1. Idea and validation
  2. Waitlist
  3. Private beta
  4. Launch
  5. First 100 users
  6. First 1,000 users
Direct outreach

Cold outreach to your first 20 users

Find 20 specific people who have the problem, write to each of them individually, and offer to set them up yourself.

What's requiredOne day to build a list of 50 names, then about an hour a day for two weeks.
Accounts
Your personal email, plus a personal LinkedIn or X account with a real photo.
Assets
A working product or beta, a two-sentence description, a Calendly or Cal.com link, a spreadsheet to track messages.
Time
One day to build a list of 50 names, then about an hour a day for two weeks.
Cost
Free.
Prerequisites
A product people can use today, even if rough.

What to do

  1. Write a one-line definition of who qualifies. Role, situation and a problem they have right now, for example “freelance bookkeepers with 10+ clients who chase receipts by email every month”. If you cannot name the situation, you will write to people who do not care.
  2. Search for people who complained about the problem in the last 30 days. Recent posts mean the pain is fresh and they still remember writing it. Good places to search:
    • X search: "[problem phrase]" min_faves:2 -filter:replies with a date filter such as since:2026-09-01, or phrases like "is there a tool that", "anyone know how to", "so tired of" plus your topic.
    • LinkedIn: search posts (not people) for the problem phrase, filter by “Past month”. Comments under those posts are full of people with the same problem.
    • Reddit: search the niche subreddit sorted by “New” for the problem phrase or a competitor name.
  3. Mine competitor review sites for unhappy users. On G2, Capterra, Trustpilot, the Chrome Web Store or app store reviews, read the 2- and 3-star reviews of the tool people use today. Reviewers often show their name, role and company. Someone who wrote “great but no way to X” is the best lead you will find if you do X.
  4. Build a list of 50 names in a spreadsheet. Columns: name, where you found them, the exact thing they said with a link, how to reach them, date sent, reply, follow-up date, outcome. You need about 50 names to end up with 20 conversations and a handful of active users.
  5. Write each message by hand, under 80 words. Say where you saw them and quote their words, name the problem, say what you built in one sentence, and ask for one small thing. Send it from your personal account with your name, never from a company page or a no-reply address. See the templates below.
  6. Ask for something small and specific. “Want me to set up an account for you?” or “Can I show you in 15 minutes?” gets more yes than “Would love your feedback”. Offer to do the setup yourself: import their data, create their first project, send a ready-to-use link.
  7. Send 10 a day, not 50 at once. Replies come in waves and you need time to answer fast. Answering within an hour or two while they are still thinking about it makes a big difference.
  8. Follow up once, after four or five days, then stop. Reply in the same thread with one new thing: a screenshot, a short Loom, or “I set up a demo account with your use case, here’s the link”. Mark the row as closed if there is no answer.
  9. Onboard every yes yourself and log what they ask. Book them into a call (see onboarding calls) and write down every question and stuck moment in your spreadsheet. The same questions from three people is your next fix.

Templates

For someone who posted about the problem:

Hi Maria, saw your post last week about chasing clients for receipts every month (“I spend two days on this”). I built a small tool that sends the reminders and collects the files into one folder. It’s early and free while in beta. Want me to set it up for your clients? Takes me 10 minutes, you just check the result.

Sam

For someone who left a competitor review:

Hi Tom, read your review of [Competitor] on G2, the part about not being able to export by client. That’s exactly what I’m building around. Would a 15-minute look be useful? If it doesn’t solve it, I’ll tell you what I’d use instead.

Follow-up, once:

Quick follow-up: I recorded a 2-minute video of it doing the receipt chase for a sample client: [link]. Happy to set it up if useful, no worries if not.

Why it works

People answer messages that prove you read what they wrote. Quoting their own words, from a post they made recently, shows the message is not a mass email. Offering to do the setup removes the effort of trying something new, and watching them use it tells you what to fix faster than any survey.

How it gets users

Slowly and one at a time. Reply rates for personal, well-targeted messages are usually low, often 5–20 percent, higher when you quote a recent post. From 50 messages, expect a handful of conversations and maybe 3–10 people who actually use the product, over one to three weeks. It does not scale, which is fine: these first users give you the feedback and the testimonials for everything that comes next.

Common mistakes

  • Sending the same template with only the name changed. People can tell.
  • Pitching features instead of naming their problem in their words.
  • Writing to people who do not have the problem, because they looked like the right job title.
  • Sending from a company account or a bulk-email tool instead of your own inbox.
  • Following up three or four times. One follow-up is polite, more is spam.
  • Asking for “feedback” instead of asking them to use it for a real task.

Next steps

This guide is connected to these guides on the map. Numbered ones come next.

  1. 103 Private beta · CommunitiesRecruit beta testers in niche communitiesPost in the communities your users belong to, describe the problem plainly, and ask for a small number of testers.
  2. 203 Private beta · Direct outreachOnboard every beta user on a callSet up the first users yourself on a short call and note where they get stuck.
  3. 305 First 100 users · Direct outreachEmail every new signup personallySend each new user a short personal email asking why they signed up, and reply to every answer.