Hidden Costs Companies Overlook When Building a GCC

Offshore business cases are rarely wrong about salary. Salary is the number that gets scrutinised, benchmarked and negotiated, and it is usually accurate within a few percent. Business cases fail on the costs that sit around salary, and specifically on the ones that never appear on an invoice.

This is what makes them hidden. They are not concealed and nobody is being dishonest. They are simply costs that arrive in a different budget, in a different period, or as consumed time rather than as spend, and standard cost modelling has no natural place to put them. By the time they become visible, the savings figure has already been announced.

This guide covers the seven costs that most reliably surface between month nine and month twenty four, why each one is missed, and what a realistic model does about it.

1. Ramp cost, which recurs

A new hire carries full cost from day one and produces useful output some months later. That gap is a real cost, and in a business case that assumes productivity from the start it simply disappears into an unexplained first year variance.

The reason it is treated as a one off is that it feels like one. In practice, any centre that grows has a ramp cohort in flight almost permanently, which makes it a running cost. A model that treats ramp as a start up expense will diverge from reality every time the centre expands.

Model it as a productivity curve rather than a switch. Assuming a new hire reaches steady state over three to six months, with a slower curve in complex or regulated domains, produces a first year projection that matches what actually happens.

2. Knowledge transfer drag

When an onshore team teaches an offshore team, the onshore team slows down. That lost output is a genuine cost of the transition, and it is almost never modelled because it lands in a different budget from the one under scrutiny.

It also lands on precisely the people you can least afford to slow down. Knowledge transfer is done by the engineers holding the most context, who are usually also carrying the most critical work. The effect is that overall delivery dips during transition by design, and because nobody modelled it, the dip gets attributed to the new team underperforming.

3. Onshore management overhead

Running a capability centre consumes onshore leadership time: hiring panels, governance, escalation, coordination, travel and the general overhead of operating across time zones. This is real capacity, it is expensive capacity, and it never appears on an invoice because it is time rather than spend.

It is highest in exactly the period when the business case is being validated, which is the first year, and it falls as the centre matures and takes on more decisions locally. That relationship is worth stating explicitly, because it means the overhead is a function of how much authority the centre has. A centre kept on a short leash generates permanent onshore overhead that a centre with real decision authority does not.

4. Fixed costs that do not scale down

Entity formation, statutory compliance, annual audit, legal counsel, payroll administration, core infrastructure and minimum viable leadership do not vary much with headcount. At five hundred people these amortise to almost nothing per head. At twenty people they are a significant proportion of total cost.

This is the single most important cost consideration for small centres, and it is the reason the old rule of thumb about needing several hundred people existed. That threshold has fallen substantially, because entity setup, compliance and infrastructure have all become far cheaper and faster, but the shape of the problem has not changed. Fixed costs still hit small centres disproportionately, and a model that assumes per head costs scale linearly will understate a small centre s cost base considerably.

5. Attrition and replacement

Recruitment is usually modelled as a one off setup cost. Attrition converts it into a permanent line. Each departure generates recruitment cost, notice period overlap where both people are paid, a fresh ramp curve for the replacement, and a period during which the team is short handed.

The compounding effect is context loss rather than cash. An engineer who leaves after eighteen months takes accumulated understanding with them, and the replacement starts from zero on a system that has grown more complex in the meantime. A model with no attrition assumption will diverge from reality permanently rather than temporarily, and the divergence widens as the centre ages.

6. Travel and in person time

Distributed teams need periodic in person contact, particularly in the first year, and it is genuinely valuable rather than discretionary. Onshore leaders travelling to the centre, centre leads travelling to headquarters, and onboarding visits for senior hires all cost real money and real time.

It is usually omitted because it feels like a travel budget question rather than a capability centre question. It is neither large nor catastrophic, but it is consistently underestimated in year one, and cutting it is a false economy that shows up later as weaker working relationships and higher escalation.

7. The cost of getting the seniority mix wrong

This is the most expensive hidden cost and the hardest to see, because it appears as a saving. Reducing the seniority mix lowers cost per head immediately and visibly. What it does not show is that a team without sufficient senior judgement escalates more, delivers less per person, consumes more onshore management time, and cannot take ownership of anything.

The result is a centre that looks efficient on a cost per head basis and is expensive on a cost per delivered outcome basis. Because almost nobody measures the second, the mistake is usually invisible until the centre has been running for two years and is still unable to own a system.

The last row is the only one that gets worse the more successfully you optimise the others.

What a realistic model does differently

None of this argues against building a capability centre. The savings are real and, in most cases, substantial. The argument is for modelling the whole number, because a centre that is genuinely delivering value should not have to defend itself against a business case that was never achievable.

• Separates fixed costs from per head costs rather than blending them into an average
• Applies a productivity curve to new hires instead of assuming immediate output
• Includes an explicit estimate of onshore output loss during transfer
• Carries a realistic annual attrition assumption from year one
• Tracks cost per delivered outcome alongside cost per head
• Is reconciled quarterly in the first year, with variance explained rather than absorbed

Frequently asked questions