Is generic AI consulting a good business?

AI consulting is the highest-probability way for most people to make money from AI right now. Faster change means more people need help. Generic consulting is the trap.

Why niche wins

When every client is different, every engagement is custom work. Nothing carries over to the next client. Go deep in one vertical and 90%+ of the solution is the same for every owner in it. Then onboarding is configuration, not construction.

That's also where a small builder beats the big AI labs. Anthropic and OpenAI aren't going to build hyper-specific, multi-step workflows for a narrow vertical like local media. It's not worth their time. It's worth yours.

How to pick the vertical

• One level deeper than the category. "Home services" is too broad. Pool service is a niche: fragmented, light on regulation, seasonal.
• Recurring revenue. I spent eight years as an insurance adjuster working with roofers. Roofing has big jobs and almost no recurring revenue, which makes it a hard software customer.
• Bring more than AI. Marketing and Meta ads skills on top of AI are rare among AI-only operators.

What to charge

Position as the full back end, not just automations: marketing, SEO, recruiting help, the works. That supports $3,000 to $5,000 a month. For a productized system that's 95% pre-built, I've priced setup at $2,000 to $10,000 plus $300 to $800 a month to manage it.

How to train a team

Mindset before tools. A hands-on half day with laptops open, building real things, beats any demo. I run it as four touchpoints: an intro call, a prep call the week before, the workshop, and a follow-up one to two weeks later.

The window is now, the next 12 to 24 months, before business owners figure this out and prices compress.