Omanisation and Your Technology Team: In-House, Outsourced, or Blended

Most companies ask this as a cost question. It is not a cost question. Over three years the two options usually land closer together than anyone expects, and the difference that actually matters is where the knowledge about your systems ends up living — inside your company, or inside someone else’s.
A second consideration sits alongside it in Oman and nowhere else in the comparison: your Omanisation position is a ratio, and the two options move it in opposite directions.
We should declare an interest before going further. ZAWAT is an outsourcing supplier. That is a reason to read the section on what outsourcing costs you with attention, not to skip it — it is the section we had the most incentive to soften, and did not.
The three models, stated plainly
In-house. You employ the people. You carry salaries, employer costs, equipment, licences, management attention and the recruitment risk. You also carry the ratio, in both directions — every expatriate hire moves it against you, every Omani hire moves it for you.
Outsourced. A supplier delivers against a scope. You carry a contract instead of a payroll, and your Omanisation ratio is unaffected by the size of the team doing the work, because they are not your employees.
Blended. A small in-house core owns the decisions, the requirements, the acceptance and the documentation; outside capacity does the building. This is the arrangement most mid-sized companies end up in after trying one of the other two, and it is usually reached by accident rather than design.
What the quota obligation actually means in practice
The Labour Law promulgated by Royal Decree 53/2023 makes employment of Omanis an obligation on the employer rather than an aspiration. Three provisions shape the decision.
The quota is set per sector by ministerial decision, not written in the Law. That is deliberate — it lets the figure move with the labour market — and it is exactly why no article should print the number for your activity. It is attached to your classification, and it is revised.
You owe the Ministry a plan. Employers are required to file an annual workforce-localisation strategy. An establishment below its quota is not simply fined; it is expected to show how it intends to close the gap.
At twenty-five employees, a second obligation appears. Employers at or above that size must have a plan for selecting and training Omani employees for leadership positions — and implement it. This is the provision most often discovered late, because it is about succession rather than headcount.
The penalty band, as summarised by counsel practising in Oman, runs from OMR 250 to OMR 500 for each Omani employee the establishment should have employed, with a six-month grace period to rectify, after which it doubles.
What changed in January 2026
Ministerial Decision 602/2025, the regulation governing work permits and work practice permits, came into force at the end of January 2026 and did something more interesting than raise a fine. It classified private-sector establishments into three categories — green, yellow and red — by Omanisation performance, and attached prices to them:
- Green — meeting or exceeding the sector quota — receives a 30% reduction on work permit fees, work practice permits and employee data registration.
- Yellow and red pay twice the standard rate on every expatriate permit and professional licence they hold.
The gap between the best and worst category is therefore not a penalty you might incur. It is a multiple on a fee you pay anyway, every year, on every expatriate you employ. For a technology function that leans on expatriate hiring, that is a line item with a coefficient in front of it.
The regime is also visibly in motion: MD 602/2025 was itself amended by Ministerial Decision 44/2026, recorded in January 2026, within months of taking effect. Plan on the mechanism, not on a specific figure.
The ICT sector specifically
MTCIT has published direction for the communications and information technology sector. In an announcement reported on 14 July 2024, the Ministry gave a 2024 baseline of 31% Omanisation in the ICT sector and set out rates for professional jobs ranging between 50% and 100% by 2026, with certain jobs allocated exclusively to Omanis, rolled out from January 2025 through 2027.
Read that range carefully rather than as a headline. It is a range across job categories, not a single figure for a company, and the binding number is the one attached to your commercial activity in the Ministry of Labour’s own classification — not the one in a news report or in this article. Confirm your current applicable rate with the Ministry before you plan hiring against it.
One more provision worth knowing because it catches small establishments: a decision reported in June 2025 requires at least one Omani employee for every commercial registration that is a year or older.
The honest version of the ratio argument
Outsourcing improves your Omanisation position, mechanically, because contracted people are not in your denominator. That is true and it is legitimate.
It is also the argument most often used to justify a decision that costs more in knowledge than it saves in ratio. If the technology function is where your operating advantage lives, moving it off your payroll to improve a percentage is optimising the measurement rather than the business. The blended model exists precisely because you can often satisfy both: hire the small core — which is the part you can realistically fill with Omani talent, and the part where the National Digital Upskilling pipeline is actually producing candidates — and contract the delivery peak.
The true cost of an in-house team of three
Three is the smallest team that holds knowledge and can still take leave. Two is a single point of failure with extra steps: one resignation and the second person becomes the only living source of truth about a system the business runs on.
Salary is the number everyone models. These are the ones that decide whether the model was right:
Employer costs beyond salary — social insurance, end-of-service, leave, air fares where applicable, medical cover. Model the loaded cost, not the offer letter.
Recruitment and the vacancy. Both the fee and the months. A role that takes four months to fill costs four months of the work not being done, and that cost lands on whoever is covering.
Equipment, licences and tooling. Per seat, per year, recurring. Small individually and consistently underestimated in aggregate.
Management attention. Someone senior must set priorities, review work and say no. If nobody does, the team drifts toward interesting problems rather than valuable ones. This is the cost that never appears on the spreadsheet and most often decides the outcome.
The bench. An in-house team is a fixed cost against demand that is not fixed. In the quiet quarter you pay for capacity you are not using; in the busy one you cannot buy more of it quickly.
Breadth. Three people cannot hold interface design, backend, mobile, infrastructure, security and data. They will be strong in two or three of those and thin in the rest, and the thin ones are where the expensive mistakes happen.
