How to Get Your First 10 Customers for a New Product
Learn how to get your first 10 customers for a new product with a step-by-step playbook: find them, reach out, close, and learn from every early sale.
If you want to know how to get your first 10 customers for a new product, start by dropping the idea that you need a big launch. You need ten people who have a real problem, who hear from you directly, and who agree to pay or to try your product in a serious way. This guide walks through how to find them, how to ask, and what to do once they say yes.
Why your first 10 customers matter more than your next 1,000
Your first ten customers are not a small version of your next thousand. They do a different job. They tell you whether the problem is real, whether your solution works, and whether anyone will pay for it.
Later customers find you through ads, search, or word of mouth. Early customers find you because you went and found them. That means you learn a lot from each one: how they describe the problem, what they tried before, and what made them say yes.
The course Launch, Price and Get Paid in AI Entrepreneur describes its aim as shipping the smallest thing that teaches you something, then winning the first ten customers by hand. If you have not yet cut your product down, read what an MVP is and how small version one should be first.
Define who your first customer is before you look for them
"Small businesses" is not a customer. "Anyone who writes a lot" is not a customer. You cannot message either group, because you do not know where they are.
Write one sentence that names a specific person with a specific problem:
A solo bookkeeper who spends Friday afternoons chasing clients for missing receipts.
A good first customer has four traits:
- They feel the problem often. It is not a once-a-year annoyance.
- They already try to fix it. A spreadsheet, a workaround, or a paid tool counts.
- You can reach them. You know a place where they gather or a way to contact them.
- They can decide. They do not need three levels of approval to try something.
If you are still choosing a problem, the lesson Problems Worth Paying For in the Basic course Business Basics for Builders is about who has the problem, which is where to start.
How to get your first 10 customers: where to find them
Start close to home and go outward. Do not start with strangers if you can start with warm contacts.
- People you already know. Past coworkers, clients, friends of friends. You are not asking them to buy out of kindness. You are asking who they know with this exact problem.
- Places your customer already talks. Forums, group chats, professional groups, and subreddits where people describe the problem in their own words. Read for a week before you post.
- People who just complained. Someone who wrote a public post about the problem last week is a better lead than someone who might have it someday.
- Customers of the tools they use now. If people already pay for a clumsy solution, they have shown they care.
- A simple page that collects interest. A landing page with a waitlist tests demand and gives you a list to write to. AI Entrepreneur's course Your First Builds includes this as Project 1: A Landing Page and a Waitlist.
Keep a plain list. Name, where you found them, what they said, and the date you last wrote. A spreadsheet is enough.
How to reach out without sounding like spam
Spam is a message that could go to anyone. A good first message could only go to this person.
Here is a short example for the bookkeeper (the person and the message are invented):
Hi Dana, I saw your post about chasing clients for receipts every Friday. I am building a small tool that sends those reminders and collects the files in one folder. I am not selling anything yet. Could I ask you three questions about how you handle it today? It should take 15 minutes.
Why this works:
- It names something they actually said.
- It says what you are doing in one sentence.
- It asks for a small, specific favor, not a purchase.
- It admits you are early, which lowers the pressure.
On the call, ask about the past, not the future. "When did this last happen? What did you do? What did it cost you?" People are poor at predicting what they will buy and good at describing what they did. The lesson Talking to Customers Without Fooling Yourself is about talking to customers without fooling yourself.
If someone does not reply, follow up once. Then move on.
Sell before you build too much: offers, pre-sales, and pilots
A kind comment is not a customer. Money, or a clear commitment, is the real signal. Ask for it earlier than feels comfortable.
Three ways to do that:
- A pre-sale. Offer the product at a founding price before it is finished, with a clear date for delivery.
- A paid pilot. Offer to solve the problem for one customer for a fixed period at a fixed price. You learn the job in detail.
- A written commitment. If money is not possible yet, ask for a short note saying they will switch to your product when it does the one thing you described.
Keep the offer narrow. One problem, one outcome, one price. If you build with AI, work out what each user costs you before you set that price. The post on how to price an AI product without losing money per user shows how, and unit economics for first-time founders explains the numbers behind it.
Do things that don't scale to win early customers
With ten customers, you can afford to be personal in ways a big company cannot. Do it on purpose.
- Onboard each person by hand. Get on a call, share your screen, set it up with them.
- Do part of the work yourself. If your AI tool gets 80 percent right, check the output before the customer sees it, and fix the rest.
- Answer fast. A reply within the hour makes a small product feel dependable.
- Write down every request. Patterns across three customers are worth more than one loud opinion.
The third project in Your First Builds, Automate One Boring Workflow, is one place to practice this. You can run a workflow by hand first, then automate only the steps that prove their worth.
If your product uses a language model, check the quality of its answers as you go. The post on LLM evals explains how to measure it instead of judging by eye.
Turn early customers into feedback, referrals, and case studies
Each of the ten should give you more than a payment.
Feedback
Book a short call two weeks after they start. Ask what they use, what they ignore, and what nearly made them quit. Their answers set your next week of work.
Referrals
Ask a happy customer one question: "Who else do you know who deals with this?" Ask for a single introduction, not a list. People are more willing to make one.
Case studies
Only publish what a real customer has agreed to, in words they have approved. Do not invent numbers or quotes, and do not stretch a small result into a big one. A plain, true story about one person's Friday afternoon is more believable than a polished claim.
Once you have ten, look at what they have in common. That is your next target group, and the bigger picture is in the roadmap from side project to AI company in 7 stages.
Frequently asked questions
How long does it take to get your first 10 customers?
It depends on your market, your price, and how many people you contact each week. Set a weekly outreach target, track replies, and adjust your message when the replies are weak.
Should my first customers pay?
Yes, if you can ask. Payment is a stronger signal than praise. If money is not possible yet, ask for a clear written commitment to switch when your product does what you promised.
Do I need a finished product to get customers?
No. You need a clear offer and a way to deliver the result, even if you do parts of the work by hand. Build more only after people show they want it.
What if nobody replies to my messages?
Check that each message names something specific about the person and asks for a small favor. If that still fails, your target group or problem may be wrong, so go back and test a different one.
Get started
The courses behind this post, from finding a real problem to launching and pricing, are short and hands-on. Each lesson takes 8 to 12 minutes.
Every lesson is free. Start learning with AI Entrepreneur: sign in on the home page and take your first short lesson today.