Few decisions in the global operations agenda generate as much confidence, yet receive as little thoughtful analysis, as the choice between establishing a captive Global Capability Centre and adopting a Build-Operate-Transfer (BOT) model. I have been in many situations where the answer was effectively decided before the question had been properly examined, often because a senior leader brought a strong preference shaped by experience in a previous organisation. After more than 25 years of building and running global operations, I have come to a different perspective. Captive and BOT are not models that can be ranked universally; they are different approaches suited to different circumstances. Organisations are more likely to make the right choice when they first understand their own needs, capabilities, constraints, and objectives, and then select the model that best fits those conditions.
The central question explored in this paper is straightforward to state but more difficult to answer in practice: how should an organisation determine whether a captive or BOT model is better aligned with its readiness, strategic objectives, risk appetite, and execution requirements? I approach this primarily as a practitioner question because the decision is ultimately made by leaders operating within real-world constraints, including time, capability, investment, risk, talent, and organisational priorities.
The decision is also far from one-dimensional. The choice between captive and BOT involves several interconnected considerations, including ownership, speed of setup, risk exposure, internal capability, control over the operation, location, transition requirements, and the long-term operating model the organisation intends to build. These factors cannot always be considered in isolation. A model that appears attractive because it offers greater speed, for example, may create challenges where direct control or internal capability is particularly important. The decision framework presented later in this paper therefore focuses on how these considerations can be assessed together rather than treated as separate or independent choices.
The purpose of this paper is to examine both models from a deliberately balanced perspective. I aim to give the captive and BOT approaches equal consideration, recognising what each does well as well as where each can present challenges, without attempting to declare one universally preferable. I have worked with organisations that have pursued both approaches successfully, and I have also seen both models encounter significant difficulties. In my experience, the outcome has often depended less on the model itself and more on how well the organisation understood its own readiness, objectives, risks, and capabilities before making the commitment. That observation forms the central thread of this paper.
This paper is a practitioner account rather than a formal research survey or empirical study. Where I draw on broader industry context, this is identified accordingly; where I present observations based on my own experience, I distinguish them as practitioner perspectives. The objective is not to provide leaders with a predetermined answer, but to offer a structured way to assess their own circumstances and make a more informed setup decision.
Industry Context and Background
The Global Capability Centre (GCC) has evolved significantly from its origins as a cost-arbitrage back-office model, and industry evidence reflects this shift. Analysts and researchers increasingly describe GCCs as strategic enterprise enablers rather than transactional offshore hubs, with centres taking on digital work, assuming end-to-end process ownership, and contributing to innovation (Everest Group, 2025; Jha and Seth, 2025). EY’s GCC Pulse research reports that approximately 92 percent of GCC leaders see their centres creating value beyond cost arbitrage (EY, 2025). The scale of this evolution is also significant. NASSCOM data cited across the industry places the number of GCCs in India at more than 1,900 by 2025, while market estimates value the global GCC market at hundreds of billions of dollars, with sustained double-digit growth projected over the coming decade (NASSCOM, 2025; Market Research Future, 2025). These figures are included to establish context rather than to form the basis of the argument. They illustrate a market that has become sufficiently large and strategically important that the way a centre is established can influence its control, culture, capabilities, and risk profile for years to come.
Two principal models have emerged for establishing a new centre, and it is important to define them clearly because the differences shape much of the discussion that follows. Under the captive model, the operation is owned and controlled by the parent organisation from day one. The parent establishes the legal entity and takes responsibility for real estate, hiring, infrastructure, and the operational build, thereby assuming the associated setup risks. External advisers may support the process, but ownership and control remain with the organisation throughout. Under the Build-Operate-Transfer (BOT) model, a third-party partner establishes and operates the centre on behalf of the client for a predefined period, typically two to four years, under agreed governance arrangements and performance standards. At the end of this period, ownership and operating responsibility, including people, processes, infrastructure, and intellectual property, transfer to the client subject to defined transfer conditions (ANSR, 2026; NeoIntelli, 2026). In a BOT arrangement, therefore, the transfer mechanism is not a secondary consideration. It is a defining feature of the model and should be treated as a central element of the decision from the outset rather than as a formality at the end of the engagement.
