The AI Adoption Framework MSPs Can Run (and Get Paid For): The AI Steering Committee
Every MSP sales team has that one prospect. The one who's been telling you no on managed IT for years. Now picture a different call. Instead of pitching managed IT for the fifth time, you ask one question: "Who is helping you with AI?"
Notice how that's worded. It assumes they've already got it handled, which gives them room to correct you: "Well... actually, we're kind of winging it right now." That answer is your opening. I've watched it turn a years-long "no" into a 90-day AI pilot.
AI is getting MSPs into rooms that used to be locked and the right AI adoption framework keeps you there. Your existing customers are bringing it up too - and most MSPs either answer it one ticket at a time (while bleeding margin) or just say no. Meanwhile, some AI consultant shows up and takes the seat that should have been yours.
And when a customer does bring AI to you, they've usually already built a Rube Goldberg machine in their head - one giant, do-everything system that's impossible to scope, price, or deliver. Your job is to take that machine apart. Break it into smaller, manageable pieces, and every piece becomes a project you can actually bill for - one that delivers immediate customer value and trust.
An AI steering committee gives that conversation a home: one meeting a month where your customer's own people decide where AI should help, and you run the process.
I walked through how to set one up and sell it on episode 14 of Inside the Pipeline. Here's the playbook, plus the kit to run it.
Key Takeaways
- An AI steering committee turns your customers' scattered AI questions into one monthly meeting. The customer owns it. You run it.
- Your job is facilitator, not AI expert. Keep the meeting on track, translate the nerd words, and keep every improvement small enough to actually finish.
- The committee surfaces problems worth paying to fix - and every one of them is a small project you can quote.
- The same offer opens doors with prospects who keep saying no to managed IT.
How to Talk to Customers About AI When You're Not the Expert
Here's the good news: you don't have to be the AI expert to lead your customer's AI conversation. You have to run the room where they figure it out.
An AI steering committee is a small group from the customer's business - an executive sponsor, an internal champion, and several people close to the work who feel the pain. The group meets monthly, picks one improvement at a time, and checks whether it worked.
I learned this running an IT steering committee for one of my largest customers, a nonprofit with about 120 employees. I didn't have to be the expert. They knew where their technology was letting them down. They knew their problems.
My job was to keep the meeting moving and handle all the nerd words. Think of it like bumper bowling: the customer throws the ball, and you keep it out of the gutter.
The Facilitator Playbook in the kit splits the work three ways:
- Customer leadership decides priorities and resources.
- The committee discovers, recommends, and measures.
- The MSP educates, facilitates, and supports the work.
That split lets you say yes to AI without having to own every answer. It also pulls AI work out of personal accounts. If an employee's best workflow lives in their personal $20 ChatGPT subscription, it walks out the door the day they do.
If you've already stopped telling customers no, this is the structure that comes next. Most of the AI adoption challenges I see come back to the same gap... nobody owns the direction.
How to Set Up an AI Steering Committee
Setting one up comes down to four decisions. The AI Steering Committee Charter in the kit walks your customer through each one.
1. Make sure the customer is ready
Start with a customer where you've already earned a business conversation. The Facilitator Playbook looks for four signs:
- The relationship is there.
- The friction is visible.
- People are already experimenting with AI.
- You can identify a corporate sponsor to drive adoption from the top.
No sponsor or no business priority yet? Hold off on the recurring meeting and work on those first.
2. Name an executive sponsor
The sponsor ties the committee to what leadership actually cares about and carries its recommendations back up the chain. Without one, good ideas die at "sounds interesting."
At Empath, my co-founder Wes plays this role for us. He made AI adoption a company direction, and the team organized around it.
3. Keep the team small and close to the work
You need a chair to run the meeting and a few people who do the actual work. I call them builders. One of the builders in my example was an entry-level sales rep who was already demoing AI tools to the rest of the company.
Size it to the customer. For an 11-20 person company, the charter suggests the owner plus one to three people. For a business of 3-10 people, skip the full charter and use the Small Team Starter.
4. Put the charter in front of the sponsor
Bring the charter with the customer's logo on it. It sets a six-month mandate and spells out who owns what. The last step is getting a 60-minute kickoff on the calendar.
A charter with their name on it turns "we should probably do something about AI" into a team with a start date.
What Happens Inside the AI Adoption Framework Each Month?
Every monthly meeting runs the same four-step loop. The charter calls it the AI Adoption Flywheel: Discover, Educate, Uplift, Measure.
5. Run the flywheel on one problem at a time
Every MSP has lived through this one: imaging Windows machines. First it was Norton Ghost and a stack of CDs. Then network boot. Then somebody figured out you could bake the RMM agent and Office right into the image.
Each round was one turn of the flywheel. You spotted the slow part (Discover), learned a better way (Educate), changed the process (Uplift), and checked the time you saved (Measure). The committee runs the exact same loop on AI. At the end of each pass, the team decides whether to keep the change, improve it, expand it, or kill it.
