How to Get Your First 10 Customers

0:09Hi, I'm Max, a visiting partner here at

0:11Y Combinator. Today I want to talk to

0:13you about something I constantly hear

0:15from YC founders, which is, "I think I

0:17know who my target customer is. Now, how

0:19do I actually go find them and start a

0:22conversation?" YC Startup School already

0:24released a lot of good advice on the

0:26strategy of early stage sales, doing

0:28things that don't scale, founder sales,

0:31making sure to charge for your product.

0:33That advice is all great and you should

0:34definitely internalize it, but today I

0:36want to talk about the tactics of

0:38getting your first 10 customers. A few

0:40weeks ago, I made a post internally on

0:42Bookface, YC's internal social network,

0:44asking founders how they actually got

0:46those first 10 customers. Dozens of

0:48founders responded and wanted to share

0:49their experience. This video is a

0:51compilation of all my learnings. So,

0:53let's get started. Before you even touch

0:55a tool or pick a channel, you need to

0:57answer one question. Where does your

0:59target customer actually spend their

1:00time? Most founders default to cold

1:03email and LinkedIn outreach. This is

1:04because it's easy. You can do it from

1:06your laptop. It feels like you're

1:08working and there's a ton of automation

1:09and tools that you can set up for all

1:11this. And for some buyers, this actually

1:13works fine. If you're selling sales

1:14software to a sales leader, for example,

1:16email and LinkedIn actually make a lot

1:18of sense. The buyer who is a sales

1:20leader lives on a computer. They spend a

1:22ton of time on LinkedIn. It's going to

1:24work for them. But a lot of times you

1:26might be selling to a school district

1:27administrator or property manager, an

1:30insurance agent or a truck dispatcher. A

1:32lot of people don't sit on a laptop all

1:34day and a lot of people don't have their

1:36inbox as the most important surface in

1:38their work life. You can blast them with

1:40cold email all you want and you're just

1:42going to conclude that outbound sales

1:43doesn't work for you. One recent YC

1:45founder spent months on email and

1:46LinkedIn outreach to a legacy industry.

1:48His open rates and reply rates were

1:50terrible. When he finally decided to

1:52actually go to an industry trade show

1:54and walk the floor, he closed more in 3

1:56days than he had in 3 months of sending

1:58cold emails. His customers were on job

2:00sites and sometimes they went to

2:02conferences. They picked up the phone.

2:04In hindsight, any amount of time he

2:05spent refining the subject line or his

2:07email copy ended up being a waste of

2:09time. So before you do anything else,

2:10sit down for an hour and answer some

2:12basic questions about your buyer. What

2:14does their average day look like? How

2:15often do they check their email? Do they

2:17go to conferences? And which ones, if

2:19so? Are they on Reddit? Are they on

2:20LinkedIn? Do they pick up the phone? Do

2:23they mostly ask their friends for

2:24recommendations for new services and

2:25software to use? Or maybe there are

2:27industry newsletters that they read. If

2:29you can't answer most of these questions

2:31concretely yet, that's a sign you

2:33haven't spent enough time with real

2:34customers. Go fix that before anything

2:36else. All right, now it's time to

2:38actually start talking to customers.

2:39Before you touch any email automation or

2:41prospecting tool, you have to start with

2:43people you already have some connection

2:45to. The first two or three customers,

2:47almost without exception in these

2:48founder stories, came from one of these

2:50sources. Friends in the industry, former

2:53colleagues, classmates from school, or

2:55more generally people one introduction

2:57away. The reason your warm network

2:58matters isn't that intros are some magic

3:01sales trick. It's that the people buying

3:02your product early are buying because

3:04they trust you as a founder, not just

3:07because of the quality of the product.

3:08If someone's going to take a bet on a

3:10new product from a new company, they at

3:12least need to trust the person building

3:13it. Your first degree and second degree

3:15network are the easiest people to take

3:17that bet on you. So before you go

3:19outbound, work through a few sources in

3:21order. First, your personal network.

3:23Former colleagues, classmates, friends

3:24in the industry. These people will trust

3:26you the most. So talk to them first.

