Your First Customer Will Not Come From Marketing: How Cursor, Lovable, Clay and Resend Got Theirs
On this page

There is a fantasy every builder has. You launch. You post the link. You go to sleep. You wake up to a graph that goes up and to the right, and a notification that says “You have a new customer”.

Here is what actually happens. You launch. You post the link. Your mom likes it. A bot follows you. The graph is a flat line with one bump, and the bump is you checking the site from your phone.

The first sale almost never comes from marketing. It comes from a founder talking to one specific person with one specific problem, in a place that person already spends time. That was true for Airbnb fifteen years ago, and it is just as true for the companies that grew fastest in the last three.

Those older stories have been told to death, so here are four recent ones, and then the patterns they share. The figures are as reported by the companies or the press, and I have kept to the parts that are well documented.

Cursor: a product good enough to be passed around

Cursor, the AI code editor from Anysphere, is widely described as one of the fastest-growing software products ever, reported to have passed $100 million in annual recurring revenue well within two years of launch, with no traditional sales team and almost no paid marketing in that period.

What they did instead:

  • Built for themselves. The founders were the target user. They knew within minutes whether a change made the product better.
  • Chose a familiar starting point. Cursor was built on the same foundation as VS Code, so developers could switch in minutes and keep their settings and extensions. The cost of trying it was close to zero.
  • Let individual developers buy. A single engineer could pay with a card. Adoption spread inside companies from the bottom up before any enterprise deal existed.
  • Spent their effort on the product. The team has said that growth experiments moved the numbers far less than product improvements did.

The lesson: if your users talk to each other all day, as developers do, the product is the marketing. That only works when switching is easy and the improvement is obvious in the first session.

Lovable: the community existed before the company

Lovable grew out of GPT Engineer, an open-source project that Anton Osika published in 2023 and that collected tens of thousands of GitHub stars. The first commercial versions saw modest results. The relaunch under the Lovable name in late 2024 is what took off, with the company reporting millions in annual recurring revenue within weeks and far more within months.

  • The first customers were already there. A large open-source following and a waitlist meant the launch reached an interested crowd on day one.
  • The first attempts did not work, and they kept going. The eventual success came from a relaunch, not from the first version.
  • Users marketed it by using it. People shared what they had built, and every shared app was a demonstration.

The lesson: give something away that earns attention from exactly the people you will later sell to. An open-source tool, a free utility or genuinely useful writing all work. The audience comes first and the product second.

Clay: slow, manual and close to the customer

Clay is the useful counterexample to overnight success. It was founded in 2017 and spent years searching for the right market before becoming a multi-billion-dollar company in sales tooling.

  • They went where users already were. According to interviews with the team, their early go-to-market lead first found Clay through a Slack community for no-code enthusiasts, and the company found early customers in the same kind of forums.
  • They narrowed the audience. Clay grew once it focused on a specific user, people running outbound sales and growth operations, instead of anyone who might like a flexible data tool.
  • They watched customers use the product. Hands-on onboarding and a customer Slack community let the team see where people got stuck.
  • Their users became a channel. Agencies and consultants built businesses on Clay and taught others to use it.

The lesson: a horizontal product needs a vertical entry point. “Anyone can use it” is not a market. Pick one kind of customer, serve them by hand, and widen later.

Resend: earn trust with something free first

Resend is an email API for developers. Before the company launched, its founder Zeno Rocha released React Email, an open-source library for building emails with React, which became popular on its own. Resend then launched in 2023 to developers who already used and trusted that library.

The lesson: the open-source project solved a real annoyance and asked for nothing. When the paid product arrived, the trust was already in place.

What these four have in common

1. A narrow first user. Developers in an editor. People who wanted an app without coding. Outbound sales operators. Developers sending transactional email. None started with “everyone”.

2. An audience before a launch. Either the founders built one, through open source or public work, or they borrowed one by showing up in an existing community.

3. A small cost to try. Free tiers, familiar interfaces, payment by card, no sales call required.

4. Founders doing unscalable work. Onboarding by hand, replying in community channels, watching sessions.

5. Persistence through a version that did not work. Lovable relaunched. Clay took years. The first attempt is rarely the one that sells.

How to apply this if you are starting from nothing

You probably do not have fifty thousand GitHub stars. The same approach still works at small scale.

  1. Write down one customer. Job title, situation, the problem in their words.
  2. Find the three places they gather. A subreddit, a Slack or Discord group, a LinkedIn niche, a meetup.
  3. Be useful there for a month. Answer questions. Do not pitch.
  4. Offer to solve the problem by hand for a few people. Use their words to shape the product.
  5. Ask for money early. A small payment is stronger evidence than a hundred compliments.
  6. Ask every happy customer one question: “Who else has this problem?”
  7. Only then think about scale, including cold outreach, content and ads.

A caution about these stories

Case studies suffer from survivor bias. For each company above, many others did similar things and failed. Timing mattered: Cursor and Lovable arrived exactly as the underlying models became good enough.

So treat these as patterns that raise your odds, not recipes. The one element that appears in every version, successful or not, and that you fully control, is staying close to a small number of real customers until something clearly works. What happens after that first sale is the subject of how to ship products in 2027.

Questions people ask

How do you get your first paying customer?

By talking directly to people who have the problem. Find where they already gather, help them in person or one to one, and ask them to pay once the product solves something for them. Almost every successful company's first customers came from manual, founder-led effort, not from marketing at scale.

Do you need an audience before launching a product?

It helps more than anything else. Lovable launched to the community around its earlier open-source project, and Resend launched to developers who already used its open-source email library. If you have no audience, borrow one by being useful in communities where your customers already are.

How long does it take to get the first 100 customers?

It varies widely. Clay spent years finding its market before growth took off, while Lovable reached thousands of customers within weeks of its relaunch. The common factor is not speed. It is that the founders stayed close to early users until the product clearly worked for a specific group.