Marketing Budget for a Tech Startup Without Revenue
What a pre-revenue tech startup should actually spend on marketing: by stage, with the minimum for a real ad test and what outside help costs.
By Suren, Founder & Growth Strategist at Gufy · Published · Last updated

Most advice on this says to spend 10 to 30 percent of your budget on marketing. That rule needs revenue or a fixed budget to take a percentage of, and a pre-revenue startup has neither.
So here is the short version. Before revenue, your marketing budget should follow what you still need to learn, not a percentage of anything.
| Where you are | What you are trying to learn | Realistic monthly cash |
|---|---|---|
| Nobody outside your network has seen it | Does anyone want this, and who? | Close to zero: tools only, roughly US$0 to US$300 |
| You know who it is for and have a message to test | Which message and channel bring in the right people? | Roughly US$1,000 to US$1,500 in ad spend, on one channel |
| You want someone else to run it | The same question, answered faster and with fewer mistakes | The ad spend above, plus roughly US$1,000 to US$2,500 in fees at the small end of the market |
The rest of this article explains each row, what not to spend on, and how to put the budget together.
Why percentage rules do not work before revenue
A percentage tells you how much to spend. It does not tell you what the money is for.
When you have no revenue, every dollar of marketing is buying an answer: who wants this, what makes them act, and what it costs to reach them. Once you see it that way, the question changes from "how much should I spend?" to "what do I need to find out next, and what is the cheapest way to find out?"
Sometimes the cheapest way costs nothing. Sometimes it costs a few thousand dollars. It almost never costs a little bit of everything.
Stage 1: before anyone has shown they want it
Budget: close to zero. Your time is the budget.
If the only people who have seen your product are friends, family and colleagues, you do not have a marketing problem yet. You have a question: does someone who does not know you care about this?
Paid marketing is an expensive way to answer that. These are cheaper:
- Talk to five people who do not know you. Not friends, and not the three people who told you it was a great idea. You are listening for the problem in their words, not pitching.
- Put up one page that states the outcome. What changes for the customer, in one sentence, with one thing to do next. Not a feature list.
- Do the outreach yourself. Direct messages, emails, the communities your customers already use. It does not scale, and it is not supposed to yet.
- Get your first paying users from people you can reach directly. A first group of twenty to thirty paying users tells you more than any survey, and it is what makes later ad spend worth the risk.
What to pay for at this stage: a domain, a simple landing page, and tracking. Google Analytics is free, and Mixpanel and PostHog both have free plans that cover a pre-revenue product. Email tools such as Brevo, MailerLite and HubSpot have free tiers that will carry you to your first few hundred contacts. Check the current limits, because they change, but you should not need a paid plan for any of these yet.
Stage 2: testing a message and a channel
Budget: roughly US$1,000 to US$1,500 a month in ad spend, on one channel, for two to three months.
Once you know who the product is for and have a message you believe in, paid ads become useful. Not because they are cheap, but because they are fast. In a few weeks they can tell you which message brings in the right people.
That only works if you spend enough to learn something. This is where most small budgets go wrong.
Take a sample plan that puts $200 a month into social ads. If a click costs you $2 and one visitor in twenty signs up, $200 buys you 100 clicks and five sign-ups. Five sign-ups cannot tell you which of two messages is better. You have spent the money and learned nothing.
The same arithmetic at $1,000 to $1,500 gives you enough sign-ups to compare two or three messages within a few weeks. Your own numbers will differ, which is the point of the test, but the principle holds: a budget too small to produce a clear answer is the most expensive budget of all.
How to run the test so the money teaches you something:
- One channel. Pick the one where your customers already spend time. Splitting a small budget across Meta, Google and TikTok gives you three unclear answers.
- Two or three messages, one audience. You are testing what to say, not everything at once.
- A landing page that matches the ad. If the ad promises one thing and the page talks about another, people leave, and you will blame the ad.
- Measure customers, not clicks. Cost per click is easy to report and tells you little. Cost per activated user or per paying customer is the number that matters.
- Fix the amount and the end date before you start. Decide in advance what result would earn more budget and what would make you stop.
One warning. Paid ads do not fix a product people do not come back to. They amplify whatever is already true. If users sign up and disappear, spend the money on finding out why before you buy more sign-ups.
