How to get your first 100 customers: the routes founders used, with cited cases and a 30-day plan

Startups get their first customers by hand: the founder finds the names, writes the messages and closes the deals, starting with a personal network and then the places buyers already gather. Lenny Rachitsky found seven routes account for 99% of consumer startups' early growth, and three cover business-to-business companies. If you are starting from zero, aim for 10 paying customers first, then test whether one route reaches strangers on the way to 100.

How do startups get their first customers?

Startups get their first customers by hand, one at a time. In the first months, the founder finds the names, writes the messages and closes the deals.

There are fewer routes than most founders expect. After reviewing the early growth strategies of more than 100 consumer companies, Lenny Rachitsky found seven reliable ways to reach a first 1,000 users. He reported that those seven account for 99% of all B2C startups' early growth. For business-to-business (B2B) companies, his separate review found that only three sourcing strategies account for every B2B company's very early growth.

Route	On the consumer list	On the B2B list
Friends and former colleagues	Yes	Yes (personal network)
Targeted strangers by email, DMs, calls	Yes	No
Places your audience already gathers	Yes	Yes
Influencers, paid or organic	Yes	No
Press	Yes	Yes
Viral content	Yes	No
Physical placement such as flyers	Yes	No

In our reading, two routes carried the most weight for consumer companies. Friends and former colleagues were a large driver of early growth for about 20% of startups. Direct recruiting of strangers was important for the early growth of 30% of consumer startups, mostly marketplaces filling their supply side.

Why 10 comes first, then 100

Ten paying customers is the first milestone because, in Stripe's words, by the time you have 10 happy customers, it isn't a fluke anymore. At that point you hear repeated needs and have people who will give references. Slack reached ten through friends: We had maybe six to ten companies to start with that we found this way.

If you have reached 10, 100 is the next test in our reading, because your personal network rarely reaches that far. If you go from 10 to 100, the next 90 customers show whether a route works with strangers. The Baremetrics case shows one company's path to that number.

To start this week, open a spreadsheet with three columns:

Company | Contact name | Email

Stripe's guide says a few dozen will keep you quite busy while you learn the mechanics. Put your network at the top, because you will close more sales from warm introductions than from an equal number of cold pitches.

Pick the route that matches what you already have

Founders who got early traction rarely spread themselves across many routes. In one review of consumer apps, most startups found their early users from just a single strategy, and no one found success from more than three.

Our method follows that count. Pick the one route that matches an asset you already hold. Add a second only after the first shows a signal. Stop at three.

Match the route to your starting asset

Starting asset	Route	Cited case	Who does the work	Hours per week (our estimate)	Time to first signal (our estimate)	Number to measure
An audience, or the ability to build one	Guest posts on other people's blogs	Buffer: around 100,000 users within the first 9 months, from guest blogging alone	The founder who writes	8	Months	Signups per post
A network	Personal contacts of founders and early hires	Gong: all of the first dozen customers were some sort of personal connection	Every founder and early employee	4	Days	Replies per introduction asked
A marketplace of early adopters (a launch site)	A demo posted where your buyers browse	Dropbox: a simple video, demoing the product, published on Hacker News	The founder	6 in launch week	Days	Signups in the first 72 hours
A niche	Regular participation in enthusiast forums	Netflix: DVD owners were found in the fringe communities of the internet	A founder who knows the niche	5	Weeks	Members who try the product

On the niche row, Netflix's early staffer did not announce himself as a company employee. Skip that part. Post under your own name and say what you are building. For more on owning a niche, see own a niche keyword and the ConvertKit case.

Expect the audience route to start slowly. Buffer's founder wrote around 150 guest posts in that period, and the early ones barely drove any traffic. If writing is your asset, the founder-led content playbook covers the cadence.

How we got the hours

These are our planning numbers, not figures from the cases.

• Audience, 8 hours: two guest posts at 4 hours each, including the pitch.
• Network, 4 hours: for example, eight introduction requests at 30 minutes each, with follow-up.
• Marketplace, 6 hours: the demo asset, the post, and replying to every comment during launch week.
• Niche, 5 hours: about an hour a day across two or three forums.

