Outsourced service desk pricing: per ticket vs per user vs dedicated agent

Why outsourced service desk quotes are hard to compare

Ask three providers to price an outsourced IT service desk and you will often get three different units: one quotes per ticket, one per user per month, and one per dedicated analyst. Each number looks reasonable on its own. None of them can be compared without translating them into the same thing: what you will actually pay over a year for the work you actually have.

This guide does not give you price ranges. Published "average" rates for outsourced service desks are rarely comparable, because they mix different scopes, regions, hours of cover and contract sizes. Instead, it explains how each pricing model works, which cost drivers move the quote, and how to normalise proposals so you can compare them honestly. If you want a number, the fastest route is to give two or three providers the same scope document (we include a checklist below) and compare their answers line by line.

One framing helps throughout: every pricing model is a way of splitting risk between you and the provider. Per-ticket pricing puts volume risk on you. Per-user pricing puts volume risk on the provider. Dedicated-agent pricing puts utilisation risk on you but gives you control. Once you know which risk you would rather hold, the right model is usually obvious.

Model 1: per-ticket (per-incident or per-contact) pricing

How it works. You pay a fixed rate for each ticket, contact or incident the provider handles. Some providers charge one rate for every ticket; others tier it by priority, channel (phone usually costs more than email or portal) or level (a Level 1 fix versus an escalation that needs a Level 2 engineer). Contracts often include a monthly minimum so the provider can staff for you.

Where per-ticket works

• Low or unpredictable volume. A company with a small IT team that only needs help at peaks, during a rollout, or for after-hours cover.
• Overflow and out-of-hours. Your internal team handles business hours; the provider takes nights, weekends and spillover.
• Trying an outsourcer before committing. Per-ticket contracts are easier to start and exit than per-user or dedicated contracts.

Where it hurts

• The provider earns more when you have more tickets. There is no commercial reason for them to fix recurring problems, write knowledge articles or push self-service. Expect to have to contract for that explicitly.
• Disputes about what counts as a ticket. Is a reopened ticket a new one? Is a follow-up phone call on the same issue a second contact? Is a duplicate email a ticket? Define this in the contract or you will argue about it on every invoice.
• Budget swings. A bad patch, a phishing wave or an office move can double a month's bill.

Model 2: per-user or per-device (subscription) pricing

How it works. You pay a flat monthly fee for each supported user (employee) or each supported device (laptops, desktops, sometimes servers and mobile devices priced separately). The fee covers an agreed scope of service desk work, typically unlimited Level 1 contacts within the hours of cover, plus defined Level 2 tasks. Out-of-scope work is billed separately, by project or by hour.

Where per-user works

• Steady volume and a known headcount. Most mid-market companies with 100 to 1,000 employees fit here.
• Finance wants a predictable number. The cost moves with headcount, which finance already forecasts.
• You want the provider to reduce tickets. Because they are paid the same whether you raise ten tickets or a hundred, the provider profits by preventing repeat issues, improving knowledge articles and automating common requests.

Where it hurts

• You pay for quiet months. If your users rarely contact IT, per-ticket might be cheaper.
• Scope creep runs both ways. The provider will try to push work into "out of scope" (projects, onboarding of a new office, application support). You will try to pull it in. A precise scope schedule is the whole contract.
• Per-user vs per-device mismatch. If many staff share devices (warehouses, clinics, retail), per-device is fairer. If one person has a laptop, a phone and a tablet, per-user is simpler.
• "Unlimited" is never unlimited. Look for fair-use clauses, after-hours surcharges and limits on onsite visits.

Model 3: dedicated agent (FTE or staff augmentation) pricing

How it works. You pay for one or more named analysts who work only on your service desk, usually priced per full-time equivalent (FTE) per month. The provider handles hiring, payroll, management, training cover, and holiday and sickness backfill. Some contracts add a team lead or service delivery manager as a fractional role.

Where dedicated agents work

• Complex or bespoke environments. Line-of-business applications, regulated processes, or ERP support where knowledge builds up over months.
• You want the analysts to feel like your team. They use your tools, sit in your Slack or Teams channels and follow your processes.
• You are replacing in-house hires. The comparison is straightforward: a fully loaded internal analyst versus the FTE rate.

Where it hurts

• You carry utilisation risk. If ticket volume drops, you still pay for the seats.
• Coverage needs more people than you think. Covering one seat for 24 hours a day, seven days a week, requires several people once shifts, leave and training are included. Ask the provider to show the staffing maths behind any 24x7 quote.
• Key-person risk. When a strong analyst leaves, knowledge leaves with them unless documentation is a contractual deliverable.

