Landing your first ten customers

The first ten customers are not a smaller version of the next hundred. They are a different job. Later, customers find you through channels. Now, you find them one at a time, by hand, and each one teaches you something you cannot learn any other way. This playbook is about doing that job well and fast, using what founders told The Product Market Fit Show about how they got theirs.

Why the first ten are different

Three things are true at zero that stop being true later.

Nobody is looking for you. There is no search volume for a product that does not exist yet, so ads and SEO have nothing to catch. The founders on the show who tried paid acquisition before ten customers describe the same result: money spent, a few sign-ups, nobody who stayed.

You do not yet know who the customer is. You have a hypothesis. The first ten conversations are how you find out whether it is right. If you buy attention instead of earning it, you skip the part where you learn.

Trust has to come from somewhere other than proof. You have no logos, no case studies, no reviews. The only substitute is a person the buyer already trusts vouching for you, or you spending enough time with the buyer that they trust you directly.

So the first ten come from three places: people you already know, people they introduce you to, and people you go and find one at a time. Across 200-plus interviews the pattern holds. The first one to three customers come from the founders' own network or their own past job. The rest come from outreach the founders run themselves.

Step 0: Research before you build

Most founders read The Lean Startup and jump straight from idea to MVP. The founders who built the biggest companies did something first. Pablo Srugo, who hosts the show, calls it research mode.

The MVP is not step one. Research is step one.

The strongest form of research is being the customer. Sanish Mondkar left his job as chief product officer at SAP with no plan, noticed "for hire" signs everywhere, and started Legion, workforce management for hourly staff. The founder of Philz Coffee agreed to use whatever he built, on one condition: he had to be trained as a barista first. He did five days of barista training, then visited Philz stores every week for a year, watching managers use early versions in silence. Legion partnered with that one chain for a year before selling to anyone else. It now serves a million paid users across 60,000 locations.

People still give you something new at 500 conversations.

The second-best form is embedding beside the customer. Legora, the legal AI company that became a $1.8B business in about two years, built its first version by working out of a law firm's office. Peregrine, now valued at $6.8B, spent 18 months inside a police department as unpaid crime analysts. The signal that they had something was a homicide detective calling them for help on a cold case. How Peregrine built a $6.8B company by doing their customer's job

The third-best form is hundreds of calls with the right questions. Shahar Peled of Terra Security ran about 100 conversations with chief information security officers in two phases. The second phase asked harder questions: how much would you pay, would you sign something, will you be a design partner and give us your time. All six design partners converted to paying customers within six months. How Terra Security hit $1M ARR in one quarter

Step 1: Narrow the customer until it hurts

The most common reason founders cannot land ten customers is that they are trying to land any customer. "Small businesses" is not a customer. "Owner-operators of independent HVAC companies with 5 to 20 field technicians in Ontario who still dispatch by phone" is a customer. You can list them. You can find their phone numbers. You know what their Tuesday looks like.

Gusto is the clearest case on the show. It is now a $1B-revenue company serving 500,000 businesses with payroll, benefits and HR. It launched in December 2012 for one segment: new technology companies in California with full-time employees who had never run payroll before. That excluded almost every company in the country. For the few it served, it was the obvious choice. About a hundred companies the founders had never spoken to signed up in the first month. How Gusto got to $1B in revenue: launching in one state with six people

Write down a definition that answers these questions:

• Who has the problem badly enough to change how they work? Not who has it at all. Who loses money, sleep or customers because of it this month.
• Who has the authority to say yes? In a 15-person company that is the owner. In a 500-person company it may be a director who needs two other people to agree.
• Who can you reach in the next four weeks? A perfect segment you cannot reach is worthless. A good segment where you already know three people is where you start.
• What do they already pay for to solve this? Spreadsheets, a contractor, an old tool, an intern. That is your real competitor and your pricing anchor.

Then write the second list: who you will not sell to. The show's founder-led sales roundup found that the founders who reached $1M ARR fastest kept a written disqualification list of industries, company sizes and titles they turned away.

The biggest mistake is failing to disqualify. Founders are biased toward saying yes because they need revenue.

