The India Global Capability Centre boom is no longer a question of whether, it is a question of how. Most multinationals of scale either operate a GCC in India today, or are weeks away from approving one. What separates the centres that succeed at multi-thousand-seat scale from those that re-paper themselves eighteen months in is rarely talent or real estate. It is the legal architecture set at the moment of incorporation.
Four design choices, made at the entity stage, compound over the life of a GCC. Get them right and the next decade of growth proceeds on plan. Get them wrong and every later move, a sub-function expansion, a leadership hire, a customer-facing service offering, an acquisition, is taxed by the original error.
The first choice is entity form. Most GCCs are incorporated as private limited companies, wholly owned by the foreign parent. This is the default and, for most operating models, the correct answer. But the default is not universal. GCCs that anticipate offering services to third-party customers in India, GCCs that will hold material intellectual property, and GCCs that may be carved out or independently financed in future, each have reasons to consider alternative structures, including limited liability partnership forms, holding-company layers, and IP-owning sub-entities. The right answer depends on a roadmap that often is not yet written when the entity is incorporated. The discipline is to write enough of the roadmap to choose deliberately.
The second choice is employment infrastructure. A GCC's employment paper, contracts, handbooks, IP assignment language, non-solicit terms, performance frameworks, is the single most-litigated artefact of its existence. It is also the artefact most commonly inherited, unmodified, from the parent's home jurisdiction. Indian employment law does not accept this. Restrictive covenants, particularly post-termination non-competes, are largely unenforceable. IP assignment requires specific consideration. POSH compliance is statutory, not optional. A GCC built on imported paper is a GCC that will discover, in its first material exit dispute, that its paper is not the paper its parent thought it was.
The third choice is data and IP architecture. Most GCCs are, in substance, processing personal data on behalf of the parent and developing intellectual property that is intended to vest in the parent. Both of these positions require explicit, written, well-papered arrangements, a data processing agreement between the GCC and the parent, and an IP assignment regime that survives both Indian tax scrutiny and parent-jurisdiction patent prosecution. Neither is exotic, but both are commonly under-papered in the rush to operationalise. The cost of fixing them after the fact, particularly the IP regime, is materially higher than the cost of getting them right at the start.
The fourth choice is governance. A GCC operating at fifty seats with informal weekly calls between the India MD and the global function head is a GCC governed by personal relationships. A GCC at five hundred seats requires a formal governance framework, board composition, delegation matrices, reporting cadence, escalation paths. The transition between these states does not happen at a moment; it happens through accretion. The GCCs that scale well are the ones that put formal governance in place before they need it. The ones that struggle are the ones that try to retrofit governance during a crisis.
None of these four choices are individually difficult. The difficulty is that they must be made together, and the people who make them well are the people who have made them before. The compounding logic is unavoidable: a GCC built on a coherent set of foundational choices accumulates value for a decade. A GCC built on inherited defaults accumulates remediation work for the same period.
Privileged commentary · Not legal advice · © Zuber & Partners