In House vs Managed OCI Operations

Once an OCI estate is live, someone has to run it, and the choice is between building that capability inside the organisation or contracting it to a managed service. This is one of the more consequential decisions a technology leader makes, because it shapes cost, risk and where attention goes for years. It is also a decision often made badly, either by defaulting to in house out of a vague preference for control, or by outsourcing to save money without understanding what good operation actually requires. This guide lays out both paths honestly so the choice can be made on the facts.

What running OCI well actually requires

Before comparing the two models, it is worth being clear about what the work involves, because the cost of in house operation is routinely underestimated. Running an estate to a reliable standard requires people who are available when problems occur, which for any serious estate means cover outside business hours. It requires skills that stay current on a platform that changes constantly. It requires enough people to cover holidays, illness and the inevitable departures without leaving gaps. And it requires the discipline to do the quiet, unglamorous work of patching, restore testing and capacity tuning even when nothing is on fire. None of this is exotic, but all of it is real, and the sum is more than a single hire.

The two models compared

The true cost of an internal team

The case for in house often rests on a cost comparison that counts only salaries, and that comparison is misleading. A credible internal operations capability is not one person, because one person cannot provide cover, cannot go on holiday, and represents a single point of failure if they leave. It is a small team, with the overhead of recruitment, retention in a competitive market, training to stay current, and management. The work is also lumpy, quiet for long stretches and then intense during an incident or a migration, which means an internal team is either over staffed for the quiet times or under staffed for the busy ones. When the full cost is counted honestly, the fixed expense of a capable internal team is high, and it is paid whether or not the estate needs much attention in any given month.

The case for in house

None of this means in house is wrong. For some organisations it is clearly right. A large enterprise running many estates can build an operations team that is busy enough to justify its fixed cost and large enough to provide genuine cover and depth. An organisation whose estate is so central to its business that it considers operational capability a core competence may rightly want it inside. And some regulated environments have constraints that favour internal control. The honest test is scale and centrality: if the estate is large enough to keep a real team busy and central enough to justify owning the capability, in house can be the right call. Below that threshold, the economics usually point the other way.

The case for managed

For most organisations the managed model fits better, for a simple reason: it converts a high fixed cost into a predictable variable one and spreads the burden of cover, skills and resilience across a provider serving many clients. The provider can afford continuous coverage because it is shared. It keeps skills current because operation is its core business. It has the depth to absorb an absence or a departure without a gap. And it does the quiet disciplines as a matter of routine rather than as work that competes with everything else. The trade off is a degree of control, exchanged for a level of reliability and economy that most internal teams struggle to match below enterprise scale.

The skills problem nobody plans for

The hardest part of in house operation is not standing the team up, it is keeping it. Cloud skills are in high demand, and an engineer you train on OCI becomes more valuable to the wider market exactly as they become more valuable to you. Retaining them means paying the market rate, offering interesting work, and accepting that some will leave anyway. When a key person goes, they take operational knowledge with them, and if that knowledge lived in their head rather than in documentation, the gap is sharp and immediate. This is the quiet risk that in house plans underweight: not the cost of the team on a good day, but the fragility of the team on the day someone resigns. A managed provider absorbs this risk because its depth means no single departure leaves a client exposed, and its knowledge lives in shared runbooks rather than individual memory. For an organisation whose estate is critical but whose team is small, this resilience is often the deciding factor on its own.

Control is not all or nothing

The instinct toward in house is often really an instinct for control, and it is worth examining because control and ownership are not the same thing. Handing the daily run to a provider does not mean losing control of the estate, provided the relationship is set up well. You still own the architecture decisions, the strategy, the budget and the standards the provider operates to. What you delegate is the execution of the run, governed by an agreement that defines what good looks like and reporting that tells you whether it is being achieved. The control that matters, over direction and standards, stays with you, while the control that costs the most to retain, over staffing a round the clock operation, moves to a provider who can do it more efficiently. Seen this way, the choice is less about surrendering control and more about deciding which kind of control is worth paying to keep, which usually points toward owning the strategy and delegating the execution.

A decision framework

The hybrid middle ground

The choice is not strictly binary. A common and effective arrangement keeps a small internal capability focused on architecture, strategy and vendor oversight, while delegating the daily run to a managed service. This gives the organisation enough internal knowledge to govern the relationship intelligently and make architectural decisions, without carrying the full fixed cost and resilience burden of an internal operations team. For many organisations this hybrid is the most pragmatic answer, combining the control of in house oversight with the economy and reliability of a managed run.

This decision sits at the centre of the managed services question. For the full context see the complete guide to OCI managed services, and weigh the coverage and service level dimensions in support models and SLAs explained. When you want to explore what a managed run would cost and cover for your estate, our OCI managed services practice will scope it honestly against your alternatives.