Jon Yoo at Suger, which sells cloud-marketplace billing, went further. He did not let every interested company in. He curated the first 20 customers on purpose, because that set would decide what the product became for the next three months. Suger reached $2M ARR with five people. Suger: $2M ARR with five people

Step 2: Build the list of 100

You need a list, not a strategy. One hundred named people who match the definition, in a spreadsheet, with a column for how you can reach each one.

Sources, in the order you should use them:

1. Your own contacts. Go through your phone, LinkedIn connections, email and every previous employer. Anyone who matches the definition or clearly knows someone who does.
2. Your team's and your investors' contacts. Ask each person for ten names. Ask specifically: "Who do you know who runs [segment]?" A vague ask gets a vague answer.
3. Communities where the segment already gathers. Trade associations, Slack and Discord groups, subreddits, local meetups, industry newsletters. Join them and read for a week before you post anything.
4. Public lists. Directories, award lists, conference speaker lists, review sites, job postings that mention the problem. A company hiring for the role you replace is a company with the problem.
5. Cold search. LinkedIn Sales Navigator, Apollo or plain LinkedIn search by title and company size. This comes last because it converts worst, but it is unlimited.

For each name, record: how you can reach them, your best guess at their current workaround, and one specific fact about them you can open with. The specific fact is what separates a message that gets read from one that gets deleted.

Step 3: Get in the room

Warm introductions convert five to ten times better than cold outreach, so spend them well. Make introductions easy to give: send the introducer a three-line blurb they can forward unchanged, with your one-sentence description, why you want to talk to this specific person, and what you are asking for (20 minutes, not a sale).

For cold outreach, frame it as learning, not selling. The founders on the show who converted best asked for 15 minutes to hear about the buyer's work, with the founder's name on the message.

Sell the meeting with me, not the product.

Omar Haroun at Eudia, which sells AI to corporate legal departments, says 90 percent of the company's customers came from cold emails he sent himself. He positioned the email as research, leaned on the credibility of a previous exit, and asked chief legal officers for very little of their time. Five Fortune 500 chief legal officers became design partners that way, and the company went from $2M to $20M ARR in twelve months. How Eudia grew from $2M to $20M ARR

The rules for the message itself:

• Short. Under 80 words. Mobile screens are small and you are an interruption.
• Specific. Reference something true about them: a hiring post, a product launch, a talk they gave, a review they left of a competitor.
• One ask. A 15 or 20-minute call about how they do the thing today. Not a demo. Not a trial.
• No attachments, no links, no calendar link in the first message. Those all read as automation.
• Follow up three times, three to five days apart, each shorter than the last. Most replies come on the second or third message.

Volume matters more than founders expect. Pylon's three founders sent 120 personalised LinkedIn messages a day between them for three months. That is about 10,800 messages and roughly 600 booked meetings, run as seven 15-minute calls each per day. Mark Hughes at SolidRoad sent around 500,000 cold emails, got about 5,000 replies and 250 demos, and closed 40 customers at $25,000 a year: $1M ARR. How SolidRoad hit $1M ARR with cold email outbound

Expect 5 to 15 percent of cold messages to get a reply when they are targeted, and 1 to 3 percent when they are not. If you are below 5 percent, the problem is the list or the message, not the volume.

Step 4: Run the conversation so it ends in a sale

The first conversation has two goals: learn how they do it today, and find out whether the pain is real enough to pay for. Do both before you show anything.

Spend the first 15 minutes on questions:

• Walk me through the last time you did [the task]. What happened, step by step?
• What did that cost you, in hours, money or things that went wrong?
• What have you tried to fix it? Why did that not stick?
• Who else is involved when this goes wrong?
• If this disappeared tomorrow, what would you do with the time?

Listen for numbers and for emotion. "It's annoying" is not a customer. "We lost a deal last quarter because it took three weeks" is. The show's roundup on first enterprise customers puts it simply: if a prospect calls your product a "nice to have", they are not your first customer. Getting your first enterprise customer

Then, only if the pain is real, show the product for five minutes, framed around their exact task. Not a tour. Their questionnaire, their dispatch board, their data if they will give it to you. Legora's demo lets a lawyer upload their own old memos and watch the AI redraft them. It converts 55 percent of the people who see it. Jay Madheswaran at Eve, legal AI for plaintiff lawyers, gets 40 percent of cold outreach to a demo for the same reason: in an industry that has never seen this, the demo does the selling. Jay Madheswaran: from $3M ARR to unicorn

End with a direct ask. The founders who close their first ten all asked. "Would you pay for this? What would it need to do for you to start using it this month?" Silence after the question is fine. Let them answer.

