The Economics of a Nano GCC: True-Up Cost vs Sticker Price

The economics of a small capability centre are not a scaled down version of a large one. Three things behave differently, and each one inverts an assumption that holds perfectly well at five hundred people: fixed costs matter enormously, the seniority mix must be heavier rather than lighter, and cost per head becomes actively misleading as a measure.

Getting this wrong is the most common reason small centres disappoint. A twenty person centre modelled with large centre economics will be underfunded on fixed costs, understaffed on seniority, and measured on a metric that rewards exactly the decisions that will prevent it from working.

This guide sets out how the cost structure actually behaves at small scale, where the break even point sits, and what to measure instead.

The fixed cost base, and why it dominates

Every capability centre carries a set of costs that barely vary with headcount: entity formation and maintenance, statutory compliance and audit, legal counsel, payroll administration, core infrastructure and security tooling, and a minimum viable leadership layer. At five hundred people these disappear into the per head average. At twenty they are a substantial share of total cost.

This is not a reason to avoid building small. It is a reason to model the two cost types separately. A model that blends fixed and variable costs into a single per head figure will understate the cost of the first ten people and overstate the cost of the next ten, which makes every subsequent planning decision slightly wrong in a direction that compounds.

The practical consequence is that the marginal economics of a small centre improve rapidly with each addition in the early range, then flatten. Going from eight people to twenty improves the per head economics substantially. Going from eighty to a hundred barely moves them.

The seniority premium, and why it pays

The second inversion is the one that causes most of the damage. Cost models built on large centre assumptions apply a seniority pyramid with a broad base of junior and mid level engineers, because that minimises average cost per head. Applied to a twenty person centre, it produces a team that cannot make decisions.

The reason is that a large centre carries ambient context. There are architects in the building, precedent is visible, and a mid level engineer absorbs judgement from the environment. A twenty person centre has none of that, so judgement has to be present in the headcount. A small centre staffed to a large centre seniority ratio will escalate constantly, which converts its cost advantage into onshore management overhead and delay.

So a Nano GCC costs more per head than a conventionally structured offshore team, and that is the design working rather than failing. The question is whether it costs less per delivered outcome, which it generally does, because a senior team resolves its own ambiguity, needs less supervision and can own systems outright.

Where the break even point sits

Small centres reach break even faster than large ones, which is counterintuitive if you are thinking only about fixed cost amortisation. The reason is that the timeline is shorter at both ends. There is no facility build, no multi year hiring programme and no long governance construction phase, so the centre reaches productive output in months rather than in years.

A realistic path has entity and compliance work running in parallel with senior hiring, first productive output within three to four months, and the full team contributing by around month six. Against that, the costs incurred are the fixed base plus a ramping salary line, which is a much smaller commitment than a large centre carries before it delivers anything at all.

The risk profile is different too. A small centre that turns out to be the wrong decision can be wound down at a cost that is recoverable. A large centre cannot, which is why large centre decisions attract scrutiny that small ones sometimes escape, occasionally to their detriment.

Why cost per head is the wrong measure

If a small centre is judged on cost per head, the rational response is to reduce seniority, because that is the only lever with immediate effect. Doing so removes the ability to decide locally, which increases escalation, which increases onshore management overhead and cycle time. The measure improves while the thing being measured gets worse.

The alternative measure is cost per delivered outcome, which is harder to calculate and considerably more honest. It captures the fact that a senior team delivering three owned systems is worth more than a junior team consuming the same budget while implementing specifications written elsewhere.

The fourth row is the most underused. Onshore hours consumed per unit of offshore output is one of the clearest signals of whether a small centre is actually working.

What actually makes the economics work

The case for a Nano GCC is not that it is a cheap way to get engineers. On a pure rate basis, a staffing arrangement is usually cheaper and involves no fixed cost base at all. The case is that a small owned team with senior depth can hold context and own outcomes, and that ownership is what produces value that compounds.

Common modelling mistakes

Each of these produces a model that is internally consistent and wrong in a way that only becomes apparent in year two.

• Blending fixed and variable costs into a single per head figure
• Applying a large centre seniority pyramid to a team of twenty
• Assuming productive output from month one rather than modelling a ramp curve
• Omitting onshore management overhead, which is the cost the centre is supposed to reduce
• Carrying no attrition assumption, which makes year two look like a failure rather than a gap in the model
• Comparing against an unloaded domestic salary rather than a fully loaded one

Frequently asked questions