How to find customers for a new startup

How do you find customers when nobody knows you?

To find customers for a new startup, look for people who recently did something because of the problem you solve, write their names down, and contact them one at a time. A recent action is easy to search for. Someone who posted a question about it last week, hired help for it, or complained about a tool is easier to reach than "small businesses that need marketing."

Most advice on how to find customers gives you a list of channels: social media, SEO, ads, cold email, events. A list of channels does not tell you who to contact on Monday. This guide starts one step earlier. You turn your customer description into something you can search for, build a list of 30 people, and then decide on a channel from where those people turned up.

Describe your customer by something they did

"Freelance designers" is a group. You cannot search for it and get a list of people who have your problem. "Freelance designers who asked in a forum how to chase a late invoice in the last month" is a description you can search for, and every result is a person with the problem right now.

The second column always contains a verb in the past tense. That verb is the trigger. It marks the moment the problem cost them enough to do something about it. People who just acted on a problem are more likely to talk about it, because they remember the details.

If you cannot write a sentence like the ones in the second column, you are not ready to look for customers yet. Fill in the first four lines of the go-to-market strategy template first, so you know who the first customer is and what they use today.

Where to find your first customers

Start where the trigger leaves a trace. These are the places worth checking, roughly in the order we would try them:

1. Your own network. Former colleagues, clients, and classmates who match the description. Ask each of them for one introduction to someone else who does. People read a message from someone they know.
2. Communities where people ask for help. Forums, Slack and Discord groups, subreddits, and LinkedIn groups where your customer asks practical questions. Read the questions first. They show you the words customers use for the problem, which you will need for your message and your landing page.
3. Public work. Job ads, project boards, open issues, conference talks, and portfolio pages. A job ad is a strong trigger, because someone is about to spend money on the problem.
4. Organizations your customer belongs to. Professional associations, trade groups, meetups, and alumni networks. The UK government's guide to finding user research participants recommends approaching "relevant professional bodies, charities and community groups" to reach specific kinds of people. The same advice works for a startup with no audience.
5. Events. Small events where your customer goes to learn something. Talk to people about their work, not your product.

Paid ads and SEO are missing from this list on purpose. They help once you know which message gets a reply. Before that, you are paying to find out that your message is unclear. Search traffic also takes months to build. A list of 30 names takes an afternoon.

Build a list of 30 names before you write a message

Make a spreadsheet with one row per person and these columns:

Name | Where I found them | What they did (the trigger) | Date of the trigger | Contacted on | Reply | Notes

Fill in the first four columns for 30 people before you contact anyone. Two things happen while you build the list.

• You find out whether the customer exists in reachable numbers. If an afternoon of searching turns up five people, the description is too narrow, or the problem is rarer than you assumed. Both are useful to know before you write a line of code.
• You see which place produces the most matches. If 20 of your 30 names come from one community, that community is your first channel. You found the channel by counting people, not by guessing.

The trigger column also keeps you honest. A row without a trigger is someone you hope has the problem. Leave them off the list.

Contact people one at a time

Write each message to one person and mention their trigger: "I saw your question about chasing a late invoice." Then ask about their experience, not about your idea. "How did you handle it in the end?" gets more useful replies than "Would you use a tool that does this?" Our guide to customer discovery interview questions covers what to ask once someone agrees to talk, and our guide to writing a cold email that gets replies covers the first message.

This is slow on purpose. Paul Graham wrote in his 2013 essay Do Things That Don't Scale that "the most common unscalable thing founders have to do at the start is to recruit users manually." The same essay describes the Airbnb founders "going door to door in New York" to meet their hosts. Your first customers come from conversations you start yourself. Once you have had enough of them, you will know what a channel that scales needs to say.

Two rules make the outreach easier to learn from:

• Send the same message to everyone in one batch. If you rewrite it for every person, a reply tells you nothing about the message.
• Write a decision rule before you send. For example, "If 5 of 30 reply and describe the problem in their own words, offer a call. If 1 or 2 reply, change the first line. If nobody replies, change the customer description." These numbers are an example, not a benchmark. Pick numbers that would change what you do next.

A worked example

Here is the method applied to a hypothetical product. The numbers are illustrations, not results.

A founder is building a tool that drafts follow-up emails for consultants after they send a proposal. The searchable description is "independent consultants who asked in the last 60 days how to follow up on a proposal." An afternoon of searching two consulting communities and LinkedIn turns up 34 people. Of those, 22 come from one Slack community, so that is the first channel.

The founder sends the same short message to 30 of them, mentioning each person's question. After ten days, the spreadsheet shows 30 sent, 7 replies, and 4 replies that describe losing a deal because of slow follow-up. Under the rule written beforehand, that is enough to offer calls. Two of the replies also say the real problem is writing the proposal, not following up. That note goes into the next test instead of being forgotten.

Keep the list attached to the decision

After a few rounds, the hard part is not finding people. It is remembering which description you used, which message you sent, what the rule was, and what you decided. The spreadsheet holds the names. The reasoning usually ends up in your head.

We are building Boring Marketing Loop to keep that reasoning in one place. In the workspace, you answer nine Customer & offer questions, and each answer stays marked as a guess until people respond. A direct outreach test records who and where, the message, the sample size, an end date, and the signal you wrote down before you started. When you record the result, including silence and objections, you choose to continue, change direction, stop, or gather more evidence, and the next test is planned from that decision.

Marketing Loop is in development and not open yet. If you want to follow along, join the build updates.

FAQ

How many people should I contact to find my first customers?

Start with about 30. That is enough to see whether replies come from a description or by luck, and few enough that you can write to each person yourself. It is a working size, not a statistical sample, so treat the result as a direction rather than proof.

Should I offer my product for free to get the first customers?

A free offer is fine for learning, but it tells you little about demand. If you can, ask for something that costs the person a little: a call, an introduction, a deposit, or a pre-order you can actually fulfill. Each of those shows more than a free signup does. Our guide on how much to charge for your first product covers setting that first price.

Can I find customers online without an audience?

Yes. Communities, public posts, job ads, and open issues are all online, and none of them need an audience. You need a searchable description of the customer and the patience to contact people one at a time. Our guide on how to market a startup with no money covers the other free channels.

When should I start with SEO or ads?

Start once a message reliably gets replies from direct outreach. At that point you know the words customers use and the offer they respond to, and you can put both into a page or an ad. Before then, direct outreach shows you more in a week than a new page will in months.

Start with the trigger

Write one sentence that describes your customer by something they did recently. Spend an afternoon finding 30 people who did it, and write down where each one came from. When the list is done, turn it into a written test with the marketing experiment template and send the first message.