What Service Fees Should an IT or MSP Company Charge?

The job these fees are actually hired to do

Most managed-services contracts are priced to cover a predictable, steady-state workload: patching, monitoring, backup verification, a known volume of routine tickets, and a support relationship that behaves roughly the same in month nine as it did in month three. That is the part of the business the per-seat or per-device number is engineered around, and when a provider prices it well, it produces a real but thin operating margin — the kind that survives a normal month and evaporates in an abnormal one.

The problem is that a meaningful share of an MSP's actual labor is not steady-state. New-client onboarding is the clearest example. Standing up a new company means discovery of an environment nobody documented, deploying agents to every endpoint, taking over domains and tenants from a predecessor who may or may not cooperate, importing users, rebuilding backup jobs, correcting whatever the last provider left half-finished, and absorbing a spike of "how do I" tickets during the first six weeks as an unfamiliar user base learns a new help desk. That work is front-loaded, expensive, and completely uncorrelated with the monthly seat count. An MSP that folds onboarding into the recurring rate is financing its own client acquisition, and the faster it grows the worse the cash position gets.

So the job a service fee is hired to do is narrow and specific: capture the labor that the recurring contract was never sized to absorb, and do it in a way the client can see coming. That framing matters, because it separates a legitimate fee from a surcharge. A fee that funds identifiable work — a dispatcher who answers at 2 a.m., a procurement coordinator who spec's and tracks hardware, a project engineer running a migration — is defensible in a renewal conversation. A fee that exists to backfill an underpriced contract is not, and clients eventually notice.

Four fee categories do most of the work in the channel:

• Onboarding / setup. One-time, charged at signature or in two milestones (signing and go-live). Covers discovery, documentation, agent deployment, tenant takeover, and the first-30-days ticket surge.
• After-hours and emergency support. An hourly premium, commonly 1.5–2× the standard rate, applied to work outside the contracted support window. Holidays often carry a higher multiplier still.
• Project and change-order fees. Fixed-price or T&M charges for anything the MSA defines as out of scope — a migration, an office move, a compliance buildout, a firewall replacement, a new-site standup.
• Procurement handling. A 5–15% markup or a flat coordination fee on pass-through hardware and licensing, covering quoting, sourcing, receiving, staging, imaging, asset tagging, and RMA management.

There are two secondary categories worth considering once the first four are clean: per-incident overage on plans that cap ticket volume, and compliance or attestation fees for clients in regulated environments who need evidence packages, annual reviews, or auditor support that a general support contract does not include.

The contribution margin on all of these is unusually high — commonly in the 80–95% range — for a structural reason, not a magical one. The technician, the dispatcher, and the PSA license are already on the books, paid for by the recurring contract. When an after-hours ticket bills at a premium, the incremental cost is the labor hour itself plus whatever shift differential you pay; the overhead is already carried. That is the entire economic argument for the fee schedule: it converts already-committed capacity into revenue that funds the next hire.

How the fee schedule fits the RevOps stack

Fees are not really a pricing decision. They are a systems problem, and this is where the RevOps discipline earns its keep, because a fee that lives in a quote but not in the ticketing rules will be discounted into nonexistence by the second month.

Trace one after-hours incident end to end. A client calls at 11 p.m. The ticket has to be created with a work type that the PSA recognizes as after-hours. The technician's time entry has to inherit the premium rate automatically — if a human has to remember to change a dropdown at midnight, the premium is lost roughly half the time. The agreement on the client record has to say that after-hours is excluded from the covered block, or the billing engine will silently absorb the hours. The invoice has to show the line clearly enough that the client's controller doesn't open a dispute. And the general ledger has to book that revenue to its own account so the owner can see, at year end, exactly what the fee schedule produced.

Break any one of those links and the fee exists on paper only. The most common failure is the third one: an MSA that says "reasonable after-hours support included" while the quote says after-hours bills at a premium. The contract wins that argument.

The upstream side matters just as much. Sales has to quote the fee, not apologize for it. If your proposal template lists the onboarding fee in a footnote, reps will treat it as the first thing to trade away when a deal stalls, and you will discover — usually a year later — that the fee has a 60% realization rate and nobody decided that. The fix is structural: make the fee a non-discountable line, or route any waiver through an approval, the same way you'd handle a discount on the recurring rate.

The reporting layer at the bottom is the part most providers skip, and it is the part that makes the fee schedule improvable. If onboarding, after-hours, project, and procurement revenue all land in one undifferentiated "services" account, you cannot answer the only questions that matter: which fee is actually collecting, which one is being waived away in sales, and whether the after-hours premium is covering the cost of staffing the shift. Separate accounts — and separate the hardware pass-through cost from the handling margin — and those questions answer themselves from the P&L.

