In-house vs outsourced IT service desk: total cost for 50 to 500 users
Short answer
For 50 to 500 users, the in-house versus outsourced decision usually turns on one thing: coverage, not workload. At the small end, ticket volume rarely needs more than one or two analysts, but covering business hours every day, through holidays and sick days, needs at least two people, plus someone to manage them. That fixed floor is what makes outsourcing attractive for smaller companies. As you grow toward 500 users, workload starts to drive headcount, the floor matters less, and an in-house team becomes easier to justify, especially if your environment is specialised. Do not compare a provider's quote with one analyst's salary. Build a full annual in-house cost (loaded salaries for every analyst coverage requires, management time, tooling, recruitment and training), subtract the costs you keep either way, and divide by your users and 12. That gives you a break-even price per user per month: any outsourced quote for the same scope below it saves money; above it, you are paying for something else, such as extended hours or flexibility. This guide gives you the formulas and a worked example built only on official U.S. statistics and labelled placeholders.
Contents
- Why most in-house vs outsourced comparisons mislead
- Step 1: define the scope you are pricing
- Step 2: work out how many analysts in-house coverage really needs
- Step 3: calculate the loaded cost of each person
- Step 4: add the in-house costs people forget
- Step 5: build the outsourced total on the same scope
- Worked example: 300 users, business-hours cover
- How the answer shifts from 50 to 500 users
- What the cost model does not capture
- Five mistakes that skew the comparison
- Decision checklist
- Frequently asked questions
Why most in-house vs outsourced comparisons mislead
The usual comparison goes like this: a provider quotes a monthly fee, someone compares it with the salary of one help desk analyst, and the provider looks expensive. That comparison leaves out most of what an in-house desk costs. The reverse mistake happens too: a provider's low per-user quote is compared with a fully loaded internal team, but the quote covers fewer hours, fewer channels or less scope than the internal team provides.
A fair comparison needs three things: the same scope on both sides, every cost on both sides, and your own volumes rather than someone else's averages. We deliberately do not quote "average" outsourcing rates or "typical" tickets per user here. Those numbers vary enormously by industry, region and how you count, and a borrowed benchmark can easily swing the answer. Instead, this guide gives you formulas to fill in. Where we use real figures, they come from official sources and are dated. Where we need example inputs to show the arithmetic, they are labelled as placeholders.
Placeholders are not benchmarks
The worked example below uses placeholder ticket volumes, handle times and productive hours to show how the formulas work. Replace every placeholder with your own data before you make a decision.
Step 1: define the scope you are pricing
Both sides of the comparison must deliver the same service. Write it down before you calculate anything:
- Users and devices supported. Employees, contractors, shared devices, locations.
- Hours of cover. Business hours in one time zone, extended hours, weekends, 24x7.
- Channels. Phone, email, portal, chat, Microsoft Teams or Slack.
- Support levels. Level 1 only (triage, password resets, how-to), or Level 1 and Level 2 (device fixes, access management, application configuration).
- Service levels. Response and resolution targets by priority.
- Onsite work. Desk-side support, hardware swaps, new starter set-ups.
- What stays in-house regardless. Infrastructure, security, projects, vendor management.
If you outsource, you will almost certainly keep someone internal who owns IT, manages the provider and handles Level 3 or project work. Those retained costs exist in both scenarios, so they drop out of the comparison; we handle them explicitly in Step 5.
Step 2: work out how many analysts in-house coverage really needs
In-house headcount is the larger of two numbers: the analysts needed to handle the workload, and the analysts needed to keep the desk open for the hours you promise. For 50 to 500 users, the second number often wins.
Workload FTE
Workload hours per year = annual tickets × average handling hours per ticket
Productive hours per FTE = contracted hours − leave − public holidays − training − expected sickness
Workload FTE = workload hours ÷ (productive hours per FTE × target utilisation)
Take annual tickets and average handling time from your ITSM tool. Handling time should include the work after the call or chat (notes, follow-ups, escalations), not just talk time. Target utilisation is the share of productive time you expect analysts to spend on tickets; the rest goes to knowledge articles, meetings, projects and gaps between contacts. Pick a figure you believe for your team and write it down as an assumption.
