In-house IT
One or more employed technologists carry the whole estate. Deep business context, immediate presence, and strong ownership.
- Context and proximity
- Single points of failure, thin cover, limited specialist depth
Guide · Strategy and operating model
Managed IT services vs in-house IT: a guide for SMEs
A practical comparison of managed IT services and an in-house IT team for growing SMEs: cost, cover, risk, capability and the strategy work that support tickets never touch.
- 9 minutes
- Owners, MDs and finance leads
- 10 to 500 users
At a glance
The question is rarely "internal or outsourced". It is which model gives the business continuous cover, defensible security and a technology plan it can fund.
Most SMEs arrive at this decision the same way. Someone leaves, an incident lands, an insurer asks a question the business cannot answer, or growth outpaces the person who has quietly kept everything running. At that point leadership compares a salary against a monthly fee, which is the narrowest possible version of the choice.
The wider version is this: technology now decides how fast you can hire, trade, acquire and prove compliance. Whoever runs it has to do two different jobs. The first is keeping the estate stable and secure every day. The second is turning technology into decisions the business can act on: what to fund, what to retire, what risk to accept. In-house teams are usually excellent at the first and starved of time for the second.
The three models
Three ways to run the estate.
Managed IT services
An external partner runs the estate to agreed service levels, with tooling, process and a bench of specialists behind the service desk.
- Cover, breadth and predictable cost
- Needs clear scope and business context built deliberately
Co-sourced
Internal ownership of business systems and vendor relationships, with the partner carrying operations, security and out of hours cover.
- Context plus capacity
- Requires a disciplined responsibility matrix
Side by side
Eight dimensions that decide the answer.
Scroll sideways to compare.
| Dimension | In-house IT | Managed IT services |
|---|---|---|
| Hours of cover | Business hours, reduced during leave and sickness | 24/7/365 monitoring and response as standard |
| Specialist depth | Generalist skills, deep in one or two areas | Cyber, cloud, data, networks and modern work under one agreement |
| Cost profile | Salaries, tooling and training, variable and rising | Predictable monthly subscription per user or device |
| Tooling | Purchased per seat at SME pricing | Enterprise-grade stack amortised across the client base |
| Security posture | Depends on individual expertise and available time | Baselined controls, continuous monitoring and evidence |
| Scalability | New hires needed to grow, slow to unwind | Scales with headcount, sites and acquisitions |
| Business context | Excellent, held by the individual | Built through governance rhythm and named team |
| Key person risk | High, knowledge often undocumented | Low, documented and covered by a team |
The real cost
A fair comparison counts everything the internal model carries: salary and employment costs, recruitment, training and certification, monitoring and security tooling, backup licensing, out of hours cover, and the productivity lost while one person triages a queue alone. It also counts the cost of absence, because a single week of leave changes the risk profile of a business that depends on one technologist.
On the managed side, the fee is predictable and scales with users or devices, and the tooling is enterprise grade because it is amortised across many clients. The saving that matters most, though, is rarely on the invoice. It is the incident that did not happen, the migration that did not overrun, and the licensing that stopped being paid for twice.
- Count tooling and licensing separately from headcount, they are usually understated.
- Price out of hours cover honestly, goodwill is not a service level.
- Attach a number to downtime per hour, then test both models against it.
- Include the cost of documentation that does not exist today.
Beyond tickets
Support keeps the lights on. Business technology strategy decides where the building goes.
The weakest managed services and the most overloaded internal teams share the same failure: everything is measured in tickets. Ticket volume tells you how busy someone was, not whether the business got better. Four things separate a strategic model from a reactive one.
A costed roadmap
What is being fixed this quarter, what is being funded next year, and what the business gets for it.
Risk in business terms
Cyber and continuity exposure expressed as trading impact, not a vulnerability list.
Spend that is defensible
Licensing, hardware and cloud consumption reviewed against how the business actually operates.
Decisions, not options
A recommendation with a trade-off stated, so leadership can decide in one meeting.
This is the distinction we build our own service around. NumataOne™ governs six connected categories, strategy, cyber, data, operations, modern work and AI, as a single managed model, so the operational work and the strategic work are run by the same accountable team rather than bought separately.
How to choose
Five questions to answer first.
- 01
Does the estate stop when one person is on leave?
If a single absence changes your risk profile, cover is the first thing to solve.
- 02
Are the same issues recurring every month?
Recurring tickets are a root-cause problem, not a capacity problem.
- 03
Can you evidence your security position today?
If you cannot show controls, backups and recovery testing, an assessment comes before any contract.
- 04
Is technology on the leadership agenda with numbers attached?
If not, you are buying support rather than strategy, and the two are priced differently for a reason.
- 05
What changes in the next 24 months?
New sites, acquisitions, headcount growth or regulatory obligations all change which model fits.
Transition
What a well-run move looks like.
Nothing here requires a big-bang cutover. Most SMEs start with an assessment, move monitoring and security first, then transfer the service desk once documentation is in place.
Assess
Baseline the estate, risks, licensing and recurring issues before anything is signed.
Design
Agree scope, service levels, security baseline and the responsibility matrix.
Onboard
Deploy tooling, document the environment and transfer knowledge from incumbents.
Stabilise
Clear the backlog of root causes, then run a hypercare period with daily review.
Govern
Move to a standing rhythm: service review, risk review and roadmap decisions.
FAQs
Managed IT services, answered.
What are managed IT services?
Managed IT services means an external provider takes accountability for running and improving an organisation's technology: service desk, device and identity management, network and cloud operations, backup, patching and cyber monitoring, delivered against agreed service levels for a predictable monthly fee. A strategic provider adds roadmap, budget and risk governance on top of day to day operations.
Is managed IT cheaper than hiring an in-house IT team?
For most SMEs below roughly 250 users, a managed service costs less than an equivalent internal team once you count salaries, recruitment, training, tooling licences, out of hours cover and holiday or sickness absence. The larger saving is usually avoided downtime and avoided rework, not the headcount line itself.
Can managed IT work alongside an internal IT person or team?
Yes, and this co-sourced model is common. The provider carries the service desk, monitoring, patching and 24/7 cover, while the internal team focuses on business systems, vendor relationships and projects specific to the organisation. The split is set out in a responsibility matrix so nothing sits in a gap.
What should be in a managed IT services agreement?
Scope by service category, response and resolution targets, hours of cover, named contacts and escalation, security baselines, backup and recovery objectives, reporting cadence, change and offboarding terms, and a stated roadmap or improvement commitment. Insist that reporting shows business outcomes, not just ticket counts.
How long does it take to move from in-house IT to a managed service?
A typical SME transition runs four to eight weeks: discovery and documentation, tooling deployment, security baseline, knowledge transfer, then a hypercare period. Environments with significant technical debt or unresolved licensing take longer, which is why the assessment comes first.
How do I know whether my current IT support is strategic or reactive?
Ask for the last three months of reporting. Reactive support shows tickets opened and closed. Strategic support shows recurring root causes removed, risk positions changing, a costed roadmap, and decisions the business made as a result.
Compare the models
against your own estate.
A Business Technology Assessment baselines cover, risk, spend and roadmap, so the choice between in-house and managed is made on evidence.