Spent Saturday and Sunday going through my outreach list from the last 30 days. Not the messages. Not the niches. Not the emails I sent. Just one column: how many employees each business had. Took me 4 hours. Worth every minute. Here's what I found: Every prospect that actually replied had between 5 and 30 employees. Every prospect that ghosted me was either a solo operator or a company with 100+ people. Zero exceptions in 200+ contacts. So I went back through the "First Win" posts in this community to see if the pattern held outside my own list. It did. Joel Rivera's first client → real estate broker, 10 agents, scaling to 30. $25k pipeline. Yagyang Shah's small win → £1k UK website client, small team behind it. Meanwhile, most "still no clients after 100 cold emails" posts I read? Solo plumbers, individual agents, family shops with 1-2 people. When you slow down and actually think about it, it makes complete sense: · Solo business owner → They ARE the operations. They don't buy AI. They work another late night. · 100+ employee company → Has IT, legal, procurement. You'll get stuck in "exploration call" hell for 6 months. · 5–30 employees → Enough chaos to need automation. Enough cash to afford it. And the owner can say yes on the call — no committee, no legal review, no vendor onboarding. The wild part — this changes almost nothing about your work. Same niche. Same voice agent demo. Same proposal. You just point your existing message at a slightly different list. Now I want to stress-test this, because my sample is only my own data. Two quick asks for anyone reading: 1. If you've closed a paying client in the last 90 days — drop their team size in the comments. Rough estimate is totally fine ("around 15," "20-ish," whatever you remember). 2. What niche were they in? Real estate, restaurant, dental, trades, ecommerce, something else? If 10+ people share their numbers below, I'll come back next week and post the full breakdown so we can all see if the 5–30 rule holds across the whole community.

I did it! That was super fun. It was my first time using Claude Code to build something and I was extremely impressed. One thing I learned is that it's much easier to not give up because if you get stuck, you can always ask AI. One thing I'd improve about the newsletter on the next iteration is adding a step where I can review the newsletter before it gets sent. Another would be using GHL instead of Gmail to send so I can send to more than just one person. I would have the automation push a draft to GHL where I could then choose the audience and send.