I never set out to build an agency. I'd hired enough of them over the years to know how that story usually ends. As a founder, I'd watched agencies miss obvious moves, charge for templated thinking, and disappear behind retainers without ever feeling like they were on my side of the table.
GTMLab happened by accident.
A couple of founder friends asked for growth help — the kind they couldn't get from any agency they'd hired. I'd co-founded Peeba through Y Combinator, raised $10M, scaled it to $20M+ in revenue across six countries. Before that I'd been a VC partner writing $500K cheques into 50+ AI startups around the world. The growth math was something I'd lived from both sides. So I helped. No proposal, no scoping doc, no agency theatre.
The referrals didn't stop. One founder became three. Three became thirty. Thirty became a hundred-plus specialists across Asia in two years — entirely bootstrapped. The shape of GTMLab emerged from what the demand actually wanted, not from a slide I'd drawn on a whiteboard.
Everyone is scratching the surface with AI
We're still in the very early innings of what AI does for growth. The capabilities are real: real-time audience analysis, rapid creative iteration, predictive modelling on cohort behaviour, content generation, attribution stitching across channels. None of this is theoretical anymore.
But here's what I see when I open most companies' AI stack: someone has a ChatGPT tab open. Maybe they generate variants of an ad headline once a week. Maybe they're playing with Midjourney for thumbnails. That's not AI-powered marketing — that's a productivity hack.
Real AI integration is structural. It runs across strategy, creative, performance, and reporting as a connected system. It's not an add-on someone in the team uses occasionally; it's how the work gets done. That's what we've spent the last two years building.
And the thing nobody wants to talk about
AI has very real limitations, and I'd argue I've seen more of them than most. I've sat on the other side of the table funding 50+ AI startups. I've watched models do impressive things and I've watched the same models hallucinate, loop, and produce confident output that misses the actual point of the question.
AI cannot do these things, today or for the foreseeable future:
- Establish brand positioning that actually resonates in a specific market
- Read emotional context in a customer interview
- Understand your stage, your cash situation, your competitive dynamics, the politics of your team
- Know what to optimise toward — the model can optimise; humans choose the objective
- Recognise when a "winning" output is actually wrong for the business it's serving
A Formula 1 car is the most advanced machine on the track. Without a great driver, it's just an expensive way to crash into a wall.
The companies winning with AI right now aren't the ones with the most sophisticated stack. They're the ones who pair AI with operators who know which problems to point it at — and which outputs to throw out. We call this Artificial Intelligence × Human Expertise, and we treat both halves as equally non-optional.
What we actually spent the last two years building
Three interconnected things, in the order they had to be built:
- Proprietary AI tooling for marketing. Not ChatGPT wrappers. Custom systems for the specific workflows growth teams actually need — signal monitoring, brief generation, creative critique before launch, attribution stitching. The tooling lets small senior teams operate at the scale of much larger agencies.
- Human-AI workflows that actually work. Two years of figuring out where AI leads with human validation, where humans lead with AI support, and where AI just gets in the way. Most teams skip this work and bolt AI onto existing processes. That's why their AI stacks don't compound.
- Integrated teams. Operators who understand both marketing and engineering — running creative, performance, and strategy as one connected system rather than three siloed deliverables. AI as a force multiplier across all three, not a chatbot sitting next to one of them.
The growth partner I never had
When I was running Peeba, I burned through agencies and failed hires. The pattern was always the same: strategies that sounded right in a deck and collapsed the moment they had to ship. Recommendations that ignored my cash position. Senior pitches followed by junior execution. The thing I wanted — a partner who thought like an operator, lived for revenue outcomes instead of retainers, and would tell me the truth — didn't exist in the form I needed it.
GTMLab is the firm I wish I'd had. Built for founders specifically. Operator mentality. Revenue accountability. AI integration not as a marketing line but as a real structural advantage we've put two years into.
The gap between companies that use AI and companies that perform with AI is widening every quarter. We're on one side of that gap, and that's the side I want every founder we work with to be on too.
If that's the partner you've been looking for, you know where to find us.