Pricing, engagement models, and typical ranges

There is no universal price sheet, and anyone who hands you one is selling something. What does exist is a set of structural conventions the channel has converged on, plus a method for landing your own numbers inside them.

Onboarding. Two common structures. The first is a multiple of the first month's recurring fee — often something in the range of one to three months of MRR, scaled to environment complexity. The second is per-endpoint or per-user: a flat amount per workstation and a higher amount per server or per tenant. The multiple-of-MRR approach is easier to quote and self-scales with client size; the per-endpoint approach is easier to defend line by line when a prospect pushes back. Providers moving upmarket tend to drift toward per-endpoint because enterprise buyers want the arithmetic.

Whichever you pick, price it against the actual hours. Track the engineer time on your next three onboardings — including the ticket surge in weeks two through six, which is the part everyone forgets — and set the fee so it covers that labor at something close to your standard rate. If the number that produces is unsellable in your market, that is real information: it means your onboarding process is too manual, and the answer is automation and templating, not a fee you can't collect.

After-hours. The 1.5–2× multiplier on your standard hourly rate is the channel convention, and it exists because it mirrors how overtime works on the cost side. Define the window explicitly in the MSA — "outside 8:00 a.m. to 6:00 p.m. local time, Monday through Friday, excluding published holidays" beats "after hours" by a wide margin. Consider a separate, higher holiday rate, and consider a minimum billing increment (one hour, or two) for callouts, because a 10-minute 2 a.m. ticket still costs you a woken technician and a ruined night.

A structural alternative worth knowing: sell an after-hours *block* rather than billing incident by incident. Clients who genuinely operate at night — manufacturing with a second shift, logistics, healthcare, hospitality — often prefer a predictable monthly adder that includes a defined number of after-hours hours, with overage billing above it. That converts a spiky fee into recurring revenue, which is worth more per dollar at valuation time.

Project and change orders. Fixed-price is better for both sides when the scope is genuinely known: a mailbox migration, a firewall swap, a workstation refresh. Time-and-materials is better for discovery-heavy work where a fixed price would either gouge the client or bankrupt you. A reasonable default is fixed-price with a stated assumption list and a change-order clause — "this price assumes 40 mailboxes under 50 GB; additional mailboxes bill at X."

The number that matters here is not the rate, it is the realization. Project rates in the channel are usually set at or slightly above standard hourly, but the profit lives in scoping accuracy. A project quoted at 40 hours that takes 65 has a negative effective rate no matter how well you priced the hour. Before you optimize the rate, measure quoted-vs-actual on your last ten projects.

Procurement handling. The 5–15% band covers real work: sourcing, quoting, purchase orders, receiving, staging and imaging, asset tagging, warranty registration, and RMA handling when a unit arrives dead. The low end applies to high-dollar, low-touch items where a percentage on a large invoice already pays for the labor; the high end applies to small, fiddly, many-SKU orders. Some providers replace the percentage with a flat per-order coordination fee plus a lower markup, which prices the labor honestly and stops a single expensive server from generating an absurd handling charge.

Two hard rules on procurement. First, disclose the model — whether you're marking up or charging a fee, say so, because a client who discovers a hidden markup by comparing your invoice to a retail price has grounds to distrust everything else you bill. Second, keep the pass-through cost in its own ledger account. Hardware revenue is high-volume and near-zero-margin; if it mixes with service revenue, your gross margin percentage becomes meaningless and so does every decision you make from it.

The arithmetic. Model each fee as: *(volume that triggers it) × (fee amount) = monthly revenue*, then check it against the role it's supposed to fund. Three new clients a month at a one-month-MRR onboarding fee, thirty after-hours incident-hours across the base at a premium rate, and a 10% handling markup on your typical monthly hardware volume — run those three lines and compare the total to the fully loaded cost of a dispatcher or a procurement coordinator. If the fees don't cover the role, either the fee is too low or the role isn't justified yet. That comparison, not a benchmark table, is what tells you whether the schedule is right for your business.

How to evaluate and shortlist the systems that bill this

The fee schedule is a policy; the PSA is what makes it real. When evaluating platforms — ConnectWise PSA, Datto Autotask, HaloPSA, Atera, Kaseya BMS, or whatever your peer group is running — the questions that predict whether your fees actually collect are narrower than a standard feature comparison.

Can it apply rates by work type and time of day, automatically? This is the single highest-value capability for a fee schedule. If the platform can look at a time entry, see that it was logged at 11:40 p.m. against an after-hours work type, and apply the premium rate without a human choosing it, your premium realizes. If it can't, it won't. Ask for a live demo of exactly this, with a ticket created outside business hours — not a slide.

Does the agreement engine handle recurring plus one-time plus out-of-scope on the same client? Some tools are excellent at recurring billing and clumsy at everything else. You need a client record that carries a monthly agreement, a one-time onboarding charge, ad-hoc project invoices, and hardware pass-through simultaneously, with each landing in a different revenue category.