A worked choice: a founder with a strong network and active forum membership takes the network and niche rows. That adds up to 9 hours a week. If introductions produce no replies after two weeks, swap in the launch row rather than adding a third route on top.

Should you do things that don't scale?

Yes. In the first weeks, the work that cannot scale is the work that gets customers. Paul Graham's essay for Y Combinator (YC) put it plainly: at least one founder (usually the CEO) will have to spend a lot of time on sales and marketing.

Two cases worth copying

Stripe's founders refused to let a yes go cold. Graham described the timid version: founders ask "Will you try our beta?" and if the answer is yes, they say "Great, we'll send you a link." The Collison brothers did the opposite. When anyone agreed to try Stripe they'd say "Right then, give me your laptop" and set them up on the spot. YC named it the Collison installation.

Both moves remove a step between interest and use. For a software-as-a-service (SaaS) product, that means you do the setup during the call, import the data yourself, or hand-migrate switchers from the tool they already use.

Why small numbers are enough

Manual recruiting feels pointless because the counts are tiny. Graham's math answers that: If you have 100 users, you need to get 10 more next week to grow 10% a week. Held steady, After a year you'll have 14,000 users, and after 2 years you'll have 2 million. Ten installs by hand in a week is a realistic founder workload.

Where manual work stops paying off

Manual work earns its cost while it teaches you something. Each hand-done setup, order or migration should change the product, the pitch or your list of who to target. When a step repeats and the notes stop changing, it has become labor, not research. That is the point to script it, template it or build it.

Keep a log for every manual task and review it each Friday:

Task: ______________________   (e.g. onboarding call with live setup)
Date | Customer | Time spent | What surprised me | Change made (product / pitch / targeting)

Friday review:
- Last 5 entries with "nothing new" in "What surprised me"?  -> automate or template this step
- Last 5 entries produced a change?                          -> keep doing it by hand
- Same objection 3 times?                                     -> fix it in the product or the pitch

The thresholds in that template are our working rule, not a published finding. Adjust them to your volume, but keep the principle: stop doing it by hand when it stops teaching you.

Which channel works first for B2B and for consumer?

The two motions start in different places. In business-to-business sales, the first customers came from people who already trusted the founders. In consumer, the first users came from places where they already gathered.

B2B: network, then targeted outreach, then investors

In Lenny Rachitsky's review, almost every B2B business BOTH hits up their personal network AND heads to the places their potential customers were spending time. His framing of the choice: "The question isn’t which of these two routes to pursue, but instead how far your own network will take you before you move on."

Figma shows the outreach route done with a filter. The list was short and chosen. It was not a blast.

Investors are the third source. At eShares, Andy Palmer from Tamr was an investor in eShares, and brought it into his company early on.

In our reading, the network carries more weight when a person has to buy. Tapping your personal network is even more important if you need to “sell” your product, because early buyers need a reason to trust you.

Set a cadence and track it. One founder in the series had a target of 15 to 18 net new IT folks at different companies to talk to every month for the first six months.

Consumer: go where users already are

Consumer founders rarely have a network of buyers. Online, that means a thread, a subreddit or a launch board: see answering questions in other forums, the Show HN launch and the Reddit marketing playbook. Offline, it means the meetup, campus or club where your user already spends time.

The order to try

Press comes last: PR can work, but rarely does.

1. Business-to-business: former colleagues and friends who fit your customer profile.
2. Business-to-business: a filtered cold outreach list, built the Figma way.
3. Business-to-business: your investors' contacts and sister portfolio companies.
4. Consumer: the forum, board or physical place your users already visit.
5. Both: communities and content, then press.

A worked example of when to move on: write down 40 names from your network. In this example, if 30 calls produce no trials, the network has run out for this product, and step 2 starts this week.

If you run a marketplace, recruit the supply side by hand first

A marketplace has two sides to recruit, and one of them has to come first. The Thumbtack team understood that to build a marketplace, you have to begin with supply. For them, supply meant local professionals.

The timeline was short and manual. The two co-founders committed to the project full time in August 2008, and the company launched in 2009 with six people. What it built was a broad base of professionals who are vetted, ranked and filtered. In the early version, a customer had to fill out a form and the pros would write back with price quotes.

