K7 Insights
MSP Org Chart: The Seats to Add at $2M, $5M and $10M
Part of Grow Your MSP
Key takeaways
- At about seven to 10 employees, hire a real service manager: someone who runs the margin on the service and how fast it reaches clients.
- On the way to $5M, split managed services and projects into two teams under one operations leader, and add a tier three escalation seat with a six-to-nine-month end date.
- Between $5M and $10M, add middle managers before it feels comfortable: pods of about $3M in revenue, a central tech team, a controller. A lot of MSPs stall at the same revenue numbers when they wait.
- A true COO takes about $10M to $12M in revenue. An outside CEO fits around $15M to $20M, once you have the cash to pay one.
An MSP org chart should change at a few predictable revenue points, and each change moves a job off the owner. The first seat to add is a real service manager, at about seven to 10 employees. A true COO usually waits until $10M to $12M. In between sit the middle managers, and that’s where a lot of MSPs stall: those seats don’t bill, so the owner puts them off, and growth stops at the same few revenue numbers. Every one of those hires is a line in the plan to grow your MSP.
What does an MSP org chart look like at $2M, $5M and $10M?
It looks like the owner holding fewer boxes at every stage. The headcounts are for MSPs of median profitability in each band (the middle half by profit), from Service Leadership’s 2026 Annual IT Solution Provider Industry Profitability Report (page 272). The headcount charts are only in the full report, but its executive summary is free. Here’s where the seats usually land.
| Revenue | Median headcount | Seats to add | What the owner still holds |
|---|---|---|---|
| Around $2M | 10.5 people ($1M to $3M) | A service manager over tier one (reactive tickets) and tier two (escalations and site visits). | Sales and finance |
| On the way to $5M | 18.6 people ($3M to $6M) | An operations manager over two teams: a managed services manager and a pro services manager who runs projects too. A tier three seat with a six-to-nine-month end date. The first sales hire, usually a BDR who books the meetings while you keep closing. | Closing deals, finance and the top of operations |
| $5M to $10M | 34.5 people ($6M to $9M) | Pods of about $3M in revenue, each with a player-coach lead. A central tech team for tools and vendors. A controller. | Sales and the top of operations |
| $10M and up | 43.1 people ($9M to $12M) | A VP of operations or COO. An outside CEO around $15M to $20M. | The CEO seat, until an outside CEO makes sense |
Two things about those headcounts. The top 25% of MSPs by profitability run leaner: 15.9 people in the $3M to $6M band against 18.6 at the median. The biggest step in the table is between two bands: 18.6 people at $3M to $6M, then 34.5 at $6M to $9M.
Why do MSPs stall at the same revenue numbers?
Because the owner waits too long to hire middle managers. I see MSPs hit walls at the same handful of numbers, about $2.5M, $4M, $5.5M, $7.5M and $11M, and a lot of those ceilings exist because nobody hired the next layer of management fast enough.
The hesitation is usually about money. The owner’s head goes something like this: if I bring in another manager who isn’t billing customers or working tickets, my bottom line drops, and EBITDA is the most important number in the world. So the owner keeps carrying the team. Then nobody has room to manage the next seven to nine people, and the business runs into a capacity ceiling. Attrition does the rest. Employees leave, or clients do, and you spend the next year refilling the same seats.
A manager is one step back for two steps forward. You take the hit to margin now so the business has room to grow into the seat.
Service Leadership’s numbers describe the same trade. Service management grows from 2.9% of headcount under $1M to a peak of 6.6% between $6M and $10M (page 160). The report calls $1M to $6M the investment phase: adjusted EBITDA sits around 10% to 10.4% while MSPs add sales and service leadership, then recovers to 13.6% at $10M to $15M as that investment pays off (page 161). If you’re building the budget for that hire, the MSP business plan shows how to set headcount against gross margin and hold back profit for growth.
How do you draw an MSP org chart before you name a single seat?
Start with three parts before you write a single title: find work, do work, get paid. A mentor told me early on that every business comes down to those three things. You find the work, you do the work you found, and you get paid for the work you did. I’ve built org charts off that idea ever since.
When I run it with a leadership team, it takes four flip charts. The first is a funnel cut into three parts. Then there’s one chart each for find work, do work and get paid, and the team calls out everything that part of the business requires, plus the KPI that proves each thing got done. Only after the walls are full do we name departments: sales, service, finance, whatever the labels end up being.
The exercise kills the urge to split too early. Somebody wants pro services as its own department. Is it doing work? Then it goes in the do-work chart. At 12 people, that’s the whole answer.
There’s a fourth box most owners never draw: who owns the funnel itself. That’s internal IT, your tools, your vendors and the products you build, the work that supports the work. In an MSP I call it the technology department. When I was operating an MSP, the vendor phone calls were what ate my time, and that’s what an empty fourth box feels like.
Which seats do you add around $2M?
Start with a strong service manager. A strong, strong, strong one makes a world of difference. You’ll want one at about seven to 10 employees, the point where the owner can’t be the person everyone calls anymore, and you’ll probably go through two or three of them on your way to $5M or $6M.
