Most advice about getting your first customers assumes you already have an audience. You don't. That is the whole problem.
What follows is the version that assumes nothing: no list, no following, no budget. It is split by stage, because the tactic that gets you customer number three actively stops working by customer number fifty.
Stage 1: your first five, recruited by hand
Your first customers are not acquired. They are recruited, one conversation at a time, and it will feel nothing like marketing.
Get specific about who. "Small businesses" is not a customer. "B2B SaaS founders with 3–10 employees losing hours a week to manual CRM cleanup" is. The test: could you name five real people who fit? If not, the definition is still too broad to act on.
Find where they already gather. Slack communities, subreddits, niche forums, industry newsletters, local meetups. Join, read for a week, answer questions before you ever mention what you built. People can tell the difference between someone who showed up to help and someone who showed up to sell.
Then reach out individually. Not a template. Reference something specific — a post they wrote, a problem they mentioned publicly, a change at their company. Lead with their problem, not your solution:
I saw your post about spending a day a week reconciling invoices. I built
something that does that part automatically. Worth ten minutes?
Expect a 10–30% reply rate on genuinely personalised outreach, and near zero on anything that reads like a template. Fifty good messages beat five hundred generic ones, and the fifty teach you something.
Stage 2: five to fifty, finding one repeatable channel
Once five people are using it and at least two would be annoyed if you took it away, stop recruiting one at a time and find the one channel that works.
The mistake here is trying six channels at once and learning nothing from any of them. Pick one — a community you are already trusted in, a content angle only you can write, a directory where your buyers browse — and give it a month of real effort before judging it.
This is also where launch platforms earn their place. Not as a lottery ticket, but as steady, compounding distribution. That's why Lifto has no launch day and no upvote race: a demo posted today keeps resurfacing to new viewers for weeks, which is the opposite of the twenty-four-hour spike that leaves you back at zero on Thursday.
Stage 3: fifty to a thousand, and what changes
Everything above is manual by design. Past roughly fifty customers, manual stops scaling and the job changes from recruiting to removing friction.
What matters now: how people find you without you (search, referrals, integrations), how quickly a new signup reaches the moment the product becomes useful, and why the ones who leave leave. The first thousand customers rarely come from one channel — they come from two or three that each work modestly and compound.
Realistically this takes six to eighteen months. Anyone promising faster is selling something.
What to measure, and what to ignore
Ignore signups. Signups are the easiest number to move and the least informative.
Watch instead:
- Activation — what fraction reach the point where the product does its job
- Week-four retention — of people who activated, how many are still there
- Qualitative churn reasons — ask every single person who leaves, and read
the answers rather than counting them
At this stage, ten users you have spoken to are worth more than a thousand signups you have not.
The mistakes that cost the most
Building more features because nobody signed up. If nobody uses what you have, more of it will not help. The problem is almost never scope.
Treating silence as rejection. No reply usually means your message did not land, not that the product is wrong. Change the message before you change the product.
Waiting until it's ready. The first version is meant to be embarrassing. The alternative is finding out you built the wrong thing after six months instead of six days.
Going free to get numbers. Free users tell you what people will accept, not what they will pay for. See below.
Common questions
How do I get my first customer with no audience at all? Direct outreach to people you can name. Not a launch, not content, not ads — twenty individual conversations with people who have the problem.
Should I offer it free at first? Free is a fine way to get feedback and a poor way to validate a business. If you go free, do it for a defined window with a small group, and ask for payment at the end. What people say they would pay and what they pay are different numbers, and only one of them counts.
How many people do I need to contact to get ten customers? At a 20% reply rate and a 20% close rate on those replies, roughly 250 genuinely personalised messages. If your numbers are far worse, the problem is the targeting or the message, not the volume.
What if nobody replies? Your message is about you. Rewrite it to be about a specific problem they have already said out loud. If ten rewritten messages also get nothing, your audience definition is wrong.
What if they say it's interesting but don't sign up? "Interesting" means you have described a real problem they do not feel urgently enough to act on. Ask what would have to be true for them to switch this month. The answer is your roadmap.
Should I launch on Product Hunt before I have customers? No. A launch amplifies whatever is already working. With nothing working, it amplifies nothing, and you have spent your one launch. Get to ten paying customers first.
How long should this take? First customer: days to weeks. First ten: one to three months. First hundred: six to twelve. Slower than that usually means the audience definition, not effort.
What if my product is too niche? Niche is an advantage at this stage. A small, precisely defined audience is easier to find, easier to write for, and more likely to care. Broad is what kills early products.
Where to go from here
If you have something to show, put a demo where founders are already looking: publish on Lifto — free, no launch day, and you keep the backlink. If you are still deciding where to launch, we compared 22 places to launch besides Product Hunt.