Give supply a reason to join before you have demand

Pros had no customers to find on day one, so Thumbtack offered them something useful anyway: a tool to repost their Thumbtack profiles to Craigslist, create more professional-looking ads and pictures, and import reviews. Once customers arrived, the pitch changed. The company could attract pros on the basis of having customers for them.

The same pattern runs through our cases on emailing pros at Thumbtack, Etsy's craft-fair sellers and Airbnb's New York photos. In our reading, the common thread is a founder making the supply side look good enough to buy from. Our tactic pages on recruiting supply by hand and professional photos for supply set out the steps.

Who does it and what you measure

The founders do this outreach themselves. They learn which sellers care, what objections come up and which listings get chosen. Hand it off later, once the script stops changing.

Count supply that responds, not supply that signs up. A listing that never answers a request makes the marketplace look empty to the buyer who tried it. Track every early request and whether a supplier replied in time.

Supply response rate = requests answered by at least one supplier within your target window
                       / total requests received

Weekly log, one row per request:
Date | Category | Area | Suppliers notified | First reply (hours) | Hired? (Y/N)

A worked example with our own illustrative numbers: you hand-recruit 40 cleaners in one city and receive 20 requests in week one. In this example, if 12 requests get a reply within your window, your response rate is 60%. The fix is rarely more sign-ups. Call the cleaners who stayed silent, ask why, and recruit in the categories and areas where requests went unanswered.

This week, pick one category and one city, list the suppliers you will contact by name, and start the log before the first request arrives.

Should you charge your first customers?

Yes, in our view, and earlier than feels comfortable. A free signup tells you someone was curious. A payment tells you the problem cost them enough to spend money on it, and that gap matters most while you are counting your first ten.

The same guide says you will then have nascent but real goodwill to turn into references and referrals. We think a payment is what makes that goodwill believable to the next prospect. A paying reference has already defended the purchase to a budget holder. A free user has not.

Charge the right people. The guide advises selling to someone who has bought similar things before, because that buyer knows how to get a purchase approved inside their company. It also notes you don’t need to have a perfectly polished product or superior sales skills to count to 10.

Two ways to charge before you feel ready

• Pre-orders. Sell access before the product is finished and build against the commitments. Our pre-order case and the pre-selling tactic walk through it.
• A paid tool from day one. Ship a narrow tool to an audience that already pays for software and skip the free tier entirely. See the paid-tools case.

Subject: Founding customer offer for [product]

Hi [name],

You mentioned [problem] when we spoke. We are building [product] to fix it.
Founding customers pay [price] for the first [period], locked for [term].
In return you get a weekly call with me and a say in what we build next.
If it does not solve [problem] by [date], I refund you in full.

Want in? Reply yes and I will send the invoice today.

When a free tier makes sense instead

Situation	Charge now or free tier (our reading)
The user is also the buyer and has paid for similar tools	Charge now
A user loves it but someone else signs the check	Free tier for the user, charge the team
Value appears only after many people join	Free tier until the network fills in
Users will never pay but bring in people who will	Free tier for those non-buyers

A free tier works for people who will never be your buyer but who spread the product to those who will. Our free tier for non-buyers tactic covers how to set one up, and the SaaS pricing models guide covers what to charge once you do.

How to know your first customers are the right ones

If you get to 100 customers, that proves people will sign up. It does not prove they are the same kind of people, or that they stay. Run this check before you scale any route from the earlier sections.

Start with a narrow target user

Write one sentence naming the user: role, company type, the trigger that makes them look. At the start you may not have an intuitive sense for your market and ideal customer profile yet, so treat the sentence as a guess you will test against every account.

Watch for a skewed early base. LinkedIn's first-week signups were virtually all of the people who signed up in the first week were part of the startup ecosystem (so, predisposed to try out new products). Early adopters are useful, but a base made only of friends and tinkerers can hide the fact that your real buyer has not shown up. Tag each account as "target" or "other" from day one.

Check that the problem repeats

The signal you want is that you start to hear recurring themes about their needs. After each call, log the problem in the customer's own words. If the top problem appears across most target accounts, you have a pattern. If every account wants something different, you have a services business.