The same arithmetic discipline applies here as in what custom software really costs: the build is the visible number and the smaller half of the total.
What outsourcing genuinely costs you
Not money. Knowledge — and it leaves in three specific pieces.
The rationale behind the domain model. Why the system treats a customer this way and not that way, which edge case forced that field, which rule came from a regulator and which from a preference. This lives in conversations, and if the conversations happen in someone else’s building, so does the knowledge.
The operational runbook. What breaks, how often, what you do about it, what the workaround is at two in the morning. Suppliers accumulate this and rarely write it down unless required to.
The dependency map. What depends on what, and what will break if you change this. This is the knowledge that makes a change cheap or expensive, and it is invisible until you need it.
Three further costs are real but more visible: latency on small changes, priority competition with the supplier’s other clients, and staff turnover inside the supplier that you do not see until the person who knew your system is gone.
The failure mode this produces is well documented and has a name: no owner on the client side. It is one of the recurring patterns in why ERP and CRM projects fail, and it is not caused by outsourcing. It is caused by outsourcing without keeping an owner.
The blended model, and the handover that makes it real
The in-house core owns the requirements, the decisions, the acceptance, the credentials, the supplier relationship and the documentation. It does not necessarily write code.
The minimum viable core is one person who owns the systems — who can say what the business needs, decide when something is done, and hold the vendor to the contract. That person is more valuable to you than a third developer, and the role is one you can realistically fill locally, because it rewards business understanding over years of framework experience.
The thing companies get wrong is treating handover as an event at the end. Handover is a deliverable produced continuously, and it consists of four things: source code in a repository your company owns, documentation written for the next engineer rather than for the invoice, an operational runbook, and recorded sessions where decisions are explained rather than only described.
If those four exist, switching supplier is a manageable project. If they do not, you are not outsourcing — you are dependent, and the price of that dependency is discovered on the day you want to leave.
Contract clauses that protect knowledge transfer
Six. They belong in the agreement, not the kick-off meeting:
- IP assignment on payment, stated explicitly, covering source code, designs and data.
- Your repository from day one. Not a delivery at the end — commits into infrastructure you own, from the first week.
- Documentation as an acceptance criterion. If the definition of done does not include it, it does not get written; nobody has ever been paid extra for it.
- Named individuals, with notice of change. You are buying particular people, whatever the contract says about resources.
- Credentials and accounts in your company’s name — hosting, domains, repositories, third-party services. This is the single most common way a company discovers it does not control its own system.
- Exit assistance, with a defined period and an agreed rate, so leaving is a clause rather than a negotiation.
These overlap deliberately with the criteria in how to choose a software partner in Oman, because they are the same question asked at a different moment.
How to actually decide
Answer three questions in order.
Is the system a differentiator or a utility? If competitors could buy the same thing and you would be no worse off, outsource it. If the way it works is part of why customers choose you, keep the decisions in-house even if the building is not.
How often will it change? Continuous change rewards an in-house core. A system that is built and then largely left alone does not.
Can you hire the core role? Not three developers — the one owner. If you can, the blended model is available to you. If you genuinely cannot, be honest that you are buying a dependency, and price the exit clauses accordingly.
Frequently asked questions
Does outsourcing affect our Omanisation percentage? Contracted personnel employed by a supplier are not your employees, so they do not sit in your ratio. That is a genuine effect, and since January 2026 it also touches fee category under Ministerial Decision 602/2025. It is not a reason on its own to move a function you should own — confirm your position with the Ministry of Labour rather than reasoning from a general rule.
What is the Omanisation percentage for IT companies? It is set by ministerial decision for your sector and activity and it is revised, so no article should give you a number to plan against. MTCIT has published direction for the ICT sector with rates for professional jobs ranging between 50% and 100% by 2026, including jobs reserved entirely for Omanis. Get the figure that applies to your commercial registration from the Ministry.
Is it cheaper to hire developers or use an agency? Over three years, usually closer than expected. In-house looks cheaper per hour and costs more in idle capacity, management attention and recruitment risk; outsourcing looks more expensive per hour and costs more in knowledge. Compare total cost over three years, and include the cost of a vacancy and of the wrong hire.
We already outsource everything and feel locked in. What now? Start with the six clauses above, applied to what you have: get credentials into your own name, get the code into a repository you own, and commission documentation as a paid piece of work rather than a favour. Then hire the owner. Doing this while the relationship is good is far cheaper than doing it while it is ending.
How large does a company need to be before an in-house technology team makes sense? Size matters less than change frequency and criticality. A company where systems change every week and an outage stops revenue needs an in-house core well before a larger company whose systems are stable. The trigger is usually the first time a delay in a small change causes a real business problem.
This article summarises published positions as at 21 August 2026 and is not legal or employment advice. Omanisation rates, decisions and penalties change; confirm your obligations with the Ministry of Labour or a qualified adviser before acting.
Sources: Royal Decree 53/2023 — the Labour Law · Ministry of Labour — Labour Law · Ministerial Decision 602/2025 (record) · Ministerial Decision 44/2026 amending MD 602/2025 (record) · Muscat Daily — fees doubled for firms failing Omanisation targets, 16 February 2026 · Oman Observer — MTCIT announces new Omanisation targets, 14 July 2024 · Paul Hastings — Oman employment law overview
Deciding how to staff a technology function? ZAWAT builds custom software and works alongside in-house teams rather than replacing them, with handover written into the contract. Book a call.