Getting your first ten customers is a completely different problem to getting your first hundred, and most of the advice written about customer acquisition assumes you are already past ten.
I want to write about the specific channels I tested in the first six months of building, what the results actually were, and what I concluded from them. I am being precise about numbers because vague claims about channel performance are useless to anyone trying to build.
What I tested and what happened
LinkedIn content. I published two to three posts per week for eight weeks, focused on startup validation topics relevant to my target customer. Total reach across the eight weeks was around 45,000 impressions. Total inbound leads from LinkedIn: four. Total paid conversions from those leads: zero. The engagement was good — lots of likes and comments from other founders sharing their own experiences. None of those founders were in the market for a tool. They were networking.
Cold email to founder communities. I purchased access to two alumni lists from regional accelerator programmes and ran a cold email sequence to 340 founders. Open rate was 31%, which is respectable. Click rate on the product link was 4.2%. Trials started: 11. Paid conversions: 1. The conversion rate was not the main problem. The main problem was that accelerator alumni, by definition, have already passed through a validation stage. Many of them had moved into execution or were raising. I was reaching the wrong moment in the founder journey.
Product Hunt launch. I launched on a Tuesday, which I had been told was optimal. I ranked 7th for the day. Traffic to the product was significant — around 800 unique visitors over 48 hours. Signups: 23. Paid conversions within 30 days: 2. Product Hunt drove a very specific type of user — people who browse launches for interest and novelty rather than people actively searching for a solution to a problem they have right now.
Paid search. I spent £400 testing Google Ads against three keyword clusters related to startup validation, business model canvas, and idea testing. Cost per click ranged from £1.80 to £4.20 depending on keyword. Paid conversions: 3. Cost per acquisition: £133. Not viable at that scale, but the data on which keywords converted was useful for organic content strategy.
Direct outreach in community forums. This is where everything changed. I spent time in several founder-focused communities — specific Slack groups, Discord servers, subreddits — not posting about my product, but reading. I identified threads where founders were describing, in their own language, the exact problem my product solved. I then sent individual direct messages: I saw your post about X. I have been building something that addresses exactly that. Would you be willing to try it and give me honest feedback?
Conversion rate from those conversations to trial: 38%. Conversion rate from trial to paid: 44%. Those numbers are not a typo. The context was perfect — I was reaching a person at the exact moment they were actively experiencing the problem and had just articulated it publicly.
Referrals from early users. The second channel that actually worked was not something I engineered — it emerged. Every founder I converted in the first month referred at least one other founder within six weeks. Not because I had a referral programme. Because the product was solving something they actively wanted to tell their peers about.
What the data says about my growth model
The two channels that worked — targeted direct outreach at the moment of expressed need, and word-of-mouth from satisfied users — have one thing in common: they depend entirely on the product creating genuine value. They cannot be optimised through better copywriting or a higher ad budget. The lever is product quality.
This is both encouraging and constraining. Encouraging because it means the product is doing something real. Constraining because it means growth is bounded by the depth of the user experience rather than the size of the marketing budget. I am spending most of my time on product rather than acquisition for exactly this reason.
At some point the distribution model will need to evolve. But the foundation has to be a product that founders actively want to tell other founders about. Everything else is a multiplier on that. Without it, there is nothing to multiply.