The First Paying Customers: What Founders Did Before They Had an Audience

0:00Before the launch, find someone willing to pay

What should you do when the product is shipped, the landing page is live, and nobody is coming?

The first useful move may not be marketing at all. It may be finding one sharply defined person with one painful problem, then reaching out directly. Rob Walling puts the shape of it simply: “You don’t need an audience to get your first customers. You need (1) a specific person with (2) a specific problem and (3) a direct way to reach them.”

If I were starting from zero, I’d make the first week smaller than feels comfortable. I’d choose one kind of customer I can actually name. Then I’d ask for a conversation, not broadcast a launch announcement. The goal is not to collect compliments. It’s to find out whether this person will spend time, make an introduction, give access to their workflow, or pay for an early version. Those are different signals from a like, a signup, or “keep me posted.”

Brian Casel offers the clearest version of this approach. Before building his product, he sent an idea to about 25 people. Around 15 or 16 scheduled calls. On those calls, he used a short presentation to describe the ideal customer, lay out the pains he thought mattered, gather feedback on wireframes, and explain a possible roadmap. He wasn’t pretending the product was finished. He was testing whether the problem was important enough to support a paid beta. As Casel says, “I created a quick slide presentation, probably lasts about ten minutes or so.” He asked people whether they recognized themselves in the problems he described. Twelve people paid. Casel asked for $200 from a single team or $500 from an agency, in exchange for early access, a lifetime discount, and free usage. In his words: “I sent it to like 25 people or something. Like 15 or 16 scheduled calls, 12 ended up paying.”

That conversion is a result from one pre-sale test, so I wouldn’t copy the percentage. I’d copy the test. Put a price and a concrete promise in front of a narrow group before spending weeks polishing every feature. If people say the pain is real but won’t commit, that’s information. You can change the offer, the customer, or the product while the cost of changing is still low. A simple first experiment is ten to twenty conversations. Show the smallest version that can communicate the value. Ask what they do today, what the problem costs them, and whether they want to become an early customer. Track payments separately from interest. The number to watch is not how many people enjoyed the conversation. It’s how many took a real step.

There are less dramatic versions of the same pattern. James Hawkins and his co-founders got about ten initial users mostly through personal networks. They created accounts for people and onboarded them through Slack, WhatsApp, or in person. Then they expanded through acquaintances, introductions, group-based outreach, and LinkedIn messages, reaching a couple dozen users within a few weeks. That story matters, but with an important distinction. These were users, not necessarily paying customers. The manual work was part of validating the product with those early users. It did not, by itself, prove that strangers would pay.

Uri Haramati’s story shows how the circle can widen. His team first mapped people in their first- and second-degree networks and reached out to them. Those contacts became the first customers. Then they targeted a company outside that circle. Pipedrive became their first paying customer from completely outbound outreach. The approach was specific rather than generic. They used Pipedrive, liked it, and explained why their own product might be relevant. Haramati remembers the message as: “we are using your tool, we love it. And we thought you might want to see Tori.”

So the sequence is not “friends forever.” It’s a bridge from people who will give you a chance to people who have no reason to.

For a solo founder, that suggests a practical order. Start with reachable people who resemble the customer you want. Ask for a short conversation or offer something useful before asking for a meeting. Then make a direct paid offer to the people who describe the problem in concrete terms. If the product serves businesses, look for a narrow set of companies where the problem is visible and your message can mention their actual tool, workflow, or situation. The offer may need founder help. You might create the account, import the first data, or sit with the customer while they try it. That work is not a failure of product-led growth. At this stage, it is how you learn whether the product creates enough value to earn payment.

The stopping rule is just as important as the outreach. Jason Cohen argues that the first couple dozen customers can come through manual founder effort, including LinkedIn, guest posts, podcasts, and personal networking. He treats failing to reach roughly 20 customers after substantial manual effort as a possible warning sign about the product itself. That doesn’t mean twenty is a universal law. It means a founder should not keep expanding the contact list forever while avoiding the harder question. After a meaningful round of conversations, look for payment, a concrete commitment, or the same urgent problem appearing again and again. If all you have is polite interest, pause the promotion and revisit who the product is for or what it promises.

