- A service desk is the main point of contact between users and IT: it receives, logs, tracks and communicates incidents and service requests, and routes them to the right team or supplier
- Support tiers describe who has the skills; priority describes impact and urgency. Set response targets by priority, not by tier
- Password resets, account recovery and MFA changes are identity functions that attackers target. Verify, approve and log every one
- SLAs should separate response, restoration, resolution and update targets, and match the capacity you have actually staffed or contracted
- Outsourcing support can work well, but your business keeps accountability for priorities, risk and supplier performance
What an IT service desk actually does
An IT service desk is the main point of contact between the people who use IT and the people who run it. It receives incidents and service requests, logs each one, tracks it through to closure, keeps users informed, and routes the work to the right internal team or external supplier. In ITIL 4, the service desk is one practice among several: incident management, service request management and problem management are separate practices that the service desk works alongside. That distinction helps in a small business too. The desk owns the contact and the communication; it does not have to carry out every fix itself.
Many organisations that say they have a service desk are running something looser: an email address, a group chat, and a few people who respond when they happen to see a message. A real service desk receives requests through a defined channel, logs them as individual items, assigns ownership, tracks progress, and closes them with confirmation that the problem is resolved. No logging means no backlog visibility. Unowned tickets get ignored. Without closure confirmation, your queue fills with issues marked "probably fine now" that nobody has verified.
It also needs to handle two distinct flows: incidents (something is broken and needs fixing now) and service requests (a user needs something set up, changed, or explained). Treating all requests the same way creates priority confusion. A laptop that will not boot is not in the same category as a request to add a new user to a shared drive. Your desk needs to distinguish between the two, even at a basic level.
Generic "minutes lost per incident" figures tell you little about your own business. For each significant incident, record how long the interruption lasted, how many users were affected, and which business processes stopped or slowed (invoicing, customer service, production, payroll). Illustrative scenario, not a benchmark: if email is unavailable for 90 minutes for 30 staff, that is 45 hours of disrupted working time, before counting delayed customer replies.
Service desk, help desk, MSP or IT management?
These terms are often used interchangeably, but they describe different things. Knowing which one you are missing stops you buying a support contract when what you actually need is someone to own IT decisions.
| Term | What it is | What it does not cover |
|---|---|---|
| Service desk | Single point of contact for incidents and service requests, with logging, tracking, routing and user communication | Setting IT strategy, budgets or risk decisions |
| Help desk | A narrower, fix-focused function that answers questions and resolves straightforward faults | Usually not structured service requests, supplier coordination or service reporting |
| Managed service provider (MSP) | An external supplier delivering a defined scope of services under contract, often including the service desk, monitoring and patching | Your business priorities and risk ownership, which stay with you |
| IT management (IT lead or vCIO) | The accountable role that sets IT priorities, approves changes, manages risk and budget, and governs suppliers | Day-to-day ticket handling |
For a deeper look at the sourcing options, read in-house IT vs. managed service provider: how to decide. If you also need to decide where IT is heading, start with an IT roadmap aligned with the business.
Once you know what you need, the next job is setting up the desk itself. Our companion guide, Setting up an IT service desk: support tiers, ticketing and SLAs, covers how to define support tiers, why password and MFA changes need identity checks, the minimum viable ticketing setup and how to set SLA targets you can meet.
When to bring in a managed service provider
A managed IT service provider (MSP) delivers a defined scope of IT services under contract, usually for a monthly fee. That scope often covers the service desk, proactive monitoring, patch management and sometimes security services. The model can work well because the provider spreads specialist staff across many clients, which gives you cover for holidays, illness and staff turnover that a single internal hire cannot provide.