It is equally important to clarify what BOT is not. A model in which an organisation remains with a managed-service provider indefinitely, with no planned transfer of ownership, is a different arrangement and falls outside the scope of this comparison. This paper focuses specifically on the captive-versus-BOT decision where eventual ownership of the centre is intended to rest with the client. The distinction matters because BOT, outsourcing, and managed services are sometimes used interchangeably in practice, even though they represent different operating arrangements. BOT is not simply outsourcing followed by a later exit. Its defining characteristic is the planned transfer of ownership and operating responsibility to the client, with the transition designed from the beginning rather than negotiated at the end (ANSR, 2026). Put simply, in a traditional outsourcing arrangement, the service provider continues to own and operate the service while the client purchases defined outputs or services. In a BOT arrangement, ownership is intended to move to the client. Maintaining this distinction is important to ensure that the comparison between the two models remains clear and meaningful.
The importance of this decision has also increased as the scope and strategic value of work performed through GCCs have expanded. Establishing the wrong setup model is no longer simply a question of accepting a somewhat higher cost base. It can affect the speed at which capabilities are developed, the degree of control over critical work, organisational integration, and the success of the eventual transition. In my experience, the industry conversation has consequently evolved from a relatively narrow focus on cost toward a broader assessment of speed, risk, control, capability, and strategic fit. It is within this broader context that the captive-versus-BOT decision needs to be considered.
Exhibit 1. Captive and BOT at a Glance (Illustrative Framework)
| Dimension | Captive Model | Build-Operate-Transfer Model |
| Who builds | The organisation itself | A specialist partner, then transfers |
| Ownership from day one | Yes | No; ownership on transfer |
| Speed to stand up | Slower; full build owned internally | Faster; partner has ready infrastructure |
| Control during build | Full | Shared with partner |
| Upfront risk to organisation | Higher | Lower during build phase |
| Long-term dependency | Low | Concentrated at transition |
Interpretation: This overview is intentionally even-handed. Neither column is the ‘good’ column. Each row is a trade-off, and which side is preferable depends on the organisation’s priorities and readiness rather than on any universal ranking.
My Experience and Practical Observations
Drawing on experience with both approaches, I have developed a set of observations that have remained consistent across geographies and sectors. These observations are presented from a deliberately balanced perspective, recognising the strengths and limitations of both models rather than favouring one over the other.
The captive model rewards those who are ready
When an organisation had a clear understanding of the work it wanted to locate in the centre, the capability to hire and manage talent locally, and the patience to invest in the build, the captive model offered significant advantages: full control and ownership from the outset, with the organisation’s culture and standards embedded from day one. I have seen captive centres become deeply integrated, high-performing extensions of their parent organisations precisely because there was no handover or transition seam to manage. In my experience, however, these benefits depend heavily on organisational readiness. When organisations underestimated the complexity of the build or lacked the capabilities required to manage it effectively, the same model could result in delays, frustration, and a centre that took considerably longer than anticipated to deliver value.
The BOT model rewards those who need speed and want to de-risk entry
When an organisation needed to move quickly, lacked deep local knowledge of the target location, or wanted to limit its exposure during the uncertain early stages, the Build-Operate-Transfer model could offer significant advantages. A capable partner could establish and operationalise the centre considerably faster than the organisation might be able to do independently, while absorbing much of the early-stage execution risk and providing established infrastructure, local expertise, and operational capabilities. I have seen this approach deliver a functioning centre in a fraction of the time that a comparable captive build would have required. However, these benefits come with a corresponding consideration: BOT concentrates significant risk and effort around the transition to ownership. Organisations that treated the transfer as a formality rather than as a carefully planned phase often encountered challenges later in the process.
Illustrative Case A: A Captive Build That Rewarded Readiness
Organisation profile. A large, established multinational operating in a regulated industry, with an existing offshore footprint and internal teams experienced in managing distributed operations.
Situation. The organisation wanted to establish a new centre to accommodate a growing volume of engineering and analytics work, with the intention of retaining these capabilities in-house over the long term.
Why the captive model was selected. Control was considered more important than speed. The work was closely aligned with the organisation’s core capabilities, the organisation had strong knowledge of the local market, and leadership had both the appetite and patience to take full ownership of the build.
Major challenge. Entity establishment, hiring, and the operational ramp-up took longer than initially anticipated. Early-stage attrition also tested the team’s resilience before the centre reached a stable operating position.
Outcome and lesson. Because the organisation was genuinely prepared for the demands of a captive build and had planned and budgeted for a challenging initial phase, the centre matured into a deeply integrated extension of the parent organisation, with no handover or transfer phase to manage. The key lesson I draw from cases such as this is that the captive model rewards organisational readiness but exposes organisations that underestimate the complexity of the build.
