Quit dev job for AI SMB platform. Landed 1st customer. Unexpected.

funnelhacker

Eighteen months ago I walked away from a steady dev gig. No more salary, no more predictable pipeline. I went all in on an agentic AI platform for Australian SMBs - not the chatbot garbage you see everywhere, but actual agents that handle voice, chat, WhatsApp, and tie into existing business systems.

The ROI thesis was simple: small businesses here bleed revenue every time a call goes unanswered. Medical clinics, tradies, restaurants - they're losing jobs daily because they can't scale human reception. The US vendors ignore AU compliance and accents. That's a gap worth exploiting.

What caught me off guard wasn't the tech - it was the market education cost. I spent months cold emailing and barely moved the needle. LinkedIn DMs and in-person conversations were the only channels that converted. The objection wasn't price or performance; it was pure disbelief that AI could actually do this now. That's a CAC problem I didn't model.

First paying customer signed: a Sydney medical centre using our AI voice receptionist. Hearing it handle real patient calls in production? Surreal. That single customer flipped my entire product philosophy. Stop building for hypothetical users, start optimising for the one who's already paying.

Sales lesson I'd hammer home for anyone doing this: skip the 'interesting, maybe later' crowd. Focus on the desperate ones - businesses that feel the cost of missed calls in their P&L every week. They educate themselves. The curious segment is a 10x harder sell and burns your runway.

Happy to nerd out on voice AI deployment, no-code agent setups, or the AU SMB go-to-market. Metrics-first.

emailwiz

Oh mate, I've been there - not with building the platform myself, but with the whole "first customer high" followed by the cold reality of who actually buys. Your point about desperation is spot on. I spent years nurturing leads for a SaaS that helped local service businesses, and the ones who signed up fastest were the ones who'd just lost a £500 job because their phone rang while they were elbow-deep in a drain. Small clinics, solo trades, that one-woman-show hotel receptionist - they bleed money every time they miss a call. The "oh interesting" crowd? They'll take six months of follow-ups and three case studies to even book a demo. Not worth your energy right now.

And that bit about market education building a moat - underrated as hell. Every independent GP practice in Sydney that's heard your pitch is now primed for the category. They might not buy today, but when their mate down the road starts using an "AI receptionist", guess who they'll call? You've already paid the cost of teaching them what it is. That compounds. Keep going after the ones who can't afford to ignore the problem.

cpaoptimizer

Love seeing this kind of grit. Quick question on your ICP - I'm assuming these are non-technical users? If so, how straightforward is it for them to describe, build, and deploy an agent? Getting the customer-specific context right is the hard bit, and that usually falls on the customer to configure. How's the onboarding flow handle that? And when things glitch in production (they always do), what's the support chain look like for someone who can't debug a logs file?