Whether it is cheaper depends on the details. Compare the total cost and the expected quality of three options: internal, outsourced and hybrid. On the provider side, look at the scope and exclusions, the SLAs and coverage hours, how project work and on-site visits are charged, the contract term, exit costs and the time your own staff will spend managing the supplier. On the internal side, use the fully loaded cost of an employee, not just the salary.
| Cost element | Internal employee | Managed provider (defined scope) |
|---|---|---|
| Core cost | Gross salary | Monthly fee per user or device for the agreed scope |
| Employment costs | Employer National Insurance, pension contributions and benefits | Included in the fee |
| Tools and training | Ticketing, remote support and monitoring licences; certifications | Often included; check which licences you still pay for |
| Cover | Holidays, sickness and vacancies need a backup arrangement | Provided by the team, within the contracted hours |
| Extras | Overtime or on-call allowances | Project work, on-site visits and out-of-scope requests, often charged separately |
| Management time | Line management and recruitment | Supplier management, SLA reviews and invoice checks |
| Exit | Notice period, handover and recruiting a replacement | Notice period, handover of data, documentation and access, transition to a new provider |
Illustrative cost model. Fill it in with your own salary data and supplier quotes; figures vary by region, seniority and scope.
A hybrid model is common: a provider handles the day-to-day queue while someone internal owns priorities, procurement and supplier relationships. To see how support costs fit your overall spending, use our IT budget guide.
Even with a full-service provider, your business stays accountable for IT. Someone internal must own priorities, approve significant changes, manage IT and security risk, review KPIs and hold the supplier to its contract. A support provider resolves tickets; an accountable IT lead or virtual CIO (vCIO) decides what IT should achieve and checks whether the supplier is delivering it. Some providers offer both, but they are different roles and should be named separately in the contract.
Before you outsource or launch a ticketing tool
Check these minimum requirements first. Most cost little to set up and are much harder to add once a service is live.
| Area | Minimum requirement |
|---|---|
| Identity verification | Documented procedure for password resets, account recovery and MFA changes, with stronger checks for privileged and executive accounts |
| Access control | Support staff, internal or external, get only the permissions their role needs; admin rights are separate, logged and reviewed |
| Ticket confidentiality | Restricted categories for HR, legal and security tickets; no passwords, recovery codes or secrets in tickets or attachments |
| Supplier governance | Written scope, named escalation owners on both sides, and an exit plan covering data, documentation and access |
| Major incidents | A separate escalation path and communication plan for outages and security incidents, not just a high-priority ticket |
| Knowledge base | A named owner and review cycle for articles, including the procedures a provider follows for your systems |
| Reporting | Monthly figures for response, restoration, backlog and repeat incidents, reviewed by the internal owner |
| Continuity | Cover for the absence of key people, and a fallback if the provider or the ticketing tool is unavailable |
Privacy matters too. A ticketing tool holds personal data about employees and often customers, along with screenshots, attachments and sometimes remote session recordings. Under UK GDPR you need a lawful basis, sensible retention periods and appropriate access controls for that data. When an external provider handles tickets, assess for each activity whether it acts as your processor or as a controller in its own right, and put the right contract terms in place.
Measuring whether your service desk is working
A service desk you cannot measure is a service desk you cannot improve. Most ticketing tools provide basic reporting by default, and you do not need complex analytics to get useful insight. Start with a small set of metrics and review them monthly. Once those are stable and understood, add more depth. Chasing ten metrics from the start usually means none of them get proper attention.
The metrics that matter most in the early stages are: first response time (how long before a person acknowledges the ticket), time to restore (how long until users can work again), resolution time (from ticket creation to confirmed fix), backlog (how many tickets are open and how old they are), SLA compliance rate and repeat incident rate (how many tickets this month concern a problem that appeared before). The last one is particularly valuable. A high repeat rate tells you that your service desk is fixing symptoms, not causes. Recurring printer issues that appear week after week are a hardware procurement problem, not a support problem.
Beyond the numbers, a brief satisfaction survey sent at ticket closure is worth implementing early. A single question ("Was your issue resolved to your satisfaction?") with a yes or no answer and an optional comment field gives you qualitative signal that metrics alone cannot provide. Users who are dissatisfied despite a ticket being marked resolved are often pointing to something important: the fix was temporary, the communication was poor, or the resolution time was technically within SLA but felt unacceptably slow given the impact. That feedback is valuable if you act on it.
The service desk practice is described in ITIL 4, the IT service management framework now owned by PeopleCert. For an overview of the practice, see PeopleCert's ITIL 4 Practitioner: Service Desk page.