170 cold emails. 0 replies. Here's the ICP mistake I made — and what finally worked.
I've been building Velor — an AI customer support agent for small B2B SaaS teams — for the past several months. The product works. Getting first customers has been harder.
Here's what I got wrong.
The mistake: targeting up-market
I got excited about logos. Sent cold email to founders at Rippling, Gusto, Deel, Mercury — 50+ emails to companies with 100+ employees. Zero replies. Then I tried the "50-200 employee" band. Still zero.
The issue wasn't the pitch or the copy. It was the target. Companies that size have:
- Dedicated support/CS teams
- Intercom contracts already signed
- Procurement processes that take months
- A "head of CX" who deletes cold email
170 emails, 0 replies confirmed what I should have known from the start: these weren't my buyers.
What actually works: 2-10 person founder-run teams
When the FOUNDER is personally answering support tickets, the dynamic is completely different:
- They feel the pain every single day
- They make purchasing decisions in hours, not months
- A $500/mo tool that saves them 4 hrs/week is an obvious yes
Our first trial customer: a 2-person fintech company. The founder was spending 4 hours a day in their support inbox. Velor deflected 70% of those tickets in week one without a single wrong answer.
The product:
Velor trains on your docs and answers customer questions grounded in what's actually in your knowledge base — not hallucinated answers. It also takes real actions (looks up orders, escalates tickets, files disputes) instead of saying "please contact support."
If you're a 2-10 person team with paying customers and you're the one answering support questions: velorplatforms.com → "speak with an agent" (no signup, live demo, 30 seconds).
Happy to talk or share more about what I've learned. Posting here partly because IH is full of exactly the founders I'm looking for.
The founder-signal problem. Large companies already outsourced the pain. When you're addressing a 50+ person company, support is someone else's job - they're not motivated to change because it's not their time cost. But a founder answering support tickets themselves? That's immediate, felt pain. The 2-10 person founder-run band works because the pain is personal and the solution is obviously worth its weight. This is why broad cold outreach fails - you're fishing in a sea of people who don't actually feel the problem. Narrowing to "founder answering their own support" filters for founders who need this today, not someday.
The part that stands out is not the company size, it is who is in the inbox. A founder answering their own tickets feels the cost every day, so they buy fast. Someone managing a team just files it as a backlog item. I have been doing direct one on one onboarding with early founders for a similar reason, when the buyer is also the user you skip the whole internal approval chain entirely. Would be curious whether your win rate holds once a couple of these 2 person teams grow to 10 or 15, since that is usually where the founder stops touching support directly.
The contrast between 170 unanswered emails and the 2-person fintech is pretty stark.
Have you seen the same response from several other 2–10 person teams yet, or is that first trial customer still carrying most of the evidence behind the ICP shift?