Why your MSP pricing model is leaving money on the table

When margins get tight, the instinct is to look at costs. Cut a tool, squeeze a supplier, push the team a bit harder. I did all of that when I was running the operation, and it helped at the edges. But the biggest leak was never on the cost side. It was in how we priced.

The model itself was quietly handing money back to clients every month, and no amount of cost-cutting was going to fix a model that was built to lose margin. If your profitability is drifting the wrong way and you can't quite explain it, this is usually why. Not your rates. Your model.

Here are the five leaks I see in almost every MSP I look at, roughly in order of how much they cost.

1.You're still selling time, not outcomes

Most MSPs grew out of break-fix, and the break-fix mindset never fully left. It shows up in the language: "we'll take a look," "we'll sort it out," "we'll just do it as a favour." Every one of those is unpriced work, and unpriced work is free work.

The deeper version of this leak is charging for effort instead of value. When you price by the hour, you are capped by the clock and punished for being good at your job. The engineer who fixes a problem in two hours earns you less than the one who takes six. Your best people literally lose you money.

Clients don't buy hours anyway. They buy uptime, security, and not having to think about their IT. Price the outcome, not the input.

2.Per-user pricing that ignores cost-to-serve

Per-user (or per-device) pricing is clean and easy to sell, which is exactly why it hides so much leakage. A flat per-seat rate assumes every user costs the same to support. They don't. The 15-person law firm with two partners who fat-finger everything and refuse MFA costs you three times what the 15-person design studio does, and you are charging them the same.

Worse, per-user pricing rarely tracks what actually drives your cost: tickets, site visits, after-hours work, the complexity of the estate. So your heaviest, most demanding clients are being subsidised by your easy ones. Over time the easy clients leave, because they are paying for service they don't use, and the painful ones stay, because they are getting a bargain. You end up with a book of business optimised to destroy your margin.

The point isn't that per-user pricing is wrong. It's that per-user pricing with no view of cost-to-serve is flying blind.

3.Tiers that don't mean anything, or no tiers at all

Two failure modes here. The first is one-size-fits-all: a single "IT support" plan, take it or leave it. That leaves money on the table at both ends. The client who would happily pay more for a premium, proactive service can't, and the price-sensitive prospect who would take a lighter package walks away.

The second, more common, is fake tiers. Good, Better, Best where nobody can tell what is actually different, so everyone buys the cheapest and you have anchored yourself low. Real tiers are built around meaningful lines: response times, proactive versus reactive, security posture, strategic input. Each tier should answer a clear question for the buyer, "what do I get if I pay more, and do I want it?"

Done well, tiers do two jobs. They let good clients self-select up, and they anchor the conversation. Put a genuinely premium tier at the top, even if few buy it, and your middle tier suddenly looks like the sensible choice rather than the expensive one.

4.You bundle everything, so nothing has a price

This one is subtle, and right now it is costing you the most. Somewhere along the way, security, backup, patching and increasingly compliance got folded inside "IT support" as things you just do. They became invisible. And you cannot sell, or raise the price of, something the client cannot see.

The clearest example today is security and accreditation. Your clients are being told by their own customers, their insurers and their regulators that they need Cyber Essentials, and increasingly ISO 27001 and real evidence. That demand is landing on your desk. If security is buried inside your base plan as a cost you absorb, you are eating a rising cost and missing a rising revenue line at the same time. Pulled out as a named, priced service, it becomes something you sell, uplift, and differentiate on.

The rule is simple: anything a client would recognise as valuable, and would consider buying separately, should be a line item, not a giveaway.

5.You never re-price

Most MSPs set a client's price on day one and never touch it again. Meanwhile your cost base rises every year: tooling, salaries, the security stack, insurance. If your prices are static and your costs are not, your margin is shrinking on autopilot. And your longest-standing, most loyal clients are usually your least profitable, because they are on the oldest rates.

Two fixes. First, an annual price review built into the contract from the start, with a stated uplift, whether CPI-linked or a fixed percentage. Clients accept this when it is expected and small. They resent it when it is a surprise. Second, a deliberate reprice of the legacy clients who have drifted well below your current rate card. It is uncomfortable, but they mostly won't leave, as long as you frame it around what has changed, the threat landscape, the tooling, the value they now get, rather than "we need more money."

What this actually costs you

Let me make it concrete, because "leaking margin" is easy to nod along to and just as easy to ignore.

And it compounds. Higher margin funds better people and better tooling, which improves service, which supports higher prices. Pricing is not a spreadsheet exercise. It is the flywheel.

How to fix it, in order

You don't fix all five at once. Sequence it:

1. Get visibility of cost-to-serve. You cannot price well while you are blind. Even a rough view, tickets and hours per client per month against what they pay, tells you which clients are underwater. This is uncomfortable and essential.
2. Unbundle the sellable services. Pull security and compliance, backup and disaster recovery, and strategic advisory out as named line items. Give each one a price, even internally at first.
3. Rebuild your tiers around meaning. Two or three tiers, each with a clear reason to exist, and a genuine premium option at the top to anchor.
4. Shift the language from effort to outcome. Fixed scope, clear deliverables, stop quoting hours. Kill the free favours, not by being difficult, but by having somewhere to put them.
5. Add the annual review, then reprice the laggards. Build the uplift into every new contract, then work through the legacy book.

The one to start with today

If you do nothing else this quarter, unbundle and price your security offering. The demand is already there. It is landing on you whether you charge for it or not, and it is the rare pricing change that clients expect to pay for, because they understand why. It is the fastest margin win available to most MSPs right now, and it doubles as a differentiator while your competitors are still giving it away.

If more than one of those stung, your model, not your rates, is the problem. And the good news is that a model is fixable.

Nathan Carroll