Managed Services vs In-House IT for a 30–100 Seat Dubai Company: A Real Cost Model
The comparison usually gets made badly, and it gets made badly in a predictable way: someone compares an IT manager’s salary against a managed service quote, notices the salary is higher, and concludes that outsourcing is cheaper. Or notices the salary is lower and concludes the opposite.
Both conclusions are unreliable, because a salary is not the cost of an in-house function and a monthly fee is not the cost of a managed service. This is the model that captures what each actually costs.
The 30 to 100 seat range is where this decision is genuinely difficult. Below thirty, outsourcing is usually straightforward. Above a couple of hundred, an internal team generally makes sense. In between, it depends on specifics — and the specifics are what this covers.
The true cost of in-house
Direct employment cost
Base salary is where people start and where they should not stop. In the UAE, the fully loaded cost of an employee includes:
- Base salary
- Housing allowance, where provided
- Transport allowance, where provided
- Annual air ticket, where provided
- Medical insurance — mandatory
- Visa and Emirates ID costs, including renewals
- End-of-service gratuity accrual
- Annual leave — salary paid for days not worked
- Recruitment cost, amortised over expected tenure
The gratuity accrual and leave provision are the two most commonly omitted. Neither shows up in monthly payroll thinking but both are real costs.
Coverage gaps
One person provides coverage during their working hours, when present. They take annual leave. They get sick. They travel. They attend training.
A single IT employee delivers meaningfully less than fifty-two weeks of coverage, and during the absences you either have no support or you pay for external cover. For a business where downtime has real cost, this is the most consequential limitation of the single-hire model and the one least reflected in a salary comparison.
The question to ask: what happens on the day your IT person is on leave and the server fails? If the answer is “we call someone,” you are already partly outsourced and should cost that in.
Skills breadth
This is the substantive argument, and it is worth stating carefully.
A 30–100 seat business needs, at various times: end-user support, server and infrastructure administration, networking, security, backup management, cloud platform administration,vendor management, procurement, and increasingly compliance work.
That is not one skill set. One person will be genuinely strong in perhaps two or three of those areas and functional-to-weak in the rest. That is not a criticism of the individual — it is what a single role can realistically cover.
The practical consequence is that the areas outside their strength get handled adequately rather than well, or get outsourced anyway. Security and backup verification are the two that most commonly suffer, and they are the two where inadequacy is invisible right up until it is catastrophic.
Tooling
An in-house function needs tools an established managed provider already has amortised across clients:
- Remote monitoring and management platform
- Ticketing system
- Backup management and monitoring
- Endpoint security management
- Patch management
- Documentation platform
- Network monitoring
These are per-seat or per-device licensed and represent a real annual cost.
Management overhead
Someone manages the IT person. Someone handles their recruitment, review, development, and eventual replacement. If your IT hire reports to a finance director or operations manager with no IT background, that manager cannot effectively assess the quality of the work — which is a genuine governance issue, not just an inconvenience.
Key person risk
A single IT employee accumulates undocumented knowledge about your environment. When they leave — and eventually they will — that knowledge goes with them unless it was deliberately documented. Replacement takes time, handover is rarely complete, and the incoming person spends months rediscovering things.
This is a real cost with a low probability in any given month and a high impact when it lands.
The true cost of managed services
The honest version of this side of the comparison:
The contracted fee
Straightforward, and usually the only figure that appears in comparisons.
Out-of-scope work
Every managed contract has an exclusions list. Project work, major changes, and out-of-hours support are commonly billable separately.
Budget realistically for this. A managed contract at a given monthly fee will not be your total annual IT service spend. Ask the provider what a typical client of your size spends on out-ofscope work annually. A provider who cannot answer that has not looked.
Hardware, software, and licensing
Almost always excluded from a support contract. Your servers, workstations, licences, cloud subscriptions, and network equipment are separate regardless of who supports them — this is a wash between the two options, but it should be in both columns rather than neither.
Internal coordination time
Someone internally still owns the relationship, raises tickets, approves work, and makes decisions. This is usually a portion of an existing role rather than a hire, but it is not zero. Businesses that outsource IT and appoint nobody to own the relationship consistently get poorer service, because nobody is holding the provider to the SLA.
Response constraints
A managed provider serves multiple clients. Your SLA determines your priority. A four-hour response from a shared team is not the same as someone already in the building who can look at it now.
For issues that are urgent but low-severity — the kind where an internal person walks over and fixes it in five minutes — outsourced support genuinely is slower. That is a real disadvantage and should be acknowledged rather than argued away.
Contextual knowledge
An internal person knows your business, your people, your quirks, and which systems matter most on which day of the month. A managed provider builds this over time but starts without it, and rebuilds it partially each time their staff change.
Good providers mitigate this with documentation and account continuity. It is a genuine difference nonetheless.
The comparison table to build
Build both columns over a three-year horizon, not one year — recruitment costs and gratuity accrual distort a single-year view.
In-house column: – Fully loaded employment cost × 3 – Tooling and licensing × 3 – Recruitment cost – Cover for absence — external support or accepted downtime – Training and certification budget – Management time, costed – Escalation support for issues beyond their skill range – Key person transition cost, probability-weighted.