6. Start with problems, not tools
The workbook prompts do most of the discovery work for you:
- What gets done twice?
- Where are people waiting?
- What does only one person know?
The most useful question in the playbook is even shorter: "Walk me through how that works today."
One MSP I talked to hands the customer a deck of AI idea cards and has them sort every card into a keep-it pile or a chuck-it pile. It gets people talking about their own work fast.
Pick the smallest fix someone would actually notice - like a short morning read-ahead for an executive. For more on starting from the customer's workflow, see how MSPs should think about AI enablement.
7. Talk for less than a third of the meeting
Here's the hard rule: keep your speaking time under one-third of the hour. If you present for most of the meeting, it turns into a webinar... and the customer stops bringing you ideas.
A good meeting sounds like the customer talking about their own work.
What an AI Steering Committee Earns Your MSP
An AI steering committee pays you twice: a monthly fee for running it, and the projects that come out of it.
The retainer. My suggested starting point is about $500 a month to facilitate one committee meeting, sold as a 6-12 month engagement. Once your team has run a few, plan on roughly two hours of work per customer per month. This is a fee for a standing meeting - not a block of hours that rolls over. Price it and sell it as a service. Don't give it away as free consulting.
One MSP I spoke with is also seeing customers in AI engagements renew at a higher rate than the rest of their book.
The projects. The committee doesn't promise anything. It surfaces problems, and leadership decides which ones to fund. In one example, four sales reps were each spending two hours a day cleaning lead lists. That's 40 hours a week... and a very easy fix to quote.
The best moment comes later, when someone on the committee tells you they built something themselves and need you to make it permanent.
Small projects I'd start with:
- Dashboards that pull reporting from several systems
- Lead list building and cleanup
- First-draft email replies
- Call recaps that turn into tasks
- Contract review against the customer's past contracts
- A brand kit for Word and PowerPoint
The way in. This is how you sell AI services without leading with a tool. Here's my challenge to every sales team: go back through the deals you lost this year and ask each one, "Who is helping you with AI?" Assume they've got it covered and let them tell you otherwise. A prospect who keeps saying no to managed IT may still say yes to that.
Over time, the committee can grow into a broader managed AI agreement once the customer has built enough to need ongoing support. For more on where the margin in AI work actually comes from, see Stop Reselling AI.
Where Customer Training Fits in the Flywheel
Training is the Educate step of the flywheel, and it's where a lot of AI improvements either stick or stall.
The workbook includes a Team AI Learning Plan for this step: who needs to learn what, a relevant example, how they'll learn it, and who owns it.
Here's an example. The committee decides a bot in Microsoft Teams would save one department hours every week. Then it turns out half that department barely uses Teams. So the first project is Teams training - the bot comes after.
That's the gap between AI access and AI readiness. A customer can own every license and still not have people who know how to use the tools well.
Empath Grow gives you a way to deliver that training to your customers' employees. You assign courses and pathways, track completion, and bring progress back to the committee meeting. The catalog covers Microsoft 365, cybersecurity habits, data handling, and everyday technology alongside AI, so it fits whatever the committee picks next.
When the committee can see who finished the training, the Measure step has something real to report.
For more on setting a training baseline for customer teams, see AI training for employees.
Start With One Customer
The first committee is the hardest one to start, so keep it small. Pick one customer this month where the relationship is strong and the friction is visible, and put a charter in front of them.
Download the AI Steering Committee Kit for the charter, facilitator playbook, and workbook.
Want to hear me walk through the whole model? Watch the October 22 episode of Inside the Pipeline. The series is built for MSP sales and account management teams, and upcoming episodes cover turning Q4 budgets into next year's wins and handling objections. Join the next one.
Frequently Asked Questions
Do we need to sell an AI assessment before a steering committee?
No. The assessment can happen inside the committee's first few meetings as part of discovery. The recurring facilitation is what you're selling, and starting there gets the customer moving sooner.
Who on our team should facilitate the committee?
Usually a vCIO, account manager, or owner who's already having business conversations with that customer. My advice to sales reps: open the door, then hand the meeting to whoever owns the relationship. The Facilitator Playbook covers the role, including what to do when meetings start to stall.
Should we sell a monthly AI engagement before we've delivered anything?
Fair concern - customers hesitate to pay monthly for something they haven't seen work. That's why the committee is designed around one small, visible win in the first cycle. The kickoff doesn't end until one improvement has an owner and a review date.
Does the steering committee write the customer's AI policy?
No. Under the charter, the committee can recommend changes to company rules, but leadership approves them. Anything involving sensitive data, significant spending, or a broad rollout goes to leadership first.
Want to see how Empath Grow fits into the Educate step of your customers' AI steering committees? Book a demo.
