Every founder eventually hits the same wall. The product works, the landing page is live, the pricing page has three tiers, and the customer count is zero. Nothing about building the thing prepared you for the part where a stranger has to hand you money.
The first ten customers are the hardest ten you will ever get, and they are also the only ones that teach you anything durable. They are not a marketing problem. They are a conversation problem, a distribution problem, and a willingness problem, usually in that order.
What follows is the practical version: where the first ten actually come from, what to say when you find them, how to price when you have no leverage, and how to tell the difference between a slow start and a dead idea.
The first ten are a different category of problem
Growth advice is written for companies that already have customers. Optimize the funnel, improve activation, reduce churn, test the headline. All of that assumes traffic exists and a percentage of it converts. At zero customers, there is no percentage. There is only you, a list of specific human beings, and whether any of them say yes.
This is why founders who read a lot of growth content often stall. They build a content calendar, set up an email sequence, and wait for a machine to spin up around them. Machines are efficient at scale and useless at the start, because you cannot optimize a conversion rate calculated from four visitors.
The right mental model early on is closer to sales than marketing. You are not trying to build a system that finds customers. You are trying to find ten people, one at a time, by hand, and learn what makes them buy.
Your first ten customers are not the beginning of your growth engine; they are the research that tells you what engine to build.
Start where you already have permission to speak
Cold outreach to strangers is the slowest available channel and most founders start there because it feels neutral and professional. Warm channels convert far better, and almost everyone has more of them than they think.
Map the network you actually have
Sit down and write out every group of people you can contact without an introduction. Former colleagues. People from your last three jobs. Clients from freelance work. Anyone you have helped for free. Communities where you are a recognized name rather than a new account. Vendors and consultants who serve the customer type you want.
Most founders will find between fifty and three hundred reachable people in an hour of listing. That is more than enough raw material for ten customers, if the product solves something real.
Filter for urgency, not fit
The temptation is to contact the people who best match your ideal customer profile. The better filter early on is who has this problem right now, badly enough to have already tried to solve it. Someone with a spreadsheet workaround, a half-finished internal tool, or a contractor they hired to patch the gap is a live buyer. A perfect-profile company with no active pain is a long, polite conversation that ends nowhere.
Look for evidence of attempted solutions. People who have already spent money or time on a problem have proven they are willing to spend more.
Do the things that do not scale, on purpose
Manual onboarding, custom setup, a founder who personally imports the data, a shared document instead of a dashboard. Founders resist this because it looks like a business that will collapse under growth. It will, and that is fine. You are not trying to build the version that scales. You are trying to earn the information that tells you what to automate.
Concretely, in the first ten customers it is completely reasonable to:
- Do the setup yourself over a screen share instead of building a self-serve flow.
- Perform part of the service manually behind an interface that looks automated.
- Charge by invoice and bank transfer instead of integrating payments.
- Deliver a weekly report you compile by hand rather than a live analytics page.
- Answer support from your personal phone number.
Each of these gives you something a self-serve funnel would have hidden. You find out where people get confused, which step they abandon, what language they use for the problem, and what they secretly wanted the product to do instead. That information is worth more than the engineering time you saved.
The conversation that actually converts
Most early sales conversations fail for the same two reasons. The founder talks about the product instead of the problem, and the founder never asks for a decision.
Lead with the problem, in their words
Open by describing the situation you believe they are in, specifically enough that they recognize themselves. Not “we help teams collaborate better” but “you are probably tracking client revisions across email threads and losing track of which version was approved.” If you are right, they will interrupt you to confirm it. If you are wrong, you have learned that in ninety seconds instead of after a forty-minute demo.
Then shut up and let them describe it themselves. The vocabulary they use is your future landing page copy. The workaround they mention is your future feature list. The person they name as the one who complains most is your future champion.
Ask for the sale explicitly
Founders end calls with “let me know what you think” and then wonder why nobody follows up. That sentence puts all the work on the buyer, and buyers with real problems are also busy people with real jobs.
End instead with a specific, small next step and a date attached: “If this looks right, I can have you set up Thursday morning and the first month is four hundred. Should I send the invoice?” You will get a yes, a no, or an objection. All three are useful. Silence is the only outcome that teaches you nothing, and vague endings produce silence.
Treat objections as specifications
When someone says the price is too high, they usually mean the value is not yet obvious, or the risk of switching is. When they say now is not a good time, they often mean the problem is real but ranked fifth. When they say they need to check with someone, you have learned who the actual buyer is.
Write down every objection verbatim. After ten conversations, the same three will have appeared repeatedly. Those three are the real product roadmap, the real pricing page, and the real onboarding flow.
Where the first ten usually come from
Across most early-stage software and service businesses, the first handful of paying customers tend to arrive through a small number of repeatable paths. Not all of them will work for you, but one of them almost certainly will.