Managed services column: – Contracted fee × 36 months – Estimated out-of-scope work annually × 3 – Internal coordination time, costed – Onboarding and transition cost – Any capability gaps requiring separate provision.
Both columns: – Hardware, software, licences, cloud subscriptions – Cost of downtime at the level each model realistically delivers
That last line is the one that decides it, and it is the one nobody calculates.
Calculating your downtime cost
Rough but adequate: total hourly staff cost of everyone unable to work, plus direct revenue impact if operations stop, plus recovery time afterwards.
For a 60-person business where an outage stops everyone, a single day of downtime typically costs a multiple of a month’s IT support under either model. If one option demonstrably reduces downtime, that difference usually dominates the entire comparison.
This is why the comparison should not be made on cost alone. It should be made on cost and risk together.
The hybrid model, which is often the answer
The framing as a binary choice is what makes this decision harder than it needs to be. In the 30–100 seat range, the arrangement that most often works is neither pure option:
One internal person handling end-user support, on-site presence, contextual knowledge, coordination, and the fast small fixes that outsourced support handles slowly.
A managed provider handling infrastructure, security, backup verification, patching, monitoring, out-of-hours coverage, absence cover, and specialist escalation.
This addresses the specific weaknesses of each. The internal person is not expected to be strong across seven disciplines. The provider is not expected to know your office’s quirks or walk to someone’s desk. The coverage gap during leave is filled. The key person risk is reduced because the provider holds documentation independently.
It is not the cheapest option on a spreadsheet. It is frequently the one with the best outcome per dirham, particularly for businesses where downtime is expensive.
What should drive the decision
Beyond the numbers:
Choose more in-house when: you have substantial on-premises infrastructure requiring daily attention; your systems are unusual or heavily customised; you need immediate physical presence regularly; you have compliance requirements demanding direct control; you are large enough to hire two or more people, which resolves the coverage and skills-breadth problems.
Choose more managed when: your environment is standardised and largely cloud-based; you need coverage beyond one person’s working hours; you need breadth across security, networking, and infrastructure that one hire cannot deliver; you want predictable costs; you are growing and want capacity to scale without recruiting.
Consider hybrid when: you are in the 30–100 range with mixed infrastructure and real downtime cost — which describes most businesses asking this question.
Questions to ask a managed provider before deciding
- What is realistically excluded, and what do comparable clients spend on out-of-scope work annually?
- What are the resolution targets, not just response times, by severity?
- What preventive maintenance is scheduled, and does it include backup restore testing?
- What happens during your team’s absence — is there continuity of who knows our environment?
- What documentation do we receive, and do we keep it if we leave?
- What administrative access do you hold, and how is it secured and logged?
- What reporting do we get, and how often?
- Can you provide a reference from a client of similar size and profile in this market?
Questions to ask before hiring
- Who assesses the quality of this person’s work?
- What happens during their annual leave and sick leave?
- Which of the required skill areas will they be weak in, and how is that covered?
- What is our documentation requirement, and how is it enforced?
- What is our plan when they resign?
- What tooling do we need to buy, and what does it cost annually?
What tooling do we need to buy, and what does it cost annually?
Working through this decision?
KNH IT provides managed IT services and AMC across Dubai, including hybrid arrangements alongside an in-house resource where that is the better fit.
We are happy to work through the cost model with you honestly — including telling you when hiring internally is the right answer for your situation. A client who makes the right decision is worth more than a contract that does not suit them.
See our AMC tiers: Annual Maintenance Cost Contact: Al Raffa Building, Office 07, Bur Dubai , info@knhit.com , contact us
FAQ
Beyond base salary, it includes housing and transport allowances, annual air ticket, mandatory medical insurance, visa and Emirates ID costs, end-of-service gratuity accrual, paid annual leave, recruitment cost, tooling and licensing, management overhead, and cover during absence. Gratuity accrual and leave provision are the most commonly omitted.
? Not necessarily, and the comparison is unreliable if made on salary versus monthly fee alone. Both sides have costs typically excluded — out-of-scope work and internal coordination on the managed side, tooling, absence cover, and key person risk on the in-house side. Build both over three years including the cost of downtime each model realistically delivers
Coverage gaps and skills breadth. One person cannot cover annual leave, sickness, and out-of-hours, and cannot be strong
across end-user support, infrastructure, networking, security, backup, cloud, and compliance simultaneously. Security and backup verification are the areas that most commonly suffer.
One internal person handles end-user support, on-site presence, and coordination, while a managed provider handles infrastructure, security, backup verification, patching, monitoring, out-of-hours coverage, and specialist escalation. This addresses the specific weaknesses of each pure option and often produces the best outcome for businesses in the 30 to 100 seat range.
Total hourly staff cost of everyone unable to work, plus direct revenue impact if operations stop, plus recovery time afterwards. For most businesses of this size, a single day of full downtime costs a multiple of a month’s IT support under either model, which is why the comparison should weigh risk alongside cost.
