IT Help Desk Outsourcing Guide: Scope, SLAs, Seat Sizing and Go-Live Steps

The short version: Outsourcing an IT help desk works when you write down the scope (what is in, what is out), attach response and resolution targets to each priority, size the team from ticket volume and not from a headcount rule of thumb, and run a staged handover with a pilot group. The formula below turns your user count into an estimate of agent hours and seats. Treat the numbers as a planning start point, then replace the assumptions with your own ticket data.

This guide is for IT managers and business owners who are replacing an in-house help desk or adding one. After reading it you can draft the scope and SLA sections of a request for proposal, estimate how much service desk capacity you need, and plan the first 30 days of a handover.

Service desk vs help desk: is there a difference?

Many providers use the two terms as synonyms. Where they are separated, the usual distinction is scope. Wikipedia's article on help desk software (read 28 October 2026) describes help desk software as part of the broader service desk category, which also includes asset management and IT service management.

Use this working rule when you write a contract:

1. Help desk: the user-facing front line. It receives incidents and questions, fixes the routine ones and escalates the rest.
2. Service desk: the help desk plus the processes around it: service requests, change requests, asset records, a knowledge base and reporting.

Ask every provider which one they mean, and list the processes you expect in the contract. The label alone settles nothing.

What should the scope list contain?

Write the scope as two lists, "in" and "out", and attach each line to a support level. For how levels split, see L1, L2 and L3 support levels.

In scope, typical items:

1. Contact channels: phone, email, chat, portal. State which are monitored in which hours.
2. Incident handling: logging, triage, priority setting, resolution or escalation.
3. Service requests: new starters, leavers, access changes, software installs, equipment requests.
4. Account and password work: resets, unlocks, MFA re-registration, mailbox and licence changes.
5. Endpoint support: laptops, desktops, printers, mobile devices; remote fixes first.
6. Application support: which applications at which level. Name each business application, and say who calls the vendor.
7. Knowledge base: who writes and maintains articles.
8. Reporting: monthly ticket report, SLA report, repeat-issue list.
9. Vendor liaison: internet, telephony, software and hardware suppliers.

Out of scope, decide explicitly:

1. On-site visits: included, charged per visit, or via a separate dispatch arrangement. See field IT dispatch.
2. Projects: migrations, rollouts and office moves.
3. Hardware purchase and warranty handling.
4. After-hours cover: if you need it, say so. See 24/7 help desk IT support.
5. Custom application development and database administration.

Which SLA items should be in the contract?

Ask for these as a list, per priority, not as one blended number. The reference points for priority definitions are in P1 to P4 priority levels, and the contract view is in what an SLA means in IT support.

1. Service hours: for example 8x5 or 24x7, with the time zone and public holiday rule.
2. Priority definitions: impact and urgency rules and an example for each priority.
3. First response target: per priority, measured from ticket creation.
4. Resolution or restoration target: per priority, and what stops the clock (waiting for the user).
5. Update frequency: how often the user is told progress on P1 and P2 tickets.
6. Escalation triggers: time-based and risk-based. See the escalation matrix template.
7. Measurement method: which tool is the source of truth, and who can audit it.
8. Reporting cycle and review meeting: monthly is normal.
9. Remedies: what happens if targets are missed repeatedly. Agree this at signing.
10. Quality measure: first-contact resolution and satisfaction survey, see first call resolution.

Brocent publishes its response targets on the SLA page.

How do you estimate how many agent seats you need?

Start from tickets, not from users. This is an example method with stated assumptions, not Brocent data or market statistics. Replace every assumption with figures from your own ticket history.

The formulas:

1. Tickets per month = users x tickets per user per month.
2. Workload hours per month = tickets per month x average handling minutes / 60.
3. Productive hours per agent per month = paid hours x utilisation rate.
4. Agents needed for workload = workload hours / productive hours per agent.
5. Agents needed with peaks = agents for workload x peak factor.
6. Coverage floor: the minimum number of people needed to keep one seat staffed for all service hours, after leave and breaks.

Worked example (all inputs assumed for illustration):

1. 200 users, assume 1.0 tickets per user per month: 200 tickets.
2. Assume 25 minutes average handling including follow-up: 200 x 25 / 60 = about 83 hours.
3. Assume 160 paid hours and 60 percent utilisation: 96 productive hours per agent.
4. Workload = 83 / 96 = about 0.9 of an agent.
5. Assume peak factor 1.5 (Monday mornings, after incidents): about 1.3 agents.
6. Coverage floor for 8x5: one seat is staffed 40 hours a week, so at least two people, to cover leave. For 24x7 coverage of one seat, 168 hours a week divided by about 40 hours per person is 4.2 people before leave, so plan on five or more.

What the example shows: the workload for 200 users is under two agents, but the coverage rule decides the staffing. That is why many smaller companies use a shared service desk rather than a dedicated team.

Inputs to measure, in order of importance: tickets per user (from the last 3 to 6 months), average handling time per ticket type, the share of calls at peak hours, and the share of requests that need an on-site visit.

What are the steps to hand over and go live?

1. Week 0, collect: user list, application list, current ticket history, existing runbooks, vendor contacts, admin access model.
2. Week 1, agree: scope list, priority rules, escalation contacts, report format, change approval rules.
3. Week 1 to 2, prepare: the provider builds its ticket queues, forms and knowledge articles; you set up a mailbox or phone number forwarding and the access it needs.
4. Week 2 to 3, pilot: one department or site goes first. Review every ticket in a weekly meeting.
5. Week 3 to 4, extend: all users switch. Publish the new contact details and the "how to raise a ticket" one-pager.
6. Day 30, review: first SLA report, repeat-issue list, adjust the knowledge base and the scope.
7. Day 90, review again: compare to the baseline you recorded before the switch.

Common mistakes

1. Choosing by headline price per seat and ignoring what is out of scope. Compare scope lists first.
2. No stop-the-clock rule, so waiting-for-user time distorts the SLA report. Define it.
3. Switching everyone on one day with no pilot. Pilot with one group.
4. No knowledge handover. Existing fixes live in people's heads; capture them in week 0.
5. Sizing from user count alone. Use tickets and handling time.
6. No exit data clause. You should be able to export your tickets and knowledge base when the contract ends.

How to check it worked

1. After 30 days, compare first response times with the SLA, per priority.
2. Check that repeat tickets fell, using the repeat-issue list.
3. Sample ten closed tickets: are notes clear enough that you could understand them without asking?
4. Ask pilot users two questions: did you reach someone quickly, and was the problem solved?
5. Confirm you can export your own ticket data.

FAQ

Is a help desk an IT job?

Yes. A help desk role is usually the entry point of an IT career: logging tickets, fixing routine issues and escalating the rest. It is level 1 work in the tiered model.

What are IT outsourcing services?

They are IT functions run by an external provider under a contract, such as the help desk, device and network support, monitoring and security. Read what to outsource first for the order we suggest.

What is a help desk platform?

It is the ticketing software that logs requests, routes them, tracks SLAs and holds the knowledge base. Ask the provider which platform they use and whether you get a portal login and data export.

Who has the best tech support for computers?

No single answer holds for every company. Judge a provider on these checks, not on claims:

1. A written SLA.
2. A sample monthly report.
3. References you can call.
4. A pilot period.

If you want a help desk that runs 24 hours, start with our 24/7 help desk IT support page.

Brocent publishes its managed IT support plans with per-user monthly prices for Hong Kong, Singapore, mainland China and the United States, so you can compare before you talk to anyone: see the plans and pricing.