Coverage FTE
Cover hours per year = hours open per day × days open per year × seats you need staffed at once
Coverage FTE = cover hours per year ÷ productive hours per FTE
Then apply a practical floor: a desk that must stay open when one person is on leave or ill needs at least two people who can do the work, even if the formula says 1.2. For extended hours or weekend cover, the coverage number rises quickly because every extra hour must be staffed by someone who is also taking leave.
In-house FTE
In-house analysts = the larger of (workload FTE, coverage FTE, your minimum for continuity), rounded up to whole people
Rounding up matters. You cannot hire 1.3 analysts, and part-time arrangements for help desk cover are hard to sustain. The gap between the FTE you need and the FTE you must hire is one of the main reasons outsourcing wins at the small end.
Step 3: calculate the loaded cost of each person
Salary is not cost. Employers also pay for benefits, paid leave and legally required contributions. Use your own payroll data if you have it. If you need a reference point and you are in the United States, two official sources help:
- Wages. The BLS Occupational Outlook Handbook: Computer Support Specialists reports a median annual wage of $61,860 for computer user support specialists (help desk technicians) in May 2025. The lowest 10 percent earned less than $40,980 and the highest 10 percent more than $100,540. The same page lists a May 2025 median of $175,140 for computer and information systems managers. Checked on Oct 31, 2026.
- Benefits load. The BLS Employer Costs for Employee Compensation release (June 2026 data) reports that for private industry workers in June 2026, wages and salaries accounted for 70.0 percent of employer compensation costs and benefits for the remaining 30.0 percent (68.5 and 31.5 percent for full-time private industry workers). Checked on Oct 31, 2026.
Loaded annual cost per person = annual salary ÷ wage share of total compensation
Using the BLS private-industry wage share of 0.70: loaded cost = salary ÷ 0.70 (about 1.43 × salary)
Two cautions. The BLS benefits share is an average across all private industry, not specific to IT roles or to your company. And national medians hide large regional differences; the OOH page links to state and metro wage data. Outside the U.S., use your national statistics office or your own payroll. The formula is the same.
Step 4: add the in-house costs people forget
| Cost line | How to estimate it | Notes |
|---|---|---|
| Management time | Share of a manager's loaded cost spent supervising the desk | Rotas, one-to-ones, escalations, reporting, hiring |
| ITSM tool licences | Agents × price per agent per month × 12 | Public examples: Freshservice pricing page (Starter $19, Growth $49, Pro $99 per agent/month billed annually) and Atlassian JSM pricing page (Standard $21.50, Premium $55.28 per agent/month, monthly billing at the default team size), both checked on Oct 31, 2026 |
| Telephony and contact tools | Your current per-seat cost | Phone queues, call recording, chat |
| Recruitment | Expected leavers per year × cost per hire | Agency fees or internal recruiter time, plus the vacancy gap |
| Onboarding and training | New starters × training days × daily loaded cost, plus courses | Include the time of whoever trains them |
| Equipment and space | Laptop, headset, desk or remote-work allowance per analyst | Often small but real |
| Cover for gaps | Overtime or contractor days during vacancies and long absences | Use last year's actuals if you have them |
In-house annual total = (in-house analysts × loaded cost) + management share + tooling + telephony + recruitment + training + equipment + gap cover
Step 5: build the outsourced total on the same scope
An outsourced quote is rarely the whole outsourced cost. Add:
- The service fee for your scope, in whatever model the provider uses (per user, per device, per ticket or per dedicated analyst). Our guide to outsourced service desk pricing models explains how each works and how to normalise them.
- One-off transition costs: knowledge transfer, documentation, tool set-up, parallel running.
- Out-of-scope charges: onsite visits, projects, after-hours callouts, new office set-ups.
- Tool licences if the provider works in your ITSM tool, or the cost of exporting data from theirs at exit.
- Retained internal effort: the time your IT lead spends managing the provider, approving changes, handling escalations and running service reviews.