- Direct warm outreach. Personal messages to people who already know you, referencing something specific about their situation.
- Communities you already belong to. Slack groups, forums, and industry associations where you have a track record of being useful before you ever mentioned the product.
- Helping in public. Answering questions thoroughly in the places your customers already ask them, with no pitch attached. Slow to start, compounds reliably.
- Referral from the first customer. Once one person is happy, ask them directly who else has this problem. Do not wait for organic word of mouth.
- Adjacent service providers. Consultants, agencies, and freelancers who serve your customer and have no competing product often make excellent introductions.
- Narrow, specific cold outreach. Not volume email. Twenty carefully researched messages that reference something only a human could have noticed.
Notice what is missing: paid ads, SEO, and social media growth. Those channels can work, but they need budget, time, or an audience you do not yet have. They are the phase-two answer to a phase-one problem.
Pricing when you have no leverage
The instinct is to price low or go free, because you feel you have not earned the right to charge. This backfires in a specific way: free users give polite feedback and no signal. Someone who pays will tell you the truth, because they now have something at stake.
Charge from the first customer, even if the number is small and even if you feel unqualified. If the price feels uncomfortable to say out loud, that is a sign you should practice saying it, not lower it.
Discount the terms, not the value
If you need to make the first ten easier to close, do it with structure rather than a permanently low price. Offer a shorter commitment, a month of setup included, a pilot with a clear exit, or a founding-customer rate locked for a year in exchange for a case study and regular feedback. All of these preserve the perceived value of the product while lowering the buyer’s risk.
What you should avoid is a price so low that the customer does not care whether it works. Indifferent customers are worse than no customers, because they produce data that looks like traction and behaves like noise.
Reading the signal when nobody says yes
At some point you will have had thirty conversations and closed nothing, and you will need to decide whether to persist or change direction. The distinction that matters is between a distribution problem and a demand problem.
You have a distribution problem if people who hear the pitch respond strongly but you cannot find enough of them. That is fixable with effort and channel experimentation. You have a demand problem if the people you reach understand the offer perfectly and still do not want it. No amount of outreach fixes that.
Some practical tests to tell them apart:
- Do prospects ask unprompted questions about implementation and timing, or only about how it works in theory?
- Do they forward the conversation to a colleague without being asked?
- When you follow up, do they respond, even to say no?
- Have any of them asked what it costs before you brought up price?
- Would they be annoyed if the product disappeared tomorrow?
Consistent no across all five, over a decent sample of the right people, is information. It means the problem you chose is real but tolerable, or the solution is not obviously better than the workaround. Neither is a personal failure, and both are cheaper to learn now than after eighteen months of building.
Turning ten into a repeatable channel
Once you have ten paying customers, the work changes. Stop selling for a week and go back through everything you wrote down. You are looking for the pattern that makes the eleventh customer easier than the first.
Ask three questions of each customer in the group. How did they first hear about you? What was happening in their business the week they decided to buy? What almost stopped them? The answers cluster faster than you expect. If seven of ten came from one channel and shared one trigger event, you have found the thing to double down on.
Then write down the sales conversation as a script. Not to read from, but to see clearly what you have been doing intuitively. That document is what lets you hand selling to someone else later, and it is the raw material for the marketing you will eventually build.
Frequently Asked Questions
How long should it take to get ten customers?
There is no universal number, and the timeline varies enormously by price point and market. A low-cost tool sold to individuals might reach ten in a couple of weeks. An enterprise product with a procurement process might take most of a year. The more useful measure is conversation volume: if you are having several real prospect conversations a week and none of them convert over a sustained period, something in the offer needs to change.
Should I build the product before I look for customers?
Build enough to make the problem and the solution concrete, then start conversations immediately. You do not need a finished product to sell, but you do need something specific enough that people can react to it. Selling a vague concept produces vague encouragement, which is the least useful feedback there is.
What if my first customers are all friends or former colleagues?
That is normal and fine, with one caveat: make sure they are paying real money and using the product for real work. A friend who signs up to be supportive is a favor, not a customer. A former colleague who has the problem, pays market rate, and complains when the product breaks is exactly the customer you want.
The part that does not get easier
The uncomfortable truth about early traction is that it is mostly a willingness test. The tactics are not secret. Almost every founder knows they should talk to more customers, ask for the sale, and follow up. The gap between the ones who get to ten and the ones who do not is rarely knowledge. It is the ability to send the message, make the call, and hear no repeatedly without deciding it means something about you.
So the practical advice is boring: keep a list, contact people on it every single working day, write down what they say, and change the offer when the same objection appears three times. Do that for eight weeks and you will either have customers or a very clear reason why not. Both outcomes beat waiting for a channel to work on its own.