Does it flag out-of-scope work before the invoice? The change-order fee only works if someone catches the out-of-scope request while it's still a request. A platform that surfaces "this ticket is against a covered agreement but the work type is excluded" at dispatch time saves the conversation; one that surfaces it at invoicing time creates a dispute.

How does its pricing model interact with your growth? This is an underrated trap. Per-endpoint pricing means every new client raises your platform cost proportionally, which nibbles the margin on the very fees you're adding. Per-technician (Atera) or per-agent (HaloPSA) pricing holds flatter as the client base grows. Neither is universally right — per-endpoint can be cheaper for a small technician team managing many devices — but model it against your actual three-year trajectory rather than today's headcount.

Can it report fee revenue separately? If you can't pull a report showing onboarding, after-hours, project, and procurement revenue as distinct lines, you're flying blind on the thing you just built.

Around the PSA, two adjacent systems matter more than they look. A payments layer with card or ACH on file (Stripe and similar) turns an onboarding deposit from a collections problem into an automatic charge, and dunning tooling keeps the fee from quietly aging into a write-off. And the accounting system — QuickBooks for most providers at this size — is where the fee strategy becomes visible or invisible depending entirely on your chart of accounts.

One more evaluation note that has nothing to do with software: check what your peer group charges before you finalize. Peer groups, channel communities, and vendor-run benchmarking give you a sanity range that a vendor's marketing page never will. If your onboarding fee is triple the local norm you will lose deals you should have won; if it's a third of the norm you are subsidizing every new logo.

Comparable models in adjacent industries

The MSP fee structure is not unique, and the comparison is genuinely useful because other industries have run this experiment longer and made the mistakes already.

Professional services firms — accounting, legal, engineering — long ago separated the retainer from out-of-scope work, and the discipline that keeps their model honest is the engagement letter. It defines scope in enough detail that "out of scope" is a factual determination rather than an argument. The MSP analog is a scope table in the MSA that lists covered and excluded work explicitly, rather than a paragraph of prose that both parties read differently. Firms that do this well also do something MSPs rarely do: they notify the client *before* out-of-scope work begins, in writing, with an estimate. The fee is far easier to collect when it was authorized in advance.

Equipment dealers and field-service businesses run almost exactly the MSP structure under different names: a service contract covering scheduled maintenance, an emergency callout rate at a premium, a parts markup, and a project rate for installations. Their hard-won lesson is about the callout minimum. A dealer who bills a two-hour minimum on emergency calls isn't gouging; they're pricing the disruption, which is real and separate from the labor. MSPs that bill after-hours in six-minute increments are underpricing the same disruption.

SaaS vendors offer the cautionary tale on implementation fees. The industry spent years discounting onboarding to zero to close deals, then discovered that free implementation correlates with poor adoption — a client who paid nothing for the setup assigns nobody to it. The parallel for an MSP is direct: a waived onboarding fee often produces a worse onboarding, because the client has no financial stake in providing the documentation, access, and internal coordination the process needs. The fee buys engagement, not just revenue.

Staffing and BPO contribute the framing for after-hours: a shift differential is a cost you incur whether or not you bill it. Any provider committing to overnight coverage is either paying a differential, paying on-call stipends, or burning out salaried technicians. The after-hours premium isn't opportunism — it's the pass-through of a real cost, and describing it that way to a client lands far better than "that's our rate."

The through-line across all four: fees survive scrutiny when they map to a cost or a capability the client can name. Onboarding maps to a documented transition project. After-hours maps to somebody being awake. Procurement maps to a person who owns the order. Project fees map to scope the contract excludes. Every one of those is a sentence a client can repeat to their own CFO, which is the actual test.

Rolling it out without damaging client relationships

Introducing or raising fees is a change-management exercise, and the sequencing determines whether it lands.

Start with new clients only. Every new MSA and SOW from a chosen date forward carries the full fee schedule; existing clients keep their terms until renewal. This gives you a clean cohort to measure — realization rate, dispute rate, win-rate impact — before you touch a single existing relationship. It also gives your sales conversation a truthful frame: this is how we price now.

At renewal, bring the schedule to existing clients with the reasoning attached. The version that works is specific: "We're formalizing after-hours support at a premium rate because we've staffed a dedicated overnight dispatcher — here's the response-time commitment that comes with it." The version that fails is a revised rate sheet with no narrative. Clients accept fees attached to capabilities and resent fees attached to nothing.

Train the sales conversation deliberately. Reps need a one-sentence answer for each fee and permission to hold the line. The most useful reframe for onboarding: it's not a fee, it's a project — with a scope, a timeline, deliverables, and a documented environment handed over at the end. Quote it as a project and the objection largely dissolves.

