Your first customers won't come from ads or SEO. They come from you, doing things that don't scale. The goal isn't growth yet. It's proof, plus the lessons you only get from real users paying you. This is the whole point of Paul Graham's essay Do Things that Don't Scale, and it is the phase most technical founders try to skip.
Go to where they already are
Don't build an audience from scratch on day one. Find the watering holes where your future customers already gather and complain about the problem: subreddits, Slack and Discord groups, forums, niche communities. Be genuinely helpful there for weeks before you mention your product. The early Airbnb team went door to door in New York to sign up hosts and improve their listings by hand. That is the energy you want.
Do things that don't scale
- Hand-recruit users one at a time. DMs, emails, in person. Ten great conversations beat ten thousand impressions.
- Onboard each customer personally. Get on a call and set it up for them. When someone agreed to try Stripe, the Collison brothers would ask for their laptop and install it on the spot. Watch where people get stuck.
- Over-deliver for a tiny number of people. Word of mouth starts with a handful of users who can't believe how good the experience was.
Mine your network honestly
Skip "buy my thing." Try "I'm building something for people who struggle with X. Do you know anyone like that I could talk to?" A referral to the right person is worth more than a sale to the wrong one.
Charge from customer #1
Free users give polite, useless feedback. Paying users tell you what is actually broken and what they would pay more for. Even a small price filters for people who genuinely feel the problem. When 37signals built Basecamp they charged from launch and covered their first-year revenue target within about six weeks.
See also the Bullseye traction framework for what to do once these manual tactics start working.