Stop building. Start DMing. The first 100 users come from 1,000 personal messages, not 1,000,000 impressions. Paul Graham called it "do things that don't scale." Eight years later it's still the only reliable playbook for going from 0 to 100.
Why manual outreach works (and ads don't yet)
You don't know your messaging. You don't know your ICP. You don't have proof. Paid ads only work when all three are dialed in. Until then, every dollar you spend on ads is just buying noisy data. Manual conversations buy you signal: the exact words your customers use, the objections that kill deals, the angle that makes someone lean forward.
The 4-channel split
- 40% Twitter/X DMs. Reply to public posts where someone describes the problem you solve. Then DM with context. Never lead with a pitch.
- 30% LinkedIn. Cold message decision-makers with one specific observation about their company. Two sentences max. No "synergy."
- 20% Reddit / Discord / Slack. Find the three communities where your customer hangs out. Contribute for two weeks. Then mention your product only when relevant.
- 10% Email. Personal, short, signed with a real name. Subject line: "quick q about [their company]."
The script that works
Hey [name] — saw your post about [specific thing]. I'm building [product] which does [specific outcome]. Not pitching — would love your gut reaction to the landing page if you have 60 seconds. Here it is: [link]. Either way, good luck with [their thing]. Three things make this work: a specific reference (proves you're human), a tiny ask (60 seconds), and an exit ramp (not pitching). Reply rates are 30–50% when sent to the right people.
Volume math
- 20 DMs/day × 5 days = 100 DMs/week.
- 30% reply = 30 conversations.
- 30% of those try the product = 9 signups.
- Repeat for 11 weeks = your first 100 users. It's boring. It works.
The trap
Around user 30, you'll be tempted to "scale" by sending the same message to 500 people via a tool. Don't. The moment your DMs look automated, the reply rate falls to 1% and your X account gets flagged. Stay manual until you have product-market fit. Then automate.
What to do during the conversation
- Ask what they were using before.
- Ask what almost made them not reply.
- Ask what they'd tell a friend about your product in one sentence.
- Write down their exact words. That's your next landing page headline.

Founder of Rocket. Writing about how independent developers and vibe coders can build, launch, and grow useful apps.
All articles by Alex →