Hire for the right job. Managed service managers manage services. Their work is the margin on the service you deliver and how long it takes to get that service into a client’s hands. Picture a product manager at a software company asking whether the product is profitable, efficient and something people want. The best ones are really good at getting people to do their jobs. Writing your PowerShell scripts belongs to somebody else.
That’s why promoting your best tier three is the classic mistake. Managing and engineering are different career paths. The newly promoted tier three sees escalations stacking up and a script nobody’s written, and does them himself.
Under the service manager, keep the tiers simple. Tier one is anything reactive: triage and quick fixes. Tier two takes the escalations and the site visits. Whenever you add dispatch, put it in its own seat.
A good dispatcher gathers the information, cleans up the ticket and decides how urgent it is without diagnosing anything, so the seat stays cheap. The most beautiful question a dispatcher can ask is “We can get someone right now for you. Is this a good time?” Half the time the client says tomorrow works, and a tech just got to finish what he was doing.
Sales and finance stay with you at this size. You don’t want the people doing the work holding the company credit card and the invoicing too.
Which seats do you add on the way to $5M?
Split the work into two teams. For a $3M or $4M MSP, I build it as an operations manager or director with two people under them: a managed services manager and a pro services manager. There’s always a magic third, too. One client of mine runs an Azure department. Another runs an AV line.
Two teams, because they make money differently. Managed services runs on recurring revenue and gets staffed for steady growth. Projects swing. They go up, they go down, and your labor cost has to match that curve. Project engineers should be 95% to 100% billable, since a project team has very little non-billable work to do. Once each side carries its own gross margin, an owner can think clearly about both.
You probably don’t have enough project work yet for a dedicated project manager, so save the money. Hire a pro services manager who’s willing to wear the PM hat.
On the managed side, add a real tier three escalation seat and give it an end date. Tier three is expensive, and you need one anyway until your documentation catches up. For the first six to nine months, that person checks ticket quality, approves knowledge base articles and writes scripts. I want them basically working themselves out of a job. Leave the seat in place past that and everything starts getting escalated to it.
Send new clients to the project team until they’re onboarded, because a new client will drown your help desk. I used to route every new client to the project team only and bill every help desk hour to the onboarding project. That showed me the true cost of onboarding, and it gave the project team a reason to finish. Once the client was documented and cleaned up, the project team handed it off to managed services, so those techs weren’t stuck spending an hour on a password reset because nobody had cleaned up Active Directory yet.
Your first sales hire is usually a BDR (business development rep) who books the meetings while you keep closing. How to hire your first salesperson covers when you’re ready and what to pay.
Which seats do you add between $5M and $10M?
Middle managers, and more of them than you think you can afford. Service management peaks as a share of headcount at $6M to $10M (page 160). A rule of thumb I’ve heard caps a manager at 15 direct reports, so once the recurring team passes that, it needs structure under the managed services manager.
I break the recurring team into pods. A pod is a book of clients with its own tier ones, tier twos and a lead. Region, niche or subject matter all work, so pick whatever suits your book. The point of a pod is making each team’s world smaller. The lead is a player-coach, part senior engineer and part mentor, who answers for the profitability of the pod.
When a pod gets to about $3M in revenue, start asking whether it’s too big. A really efficient small MSP is about $3M, so that’s a big practice.
Some models put an account manager and a vCIO inside every pod. I pull them out into a central revenue team, because technical people are bad at managing quotas and somebody has to manage those quotas. If you’re working out what that advisory seat should cost and do, the vCIO job description covers when it pays for itself.
Staff the fourth box now. A central tech team owns your tools, your vendors, your RMM, your scripting and your product. I gave my central tech team a dashboard gauge that tracked how much money they saved the company. Beat a vendor down by $0.20 an endpoint, multiply that by your endpoints, and that’s your tech team’s number to brag about.
The project side changes once there’s enough project work to pay for it. At $3M or $4M, the pro services manager wears the PM hat to save money. Later there’s budget to go higher, and the same dual seat goes to a senior hire: a project lead (the MSP I ran called the seat its PMO) who is the pro services manager and the project manager at once, and a player-coach too.
It’s an expensive seat. Mine had his PMP, a project management certification. I was paying him almost $150,000 a year, and that was years ago.
Somewhere in the $7M to $10M range, bring in a controller or an accounting manager. There’s too much going on day to day for an outsourced bookkeeper to keep up with the strategy, and someone’s got to be on the inside getting that data to the decision makers. I see a lot of MSPs get stuck between $5M and $10M because their middle managers never get the numbers they’d need to make decisions without the owner.
This is the stretch I know from the inside. I ran an MSP from $2M to $20M in under four years as its operator, a true 10x, and it runs near $50M today on the same structure I built. I capped pods at about $3M. San Diego got so big I split the county in half, north and south. My service manager carried 15 direct reports, and each pod ran about 15 people.
I got one seat wrong. I let my tier three engineers do projects, and once I’d done it, I wished I could pull them back out. Keep yours on escalations and documentation.