Hybrid and less common models

Most real contracts blend models. The common combinations:

• Per-user base plus per-ticket overflow. A flat fee covers normal volume up to an agreed number of tickets; above that, a per-ticket rate applies. This shares volume risk.
• Dedicated core plus shared overflow. Named analysts during business hours; a shared pool after hours, billed per ticket or as a fixed add-on.
• Tiered by support level. Level 1 billed per user, Level 2 and Level 3 billed by the hour or as a block of hours.
• Outcome-based pricing. Part of the fee is tied to metrics such as SLA attainment, first-contact resolution or user satisfaction. Rare in the mid-market and hard to administer, but worth discussing if a provider claims strong performance.
• Block of hours / retainer. A pre-paid bank of hours used for any service desk or admin work. Simple, but unused hours usually expire.

Hybrids are often the best answer, but they make comparison harder. Insist that every proposal is restated as an annual cost under the same volume assumptions (see the comparison method below).

Side-by-side comparison of the three models

Our view: for a mid-market company with a steady flow of requests, start with per-user pricing and a tight scope schedule. Use per-ticket for overflow or after-hours only. Move to dedicated agents when you notice the shared pool repeatedly escalating issues that a person who knows your systems could fix in minutes.

The cost drivers that matter more than the model

Two providers using the same pricing model can quote very different prices because they are pricing different services. These are the drivers that move the number, roughly in the order they matter:

• Hours of cover. Business hours in one time zone, extended hours, follow-the-sun, or 24x7. Each step up requires more staff and shift premiums.
• Channels. Phone support is more expensive to staff than email, portal or chat because calls cannot be queued as easily. Ask for a per-channel breakdown.
• Response and resolution SLAs. A short answer-time target for phone calls, or a tight resolution target for priority incidents, means more staff kept available. Tight SLAs with service credits cost more.
• Scope of support levels. Level 1 only (triage, password resets, how-to questions) versus Level 2 (device fixes, access management, application configuration) versus Level 3 (infrastructure, network, cloud).
• Applications covered. Standard productivity suites are cheaper to support than in-house or industry-specific applications, which need training and documentation.
• Languages. Each additional language can mean additional staff or a different delivery location.
• Onsite support. Desk-side visits, hardware swaps and new office set-ups are often priced separately, by visit or by hour.
• Tooling. Whose ITSM platform is used? If the provider uses its own (for example, a shared ServiceNow or Freshservice instance), licensing is usually included but you lose some control over data and reporting. If they work in your tool, you pay for their agent licences. See our ITSM licence pricing comparison for what those seats cost.
• Security and compliance requirements. Background checks, data residency, regulated industries and audit evidence add cost.
• Transition and onboarding. Knowledge transfer, documentation, tool set-up and parallel running are often billed as a one-off fee.
• Contract term and volume. Longer terms and larger user counts usually lower the unit price. Short terms and small accounts cost more per unit.

Before you compare prices, compare these eleven lines. If one proposal looks much cheaper, it almost always differs on hours, channels, SLA or scope.

How to normalise proposals into one comparable number

Use this method to turn any mix of pricing models into a single annual figure you can compare.

Outsourced vs in-house: the comparison most teams get wrong

Teams often compare an outsourced quote against the salary of one analyst. That understates the in-house cost. A fair in-house figure includes:

• Salaries, benefits and employer taxes for every analyst needed to cover your hours, including leave and sickness cover.
• Recruitment and onboarding each time someone leaves.
• Management time: a team lead or IT manager supervising the desk.
• ITSM tool licences, telephony and knowledge base tooling.
• Training and certification.
• The cost of gaps: what happens when the only person who knows a system is on leave.

Equally, a fair outsourced figure includes the internal time needed to manage the provider: weekly service reviews, escalations, approving changes and onboarding new staff into the provider's processes. Outsourcing moves work; it does not remove all of it.

Outsourcing tends to win when you need extended hours, when you struggle to hire and keep Level 1 staff, or when volume is too spiky to staff efficiently. In-house tends to win when your environment is highly specialised and knowledge is the main asset. Many mid-market teams end up hybrid: an internal team for Level 2 and Level 3, and an outsourced Level 1 desk.

Contract traps to check before signing

Which model should you choose?

Whatever you choose, make the provider's reporting part of the deal: monthly volumes by category, SLA attainment, reopen rates and the top recurring issues. That data is what lets you renegotiate the model in year two, when you know your real volumes.

Frequently asked questions

Related articles