Stage 3: paying someone to run it
Budget: the ad spend above, plus fees.
Ad spend and fees are separate. The ad spend goes to the platform. The fee pays for the person or team who decides what to test, builds it, and tells you what happened.
At the small end of the market, a specialist or small agency working with early-stage startups will typically charge somewhere around US$1,000 to US$2,500 a month. Larger agencies commonly start well above that, and many expect an ad budget big enough to justify their fee. So a realistic all-in figure for a pre-revenue startup that wants help is roughly US$2,000 to US$4,000 a month, for a fixed period.
For that you should expect more than someone pressing buttons in an ads account:
- A clear view of who you are targeting and what you are testing, written down
- Tracking set up properly, so you can see what happens after the click
- The ads, and the landing page they point to
- Reporting on sign-ups, activation and customers, not just impressions and clicks
- A straight recommendation at the end: scale it, change it or stop
If you are not sure you need outside help yet, read Why Hire a Digital Marketing Agency? first. If you are comparing providers, How to Choose the Right B2B Digital Marketing Agency covers what to ask.
What not to spend on before revenue
A sensible budget is as much about what you leave out.
- A little bit of everything. $200 here and $50 there across five channels feels thorough and produces nothing you can act on.
- PR retainers and brand campaigns. Awareness is worth paying for once people who arrive can become customers.
- Annual software contracts. Stay on free tiers and monthly plans until a tool has earned its place.
- Influencer deals before your message is proven. You would be paying someone else to test your positioning in public.
- Building features "to help marketing". If the message is not landing, another feature rarely fixes it.
How to build your budget in four steps
- Start with what you can afford to lose. Look at your runway and set aside an amount that will not hurt you if the test fails. That is your ceiling.
- Name the one question. "Will operations managers at small logistics firms sign up for a free trial from a LinkedIn ad?" is a question. "Get the word out" is not.
- Set a fixed test. One channel, a fixed amount, a fixed end date.
- Decide what earns the next dollar. Write down, before you start, what result means more budget and what result means stop. Then keep to it.
Three example budgets
| Cost | Validating | Testing it yourself | Testing with help |
|---|---|---|---|
| Tools | US$0 to US$300 | US$0 to US$300 | US$0 to US$300 |
| Ad spend | None | About US$1,000 to US$1,500 | About US$1,000 to US$1,500 |
| Fees | None | None | About US$1,000 to US$2,500 |
| Monthly total | Up to US$300 | About US$1,000 to US$1,800 | About US$2,000 to US$4,000 |
| Run it for | Until strangers say yes | Two to three months | Two to three months |
These are starting points for an early-stage software product, not quotes. Your own numbers depend on your market, your price and how quickly customers decide.
To check whether a channel pays for itself, work through CAC, LTV and payback before raising the budget.
Common questions
How much should a pre-revenue startup spend on marketing?
It depends on what you still need to learn. Before strangers have shown they want the product, spend close to nothing beyond tools. To test a message and a channel properly, plan for roughly US$1,000 to US$1,500 a month in ad spend on one channel for two to three months. If someone else runs it for you, fees are on top of that.
Can you market a startup on $500 a month?
Yes, if the money goes to tools and your own time goes into conversations, outreach and a clear landing page. $500 a month is not enough for a paid ad test that tells you anything, so do not split it across channels and hope.
Should a startup without revenue run ads at all?
Only once you know who the product is for and have a message you want to test. Ads are a fast way to learn which message brings in the right people. They do not fix a product people do not come back to; they make that problem more expensive.
What is a reasonable ads budget for a first test?
Enough to compare two or three messages on one channel within a few weeks. For most early-stage products that means roughly US$1,000 to US$1,500 a month. Fix the amount and the end date before you start, and decide in advance what result would earn more budget.
Where Gufy fits
Gufy works with SaaS and mobile-app startups that have a live or near-launch product; what working with a startup marketing team covers is set out on our services page. We connect the strategy, the tracking, the paid acquisition and the landing pages, then what happens after sign-up: onboarding, activation and lifecycle messaging.
If you want a second opinion on where your next marketing dollar should go, book a growth call. Bring your numbers and what you have tried.
