Everyone's Launching a GCC in India. Most Get Month One Wrong.

A decision-maker's field guide to what a GCC is, how it is set up, how it runs, and where launches actually lose time.

The Problem GCC Launch Plans Keep Papering Over

A European industrial group approved its India centre in March with a go-live target of October. The programme plan was thorough in the places programme plans usually are: entity structure, transfer pricing, a shortlisted building in Bengaluru, an IT build with a dependency chart, a communications plan for the functions losing work.

The entity was registered in nine weeks. The lease was signed in eleven. The network was live in August. And the centre opened in February, four months late, because at go-live it had 22 of the 90 people it needed and roughly half of those had been hired against roles that turned out to be described wrongly.

This is the ordinary shape of a late GCC. The parts of the launch with contracts and named vendors behave predictably; the part that depends on finding and validating several dozen specific people does not, and it is almost always the part planned in least detail. Talent tends to appear in the plan as a single line with a headcount number and a recruitment partner against it.

The market context makes the squeeze tighter. The Zinnov–nasscom India GCC Landscape 2026 report counts 2,117 GCCs across 3,728 units in India, employing about 2.36 million people and generating roughly $98.4 billion in revenue — with around 900 GCC units in Bengaluru alone, about 35% of the country's GCC workforce. Every new centre is recruiting into a market where hundreds of established ones are already hiring.

This ReadyForRole guide covers what a GCC actually is, the four ways to set one up, the five stages of running one, and why the critical path runs through stage two.

What a Global Capability Centre Actually Is

A Global Capability Centre is a wholly-owned, in-country unit that a global company builds in order to own critical work rather than outsource it — engineering, finance, data, R&D, HR and customer operations. It is not a call centre and not a cost-cutting shortcut; it is a permanent extension of headquarters, built to last, with its own leadership, its own capability and its own claim on the roadmap.

That ownership is the whole point, and it explains why boards keep approving them. Cost efficiency without a quality trade-off: global-grade output at a fraction of onshore cost structures. Direct access to deep talent pools: engineering, analytics and product talent at a scale few markets match. Round-the-clock delivery: a time-zone advantage that shortens product and support cycles. And full ownership with no vendor dependency: the IP, the process and the roadmap stay inside the company.

The last of those is what separates a GCC from outsourcing, and it is also what makes the talent question harder. An outsourcing contract transfers the staffing problem to a supplier. A GCC keeps it — which is the correct trade for control, but only if the staffing problem is planned as seriously as the entity and the building.

ReadyForRole plans the talent pipeline against the ramp curve rather than the headcount total, because a centre that needs 90 people by October does not need 90 in October; it needs the right 20 in June so that the next 70 have somewhere to land.

One-sentence takeawaya GCC is an ownership decision, not a cost decision — and the thing you take ownership of first is the talent problem the outsourcing model used to hide.

Where This Shows Up in the Enterprise

Global HQ Leadership — Choosing the Set-Up Model

Current pain: the model is chosen on speed and legal comfort alone, without asking which one leaves capability inside the company at the end. There are four common routes. A wholly-owned subsidiary means registering the entity, hiring directly and owning the IP outright — maximum control, maximum accountability. Build-Operate-Transfer has a partner set it up and run it, then transfer full ownership once operations are stable. A third-party managed GCC leaves an external partner owning delivery long-term while you retain outcomes and governance. A joint venture or co-location launches fast inside a partner's entity or plug-and-play space with local expertise built in.

Targeted design: the model is chosen against the capability endpoint, not just the start date — and in transfer or managed models, the knowledge-retention terms are written at contract stage rather than negotiated at handover, when leverage has gone.

GCC Programme Director — Sequencing the Five Stages

Current pain: the five stages are run as a waterfall — entity setup and compliance, talent acquisition and onboarding, process transition and ramp-up, steady-state delivery, then scale and innovation — with talent starting only once the entity exists. Since hiring specialist roles takes months, that sequencing guarantees the centre is staffed after it is supposed to be running.

Targeted design: talent work starts in parallel with entity setup, not after it. Role specifications, assessment design and pipeline building happen while lawyers and landlords do their part, so offers can go out the week the entity can legally make them.

