India added more than 500 new capability centers in five years, and almost every one of them was sold by a partner whose deck looked identical to the one on your screen right now. Same maturity curve. Same city comparison slide. Same promise of “day-one delivery readiness.”
The decks are interchangeable because the differentiators aren’t in the pitch; they’re in the operating detail that nobody volunteers. Which entity structure they actually filed last time. What the twelve-month attrition was at the last three centers they built. What happens to your employment contracts on the day you transition the center in-house.
That’s the gap this guide closes. The questions to ask GCC setup partner firms are not the ones on the RFP template your procurement team downloaded; they’re the ones a weak partner physically cannot answer without exposing that they’ve built two centers, not twenty.
India now hosts 2,117 GCCs operating across 3,728 units, employing 2.36 million professionals and generating USD 98.4 billion in revenue as of FY26, a 32% increase in center count since FY2021.
That growth created a second market underneath it: enablers, advisory shops and staffing firms repositioned as GCC specialists. Some have built thirty centers. Some have built one and written forty blog posts about it. The thirty questions below are sorted so you find out which, inside a single 60-minute call.
TL;DR
This is a due-diligence checklist for enterprise leaders who have already shortlisted two or three vendors and now need to separate them. It covers entity and compliance, talent and attrition data, commercial and exit terms, governance, and reference checks.
The single sharpest filter is question 12. Ask for twelve-month attrition at the last three centers the partner built, by role band. EY's India research puts average GCC attrition at 12.5%, with roughly 60% of centers landing between 5% and 15% so a partner who cannot produce their own number against that benchmark is telling you something.
By the end you'll have a scoring framework, the four contract clauses worth negotiating hardest, and a clear read on which vendor is selling capability versus selling a slide. Working through this gcc partner due diligence list before you sign is the cheapest risk reduction available to you.
What GCC partner due diligence actually means
GCC partner due diligence is the structured evaluation of a vendor’s legal, talent, commercial and governance capability before they are contracted to establish your capability center. It verifies claimed track record against evidence entity filings, attrition data, reference centers and exit terms rather than accepting capability statements at face value.
Why the shortlist all sounds the same
The problem isn’t that partners lie. It’s that the pitch stage rewards abstraction, and abstraction is where risk hides.
A GCC setup runs 6–14 months from signature to a functioning center of 50+ people, spans three regulatory regimes, and carries a run-rate that typically lands between ₹18–35 lakh per FTE annually for senior engineering roles in tier-1 Indian cities. Get the partner wrong and you don’t find out in month two you find out in month nine, when the center is half-staffed and the leadership hire has fallen through twice.
Most dedicated development teams underestimate the transition workload by 3–4x. They budget for recruitment and office fit-out, then discover that transfer pricing documentation, payroll registration across states, and the knowledge-transfer runway for the first delivery team consume more calendar time than hiring does.
The second failure mode is quieter. The partner delivers a center that works and owns the operating scaffolding so thoroughly that taking it in-house three years later means rebuilding it. Ask the exit questions before signature, when you still have leverage, not in year three when you don’t.
The 30 questions, grouped by risk area
Use these in sequence. Each group escalates: the entity questions establish competence, the talent questions establish honesty, and the commercial questions establish alignment. Running them as gcc rfp questions in writing first, then repeating the sharpest ones live inconsistency between the written and verbal answer is itself a data point.
Entity, compliance and legal structure (Q1–7)
- Which entity structures have you filed for clients in the last 24 months: private limited, LLP, branch office and why did you recommend each?
- Who is the legal entity of record during the build phase: us, you, or an employer of record?
- How do you handle transfer pricing documentation, and do you prepare it in-house or subcontract it?
- What is your process for state-wise payroll, PF, ESIC and professional tax registration across the cities you’re proposing?
- Which statutory compliance obligations will sit with us on day one, and which stay with you?
- Show us the IP assignment clause chain: parent company → local entity → individual employment contract. Where are the gaps?
- What data-residency and DPDP Act obligations apply to our sector, and who signs off on the assessment?
Strong partners answer question 6 with a document. Weak ones answer it with reassurance. The chain matters because IP assigned to a local entity you don’t ultimately control is an asset you don’t ultimately own.
