Building a GCC in India, answered plainly

A Global Capability Centre (GCC) is your own team in India. They sit on your payroll and work on your roadmap, instead of the work being handed off to an outsourcing vendor. You will also hear it called a captive centre or a Global In-house Centre. What sets it apart is ownership. The people, the IP, and the knowledge all stay yours.

Outsourcing rents capability from a vendor. A GCC builds capability you own. In a BPO or outsourced model, the provider's team does your work, keeps the IP, and adds a margin on top. In a GCC, the team is yours, the knowledge stays in-house, and you decide how the function runs.

A GCC usually starts to make sense at around $300M in revenue. Below that, the cost of an entity, office, and leadership team is harder to justify. Four groups drive most of the demand. Small and emerging companies using lighter models, mid-market firms with $300M to $5B in revenue, PE-backed companies, and large enterprises. Mid-market is now roughly half of India's GCCs, and it is the fastest-growing segment.

Most GCCs start between 50 and 100 people. Some mid-market centres break even at just 25 to 50. Below that range, compliance and leadership overhead starts to outweigh the labour savings. That is why very small teams usually begin on an Employer of Record model, and only set up their own entity once they scale.

Here is a simple test. If the work does not need someone physically present, it can move. That covers technology, engineering, R&D, analytics, AI and machine learning, finance, HR, marketing, customer service, procurement, supply chain, and project management. Technology, engineering, and R&D usually go first. Finance, HR, and other support functions are added once the centre is up and running.

Setup cost scales with size. A 50-person pilot runs roughly $550K to $3M, and it climbs as headcount grows. Entity registration alone takes about four to eight weeks. Phased or advisory-led models can bring the realistic entry cost down to around $500K to $2M by spreading it out. Running a few-hundred-person centre then usually costs about $1.8M to $2.5M a year. These are market estimates and will vary with your skill mix and location.

Usually within 18 to 30 months for a straight build. Outsourcing tends to look cheaper in Year 1, because there is no entity, office, or hiring cost to absorb. The business case turns positive around Year 2, once the centre is staffed and ramped. By Year 3 the GCC is usually clearly ahead, and the gap widens each year after that. A phased model pulls that break-even point earlier.

Over three to five years, GCCs tend to be 20 to 35% more cost-effective than outsourcing, once you count vendor margins, attrition, and management overhead. Outsourcing, BPO, and EOR contracts usually carry an 18 to 45% vendor margin no matter how they perform, and the cost steps up at every renewal. A GCC's cost per head stays comparatively flat.

Yes. An Employer of Record (EOR) model lets you hire and operate in India without first setting up your own entity. That suits pilots and smaller teams testing the market. Many companies then follow a phased path. They start on EOR, move to Build-Operate-Transfer to scale, and set up their own entity later. That way the upfront cost and risk are spread across stages, rather than paid all at once.

A captive is fully owned and run by you. It is still the most common model globally, at around 58% of GCCs. GCC-as-a-Service means a partner runs the day-to-day under your brand and governance. It is smaller but faster-growing, and popular with mid-market firms. Build-Operate-Transfer sits in the middle. A partner sets up and runs the centre, then hands ownership over to you once you are ready.

A few things matter most. A real track record of setting up and scaling centres, genuine local market and location knowledge, speed and certainty of launch, transparent pricing, and existing trust and sector relevance. A GCC is a high-stakes, long-term decision, so peer references and proven outcomes count for more than a headcount figure or a list of centres opened.

Start with the gap between what a provider promises and what they can really deliver, because that is where a lot of them fall short. Push hard on timelines. Dates for workspace, the legal entity, and hiring are often more hopeful than real, so ask what they are actually committed to and what happens if those dates slip. Ask how change management and governance will work once you go ahead, and bring your own workstream leads in early, or the original goal and the messaging start to drift. And get an honest view of how much of your global HR policy will actually transfer, and who will own statutory, corporate, and labour compliance.

Finism is a founder-led, hands-on advisory built for the mid-market and PE-backed companies that bigger players often underserve. You work directly with experienced practitioners, not a rotating team. You get personalised support across the whole journey and one clear point of accountability. Bigger firms bring scale and cost. Finism brings focus, practitioner experience, and ownership.

Ownership on your side. The builds that go wrong are usually the ones left almost entirely to the provider, with no one internal truly accountable for the outcome. Nominate a single senior governance lead who owns the setup end to end, brings your workstream leads in early, and holds both sides to the plan. Once that person is in place, everything else, the timelines, hiring, compliance, and cost, is far easier to keep on track.