0→100 is manual labor, not marketing. Everything scalable is for later; right now, do things that don't scale:
1. Go where your users already complain. Reddit threads, Discord servers, Facebook groups, forums where people describe the exact problem you solve. Don't spam links — answer the actual question and mention you built something for this. 10 genuinely helpful comments beat 100 link drops.
2. Direct outreach, personalized: 30 DMs/emails a day to people showing the symptom publicly. 'Saw your post about X — I built something for that, want a look?' Expect 10–20% response at this scale, which is enough math for 100 users.
3. One launch moment: Product Hunt, Hacker News 'Show HN', a relevant subreddit. Won't change your life; will bring your first strangers and unfiltered feedback.
4. Onboard the first 30 personally — a 15-minute call each. Feels absurdly inefficient; is actually your highest-value activity: they convert to advocates, and you watch where the product confuses them in real time.
5. Then ask every active user: 'who else do you know with this problem?' A working referral ask at n=50 beats a referral FEATURE at n=50.
Milestone check: if you can't get 100 users manually, the problem is positioning or product — and better marketing would only have hidden that longer.