Illustrative Case B: A BOT Route That Bought Speed and De-Risked Entry
Organisation profile. A mid-sized organisation entering an unfamiliar offshore location for the first time, with limited internal experience in establishing and operating overseas operations.
Situation. The organisation needed to establish the required capabilities quickly to support its growth commitments and could not absorb the distraction, complexity, and risk associated with building a captive operation from the outset.
Why the BOT model was selected. Speed and risk reduction were prioritised over immediate ownership. A capable partner could establish and operationalise the centre more quickly, provide local expertise that the organisation did not yet possess, and assume much of the early-stage execution risk, with full ownership planned to transfer to the organisation once the operation had reached the agreed transition point.
Major challenge. As the operation matured under the partner’s management, the organisation’s own culture, standards, and ways of working had not been fully embedded. As a result, the transfer required significantly more planning, leadership attention, and change-management effort than initially anticipated.
Outcome and lesson. The centre became operational in a fraction of the time a comparable captive build would have required. However, the full value of the BOT model was realised only after the organisation treated the transfer as a significant phase of the operating model rather than as a formal handover. The key lesson I draw from cases such as this is that the BOT model can accelerate entry and reduce early-stage risk, provided the organisation plans for and actively manages the transfer from day one.
The decisive variable is knowledge, not model
The observation I hold most firmly is that, in my experience, successful outcomes have depended less on the model selected than on what the organisation understood about its own circumstances before making the commitment. Organisations that were candid about their readiness, risk tolerance, and strategic objectives tended to perform well under either model because they selected the approach that best suited their circumstances and prepared accordingly. By contrast, organisations that skipped this self-assessment often struggled regardless of the model they chose, particularly when the decision was driven by a prior preference or by the experience of a peer organisation. This observation led me to examine more closely what I mean by organisational knowledge and to define it in a way that can be assessed in practice.
If knowledge is a decisive factor, it is important to define what that knowledge actually comprises. In this context, I do not mean general awareness of the GCC landscape or familiarity with the two setup models. I mean the organisation’s readiness to understand, make, and execute the decision effectively. I refer to this as Organisational Decision Readiness, which I define across five dimensions.
The first dimension is strategic clarity: a clear understanding of why the centre is being established, what work will be moved into it, and what the desired long-term operating model should look like.
The second is operational readiness: the organisation’s capability and capacity to establish and operate a centre independently, or to assume responsibility for the operation when ownership transfers under a BOT model.
The third is local-market knowledge: an understanding of the talent market, regulatory environment, infrastructure, and practical operating realities of the chosen location.
The fourth is risk readiness: an honest assessment of the organisation’s tolerance for early-stage risk and its ability to manage the financial, operational, execution, and organisational risks associated with the setup.
The fifth is governance capability: the decision rights, performance expectations, accountability mechanisms, and management structures required to establish and operate the centre effectively under either model.
Two considerations cut across these dimensions and warrant particular attention. Talent-market readiness is a critical component of local-market knowledge, while technology and digital readiness forms an important part of operational readiness. Both become increasingly significant as modern GCCs take on engineering, analytics, product, and other technology-intensive capabilities.
Taken together, these five dimensions provide a practical way of assessing whether an organisation is ready to make and execute the setup decision. Where readiness is strong, the organisation is better positioned to select the model that fits its circumstances and prepare effectively for its implementation. Where readiness is weak, those gaps can undermine the outcome regardless of the model selected. This is why I frame the decision around informed commitment rather than model preference.
Exhibit 2. The Five Dimensions of Organisational Decision Readiness (Author’s Experience-Based Model)
| Dimension | What it means | Key question to ask |
| Strategic clarity | Clear purpose, scope of work, and target long-term operating model | Do we know what this centre is for and what it should become? |
| Operational readiness | Capability and bandwidth to build, run, or absorb a centre | Can we execute the build, or absorb it at transfer, without stalling the business? |
| Local-market knowledge | Understanding of talent, regulation, and location realities | Do we truly understand the location we are entering? |
| Risk readiness | Honest tolerance for, and capacity to manage, early-stage risk | How much early risk can we carry, and are we being honest about it? |
| Governance capability | Decision rights, performance expectations, management structures | Can we govern this centre well under either model? |
Interpretation: I use these five dimensions to turn a vague sense of readiness into an honest conversation. An organisation strong across all five can choose either model and succeed; weakness on any dimension tends to surface later regardless of the model. This construct is the intellectual core of the paper: the decision is less about the model than about the readiness behind the commitment.
