
I coded two apps for my phone while watching TV. And I am, genuinely, the least technical person you'll ever meet. If I can do it, imagine what YOU could build and how much your clients would pay for it. Here's the ABC framework for turning AI into revenue.
Hands up who wants to make more money from AI. Oh... okay... so it's everyone then? 😃
I have a small MSP mastermind group that meets here in the UK once a year. And several members of that are killing it with AI right now.
One guy in particular is very smart and very switched on with turning AI into a revenue stream. Don't worry, I'm going to be collaborating with him on a bunch of content soon because I think you need to know what he knows.
But for now let me tell you a little bit about the ABC framework that we've put together. It's a super simple way to remember the three key ways to make money from AI right now.
A: Assess
This is where it starts. Sit down with a client and properly figure out what’s actually happening with AI in their business right now.
The big thing to look for here is shadow AI. Most business owners have never considered the implications of staff putting business data into their own personal AI tools. You should be talking about this with all of your clients. Not only can you set them up with Copilot properly but you can block other AIs and put an acceptable AI policy in use.
B: Build
This is where it gets genuinely exciting. Because, thanks to Claude Code, you now have a software development arm.
Think about how many of your clients are paying a subscription for a piece of software that’s costing them $$$$$$ every year. Or they're tolerating workflow inefficiencies that you could easily code away for them.
If you've never got round to coding with Claude, honestly it's so easy. I am the least technical human on the planet and I've coded two apps for my phone. I did it while I was watching TV. Genuinely not boasting, here. If someone like me can do it, imagine what someone with your skills could do.
Would your client pay you $5,000 or more to own a piece of software outright, permanently, rather than renting it forever? For a lot of them, the answer is yes. And that’s before you factor in the recurring revenue you then generate from maintaining it, securing it, and keeping the codebase up to date.
I reckon you've got about 18 months, tops, to do as much of this kind of work as you can for your clients before the world shifts and everyone realises they can just do it themselves. Or there's a bunch of 18-year-olds selling it to your clients instead.
C: Consult
This is the big-ticket item, and it’s the one most MSPs jump to first when they think about making money from AI.
Consulting is strategic. It’s about the client’s growth plans and where AI fits into them over the next one, two, three years. You know significantly more about what AI can and can’t do than almost any client you work with.
You might feel slightly embarrassed selling a few hours of your time and charging a big fee for it but you've got to think in terms of the outcome they'll get from it, not the time.
It's like the old parable... I'm sure you've heard this one before...
A factory's critical machine broke down, halting all production, so the owner brought in a top expert who inspected it briefly and fixed it instantly with a single, light tap of a hammer. When the owner received a $10,000 bill, he was furious and demanded an itemised invoice, asking, "Why is it so expensive for just a few minutes of work and a tap?"
The expert sent back a new invoice that read: $1 for tapping the machine, and $9,999 for knowing exactly where to tap.
So. The ABC framework... which of those do you think it's going to be a big new revenue stream for you?

