MSP pricing models, explained

An MSP pricing model is the contractual framework that determines how a managed services provider charges for ongoing technology support, monitoring, and management. The structure dictates not just what you pay each month but what the MSP is incentivized to fix proactively versus what it costs them to resolve. That distinction separates a good pricing structure from one that quietly works against you.

Most mid-market companies encountering MSP contracts for the first time assume pricing is mainly about the number. It is not. A per-device contract that looks cheaper than a per-user one can be more expensive within 18 months if your device-to-user ratio rises. An all-in flat fee that looks like it covers everything may carve out the three categories of work your environment generates most.

Understanding the structure lets you compare proposals on equal terms, ask the right questions during contract negotiations, and build a realistic multi-year technology budget. It also tells you something about the MSP itself: providers who explain their pricing clearly and walk you through what is excluded tend to behave the same way when things go wrong. If you are evaluating whether your current or prospective MSP is the right fit, Seven Roots works with mid-market companies on vendor-neutral technology strategy and MSP oversight.

Per-user pricing charges a flat monthly fee for every supported user in the organization, regardless of how many devices that user runs. Typical mid-market ranges run $75 to $150 per user per month, though the spread is wide depending on what is included and how technically complex your environment is.

At most MSPs, a per-user agreement covers helpdesk access, remote monitoring of devices, security patching, and some baseline endpoint protection. The appeal is straightforward: headcount is easy to count and changes predictably, so budgeting is clean.

Where per-user pricing works best is in companies with stable headcount and a reasonably uniform device environment, one laptop and one workstation per employee, for example. The math is simple and the monthly bill is predictable.

Where it gets expensive is at the edges. Part-time employees who need occasional helpdesk access still count as a full user in most contracts. Contractors who use company systems typically do too, even if they work two days a week. Conference room systems, shared workstations, and lobby kiosks sometimes trigger per-user fees even when no individual user is assigned. Read the user definition clause before you sign, because that paragraph determines more of your total cost than the headline rate.

Per-device pricing charges a flat monthly fee for each managed endpoint: desktops, laptops, servers, and sometimes network devices, printers, and mobile devices depending on the contract scope. Mid-market ranges typically run $25 to $75 per device per month, with servers billed separately at higher rates, often $75 to $150 per server per month.

This model fits companies where devices outnumber users: manufacturing floors with shared workstations, healthcare environments with dedicated clinical terminals, or labs where equipment outruns headcount. In those settings, per-device pricing is often cheaper than per-user because you are paying for hardware endpoints rather than people who may touch many devices.

Where per-device pricing creates friction is in modern, mobile-heavy organizations. Employees who carry a laptop, a tablet, and a company-issued phone represent three billable endpoints under most per-device agreements. BYOD policies compound the problem: if personal devices connect to company systems and receive any monitoring or management, some MSPs count them. Defining exactly which devices are in scope, and which are out, is the most important negotiation in a per-device contract.

The model also disadvantages companies that are growing fast. Adding headcount means adding devices, and each new hire triggers an immediate billing increase with no softening mechanism like a user band or volume discount.

All-in flat-fee pricing sets a single monthly number that covers all users and devices up to a defined headcount or device band. For mid-market companies, this structure typically runs $10,000 to $40,000 per month depending on company size, environment complexity, and what the MSP actually includes in the fee.

The appeal is real. Budget-conscious organizations, particularly those with finance teams that dislike variable monthly bills, find flat-fee contracts easier to model. You know what you pay every month. Quarterly planning is simpler. There are no per-seat or per-device true-ups at renewal.

But the phrase "all-in" is the part that deserves scrutiny. In practice, all-in almost never means all-in. Nearly every flat-fee contract carves out major projects: network infrastructure overhauls, office relocations, new-site buildouts, merger integrations, and application migrations. These are billed separately, typically at a project rate or a high hourly rate. The monthly fee covers steady-state operations, not change events.

The other common exclusion is out-of-scope hardware. If your flat fee covers management of existing devices, a refresh cycle where you replace 80 laptops may generate a new-device onboarding fee that does not show up in the base contract. Read the exclusions list, not just the headline number.

Co-managed technology pricing is structured differently from the three full-outsource models. Here, the MSP handles specific tiers of support and specific infrastructure functions while an internal technology employee or small team retains ownership of strategy, vendor relationships, and escalation decisions.

Billing structures for co-managed arrangements vary more than the other models. Some MSPs charge a partial flat fee covering defined service areas, typically $3,000 to $8,000 per month for tier-1 and tier-2 helpdesk support plus infrastructure monitoring. Others bill on a blended hourly rate of $125 to $200 per hour with a monthly retainer for priority access. A few use a hybrid: a lower per-user fee, often $40 to $75 per user, covering monitoring and patching, plus hourly billing for everything above that.

This model works well when a company has one or two internal technology employees who are strong on operations but need reinforcement at the infrastructure or security layer. It also fits organizations that have outgrown a fully outsourced arrangement but are not yet ready to hire a full team. The internal employee stays in the role and owns the vendor relationships; the MSP provides the capacity and specialized depth they cannot cover alone.

What does not work is ambiguity about who owns what. Escalation paths, on-call responsibilities, and decision authority need to be written into the agreement from the start.

No pricing model is categorically cheapest. The right model depends on your headcount-to-device ratio, project volume, and how your environment changes over time. A manufacturing company with 150 employees and 400 devices will almost always pay less under per-device pricing. A professional services firm with 150 employees and 160 laptops will usually prefer per-user. Here is how the models stack up across the key variables that matter to mid-market buyers.

Plan a 12-month refresh review for any MSP agreement. Pricing bands, headcount thresholds, and market rates all shift. An agreement that was competitive at signing may be significantly above market by year two.

The pricing model is the frame. The hidden costs are what fills it in. Three patterns show up across all four models and catch buyers off guard more consistently than anything else.

The first is the project hourly rate. Every contract has one, and very few buyers ask about it before signing. Whether your agreement is per-user, per-device, flat-fee, or co-managed, project work above the steady-state scope gets billed at an hourly rate. For many MSPs, that rate runs $150 to $225 per hour. A mid-size network refresh, a server migration, or a new-location buildout can generate tens of thousands of dollars in project billing that no monthly fee covers.

The second is the annual escalator clause. Buried in contract boilerplate, these clauses automatically increase the monthly fee by 2 to 5 percent each year at renewal. They are often tied to a CPI index or to an MSP's internal cost increase figure. Most buyers do not notice them until year two.

The third is "reasonable use" language in all-in contracts. That phrase gives the MSP discretion to reclassify unusual support volumes as out-of-scope project work. If your team opens significantly more tickets than average during a system transition, you may receive a project invoice alongside your flat monthly fee.

If you want a structured read on your current or prospective MSP agreement before signing, Heartwood can walk you through the contract terms and flag the clauses worth negotiating.