What I take from these early stories is that your first customers are not a crowd waiting to discover you. They are a small group you can reach, help, and ask to make a decision. Before the launch gets louder, make the conversation more specific. Find the person, name the problem, and ask for the commitment that would make the product real. If nobody will pay yet, the next thing to build may not be another feature. It may be a sharper reason to say yes.

6:50Let the first customers change the product

Getting the first customer is only the beginning. The next question is harder: why did this person pay, and can they get the promised result without you standing beside them?

Jonathan Fields learned this the uncomfortable way. His first customers came through relationships, personal effort, and heavy discounts, with some luck involved. He had not initially defined a clear ideal customer or settled on whether to serve enterprise, small businesses, or the middle of the market.

The useful part came after those early sales. Customer feedback helped his team identify and build the product elements customers considered most important. They then used those elements in advertising and a more repeatable model. Fields says that once they understood the critical pieces of the business, they could leverage them into a model that was scalable and repeatable.

That suggests a different job for your first customers. They aren’t just proof that somebody will hand over money. They’re a set of clues about the result people are actually buying. If your product has several possible audiences, don’t ask only, “Who signed up?” Ask, “Who got enough value to pay, return, and describe the value in a way I can repeat?”

Gia Laudi recommends starting with customers who are already happily paying and succeeding with the product. She uses their answers to inform messaging, positioning, onboarding, and product direction. Sometimes the research reveals two or three different customer jobs. Then there’s a decision: keep serving all of them, or lean into one.

The signal isn’t a flattering comment after a demo. It’s behavior. James Hawkins watched what happened after people signed up and helped them through Slack, WhatsApp, or in person. When people kept coming back, that suggested the product was solving something real. When they disappeared, Hawkins treated it as a product problem: "If people sign up but don’t come back over and over," he says, "you’ve probably found a valid problem, but not solved it."

Dean Mathews made a similar distinction with a different kind of product. His site was getting two or three signups a day, but he didn’t count those as customers. He counted people who adopted the product and converted to payment. In his words, "Signups are different than somebody adopting or converting over to a paying customer."

So for the next few weeks, I’d track three separate events: someone starts, someone uses the core feature again, and someone pays. Then I’d ask where the chain breaks. If people pay but don’t return, the promise may be too strong or the product too hard to use. If they return but don’t pay, the free experience may already solve enough of the problem. If they sign up and vanish, the issue may be the audience, the offer, or the first five minutes.

This is where the first customer call becomes diagnostic rather than performative. April Dunford describes the first B2B conversation as discovery plus enough value to earn a second conversation, not an attempt to customize everything on the spot. And Gokul Rajaram offers a simple test: "can your customer accomplish their goal without someone having to train them through it?" Rajaram’s example comes from larger companies, but the question travels well to a solo founder.

After someone buys, ask what was happening right before they looked for a solution. Ask what result would make the purchase worthwhile. Ask what they expected to happen next, and where they got stuck. Watch them try the important workflow. You’re not collecting feature requests. You’re finding the shortest path from the customer’s problem to the outcome they paid for.

The amount of help you provide also tells you what kind of business you’re building. Arvid Kahl points out that five or ten people paying a couple thousand dollars a month can make further acquisition unnecessary for a while in a consulting or freelance-style business. Low-touch SaaS is different. Kahl says acquisition has to be continuous because customers churn and their needs change.

That distinction matters before you copy another founder’s playbook. A high-priced, service-assisted product can use onboarding calls and bespoke work as part of the offer. A low-priced SaaS product eventually has to deliver value with much less founder intervention. Early manual help is useful when it teaches you what to automate. It becomes dangerous when the product only works because you are permanently operating it by hand.