3:29Next is your secondderee LinkedIn

3:31connections. Look at who your

3:32connections know and ask for warm

3:33intros. One recent YC founder told me

3:35she got intros to about half of the

3:37customers she closed during her YC batch

3:39via LinkedIn intros. Third, there are a

3:41bunch of AI powered network search

3:43tools. Happenstance, for example, is a

3:45recent YC company that lets you search

3:47across your whole extended network in

3:48natural language. You can ask things

3:50like find people who work on

3:52notification systems at big companies

3:54and it'll surface relevant people from a

3:56much wider net than you can scroll

3:57through manually. When you ask for an

3:59intro, be specific. Tell them who you

4:01want to meet, why they specifically

4:02would care, and what to say in the

4:04forwarded email. The easier you make it

4:06for them, the more likely the intro

4:08actually happens. One pattern I want to

4:09call out, prospecting tools only start

4:12to matter once you have 10 to 20 quality

4:14customers. Tons of founders spend weeks

4:16setting up outreach tools when they

4:18actually have a lot of secondderee

4:19connections they haven't messaged yet.

4:21If you haven't worked your network,

4:22you're skipping the lowest hanging

4:23fruit. One of the most surprising things

4:25that came up in those YC founder stories

4:27is that so many of them to close their

4:29first few customers actually showed up

4:31in person. They didn't just settle for

4:32getting on a Zoom, they got into a room.

4:35They made it their mission to be in the

4:36same room as the buyer. A lot of people

4:38don't want to do this tactic. It's

4:39slower. It's awkward. It might cost

4:42money if you have to travel somewhere

4:43and you might get rejected to your face

4:45in real time. But it also happens to

4:47work better than basically anything

4:48else. First, fly out to close them. One

4:51recent YC founder flew to a single

4:53executive buyer four weeks in a row. The

4:56buyer kept rescheduling, but he kept

4:58showing up and eventually closed them.

4:59Another founder regularly showed up at

5:01customer offices uninvited and got asked

5:03to leave most of the time. One time he

5:06actually flew to Hawaii to meet a

5:07customer who kicked him out after 8

5:09minutes of conversation. And that

5:10customer through some persistence became

5:12one of his biggest accounts. Another

5:14learning was that small conferences tend

5:16to work really well. The conversion rate

5:18from one real conversation at a small

5:20industry specific conference can be much

5:22higher than trying to cold email that

5:23same person. Here's a mini playbook on

5:25conferences that multiple founders used.

5:28You can set up a calendarly and put

5:2915-minute slots back to back across

5:31every day of the conference. Before the

5:33event starts, email the attendee list a

5:35couple of times and fill as many slots

5:37as possible. Email them again during the

5:39event to catch the people who didn't

5:41open the first email. Stack meetings the

5:43whole day and that'll make sure you get

5:44the most out of the conference. Another

5:46tactic is micro events. A few founders I

5:49spoke to ran founder dinners or happy

5:51hours for their ideal customer profile.

5:53Something like 6 to 10 people, maybe $50

5:55to $100 ahead. A lot of founders

5:58consistently said that this converts way

5:59better than large events. Once someone's

6:01eating dinner with you, it's very

6:02unlikely they're going to ignore your

6:04email afterwards, and sometimes they'll

6:06even go out of their way to help you.

6:07For the first 10 customers, there really

6:09is no tool that can replace being in the

6:11same room as your buyer. If you're

6:13building a consumer product or a product

6:14for small businesses, there's often a

6:16place online where your future customers

6:17are already complaining about the

6:19problem you solve. Your job is to find

6:21it and then show up there as a real

6:23person. For a lot of founders, that

6:24place has actually been Reddit. A common

6:26Reddit story looks something like this.

6:28A founder posts a video of their product

6:29on Reddit. A lot of the comments start

6:31ripping them apart, and the person

6:33thinks their launch is a disaster. But

6:35then there's a couple hundred lurkers,

6:36the people who never would comment, who

6:38might quietly sign up for their product.

