How to get your first ten customers

First, answer where your customer actually spends their time

Most founders already know the high-level advice: do things that don’t scale, run founder-led sales, and charge for the product early. The hard part is usually more tactical. You think you know who the target customer is; now how do you actually find them and start a conversation? The pattern from YC founder stories is that the first ten customers almost never come from choosing the perfect tool. They come from understanding where the buyer already spends time, then showing up there before you automate anything.

Most founders open that search with cold email, LinkedIn, and prospecting tools. It is easy — you can do it from a laptop, it feels like work, and there is endless automation to set up. For some buyers this genuinely works: sell sales software to a sales leader and email plus LinkedIn make sense, because that buyer lives on a computer and spends real time on LinkedIn. But a lot of the time you are selling to a school-district administrator, a property manager, an insurance agent, a truck dispatcher — people who don’t sit at a laptop all day and whose inbox is not the center of their work life. Blast them with cold email and you will just conclude that outbound doesn’t work for you.

One founder I know spent months emailing and messaging a legacy industry with terrible open and reply rates, then finally went to a trade show and walked the floor — and closed more in three days than in three prior months of cold email. His customers were on job sites and at conferences; they picked up the phone. Every hour he had spent tuning subject lines was wasted. So before anything else, sit down for an hour and answer basic questions about your buyer. What does their average day look like? How often do they check email? Which conferences do they attend? Are they on Reddit, on LinkedIn, do they pick up the phone, do they ask friends for recommendations, do they read an industry newsletter? If you can’t answer most of these concretely, that is the sign you haven’t spent enough time with real customers — fix that first.

Customers one through three come from your network

Before you touch any automation, start with people you already have some connection to. Almost without exception, the first two or three customers come from friends in the industry, former colleagues, classmates, or someone one introduction away. This is not because intros are a magic sales trick. It is because early customers buy because they trust you as a founder, not because the product is polished — and your first- and second-degree network are the people most willing to take that bet on you.

So work the sources in order. First your personal network — former colleagues, classmates, friends in the industry — the people who trust you most. Then your second-degree LinkedIn connections: look at who your connections know and ask for warm intros. One founder got intros to about half the customers she closed during her batch this way. Third, the AI-powered network search tools; Happenstance, for instance, lets you search your extended network in natural language — “find people who work on notification systems at big companies” — and surfaces a far wider net than you could scroll through by hand.

The intro request itself matters. Make it effortless for the person helping you: say exactly who you want to meet, why that specific buyer would care, and what to put in the forwarded email. The less translation work you ask your friend to do, the more likely the intro actually happens. One pattern worth calling out — prospecting tools only start to matter around ten to twenty quality customers. Founders routinely burn weeks configuring outreach tooling while sitting on second-degree connections they have never messaged. If you haven’t worked your network, you are skipping the lowest-hanging fruit there is.

Customers four through ten come from doing things that don’t scale

The most striking thing in these founder stories is how many of them, to close their first few customers, physically showed up. They didn’t settle for a Zoom; they got into the room. Most people avoid this — it is slower, awkward, sometimes costs money, and you might get rejected to your face in real time — and it also works better than almost anything else.

Fly out to close them. One founder flew to a single executive buyer four weeks in a row while the buyer kept rescheduling, and eventually closed. Another showed up at customer offices uninvited and got asked to leave most of the time; once he flew to Hawaii to meet a customer who kicked him out after eight minutes, and through sheer persistence that customer became one of his biggest accounts. The point is not that you should be obnoxious. It is that, for the first ten customers, there is often no tool that replaces being in the same room as the buyer.

Small conferences work unusually well too — the conversion from one real conversation at a small, industry-specific event beats cold-emailing that same person by a wide margin. A mini-playbook several founders used: set up a Calendly with 15-minute slots back to back across every day of the conference, email the attendee list a couple of times beforehand to fill the slots, email again during the event to catch the non-openers, and stack meetings all day. Micro-events are the other move — a founder dinner or happy hour for six to ten people in your ideal customer profile, maybe fifty to a hundred dollars a head, consistently converts better than a large event. Once someone has had dinner with you, they rarely ignore your follow-up, and sometimes they go out of their way to help.

Go where the pain is expressed in public

If you are building for consumers or small businesses, there is usually a place online where your future customers are already complaining about the exact problem you solve. Your job is to find it and show up there as a real person. For a lot of founders that place is Reddit. A common arc: a founder posts a product video, the comments rip it apart, they think the launch failed — and meanwhile a couple hundred silent lurkers quietly sign up. One founder told me Reddit was the source of his first ten customers; he found old threads where people complained about his exact problem and DM’d every commenter one by one. A healthcare founder spent a couple of months where responding to Reddit and Facebook complaints was basically her whole job, two to five posts a day. She got shadowbanned from a few subreddits and also got a pile of customers.

