A startup marketing strategy that fits your stage.

Most marketing advice is written for companies that already know who buys from them and why. Early-stage startups don't, yet. That's the whole difference, and it changes which tactics are worth your time.

This guide splits startup marketing into three stages: before product-market fit, early traction, and growth. For each, it covers what to do, what to measure and what to ignore. We'll be honest about where search fits too, since we run startup SEO and know it isn't always the right first move.

Why does a startup need a different marketing strategy?

A mature company markets a known product to a known buyer. A startup is still finding out both. So the first job of marketing isn't reach. It's learning.

That has practical consequences:

• Messaging will change. Anything you produce now may need rewriting in three months.
• Channels are unproven. You don't yet know whether your buyers respond to search, communities, outbound or partners.
• Budget is small and expensive. Every dollar spent on the wrong channel is a dollar of runway gone.
• The founders are the marketing team. Nobody else knows the problem as well, and nobody else can close the first deals.

A strategy that ignores those facts usually ends up as a brand campaign nobody asked for, or a content calendar for a product that pivots before the posts go out.

What should you do before product-market fit?

Talk to people, one at a time. Paul Graham's 2013 essay "Do Things that Don't Scale" puts it plainly: "The most common unscalable thing founders have to do at the start is to recruit users manually." Waiting for people to discover you rarely works this early.

What that looks like in practice:

• Find where your buyers already talk. Industry forums, Slack and Discord groups, subreddits, LinkedIn comment threads, trade association events.
• Offer something useful, not a pitch. A teardown, a spreadsheet, an honest answer to someone's problem.
• Recruit users by hand. Personal emails, calls, demos you run yourself.
• Write down the words they use. How customers describe the problem becomes your headlines, your landing pages and, later, your keyword list.
• Look at search data as research. Free tools like Google Search Console (once you have a site) and keyword tools show what people call the problem and which competitors they mention.

What to skip: paid brand awareness, big website rebuilds, and anything that needs months to pay back.

How do you choose your first marketing channels?

Once a handful of customers stick around and some arrive without a personal invitation, you're into early traction. Now the job is to find one or two channels that can bring more of the same customers.

Pick two at most. Give each a fixed budget, a fixed time box and a single number that decides whether it continues, ideally cost per paying customer rather than cost per signup.

What rules do startup marketers need to follow?

Small companies get the same rules as large ones. Three come up constantly:

• Cold email. In the US, the CAN-SPAM Act requires accurate header information, subject lines that reflect the content, a valid physical postal address, a clear opt-out and opt-outs honoured within 10 business days. Other countries are stricter.
• Creators, affiliates and employees. The FTC's Endorsement Guides, revised in 2023, require clear disclosure of any material connection. A free product counts. "Affiliate link" on its own isn't an adequate disclosure, and employees posting about their employer need to say so in the post, not just on their profile.
• Naming competitors. The FTC encourages comparative advertising that names competitors, provided it's truthful and not deceptive. Keep feature and pricing comparisons dated and current.

None of this is legal advice, but ignoring it is how a cheap growth tactic becomes an expensive problem.

When do SEO and paid ads start to make sense?

When you know what you're selling, to whom, and roughly what a customer is worth.

Paid search works once you can say, with some confidence, what you can afford to pay for a customer. It's fast to test, and it tells you which search terms convert, which is useful for SEO later. Our Google Ads calculator helps you size a test.

SEO works when people search for the problem you solve, the job your product does or the incumbent you replace. It's slow to start and gets cheaper per customer over time. For a young domain, the trick is choosing specific searches it can realistically win, not the category term the market leader owns. That's the core of how we approach SEO for startups.

AI search is worth watching too. Buyers increasingly ask ChatGPT or Google's AI features for tool recommendations. Being cited there depends on many of the same things as SEO: clear pages, credible mentions elsewhere and specific answers. Our explainer on generative engine optimisation covers the basics.

What should you measure at each stage?

The wrong metric can make a failing channel look healthy. Match the number to the stage:

• Before fit: conversations held, demos run, and how many users are still active a few weeks later. Retention tells you more than any traffic chart.
• Early traction: cost per paying customer by channel, and how long a test takes to produce a clear answer.
• Growth: customer acquisition cost against customer value, payback period, and the share of new customers arriving from channels you own, like search and email, rather than rent.

Track signups by source from day one, even if the numbers are tiny. When you come to choose channels, a few months of honest data beats any benchmark you'll find online.

How much should a startup spend on marketing?

Anyone who quotes you a single percentage without knowing your stage, margins and sales cycle is guessing. A more useful way to set the number:

1. Work out what a customer is worth. Average revenue per customer over a sensible period, minus the cost to serve them.
2. Decide what you can pay to acquire one. That's your ceiling per channel.
3. Fund tests, not campaigns. Enough budget to get a clear answer from each channel within your time box.
4. Count founder time. Before fit, most of your marketing spend is hours, not dollars.

For reference, our own published prices are $1,199 a month for SEO and $999 a month for PPC management. Whether either is right for you depends entirely on the numbers above. The marketing budget calculator helps you split a budget across channels.

A worked example: one startup, three stages

Imagine a two-founder company building scheduling software for physiotherapy clinics.

• Before fit: the founders call clinic owners they find through a professional association directory, run demos themselves and notice that owners keep saying "no-shows", not "scheduling". They rewrite the homepage around no-shows.
• Early traction: they test two channels for eight weeks. Outbound email to clinic managers produces demos. A small paid search test on "reduce patient no-shows" produces trials but few paying customers. They keep outbound and pause paid.
• Growth: with paying customers and stable messaging, they publish pages on no-show policies, reminder templates and comparisons with the two booking tools clinics usually leave. Those pages start bringing trials from search after a few months, and the paid test gets rerun with what they've learned.

The specifics will differ for you. The pattern of learn, test narrowly, then scale what works rarely does.

What doesn't work for early-stage startups?

• Being everywhere at once. Five half-run channels teach you nothing.
• Buying email lists. Poor data, deliverability damage and legal exposure.
• Measuring signups instead of customers. Free signups can hide a channel that never converts.
• Brand campaigns before fit. Awareness of a product that's about to change is wasted.
• Copying a big competitor's playbook. They have authority, budget and brand searches you don't.

The short version

Before product-market fit, marketing is mostly founders talking to buyers. During early traction, it's disciplined tests on one or two channels. After fit, it's scaling what worked and building assets, like search rankings, that keep paying back. If you're weighing whether to build that in-house or bring in help, our comparison of outsourced vs in-house marketing is a good next read.

Sources: Paul Graham, "Do Things that Don't Scale" (July 2013); Federal Trade Commission, "CAN-SPAM Act: A Compliance Guide for Business"; Federal Trade Commission, "FTC's Endorsement Guides: What People Are Asking" (Endorsement Guides revised 2023); Federal Trade Commission, "Statement of Policy Regarding Comparative Advertising".