The last piece was the numbers. We ran an open P&L, so anyone in the company could see it. The goal was that on the 4th of every month, once the books closed, my managed services director would get a P&L by class: every dollar of cost and revenue in managed services. He could see where his team overspent, where it under-delivered and where we gave away too many credits, and decide what he’d change if he owned the place.
Then I moved to a profit-sharing model, so part of his pay came from the margin he managed. The data plus a reason to move it is how we got through the $10M ceiling.
If you’re somewhere in this band and the chart on paper doesn’t match how the work actually moves, book a call. I’m habitual with whiteboards, and we’ll draw the next version of yours together.
Who do you hire at $10M and after?
A true head of operations. Finding a real VP of operations or COO, or whatever you want to call the seat, takes about $10M to $12M in revenue. Below that, an operations director over the two teams carries the load and the owner holds the top of operations.
Pay your directors like owners of their departments. Your service and ops directors get a variable piece of their pay tied to the gross profit of the department they run, which only works if they can see that number every month. That’s the same P&L by class and profit share from the last stage.
The seat after that is the CEO. Around $15M to $20M, I usually recommend hiring an outside CEO, especially when it’s unclear which owner is really in charge. It solves the ownership problem real quick, and by then you have the actual cash to do it.
What org chart mistakes cost MSPs the most?
Structure in the wrong place, usually one of these four:
- Too many departments, too soon. I’ve watched an owner split service into specialist teams, each with its own supervisor, until the KPIs had to get so detailed that the results got blurry. Flatten it back to the two teams, managed services and projects.
- Techs shared between service and projects. Keep the two teams apart. The one exception is tier twos who occasionally go out to sites. Project labor cost has to follow the project curve, and a managed services tech who keeps getting borrowed for installs makes both teams’ numbers wrong.
- Security reporting to the help desk manager. If the people who check the work share a manager with the people who do it, that manager can sweep a help desk mistake under the rug. The only way I can have pure accountability is if they do not have a shared manager.
- A ticket board for every box on the chart. Owners mirror the org chart in the PSA, and the reporting gets worse. There’s actually better tracking when you don’t change boards, because when everything happens on a single board, SLAs don’t reset.
The next seat is usually one that doesn’t bill
Your next hire is usually a seat that doesn’t bill, which is exactly why it’s still yours.
If you’re not sure which seats you’re in, answer four questions. Who does the help desk come to when a ticket’s stuck? Who takes the vendor calls? Who decides whether a new client starts with the project team? Who reads the managed services numbers after the books close? If the answer to any of them is you, find that seat on your row of the table. That’s the next hire.
Getting out of those boxes, one hire at a time, is a big part of growing your MSP past the next ceiling. If you want a second opinion on which hire comes first, bring those four answers to a call.
- What does an MSP org chart look like?
- Under the owner, three groups: finding work (sales and marketing), doing work (service and projects) and getting paid (finance and admin), plus a technology team that owns tools and vendors once you can staff it. Around $2M the doing-work side gets a service manager. By $3M to $4M it splits into a managed services team and a project team under one operations manager or director. Between $5M and $10M the managed side breaks into pods of about $3M in revenue each, every pod with its own lead, and a controller joins finance.
- When should an MSP hire a service manager?
- At about seven to 10 employees, the point where the owner can't be the person everyone calls. Hire someone who manages the service: the margin on it and how long it takes to reach a client. Promoting your best tier three usually backfires, because managing and engineering are different career paths. Expect to go through two or three service managers on the way to $5M or $6M.
- When does an MSP need a COO?
- Finding a true VP of operations or COO takes about $10M to $12M in revenue. Below that, an operations manager or director runs the managed services and project teams, and the owner holds the top of operations. Once the COO is in, the next leadership seat to plan for is an outside CEO, which I usually recommend around $15M to $20M.
- How many employees does a $5 million MSP have?
- MSPs of median profitability with $3M to $6M in revenue had 18.6 employees, according to Service Leadership's 2026 Annual IT Solution Provider Industry Profitability Report (page 272). The top 25% of MSPs by profitability in that band ran with 15.9. The median climbs to 34.5 people at $6M to $9M.
- How many direct reports should an MSP service manager have?
- A rule of thumb I've heard is 15 direct reports, and the service manager at the MSP I ran as its operator had 15. Past that, break the managed services team into pods, each with a player-coach lead who answers for the pod's profitability.
- Should an MSP separate managed services and projects?
- Yes, once you're around $3M to $4M. At 12 people, keep them in one team. After that, split them under one operations leader, because the two make money differently: managed services grows with recurring revenue, while project work rises and falls and its labor cost has to follow that curve. Give each its own gross margin, and keep the two teams from sharing technicians. The one exception is tier twos who occasionally go out to sites.
Go deeper
The rest is in the book.
The sales boxes on this chart get their own book. Selling Without You builds the revenue side from the first hire up: the sales manager, the BDR, the account executive and the account manager, the order you hand them off and how you pay each one. It's free. Drop your email and I'll send it, then send the toolkit that goes with it.