Site Leader — Hitting Ramp Targets Without Inflating Attrition

Current pain: behind schedule, the centre hires fast against a slipping target. Year-one attrition then runs high because a meaningful share of the intake was never a fit, and each exit costs more than the vacancy did — it costs the ramp of everyone the leaver was supporting.

Targeted design: the readiness bar stays fixed while the ramp plan flexes. Candidates are validated against the role before offer, intake is banded by verified readiness, and the supervision load per band is planned rather than discovered by managers in month three.

Across all three, the ReadyForRole rule holds: infrastructure risk is bounded and contractual, while talent risk compounds — so the talent workstream deserves the same planning depth as the entity, and it should start on the same day.

The Failure Modes Nobody Puts in Their Deck

These are the patterns ReadyForRole has seen quietly delay GCC launches — paired with the decisions that survive them.

Talent planned as a headcount number. The plan says "90 FTE by October" and nothing about which roles are scarce, which are interchangeable, or what the ramp curve must look like for the centre to function at each stage. Design decision: plan talent as a sequenced curve of specified roles, with the scarce ones started first and the leadership seats started before everything else.

Recruitment started after the entity is registered. Legal and talent are treated as sequential because the entity must exist before an offer letter can issue — but specification, assessment design and pipeline building need no entity at all. Design decision: run the talent workstream in parallel from day one and hold offers, rather than holding the whole process.

Local market reality absent from the plan. A plan written at headquarters assumes notice periods, salary bands, commute tolerances and counter-offer behaviour from a different market. Bengaluru in particular, with roughly 900 GCC units competing, behaves nothing like the assumption. Design decision: build the ramp curve on local notice periods and local competitive behaviour, and validate the bands before the plan is approved.

Knowledge transfer scheduled as an event. Stage three is booked as a series of shadowing weeks, after which the India team "owns" the process. Documentation is thin, the source team is busy, and ownership transfers on paper only. Design decision: make transfer outcome-based — the receiving team runs the process unaided for a defined period before sign-off — and resource the sending team for it.

Steady state treated as the destination. The programme closes when SLAs are met, so the centre optimises the work it was given and never builds the case for more. Two years later it is still doing exactly what it was set up to do. Design decision: fund stage five from the start, with explicit capability-building goals beyond the transition scope.

Attrition modelled at the company average. First-year GCC attrition in a competitive city is not the global average, and a ramp plan built on the global figure is short of people from month six onward. Design decision: model local first-year attrition explicitly and build the buffer into the hiring curve rather than discovering it.

Actionable Takeaways

• Choose the set-up model against where capability ends up, not only against the launch date.
• Start the talent workstream on day one, in parallel with entity and infrastructure work.
• Plan talent as a sequenced ramp curve of specified roles, not a single headcount number.
• Write knowledge-retention terms into BOT and managed contracts at signing, not at handover.
• Build the ramp on local notice periods, salary bands and attrition — not headquarters assumptions.
• Make process transition outcome-based: the receiving team runs it unaided before sign-off.
• Fund stage five — scale and innovation — from the beginning, or the centre will stop at steady state.

In ReadyForRole's GCC engagements, the launches that hold their date are the ones where talent had a workstream lead with the same seniority as the entity and infrastructure leads — not a line item under operations.

Enterprise Decision Framework: The ReadyForRole GCC Launch Gate

Use this checklist before signing off a GCC launch plan.

1. Model fitDoes the chosen set-up model leave capability where you need it at the end, not just at launch?Yes · No · Partial
2. Parallel talent trackDoes talent work start on day one rather than after entity registration?Yes · No · Partial
3. Ramp curveIs talent planned as a sequenced curve of specified roles rather than a headcount total?Yes · No · Partial
4. Scarcity mappingHave you identified which roles are genuinely scarce locally and started those first?Yes · No · Partial
5. Local assumptionsAre notice periods, salary bands and first-year attrition modelled on local data?Yes · No · Partial
6. Transfer definitionIs knowledge transfer defined as an outcome the receiving team demonstrates, not a calendar of sessions?Yes · No · Partial
7. Readiness barIs the hiring standard fixed, with the ramp plan as the variable when timelines slip?Yes · No · Partial
8. Stage five fundedIs there a resourced plan for capability growth beyond the original transition scope?Yes · No · Partial

The gate ReadyForRole sees teams skip most often is the parallel talent track — and it is the one that decides whether the centre opens on time, because everything else on the plan has a contract behind it and talent does not.