6:40Another founder told me Reddit was the

6:42source of his first 10 customers. He'd

6:44find old threads where people complained

6:46about the exact problem he was solving

6:48and he would DM every commenter one by

6:50one. A healthcare founder said that

6:52responding to Reddit and Facebook

6:53complaints was basically her whole job

6:55for a couple of months. She was doing

6:56two to five posts a day. She did get

6:58shadowbanned from a few subreddits along

7:00the way, but she also got a bunch of

7:02customers. This principle is broader

7:03than just Reddit. You've got to find the

7:05place where the pain is being expressed

7:07in public. For consumer products, that's

7:09often Facebook groups, Discord, or

7:11YouTube comments. For some B2B niches,

7:13it could be industry specific forums or

7:15trade association message boards. The

7:17nice thing about Reddit is that threads

7:19get pulled into Google search and

7:20persist for years. So, the work you do

7:23today will keep paying off for a long

7:25time. Okay, you've worked your warm

7:27network. You've showed up in person.

7:28You're active in the communities where

7:30your customers complain. And at some

7:31point, you're going to need to go

7:32actually outbound the people who you

7:34don't know. Your job is to find

7:36companies that match your ideal customer

7:37profile, find the right person at each

7:39one, find their contact info, and reach

7:41out. Here, I'll talk about a few tools

7:43that are most common and that you might

7:45want to try. Apollo was the most common

7:47starting point for the founders I talked

7:48to. It's basically a lead database with

7:50email finding capabilities and a basic

7:52sequencer. Their free tier is generous

7:54enough for you to build your first list.

7:56For most founders at this stage, Apollo

7:57alone is enough. Clay is another really

8:00popular tool. It lets you build research

8:02and enrichment workflows with AI on top

8:03of your prospect list. It starts to

8:05matter when you want to qualify leads on

8:07something very specific like what

8:09software your prospects use, who's been

8:11hiring, or who posted about a topic

8:13recently on LinkedIn. LinkedIn Premium

8:15is the richest source of fresh

8:16professional data. Often times, you'll

8:18want to send a connection request

8:19without a message followed by a short DM

8:22once they actually accept you. One

8:23recent YC founder mentioned that he got

8:25his biggest customer this way and that

8:27was his highest converting cold outreach

8:29channel. Great. Now you have your

8:30outbound lead list. It's time to decide

8:32how you're actually going to communicate

8:34with these prospects. One of the most

8:36interesting patterns I saw is that for a

8:37lot of founders, the most effective

8:39early outreach wasn't actually framed as

8:41a sales pitch. It was framed as

8:43something else. Maybe asking for advice

8:44or for mentorship or asking for a

8:47product review or for a whiteboard

8:48session. To be clear, you shouldn't

8:50mislead people. If you're not actually

8:52open to learning from them and you're

8:54just trying to disguise a sales call,

8:55don't do this. But if you genuinely want

8:57to learn and invest in building a

8:58relationship with these people, this

9:00framing can often be a real foot in the

9:02door. Here's a few examples. One founder

9:04reached out to dozens of CEOs in his

9:06space asking if they'd be his mentor.

9:07His messages were short, they were real,

9:09and most people were flattered. A couple

9:11accepted, and a few of them eventually

9:13became customers. Another founder talked

9:15to 200 salespeople before her team even

9:17built out the product. Every week, she'd

9:19max out her LinkedIn connections with

9:21one specific hypothesis she was testing

9:22that week. basically user research on

9:25steroids. She said about 50% of her

9:27connection requests were accepted and

9:29she converted 20% of those into calls.

9:31By the time she had a product, she

9:33already had a pipeline of people who

9:34told her exactly what they needed and

9:36she could go reach out to them. A

9:37DevTools founder offered startups free

9:40whiteboarding sessions on their agent

9:41architecture and then offered to open up

9:43a shared Slack channel so he can help

9:44them implement it. The trick was the

9:47architecture just happened to require

9:48his exact product. And here's one really

9:51creative variation from another founder.

