The short answer
Your first ten customers almost always come from the founder doing direct, unscalable work: asking people you already know, showing up in communities where the problem is discussed, sending short personal messages to people who clearly have the problem, and onboarding each customer by hand. Ads and content rarely work at this stage because you do not yet know what message converts.
On this page (8)
Why the first ten are different
The first ten customers are not a growth problem. They are a learning problem with revenue attached. You are trying to find out who buys, why they buy, what they say when they describe the problem, and what nearly stops them. Every one of those answers feeds into your positioning, pricing and product later.
That is why founder-led sales matters here. If you hand early selling to an agency, an ad platform or a hire, you lose the conversations that tell you what is actually going on. Paid channels also need a message that already converts, and you do not have one yet. The founder's job right now is to be in every conversation.
It also means the methods that work will not scale, and that is fine. You are not building a machine yet. You are collecting the raw material the machine will eventually be built from.
Start with your warm network
The fastest route to the first few customers is usually people who already trust you: former colleagues, clients from previous work, people in your industry, friends of friends who fit the profile. The point is not to sell to people out of loyalty. Customers who buy as a favour churn as soon as the favour is repaid and give you misleading feedback.
Instead, ask your network for introductions to people who have the problem. A message such as 'I am building something for practice managers who still chase late invoices by phone. Do you know anyone who deals with this?' is easy to answer and does not put the reader on the spot. Keep a simple list of who you asked, who they introduced and what happened.
That list is your first sales pipeline. A spreadsheet with columns for name, source, date of last contact, stage and next step is enough. Review it every few days. Most early deals that are lost are not rejected; they simply drift because nobody followed up, and a simple list makes the drift visible.
Find people where the problem is discussed
The rule for communities is to give before you ask. Answer questions, share what you have learned, and become a recognisable name. When you do mention what you are building, do it where it is welcome and as a response to a stated need, not as a broadcast.
- Industry forums and subreddits — Read for months-old threads where people describe the exact problem, then help in public before you ever mention your product.
- Professional groups and associations — Local chapters, trade bodies and meet-ups often have members who share a workflow and talk to each other.
- Slack and Discord communities — Many niches have one. Follow their rules on self-promotion strictly; many have a dedicated channel for it.
- Comment sections of incumbent tools — Review sites and support forums for the software people currently use are full of specific complaints.
- Events and conferences — Smaller, niche events are often better than big ones because attendees have time to talk.
Direct outreach, done respectfully
Cold outreach works at this stage when it is personal, short and specific. It fails when it is a template sent to a scraped list. Ten carefully chosen people with a message that shows you understand their situation will usually outperform hundreds of generic emails, and they will not damage your reputation in a small market where people talk.
A good first message names why you chose them, describes the problem in their language, and asks for something small, such as fifteen minutes to learn how they handle it today. Do not attach a deck. Do not pitch features. If they reply, the conversation is the product demo. If they do not, one polite follow-up is reasonable; more than that is not.
Respect the law and the norms as well as the person. Rules on unsolicited marketing messages vary by country and by channel, and data protection regimes such as GDPR apply to how you store contact details. Keep lists small, honour every request to stop, and never buy contact data you cannot explain the origin of.
Onboard every early customer by hand
When someone says yes, set them up yourself. Get on a call, share their screen or sit with them, import their data, configure the product and watch where they hesitate. This is the single most valuable research you will do, because you see the gap between how you imagine the product is used and how it actually is.
Write down every question they ask and every point where they get stuck. Those notes become your onboarding flow, your help documentation and your product backlog. Check in a week later, then again a few weeks after that, and ask what they have used and what they have ignored.
| Channel | Speed | Effort per customer | What you learn |
|---|---|---|---|
| Warm introductions | Fast | Low | Whether people who trust you see the value |
| Communities | Slow to start | Medium | The language people use to describe the problem |
| Personal outreach | Medium | High | Which segment responds and why |
| Events | Lumpy | Medium | Objections, in person and unfiltered |
Ask for referrals at the right moment
The best time to ask for a referral is just after a customer has had a clear win: the first report that saved them an afternoon, the first invoice paid on time. Ask specifically: 'Who else do you know who deals with this?' is better than 'Would you recommend us?'. Make it easy by offering to write the introduction for them.
If nobody refers anyone, even satisfied customers, take that as information. Either the problem is not painful enough to talk about, or the people who have it do not know each other, which means you will need a different acquisition channel later.
Your action checklist
0/7 doneSaved in this browser only. Nothing is sent anywhere.
Questions
Should I sell to friends and family?
Only if they genuinely have the problem and would pay a stranger to solve it. Friends who buy as a favour rarely use the product, and their politeness hides the feedback you need. A better use of your personal network is asking for introductions to people who match your customer profile and have no reason to be kind to you.
Is cold email still worth doing for a new startup?
Personal, low-volume cold outreach can work well early on, especially in business markets where buyers are reachable and the problem is specific. Mass templated email is a different thing and tends to get ignored or damage your reputation. Keep messages short, show you understand their situation, ask for something small, and follow the marketing rules that apply where they are.
Should I offer my product free to the first customers?
It is usually better to charge something, even at a discount, because payment shows the problem matters and paying customers give more honest feedback. A short free pilot with a clear end date and an agreed price afterwards can work in business markets. Indefinitely free early users often never convert and can make your demand look stronger than it is.
How long does it take to get the first 10 customers?
It varies widely with the market, price point and how well connected you are. A simple tool for a niche you know well can find ten customers within weeks; an enterprise product with long approval cycles can take much longer. Measure progress by conversations held and lessons learned each week rather than by a target date.
Finished? Mark it so your paths and the library remember.