Measure activation and week-four retention by cohort

Define activation as the first action that delivers the core value, such as a first invoice sent or a first file shared. Group users into a weekly cohort by signup date, then track the share of each cohort that activated and the share still active in week four. Read the shape of the curve with our retention curve guide and compare your rate against our activation benchmarks.

If they are mostly "other," rewrite the target sentence.

Add a product-market fit survey

Once a cohort reaches week four, send the product-market fit survey to active users only. Ask how they would feel if they could no longer use the product, and what main benefit they get. Cross the answers with your tags.

The fit-check template

Copy this into a spreadsheet, one row per account.

Account | Signup week | Source route
Target-user match (Y/N) | Why / why not
Problem in their words | Matches top problem (Y/N)
Activated (Y/N) | Date of first core action
Retained week 4 (Y/N)
Paid (Y/N) | Plan / amount
Referred someone (Y/N) | Who
PMF survey answer | Main benefit named

Weekly summary:
Target accounts this week: __ of __
Target accounts activated: __
Target accounts retained week 4: __
Target accounts paid: __
Target accounts referred: __

A 30-day plan to your first 10 paying customers, and the number to watch

This is our method. The founder sends every message and takes every call. A first marketing hire, if you have one, builds and cleans the list. The rates below are a worked example, not benchmarks. Replace them with your own after week 1.

Work backward from 10 paid

10 paid       <- half of trials convert
20 trials     <- half of calls start a trial
40 calls      <- 2 of every 3 replies book a call
60 replies    <- 20% of messages get a reply
300 messages

Screen every name before it enters the 300

A name goes on the list only with three yeses:

1. Bought a similar tool before.
2. Has the trigger problem now: a new hire, a failed audit, a tool being retired.
3. Can approve the spend without a committee.

Two yeses go on a later list.

Write the first message, then follow up

Lenny's business-to-business research found cold outbound works when done creatively, and that "For your first set of customers, trust is the key." In our reading, homework and a free useful thing are how a stranger earns that trust.

Hi {name}, I saw {specific detail from their site, post or job ad}.
Teams like yours often lose {time or money} to {problem}.
I put together {useful thing}. It's yours either way.
Would a 15-minute call on {day} be worth it to see if we fix this?

Day 3:  "Attaching the {useful thing} in case it helps."
Day 7:  "One question: is {problem} still on your list this quarter?"
Day 14: "Last note. Reply 'later' and I'll check back in 90 days."

Filled in, for a 12-person dental practice:
Hi Dana, I saw you're hiring a front-desk manager.
Practices your size often lose hours to rebooking missed appointments.
I put together a one-page no-show recovery script. It's yours either way.
Would a 15-minute call on Thursday be worth it to see if we fix this?

When someone says yes

Do not send a link afterward. The trial starts before you hang up, and you book a day-7 check-in in the same call.

The four weeks, cumulative

Week	Founder's work	Sent	Replies	Calls	Trials	Paid	Pass mark (our reading)
1	Screen 300 names, send 100	100	20	8	4	0	20 replies
2	Send 150, run day-3 and day-7 touches	250	50	25	12	2	25 calls
3	Send last 50, run day-14 touches, day-7 check-ins	300	60	40	20	6	20 trials
4	Convert open trials, ask each payer for one referral	300	60	40	20	10	10 paid

Trials start on the call, so they track calls in the same week. Paid lags trials by about a week. In this example, if week 1 misses its pass mark, rewrite the message before sending the next 150.

The number to watch

After week 4, track weekly growth in paid customers. If you start from 10, the time to 100 depends only on that rate:

weeks to 100 = ln(100 / 10) / ln(1 + g)

g = 5%   ->  about 47 weeks
g = 10%  ->  about 24 weeks
g = 15%  ->  about 16.5 weeks

Plug your own rate into the growth model template. Log each change to the message or screen in the experiment backlog.

What this page does not cover

This page is for you if you are working toward your first 10 to 100 customers. It does not cover paid acquisition, hiring a sales team or scaling channels beyond that stage. It is not financial or legal advice, including on pricing, refund terms or contracts with early customers.