The principle is broader than Reddit: find the place where the pain is voiced publicly. For consumer products that is often Facebook groups, Discord, or YouTube comments; for B2B niches it might be an industry forum or a trade-association board. Reddit has the nice property that threads get pulled into Google and persist for years, so the work you do today keeps paying off.

Then, and only then, go outbound

Once you have worked your warm network, shown up in person, and gotten active in the communities where your customers complain, you will eventually need to reach people you don’t know. The job is simple, but not magic: find companies matching your ideal customer profile, find the right person at each, find their contact info, and reach out.

A few tools cover most of this. Apollo is the most common starting point — a lead database with email-finding and a basic sequencer, and a free tier generous enough to build your first list; for most founders at this stage Apollo alone is enough. Clay layers AI research and enrichment on top of your list and starts to matter when you want to qualify on something specific, like what software a prospect uses, who is hiring, or who posted about a topic recently. LinkedIn Premium is the richest source of fresh professional data — often you send a connection request with no message, then a short DM once they accept; one founder landed his biggest customer through exactly that, his highest-converting cold channel.

The most effective early outreach often isn’t framed as a sales pitch at all. It is framed as asking for advice or mentorship, a product review, a whiteboard session. To be clear, don’t mislead people — if you are not actually open to learning and just disguising a sales call, skip this. But if you genuinely want to learn and build the relationship, the framing is a real foot in the door. One founder asked dozens of CEOs in his space to be his mentor with short, real messages; most were flattered, a couple said yes, and a few became customers. Another talked to two hundred salespeople before her team had even built the product — every week she maxed out her LinkedIn connections around one specific hypothesis, roughly half accepted, and she converted twenty percent of those into calls, so by launch she already had a pipeline of people who had told her exactly what they needed. A devtools founder offered startups free whiteboarding on their agent architecture and a shared Slack channel to help implement it — the architecture just happened to require his exact product. And one selling to lawyers offered to pay them a hundred to two hundred dollars an hour for product feedback; about thirty percent accepted, and the conversion was high enough that his acquisition cost stayed reasonable.

The copy matters less than you think — a few things matter a lot

Once you have a framing, write the outreach. The exact wording matters less than founders believe, but a few things really do. Keep it under seventy-five words; long emails get ignored and read as LLM-generated. Make the single most important line a clear call to action — reply, call, a fifteen-minute demo — because without it people assume the worst, that you are going to eat their time. And run the one test almost nobody runs: read the email aloud to a friend, and rewrite anything that doesn’t sound like something you would actually say to a real person. That one-minute check removes most of the AI-sounding lines that ruin cold email.

One more move: the highest-converting outreach often gives the prospect something before asking for anything. An API-security company can run a quick vulnerability scan on the prospect’s public site and show them what it found; a mobile-onboarding tool can walk through the prospect’s app and send a couple of specific suggestions; a compliance startup can prepare a short audit note specific to that product, then ask for a meeting to go deeper. This work isn’t sustainable at scale — and that is the point. You are not doing it at scale; you are getting your first ten customers. Twenty minutes of work before asking for thirty minutes of someone’s time is a fair trade.

Then follow up. Something like three or four times over a couple of weeks is not rude when the message is relevant and specific. Early founders often stop after one note because they don’t want to feel pushy, but buyers are busy, inboxes are messy, and silence usually means “not now” more often than “never.”

You are the advantage

Here is the frame I find most useful for this whole phase. Customers one through three come from your personal network — friends, former colleagues, someone one intro away — and in the stories I have collected there are basically no counterexamples. Customers four through ten come from doing things that don’t scale: flying out, Reddit DMs, a dinner for six, personalized LinkedIn messages, free consulting. This part is tedious and manual, and it is a very good way to learn before you automate. Customers ten through fifty are where the playbook shifts — by then you have a refined pitch, case studies, and a real grasp of what resonates, and that is when Apollo, Clay, and email sequences start to make sense, because now you have a message worth scaling.

The reason the messy four-to-ten stretch works is that you, the founder, are the one doing it. When you show up at someone’s office, DM them on Reddit, or send an email only someone who studied the problem could have written, you signal something no automation can fake: that you care enough to put your own time in. That is your edge right now, the one thing established competitors in your space don’t have. Lean all the way into it. The first ten customers will not come from a tool. They will come from you tapping your network and showing up — and you don’t have to be a great salesperson to do it, only willing to do the unscalable things most people won’t.

Distilled from Max Kolysh’s YC Startup School talk on getting your first ten customers, itself built from dozens of YC founder stories — the tactics are theirs; the framing here, and any mistakes, are mine.

Related reading

• How to pick a startup idea — going deep enough on customers that you could run their business.
• How to build an AI-native services company — why capping early pilots protects the product you’re trying to build.