Outsourced annual total = service fee + (transition cost ÷ contract years) + expected out-of-scope charges + tool licences + vendor management time
Then compute the number that makes quotes easy to judge:
Break-even price per user per month = (in-house annual total − costs you keep either way) ÷ (users × 12)
"Costs you keep either way" are the lines that do not disappear when you outsource: the manager time that becomes vendor management, any licences you still pay for, and onsite work you will still do yourself. A per-user quote for the same scope below the break-even saves money; one above it costs more, and you should be able to say what the extra buys you.
Worked example: 300 users, business-hours cover
This example shows the arithmetic end to end. Salary and benefits figures come from the BLS sources above; tooling uses Freshservice's published Growth price. Ticket volume, handling time, productive hours, utilisation and the management shares are placeholders chosen only to make the maths visible. They are not benchmarks.
| Input | Value | Source |
|---|---|---|
| Users | 300 | Placeholder |
| Annual tickets | 3,000 | Placeholder: replace with your ITSM data |
| Average handling time | 0.5 hours | Placeholder |
| Productive hours per FTE | 1,760 per year | Placeholder: 2,080 contracted hours minus your leave, holidays and training |
| Target utilisation | 75% | Placeholder assumption |
| Cover | 9 hours × 5 days × 52 weeks, one seat | Scope decision |
| Analyst salary | $61,860 | BLS median, computer user support specialists, May 2025 |
| Manager salary | $175,140 | BLS median, computer and information systems managers, May 2025 |
| Wage share of compensation | 0.70 | BLS ECEC, private industry, June 2026 |
| Management share in-house / retained if outsourced | 20% / 10% of a manager | Placeholders |
| ITSM licences | $49 per agent per month | Freshservice Growth, billed annually (checked Oct 31, 2026) |
| Calculation | Result |
|---|---|
| Workload FTE = 3,000 × 0.5 ÷ (1,760 × 0.75) | 1.14 |
| Coverage FTE = (9 × 5 × 52) ÷ 1,760 | 1.33 |
| In-house analysts (larger figure, continuity minimum of 2) | 2 |
| Loaded analyst cost = $61,860 ÷ 0.70 | $88,371 |
| Analysts: 2 × $88,371 | $176,743 |
| Management: 20% × ($175,140 ÷ 0.70) | $50,040 |
| ITSM licences: 2 × $49 × 12 | $1,176 |
| In-house subtotal (before recruitment, training, telephony, equipment, gap cover) | $227,959 |
| Retained if outsourced: 10% × ($175,140 ÷ 0.70) | $25,020 |
| Break-even = ($227,959 − $25,020) ÷ (300 × 12) | about $56.37 per user per month, plus your recruitment, training and other in-house lines divided by 3,600 |
Read the result correctly. It is not a market price for outsourced service desks. It says that, under these inputs, a provider charging less than about $56 per user per month for the same business-hours Level 1 and Level 2 scope would cost less than running it yourself, before counting recruitment and training. Change any placeholder and the answer moves. Note also what drove it: the continuity minimum of two analysts, not the workload. That is typical of the small end of this range and is the key insight to test with your own numbers.
How the answer shifts from 50 to 500 users
| Company size | What usually drives in-house cost | What that means for the decision |
|---|---|---|
| Around 50 users | The continuity floor and management time; workload is small | Hard to justify two dedicated analysts; outsourcing or a hybrid with an IT generalist is often cheaper |
| 100 to 200 users | Still mostly coverage; workload approaches one to two FTE | Compare carefully; the break-even test is decisive here |
| 250 to 350 users | Workload and coverage start to align | In-house becomes competitive for business hours; outsourcing still helps for extended hours |
| 400 to 500 users | Workload drives headcount; a team lead becomes a full role | In-house often wins on cost for core hours if you can hire and retain; outsource overflow and out-of-hours |
These bands describe how the formulas behave, not market averages. Run the model at your real size and again at the size you expect in two years; a decision that is right at 150 users can be wrong at 400.
What the cost model does not capture
Reasons to keep it in-house
- Your environment is specialised: line-of-business applications, regulated processes or ERP support where knowledge takes months to build.