Talent, attrition and hiring data (Q8–15)
This is the group that separates operators from advisors. Anyone can describe a hiring process; only a partner who has actually run centers has the data behind one.
- What is your median time from approved requisition to offer accepted, by role band?
- What percentage of your offers are accepted, and what percentage of accepted offers actually join?
- How do you handle notice-period buyouts, and who funds them?
- What is your replacement commitment if a hire doesn’t work out, and is it contractual or best-effort?
- What was twelve-month attrition at the last three centers you built, by role band?
- How did that attrition compare to the market benchmark for those cities in the same period?
- What is your leadership-hiring track record, specifically country head, engineering director, finance controller?
- Which roles in our proposed org chart do you consider hardest to fill in our target city, and what’s your plan B?
Question 12 is the one to hold the line on. Vendors will offer a market-average number instead of their own; that’s a deflection. EY’s India GCC research reports average GCC attrition at 12.5%, with close to 60% of centers reporting between 5% and 15% so you already have the benchmark. What you need from the partner is their own performance against it, from centers they can name.
A partner who has built at scale will also volunteer role-band splits without being asked, because they know engineering attrition and shared-services attrition are different problems. If your gcc vendor evaluation questions get a single blended number in response, the data probably doesn’t exist.
Commercial model and exit terms (Q16–22)
- Which pricing model are you proposing build-operate-transfer, fee-per-hire, management fee, or hybrid and what does the total cost of ownership look like in year three, not year one?
- What triggers the transfer in a build-operate-transfer arrangement: a date, a headcount, or our discretion?
- What is the transfer fee, and is it a fixed amount, a multiple of run-rate, or negotiated at the time?
- If we transfer early, what do we pay? If we transfer late, what do you charge?
- Do employment contracts transfer to us automatically, or are they re-papered and who bears that cost and risk?
- What are your rate-card escalation terms, and are they capped or indexed?
- What is the exit-and-transition clause: notice period, handover obligations, data return, and continuity of the delivery team?
Question 19 is where most contracts are quietly asymmetric. Early-exit penalties are common; late-transfer charges rarely are. A partner confident in their own value will accept symmetry here.
Negotiate the transfer fee as a fixed amount agreed at signature, not as a multiple calculated at transfer; that single change removes the incentive to inflate your run-rate in the final year.
Governance, delivery and integration (Q23–27)
- Who is our named account owner, are they dedicated to us, and what else are they running?
- What is the escalation path and the committed response time when a hire slips or a milestone moves?
- How do you report what the weekly and monthly pack contain, and can we see a real, redacted example?
- How does the center integrate with our existing engineering and delivery tooling, and who owns that integration?
- At what point does your team step back, and what does the handover checklist contain?
Ask for the redacted reporting pack in question 25. It is the fastest way to tell whether a partner runs a delivery discipline or a relationship. Shared-bandwidth account managers are the most common source of month-four slippage; one person covering six accounts cannot chase a stalled leadership search.
References and verification (Q28–30)
- Give us two references from centers you built that are now fully transferred to the client, not centers you still operate.
- Give us one reference from an engagement that went badly, and tell us what you changed afterward.
- Can we speak directly to the India leadership of a center you built, without you on the call?
Question 29 filters hard. Every partner with real volume has a center that underperformed. The ones who describe it specifically the wrong city, the leadership hire that failed twice, the scope that changed in month five are demonstrating the kind of experience that only comes from having been there.
What this looks like in practice
Across 527+ delivered IT projects, two patterns show up repeatedly in how to evaluate gcc partner conversations.
A fintech scale-up building an India engineering hub. The initial shortlist ranked partners on price per hire. Re-scoring them on questions 12, 19 and 28 reversed the ranking entirely; the cheapest vendor had no transferred-center references and a transfer fee calculated as a multiple of final-year run-rate. The engagement that ran instead used a fixed transfer fee agreed at signature and a 7–10 day replacement guarantee written into the contract rather than the SOW.
An enterprise SaaS org staffing a cloud and platform team. The binding constraint wasn’t cost, it was leadership sequencing: the center’s first ten engineering hires kept stalling because the engineering director role was still open. Loading the leadership search first, then running the individual contributor pipeline against a 7–10 working day shortlist cycle, cleared the backlog. Where a client needs to hire cloud engineers at volume, sequencing beats sourcing speed almost every time.
