers goes into that.
The same applies to measurement. If nobody can say which channel a paying customer came from, the experiment cannot produce an answer. A small amount of setup, one form field asking how they heard about you and consistent tagging on links, pays for itself in the first month of reading results.
Review it monthly, against one number.
A seed budget should be reviewed every month against the number marketing is accountable to, usually qualified trials or demos first and paying customers after. The number a startup's marketing is accountable to covers how to pick it.
If the experiment is working, move more into it. If it has run its time box without an answer, stop it and fund the next one. If nobody can tell either way, the budget's real problem is measurement, and that gets fixed before anything else is spent.
The board question has a better answer than a percentage. It is a sentence: this is the experiment, this is who runs it, this is what it costs and this is when we will know.
Write the experiment first, fund one or two channels to run it properly and review the number every month.
Tell us what you're building.