- You want IT staff who also handle projects, onboarding and process improvement between tickets.
- Security or data residency rules make third-party access hard.
Reasons to outsource
- You need extended hours, weekends or multiple languages, which multiply in-house coverage cost.
- Hiring and keeping Level 1 staff is a recurring problem.
- Volume is spiky (seasonal peaks, acquisitions, rollouts) and you do not want to staff for the peak.
- Your IT lead spends more time managing the desk than improving IT.
The hybrid most mid-market teams end up with
An outsourced Level 1 desk for first contact, password resets, how-to questions and out-of-hours cover, with a small internal team owning Level 2, Level 3, projects and the provider relationship. Model this as a third column: outsourced fee for Level 1 plus a smaller in-house team sized only for the work you keep.
Five mistakes that skew the comparison
1. Comparing a quote with one salary
A single analyst cannot cover a desk through holidays, sickness and training. Compare against the full in-house team that your coverage hours require, at loaded cost.
2. Comparing different scopes
A provider quoting Level 1 only during business hours is not pricing the same service as an internal team that also handles Level 2, onboarding and small projects. Restate both sides against one scope document.
3. Forgetting the retained team
Outsourcing does not remove the need for someone internal to own IT. If you leave that cost out of the outsourced column, outsourcing looks cheaper than it is.
4. Using borrowed benchmarks for your own volumes
Tickets per user and handling times vary widely between companies. Pull your own 12 months of data; if you do not have it, estimate it and label it as an estimate, then rerun the model once you have three months of real data.
5. Ignoring growth and contract terms
A three-year outsourcing contract priced for today's headcount can become expensive or restrictive if you double in size, and an internal team sized for today can be overwhelmed. Model your expected headcount and check true-up and exit clauses.
Decision checklist
Before you decide
- Scope written down and identical for both options
- 12 months of ticket data: volume, channel, category, handling time
- Productive hours per FTE calculated from your own leave and holiday policy
- Loaded cost from payroll or a dated official source
- Coverage FTE and continuity minimum applied
- All in-house lines included: management, tools, recruitment, training, gap cover
- Outsourced total includes transition, out-of-scope and vendor management time
- Break-even price per user per month calculated
- Model rerun at the headcount you expect in two years
- Exit terms and data ownership checked in any outsourcing contract
If you are comparing ITSM tools as part of the same decision, see our ITSM pricing comparison for licence costs that feed into both columns.
Frequently asked questions
Is it cheaper to outsource the IT service desk?
It depends on your size, hours and scope. For smaller companies, the need for at least two analysts to keep the desk open through leave and sickness often makes outsourcing cheaper. As you approach 500 users, workload drives headcount and in-house becomes more competitive. Calculate a break-even price per user per month and compare quotes against it.
How do I calculate the true cost of an in-house service desk?
Multiply the number of analysts that both workload and coverage require by each person's loaded cost, then add management time, ITSM and telephony tools, recruitment, training, equipment and the cost of covering gaps. Loaded cost is salary divided by the wage share of total compensation; BLS data for U.S. private industry in June 2026 puts that share at 70.0 percent.
How many service desk analysts do I need for 300 users?
There is no reliable rule of thumb. Calculate workload FTE from your ticket volume and handling time, calculate coverage FTE from the hours you must be open, take the larger figure and apply a continuity minimum, usually two people for a desk that must stay open during leave.
What is a break-even price per user for outsourcing?
It is your in-house annual total, minus costs you would keep either way, divided by users and 12 months. An outsourced quote for the same scope below that figure saves money; above it, you are paying for something extra, such as longer hours or flexibility.
What do outsourced service desk quotes usually leave out?
Transition and knowledge-transfer fees, out-of-scope work such as onsite visits and projects, tool licences if the provider works in your system, and the internal time needed to manage the provider. Add all of them before comparing.
Want us to run this model with your numbers?
Send us your user count, 12 months of ticket data and the hours you need covered. We will build both sides of the comparison and scope an outsourced desk against the same assumptions.