The short answer
Build-Operate-Transfer suits companies that want to own the centre eventually but not carry setup risk now. A wholly owned subsidiary suits companies whose board has already committed and who want control and the best long-run economics above roughly 150 people. A managed capability centre suits companies that need capacity fast, have a short investment horizon, or may never want an Indian legal entity. Below 25 people, a wholly owned entity rarely pays for its own fixed costs.
Four questions that decide it
Skip the feature comparison for a moment. The model follows from four answers.
- Do you intend to own the entity in five years? If the honest answer is yes, BOT or greenfield. If it is “probably not”, a managed structure saves you an expensive detour.
- How fast do you need working people? Six weeks points to managed or extended team. Six months allows greenfield.
- What is your investment horizon? A private equity portfolio company with a three-year hold has a different calculation from a family-owned manufacturer with a twenty-year view.
- How much internal capacity do you have? Greenfield demands real time from your legal, tax, HR and IT functions. If those teams are already at capacity, you will get a slower greenfield than a BOT.
Build-Operate-Transfer in practice
The appeal is obvious: someone else absorbs incorporation, first-year compliance and the learning curve of hiring in an unfamiliar market, and you end up owning the result. The risk is equally specific — a BOT is only as good as its transfer terms, and those terms are negotiated at the beginning when the centre is worth nothing and executed later when it is worth a great deal.
Five clauses to settle before signature
- Transfer trigger. A date, a headcount, or your option at any time after a minimum period. Ambiguity here is where BOTs go wrong.
- Transfer consideration. A formula, not a future negotiation. Common approaches are net asset value plus a fixed fee, or a multiple of the build fee.
- Employee continuity. Explicit commitment that employees transfer with tenure and benefits intact, and no non-solicit that survives transfer against your own staff.
- Documentation obligation. Runbooks, compliance records and vendor contracts maintained continuously and delivered at transfer, with a defined standard.
- Post-transfer support. A named transition manager for a defined period after handover. Two quarters is a reasonable ask.
Kompass writes all five into the initial agreement as standard. If a prospective partner resists any of them, that resistance is telling you something useful.
Wholly owned subsidiary in practice
Cleanest story, best long-run economics, highest demand on your own organisation. There is no transfer event to negotiate, no intermediate employer, and candidates join your company on day one — which materially helps at the senior end.
The costs are front-loaded and real: four to eight extra weeks before the first offer, one-time setup in the USD 120,000 to 250,000 range, and sustained attention from your legal, tax and HR leadership. Below about 30 people, the fixed compliance and leadership overhead per head makes this structurally more expensive than the alternatives.
Managed capability centre in practice
The partner employs the team and holds the compliance burden indefinitely, while you direct the work and set the hiring bar. Fastest to stand up, no capital commitment, clean exit on notice.
The trade is ownership. The team is not legally yours, which affects IP assignment mechanics, long-term retention narratives, and how your auditors and acquirers view the arrangement. It is a good permanent answer for some companies and a poor one for others; the deciding factor is usually whether the work being done is core to your product.
Decision table
| If this describes you | Start with |
|---|---|
| Board approved, 150+ people planned, long horizon | Wholly owned subsidiary |
| Want ownership but cannot spare internal bandwidth now | Build-Operate-Transfer |
| PE-backed, three to five year hold | Managed capability centre |
| Need 30 people working within eight weeks | Managed, convert later |
| Still proving the concept internally | Extended team, 8–15 people |
| Highly regulated, strict data residency | Wholly owned subsidiary |
| Under 25 people with no growth plan | Managed or extended team |
Converting between models
None of these is permanent. Extended teams convert to managed centres, managed centres convert to owned entities, and BOTs are designed to convert. The thing to protect is optionality: make sure your contract permits conversion, defines how employees move, and does not price the conversion as a penalty. Design for the conversion at the start and it is administrative. Retrofit it later and it is a negotiation.