I’ve been watching a pattern play out in MSP after MSP lately, and it’s worth naming. The owner or managing director is genuinely switched on about AI. They’re up early on a Saturday tinkering with prompts. They’re subscribed to half a dozen newsletters. They’ve run a pilot on quoting, or proposal drafts, or ticket triage. They can tell you, with real confidence, what GPT-5 does differently to Claude.
Then Monday morning rolls around.
You walk past the techs’ desks and someone is printing a spreadsheet out to “have a proper look at it”. The account manager is manually copying fields from one system into another. The service coordinator is re-reading a long email thread for the third time trying to figure out what the client actually agreed to. Not one of them has opened an AI chat today. Maybe not this week.
The leader has moved. The business hasn’t.
The solo-operator trap
This is where most MSP AI journeys quietly stall. The owner is thinking AI-first. The team is still thinking the way they thought in 2022. And because the owner is the one doing all the experimenting, they can tell themselves a comforting story — we’re on it, we’re ahead of the curve, we’re investing in AI. On paper, yes. In the business, no.
Real adoption isn’t measured by how many prompts the boss has saved or how many pilots are running. It’s measured by what an average Tuesday looks like for the people doing the work. If the first instinct when someone hits a hard problem is to ring a colleague, send an email, or open a spreadsheet — AI hasn’t arrived yet. It’s just a hobby the owner has.
That’s a confronting thought, but it’s the honest one.
You are the coach now
The shift that moves the needle isn’t another tool or another pilot. It’s a change in your job description. At the next team meeting, you stop reporting on AI and start teaching it.
Walk your people through what you tried this week. Show them the prompt that didn’t work, then the one that did. Show them the output that saved you forty minutes on a scope. Let them see you thinking out loud. You don’t need to be an expert — you need to be visibly in motion. That’s what gives them permission to start moving too.
If you’re the most AI-literate person in the building and you keep it to yourself, you’re not leading. You’re collecting.
Make AI the front door
Here’s the non-negotiable I’d put in place this week, and it costs nothing. Every person in the business sets an AI chat — Claude, ChatGPT, or Gemini, pick one — as their browser homepage. Not Google. Not the intranet. Not the weather.
Every single time someone opens a browser, an AI chat window is the first thing they see. A blinking cursor, waiting for a question.
It sounds small. It’s not. Most of the friction stopping people from using AI isn’t capability, it’s habit. They forget it’s there. A homepage removes the remembering. It puts the tool under their nose, dozens of times a day, until asking it first stops feeling like a new behaviour and starts feeling like the normal one.
The real alignment test
So here’s the question I’d sit with. If I walked into your office on a random Tuesday and watched your team for an hour — not you, them — would I see an AI-first business, or would I see a business with an AI-first owner?
If the answer isn’t the same for both, that’s your next piece of work.
CIAOPS 