Step 5: Make the first ten say yes

Your first customers are taking a risk on you, so remove every other risk you can.

Charge, but make it easy to start. Free pilots produce polite users who never convert. A paid pilot, even at a small amount, produces a real customer who tells you the truth. George Kurdin at Monk, which automates accounts receivable, closed a $36,000 pilot off a one-pager and a Loom video, without ever getting the prospect on a call. Cold email alone took Monk to $1M ARR in six months. He shut down his last startup and gave the money back, then hit $1M ARR in 6 months

Add friction on purpose. This sounds backwards. Jack Newton at Clio, now a multi-billion-dollar legal software company, required early beta users to sit through a 30-minute video before they could get in. He wanted about ten engaged customers who were representative of the wider market, not thousands of casual ones, and interviewed each to check they were not asking for something esoteric. The customers who jump through a hoop are the ones with the real problem. Getting your first enterprise customer

Do the setup for them. Import their data yourself. Configure it on a call. Write the first templates. Whatever the product needs before it is useful, you do it, by hand if necessary.

Show up in person when it matters. SolidRoad's founder took 56 flights in one year to sit in customers' offices, fix onboarding, meet every user and convert month-to-month plans into annual and multi-year contracts. The company's net revenue retention is 186 percent.

Give them your phone number. Early customers should be able to reach a founder directly. The speed of your response is the product's main feature right now.

Make the paperwork tiny. Pylon, GetVocal and Eve all signed early customers on one or two-page agreements the founders negotiated themselves. Anything that goes to a legal team stalls for weeks.

What to expect from the numbers

A rough funnel for a well-targeted list, from founders who tracked it:

So ten customers from a mostly cold list means 300 to 700 names and 40 to 80 meetings. Ten from a warm list means 50 to 100 names and 20 to 30 meetings. SolidRoad's numbers, at industrial scale, are the same shape: 500,000 emails, 250 demos, 40 customers.

Budget four to eight weeks of one founder's full time. If you are past eight weeks with fewer than five, stop and change the segment, the message or the product, in that order.

Traps founders fall into

• Pitching in the first meeting. You learn nothing and the prospect learns you are not listening.
• Chasing logos. A big name that takes six months to sign teaches you less than three small companies that start Monday. Rubrik, now a $16B public company, signed its first customer at a local hardware chain in New South Wales. Soham Mazumdar, Rubrik
• Building for the maybe. A prospect who says "I'd buy it if it also did X" is usually saying no politely. Ask for a paid commitment contingent on X before you build it.
• Giving it away. Free users do not fight for your product internally, do not tell you the truth, and do not become references.
• Stopping at three. Three customers is a coincidence. Ten is a pattern. Keep going until the pattern shows you who the customer is.
• Outsourcing this to a salesperson. No one can sell a product nobody has bought yet except the person who can change it. The first sales hire playbook covers when that changes.

What ten customers should teach you

By the tenth customer you should be able to answer, with evidence, all of these:

• Who the buyer is, by title and company type, and who else has to agree.
• The exact moment they realise they need it, and what triggers it.
• The workaround you replace and what it costs them.
• Which three features they use every week and which ten they never touch.
• What they say when they describe you to a peer, in their words, not yours.
• Roughly what they will pay and how they want to pay it.

Those answers are the input to everything in the next playbook. If you have ten customers and cannot answer them, you have ten pilots, not ten customers, and the next step is to go back and ask.

Episodes and posts cited

• Founder-led sales: how 10 founders closed their first deals
• Getting your first enterprise customer
• The five steps to product-market fit, Pablo Srugo
• How Peregrine built a $6.8B company by doing their customer's job
• How Terra Security hit $1M ARR in one quarter, Shahar Peled
• How Gusto got to $1B in revenue, Josh Reeves
• Suger: $2M ARR with five people, Jon Yoo
• How Eudia grew from $2M to $20M ARR, Omar Haroun
• How SolidRoad hit $1M ARR with cold email outbound, Mark Hughes
• Jay Madheswaran, Eve: from $3M ARR to unicorn
• George Kurdin, Monk: $1M ARR in 6 months
• Soham Mazumdar, Rubrik