A scoring framework for choosing a GCC enabler
Score each shortlisted vendor 1–5 per dimension. Anything below 3 on evidence quality or exit terms should be disqualifying, not a negotiation point.
| Dimension | What a weak answer looks like | What a strong answer looks like |
| Entity & compliance | “We’ll handle all of that” | Named structures filed, documents shown |
| Attrition evidence | Market-average figures only | Own data, by role band, from named centers |
| Commercial model | Year-one pricing only | Three-year TCO with capped escalation |
| Exit terms | Transfer fee “to be agreed” | Fixed fee, symmetric triggers, contracts transfer |
| References | Current clients only | Fully transferred centers, direct access |
Weight exit terms and attrition evidence at double the others. Those two dimensions predict the outcomes that actually cost money. Entity work is table stakes: any credible partner offering GCC setup services in India can incorporate an entity, but far fewer can show you what happened to the centers they walked away from.
What most teams get wrong
Most evaluation processes over-index on the build and under-index on the exit. Teams negotiate the setup fee hard, then accept a transfer clause that defers the fee calculation to a future date, handing the partner a structural incentive to grow your run-rate in the final year rather than optimize it.
The second mistake is treating the partner’s client logos as evidence. Logos prove a contract existed. They don’t prove the center scaled, retained people, or transferred cleanly. Ask which of those logos are transferred centers, and the list usually gets much shorter.
The third is sequencing leadership hiring last because it’s expensive. An unfilled engineering director seat stalls the twenty hires beneath it, and that delay costs more than the salary premium ever did.
Before you sign
If you’re mid-evaluation and want to pressure-test your shortlist against this checklist particularly the attrition evidence and the exit clause it’s worth a conversation before the contract is signed rather than after.
Supersourcing has run global capability center builds, IT consulting services engagements and recruitment process outsourcing programs across 527+ projects, with a 98% candidate joining rate and a contractual 7–10 day replacement commitment.
Send the deck and the draft MSA to mayank@engineerbabu.com, or start here: supersourcing.com/contact-us.
FAQ
What does a GCC setup partner actually do?
A GCC setup partner handles entity incorporation, statutory and payroll registration, office infrastructure, leadership and team hiring, and operational governance until the center is transferred to the parent. Scope varies widely; some partners are purely advisory, others run the center operationally for two to three years under a build-operate-transfer model. Confirm which model is being proposed before comparing prices.
How long does it take to set up a GCC in India?
Entity incorporation typically takes 4–8 weeks. A functioning center of 50+ people usually takes 6–14 months from signature, depending on leadership hiring speed and how much of the delivery scope transfers in the first wave. Partners quoting under six months are usually excluding the knowledge-transfer runway, which is where most timelines actually slip.
What is a typical GCC attrition rate in India?
EY’s India research puts average GCC attrition at 12.5%, with roughly 60% of centers between 5% and 15%. Engineering and AI/ML roles run materially higher than shared-services roles, so a blended number tells you little. Always ask for attrition by role band and by center age.
What should be in a GCC RFP?
Beyond scope and pricing: entity structure recommendation with rationale, attrition data from prior centers, named account ownership, escalation SLAs, three-year total cost of ownership, transfer triggers and fees, and two references from fully transferred centers. The questions to ask GCC setup partner candidates in an RFP should force evidence, not capability statements.
How do I verify a partner’s track record independently?
Ask to speak with the India leadership of a center they built, without the partner on the call. Cross-check entity filings on the MCA portal. Check whether the centers they cite are still operating and at what headcount. If a partner resists direct reference access, treat that as the answer.
Should we use a partner at all, or build directly?
Direct builds work when you already have India leadership in place and a legal function that can absorb the compliance load. Below roughly 30 FTEs, or without an existing India footprint, the partner route is usually faster and cheaper on a fully-loaded basis. The decision is less about capability than about whether your senior team has 300 hours to spend on setup.
What if we’ve already signed and the answers are now uncomfortable?
Most contracts have a renegotiation window at the first milestone. Use it to fix the transfer-fee mechanism and the escalation SLA, specifically those are the two clauses that compound. If you want an outside read on where your current agreement is exposed, a short review conversation is usually enough to identify it.