9:53He was selling to lawyers and he

9:55actually offered to pay them $100 to

9:57$200 an hour, basically a portion of

9:59their hourly fee in exchange for product

10:01feedback. About 30% of people he reached

10:03out to accepted that offer. The math

10:05sounds expensive, but the conversion

10:07rate was high enough that his customer

10:08acquisition cost ended up being quite

10:10reasonable. This kind of approach can

10:12especially work in markets with high

10:13contract values. Once you've decided on

10:15your framing, it's time to actually

10:16write some outreach copy. In my

10:18experience, the exact copy matters less

10:20than people think, but a few things

10:22really do matter. First, keep it under

10:2475 words. Long emails tend to just get

10:27ignored, and they'll assume you used an

10:28LLM to write it. Second, the single most

10:31important line is a clear call to

10:32action. What do you actually want from

10:34this person? Do you want to reply? Do

10:36you want to call? Are you offering a

10:3815-minute demo? Make it really clear.

10:40Otherwise, they'll assume the worst.

10:42Basically, that you're going to eat up a

10:43bunch of their time. Third, there's a

10:45very easy test that a lot of people

10:46don't do to check whether the email

10:48sounds human or not. Find a friend and

10:50just read your email out loud to them.

10:52If anything sounds like something you

10:53wouldn't actually say to a real person,

10:55rewrite that part. This is basically a

10:57one minute sanity check and it'll help

10:59you remove most the AI sounding lies

11:01that end up ruining cold emails. One

11:03last tip here. Some founders said that

11:05their highest converting outreach gave

11:07the prospect something before asking for

11:09anything. For example, an API security

11:11company can do a quick vulnerability

11:13scan on the prospect's public website

11:15and show them what they found. If you're

11:17a mobile onboarding tool, you can walk

11:19through the prospect's app and send a

11:20couple of specific suggestions. One

11:22example from the YC founders I talked to

11:24was a compliance startup that prepared a

11:26short audit note specific to that

11:28prospect's product. And then they asked

11:30each one for a meeting to go into more

11:32detail. Now, this kind of work is not

11:34sustainable at scale. And that's kind of

11:36the point. You're not trying to do this

11:38at scale. you're trying to get your

11:39first 10 customers. 20 minutes of work

11:41before asking for 30 minutes of

11:43someone's time is actually quite a

11:44reasonable trade. And one last thing

11:46here, don't forget to follow up.

11:48Something like three or four times over

11:49a couple weeks is perfect. I want to

11:51leave you with one frame that I think is

11:53a really useful way to think about this

11:55whole phase. Customers 1 through three

11:58almost always come from your personal

12:00network. Again, friends, former

12:02colleagues, people one intro away. In my

12:04founder survey, there were basically no

12:06counter examples to this pattern of how

12:08people found their first few customers.

12:11For customers 4 through 10, they come

12:13from doing things that don't scale.

12:14Flying out to customers, sending Reddit

12:16DMs, a small dinner for six prospects,

12:19personalized LinkedIn DMs, or free

12:21consulting sessions. This part is going

12:23to be tedious and it's going to be

12:24manual and it's a really good way to

12:27learn before you automate. Then

12:29customers 10 through 50 are where the

12:31playbook starts to shift. By this point,

12:33you'll probably have a refined pitch,

12:35case studies, and you'll have a really

12:37good grasp on what resonates about your

12:38value prop. And that's when Apollo,

12:40Clay, email sequences, all these higher

12:43volume outreach tools actually start to

12:45make sense. Now you have a message

12:47that's worth scaling and will get you

12:48results. So, as a reminder, the reason

12:51this messy middle phase from 4 to 10

12:53customers works is that you personally,

12:55the founder, are doing it. When a

12:57founder shows up at someone's office,

12:59DMs someone on Reddit, or sends a really

13:01researched email that only somebody who

13:02studied this problem knows about, it

13:04signals something no automation tool is

13:06going to be able to fake. That you care

13:08enough about this problem to put your

13:10own time into it. That's your advantage

13:12right now. It's the one thing you have

13:14that established companies in your space

13:16won't. So, you have to really lean into

13:18that. So, in conclusion, the first 10

13:19customers will likely not come from a

13:21tool. They're going to come from you

13:23manually tapping into your network and

13:25showing up. You don't have to be an

13:27amazing salesperson to do this. You just

13:29have to be willing to do the unscalable

13:31things that most salespeople and

13:32founders are not willing to do. Good

13:34luck and I can't wait to see what you

13:36build.