Here’s the practical loop: sell to a small group, watch them use the product, find the moment where value appears or disappears, and change one thing that affects that moment.

Give each change a short test. If customers repeatedly ask for the same outcome, make that outcome clearer in the offer or onboarding. If they need the same explanation every time, put that explanation into the product. If a group pays, returns, and needs less help after each iteration, you may have found your starting segment. And if weeks of personal attention produce signups but no payment or repeat use, don’t respond by shouting louder. Change who you’re serving, what you promise, or what the product actually does.

Once those signals are visible, the question changes. You’re no longer asking where to announce the product. You’re asking which channel can reliably reach more people who look like the customers who stayed.

12:47Choose the next channel only after the first signal

Once a few people are using the product, the next question is where to find more people like them. The answer depends on what the first customers have already taught you.

Tyler Tringas has a useful warning here. The first paying customers for Storemapper came from the e-commerce businesses he had already served as a freelancer. That was a real acquisition channel, but it wasn’t necessarily the channel that could carry the company to a hundred or a thousand customers. As he puts it, “the things that get you your first few customers aren't necessarily what get you to 100.”

His broader model is to stay with a channel while it works, then move when it starts to flatten. “You sort of will find one or two channels that work really well to a point,” he said, and then “you have to ... switch gears.” The practical question is whether the channel is still producing results or has reached diminishing returns.

So I’d separate the first channel from the next channel. Use the first one to learn who buys and why. Choose the next one only after you can describe the signal that says the product is ready for more strangers.

Then came the public launch. Hawkins says they knew they were ready for Hacker News when “Users could sign up and use the product without our help,” and when people they didn’t know had already signed up and found it useful. The launch produced about 300 deployments within a couple of days. The team also spent roughly $2,000 promoting the repository on Twitter, and the project began trending on GitHub.

The headline is 300 deployments. The useful detail is that the product was ready for a public launch only after people could sign up and use it without help, including people the team did not know. This was an open-source product, and deployments aren’t payments. The narrower lesson is that a public launch is more useful when people can reach the product’s result on their own.

Laszlo Kiss saw the other version. His initial launch gained traction, but he launched before the product was stable and before the pricing was properly worked out. He later said, “I launched too soon,” and lost his first users. Rob Walling’s fix is to release in stages. Start with an interested list, then let in 50 people or 300 people, correcting pricing and missing features before exposing the product to a much larger crowd.

That doesn’t mean “never launch.” It means asking what the launch has behind it. Is there a waitlist, a clear use case, a product people can use on their own, or a group already asking for it? If not, a launch may produce attention without customers. If so, it may be the right next channel.

Search can be that next channel when the customer already knows how to describe the problem. Dean Mathews got his first customer through paid search within a couple of days. He used PPC for about a month, then SEO began bringing people in, and he reached his first ten paying customers in roughly two months. At that point, the product was getting two or three signups a day, but Mathews counted the people who adopted and paid, not signups alone.

Content can work on a smaller scale, too. One plugin creator got three customers and $45 in monthly recurring revenue from a YouTube video with about 370 views over four months. That result shows that a video with a small number of views can still produce customers.

If I were starting from zero, I’d run the next channel as a small test, not a new identity. Choose one route that matches the evidence: search if people are actively looking, outbound if you can name and reach the buyer, or a public launch if strangers already succeed without help. Track qualified trials, payments, repeat use, and the hours spent. Stop when the channel creates attention but no movement through those steps. Then return to the product or the customer definition, rather than simply turning up the volume.

The decision is less “Which channel is best?” It’s “What has the product earned the right to try next?”

Write down the next experiment before you start: who it is for, where you’ll reach them, what counts as a qualified response, and how many attempts you’ll make. Keep one simple record of contacts, replies, trials, payments, repeat use, time spent, and reasons people decline. After that test, you should know whether to scale the channel, change the offer, or stop. That is the practical dividing line between a launch that teaches you something and a launch that merely makes noise.