In short
Established companies typically spend 5–12% of revenue on marketing; growth-stage startups often spend 15–30% and venture-funded ones more. Before revenue, percentages are meaningless: budget instead from what a customer is worth over their lifetime, what it costs to acquire one through each channel, and how many you need to hit the next milestone. Early money should go mostly on learning which channel works, in small tests, before any channel gets a large commitment.
The benchmarks, and their limits
| Stage | Typical marketing spend | Notes |
|---|---|---|
| Established business, steady growth | 5–12% of revenue | Higher for consumer brands, lower for B2B with sales teams |
| Growth-stage startup | 15–30% of revenue | Buying growth ahead of revenue |
| Venture-funded, land-grab phase | 30–50%+ of revenue | Only sensible with a clear payback model and capital to burn |
| Pre-revenue | Not a percentage | Budget from customer economics; see below |
These figures appear in most surveys and they describe what companies do, not what any particular company should do. The percentage rule fails at the point most people ask the question: the company has little or no revenue, and any percentage of nothing is nothing.
The method that works before revenue
Three numbers, all estimable with a spreadsheet and some honesty:
What a customer is worth. Lifetime value, or a conservative proxy: average revenue per customer per month, times gross margin, times how many months they stay. If you do not know how long they stay, assume a year and revise. For a $50-a-month subscription at 80% margin lasting twelve months, that is $480.
What one costs to acquire, by channel. You do not know this yet, which is the point of the first budget. Early estimates: search ads in your category (look at the auction prices), a content programme (cost of producing pages divided by the leads they might produce), outbound (hours divided by meetings), partnerships. Guesses at first; measured after the first tests.
How many you need. To reach the next milestone: break-even, the next funding round, the point where referrals sustain growth. If you need 200 customers in a year and the blended acquisition cost is $200, the budget is $40,000, whatever percentage that is of anything.
The constraint that makes this work: acquisition cost must be comfortably below lifetime value, ideally a third or less, with payback inside a year. If the numbers do not clear that bar, more budget does not fix the business; it accelerates the loss.
How to spend the first budget
The first year's marketing money is mostly for learning, and it should be spent in a shape that produces learning.
Small tests across several channels. A few thousand dollars each on two or three plausible channels, run long enough to measure cost per customer, not cost per click. Kill the ones that do not work; concentrate on the one that does. Most startups find that one channel produces most of the growth, and the sooner it is identified, the less is wasted.
Foundations that compound. A fast, clear website with pricing on it, a Google Business Profile if local, analytics that attribute customers to sources, and a handful of pages answering the questions buyers ask. These are cheap, done once, and make every other channel work better. Our view on what analytics to track is the setup that makes the tests measurable.
Founder time, counted. In the first year the most effective marketing is often the founder talking to customers, writing, and showing up where buyers are. That is real cost, and it should be in the budget as hours, so its return can be compared with paid channels.
The split at different stages
A rough guide once a working channel is known:
- Finding the channel: 60% tests, 30% foundations, 10% brand.
- Scaling the channel: 70% the channel that works, 20% the next candidate, 10% foundations and brand.
- Established: balanced between the proven channel, content and search that compound, and brand.
Brand spending, meaning awareness with no direct response, is the last thing to fund and the first thing agencies propose. It works for companies that already have distribution. It is a bonfire for companies that do not.
Mistakes that waste the first year
Spending on everything at once. Five channels at $500 each produce five inconclusive results. Two channels at $2,500 each produce a decision.
Measuring the wrong thing. Impressions, followers, clicks. The only number that justifies marketing spend is customers, and their cost and value. Anything else is a proxy that may or may not connect.
Hiring an agency before knowing the channel. Agencies are good at executing a known channel at scale. Paying one to discover the channel is expensive discovery.
Buying the website twice. A cheap site that has to be rebuilt in a year, or an expensive one built before the positioning was known. A modest site done well that can grow is the right first purchase.
Scaling before payback is proven. A channel that looks good at $2,000 a month often does not at $20,000, because the cheap customers were the first ones found. Increase in steps and watch acquisition cost at each.
A worked example
A B2B software startup with a $200-a-month product, 85% margin, and an estimated 18-month customer life values a customer at about $3,000. It needs 100 customers in the next twelve months. It tests search ads and a content programme at $3,000 a month each for three months, finds search produces customers at $900 and content at $1,500 but rising in volume, and concentrates $6,000 a month on search with $2,000 continuing on content. Annual budget around $100,000, or roughly a third of first-year revenue, arrived at without ever consulting a percentage.
If you would like help building the customer-economics model for a specific business, book a call; it is usually a spreadsheet and an hour.
Common questions
What percentage of revenue should a startup spend on marketing?
Growth-stage startups typically spend 15–30% of revenue, established companies 5–12%. Before revenue the percentage is meaningless; budget instead from what a customer is worth, what one costs to acquire through each channel, and how many you need to reach the next milestone, keeping acquisition cost well below lifetime value.
How much should a pre-revenue startup spend on marketing?
Enough to test two or three plausible channels long enough to measure cost per customer, typically a few thousand dollars each over two or three months, plus the cheap foundations: a clear fast website with pricing, analytics that attribute customers to sources, and pages answering buyers' questions. Concentrate spend only once a channel has shown a payback.
What is a good customer acquisition cost for a startup?
One that is comfortably below the customer's lifetime value, ideally a third or less, and that pays back within about a year. The absolute figure varies enormously: $20 for a consumer app, $2,000 for a B2B subscription. The ratio and the payback period are what matter, and they are the test any channel must pass before it gets more budget.
Should a startup hire a marketing agency?
Usually not until a working channel has been identified. Agencies excel at scaling a known channel; paying one to discover which channel works is slow and expensive discovery. Early on, founder time, small paid tests and a few well-made pages find the channel faster; an agency or a hire then scales it.
Is content marketing or paid advertising better for a startup?
Paid advertising gives fast, measurable results and stops when spending stops; content and search take months and then compound. Most startups need paid first to learn which messages and terms convert, then content targeting exactly those, so that acquisition cost falls over time. With very limited money, paid to find the channel, then content to make it cheaper.