Decide your waiver policy before anyone asks. Some waivers are strategically correct — a multi-year commitment, a marquee logo, a large seat count. What kills the schedule is unlogged, unapproved, case-by-case waiving. Route exceptions through the same approval path as a discount on recurring revenue, and track realization as a metric you actually look at.

Finally, watch the leading indicators for six months: fee realization rate by category, dispute rate per invoice, days-to-collect on fee lines versus recurring lines, and win rate before and after. If disputes spike on one category, the problem is almost always disclosure or invoice wording rather than the fee itself. If win rate drops materially, the fee may be out of line with your market — but check first whether it's actually a sales-confidence problem, which is far more common.

Related questions

Should an MSP charge an onboarding fee if the prospect is switching from a competitor?

Yes — competitive takeovers are usually the *most* expensive onboardings, because you inherit undocumented environments and a predecessor with no incentive to cooperate. If you need a concession to win, discount the first months of recurring rather than waiving the project fee.

Is a hardware markup ethical if the client can see retail prices?

Yes, when disclosed. The markup covers sourcing, staging, imaging, asset tagging, and RMA handling — not the box. Providers who state the model up front, or charge a flat coordination fee instead, rarely face objections. Undisclosed markups discovered later are what damage trust.

How do we handle after-hours work for a client who operates 24/7?

Sell a defined coverage block rather than billing incident by incident. A monthly adder that includes a set number of after-hours hours, with overage above it, matches their reality, smooths your revenue, and avoids a bill-shock conversation every month.

What percentage of revenue should come from fees?

There's no correct target. The useful test is whether fee revenue covers the roles it was meant to fund — dispatcher, procurement coordinator, project engineer. If it does, the schedule works regardless of the percentage.

Do fees hurt renewal rates?

Disclosed, scope-defined fees generally don't. Surprise charges do. The predictive factor is whether the client saw the fee coming and can explain what it bought — not the fee's existence or size.

FAQ

What service fees should an IT or MSP company charge?

Four categories cover most of the channel: a one-time onboarding or setup fee, an after-hours and emergency support premium at roughly 1.5–2× the standard hourly rate, project or change-order fees for out-of-scope work, and a procurement handling markup in the 5–15% range on pass-through hardware. Add per-incident overage and compliance or attestation fees once those four are billing cleanly.

Why are these fees such high-margin revenue?

The overhead is already carried by the recurring contract. Technicians, dispatchers, PSA licenses, and management are paid for by the monthly seat or device fee, so an incremental after-hours hour or a procurement markup only incurs its direct labor cost. That's what puts contribution margin in the 80–95% range — capacity utilization, not a pricing trick.

How many separate fees should we list?

Three to five. A focused schedule is easy for a rep to explain and easy for a client to accept. Beyond that, each additional line adds invoice complexity and starts reading as nickel-and-diming, which costs more in trust than it collects in revenue.

Can we waive fees for large clients?

Yes, with discipline. Multi-year commitments, high seat counts, and strategic logos are legitimate reasons. The requirement is that waivers get approved and logged the same way a discount on recurring revenue would be — unlogged, case-by-case waiving is what quietly erodes a schedule until it exists only on paper.

How do we make sure the after-hours premium actually gets billed?

Automate it. Configure the PSA to apply the premium by work type and time of day so the rate attaches to the time entry without anyone choosing it at midnight. Then confirm the MSA excludes after-hours from the covered block — a contract that promises "reasonable" after-hours support will override your rate sheet in any dispute.

How often should we review the fee schedule?

Annually at minimum, and immediately after any material cost change — a wage adjustment, a new shift differential, a platform price increase, or a shift in your client mix. Check realization by category at the same time; a fee that's being waived 40% of the time is a sales or approval problem, not a pricing one.

Sources

• https://www.connectwise.com/platform/business-management
• https://www.datto.com/products/autotask-psa/
• https://www.atera.com/pricing/
• https://halopsa.com/pricing/
• https://www.ninjaone.com/pricing/
• https://stripe.com/pricing
• https://www.pax8.com/
• https://quickbooks.intuit.com/pricing/
• https://www.dol.gov/agencies/whd/overtime
• https://www.sba.gov/business-guide/manage-your-business/pricing-your-products-services

Related on PULSE

• [How Many Attendants Should I Schedule Each Day at My Car Wash?](/knowledge/tl0067)
• [How Many Sales Reps Do I Need to Hire for My Logistics Company?](/knowledge/tl0058)
• [How Many Salespeople Do I Need to Hire for My Car Dealership?](/knowledge/tl0052)
• [How Many Producers Do I Need to Hire for My Insurance Agency to Grow My Book?](/knowledge/tl0015)
• [How Do I Figure Out How Many People to Schedule Each Day and at What Times for My Single Store?](/knowledge/tl0002)