Exhibit 3. Balanced View of the Key Trade-Off Factors (Author’s Experience-Based Model)
| Factor | Captive | BOT | Underlying consideration |
| Speed to stand up | Slower | Faster | Internal versus partner execution capability |
| Control | Higher from day one | Shared initially | Governance and decision rights during the build |
| Early-stage risk | Higher | Lower | Who absorbs the setup and ramp risk |
| Ownership | Day one | On transfer | When the organisation wants to hold the asset |
| Transition | None required | Required, and central | Transfer readiness and change effort |
| Local capability | Required internally | Partner-supported | Depth of local-market knowledge available |
| Culture and standards | Embedded from the start | Must be actively transferred | How the parent’s ways of working take root |
Interpretation: Reading across each row, the two models are close to mirror images: the captive model’s strengths are the BOT model’s costs, and vice versa. The added column names the underlying consideration that actually drives each trade-off, so a leader can see not just which model scores better on a factor but why. This symmetry is the heart of my argument that neither model is superior in the abstract.
4. Key Challenges and Common Pitfalls
Some pitfalls are specific to a model; many are shared. I have grouped them so leaders can see which apply to the path they are considering and which apply regardless.
Pitfalls more common in captive builds
- Underestimating the build. The entity, real estate, hiring, and operational ramp take longer and cost more than the plan assumes, especially in an unfamiliar location. I have seen this delay contribution by many months.
- Overestimating internal capability. Building and running a centre in a new geography is a distinct skill; organisations that assumed their existing teams could simply extend into it were often surprised by the gap.
Pitfalls more common in BOT arrangements
- Treating the transfer as a formality. The transition is the moment of maximum risk, and organisations that did not plan and resource it seriously inherited an operation they did not fully understand.
- Neglecting culture and standards during the build. When the partner’s ways of working take root, the organisation can transfer the operation and find its own culture never took hold. This has to be actively managed from the start, not retrofitted at transfer.
Pitfalls common to both models
- Choosing the model before understanding the need. One of the most damaging patterns I have observed is selecting a model based on prior assumptions or a peer organisation’s experience rather than on the organisation’s own readiness, objectives, and circumstances. A model that worked well for one organisation may not be appropriate for another with different capabilities, priorities, or constraints.
- Weak governance. Regardless of the model selected, a centre without clear governance, decision rights, accountability, and performance expectations can quickly lose direction. The choice of setup model does not replace the need for effective management and governance.
- Treating location as a separate decision. Talent availability, the regulatory environment, infrastructure, cost dynamics, and the long-term viability of the location matter under both models. I have seen otherwise well-considered setup decisions undermined because the location was assessed independently of the operating model and long-term capability requirements.
- Underestimating talent-market risk. Talent deserves greater attention in the setup decision because GCC performance depends heavily on the availability and quality of skilled talent, competition for talent, attrition, local leadership capability, specialised skill pools, wage dynamics, and employer brand. Both models are exposed to these factors, although the timing and nature of that exposure can differ. A captive model assumes more direct responsibility for navigating these dynamics from the outset, while a BOT partner may provide greater support during the build and operate phase, with some of the underlying talent considerations becoming more visible as the organisation prepares for transfer.
- Underestimating technology and digital readiness. As GCCs increasingly take on engineering, analytics, product, and other technology-intensive work, the setup decision must also consider technology infrastructure, cybersecurity, data access, enterprise architecture, automation maturity, and AI and digital capabilities. These considerations apply to both models, although the responsibilities differ. In a captive model, the organisation must establish and manage the required environment itself. In a BOT model, it must ensure that the partner’s technology environment can ultimately be integrated into the organisation’s own architecture and that data, security controls, systems, and processes can be transferred without disruption.
Economic and Commercial Pitfalls
A final group of pitfalls relates to the economic and commercial dimensions of the decision. I treat these separately because, in my experience, they are among the most frequently overlooked aspects of the captive-versus-BOT comparison. The choice between the two models is, among other things, a financial decision, yet the models are often evaluated using different cost assumptions or over different time horizons. The most common mistake is to compare them on the wrong cost basis.
I do not present specific financial figures in this paper because the economics vary significantly by organisation, location, scope, scale, and operating model. The underlying principle, however, is consistent: the decision should be evaluated on a total cost of ownership basis across the full lifecycle rather than on headline setup costs alone.
The recurring economic and commercial pitfalls I have observed include:
- Comparing only the initial setup cost. A captive build may appear more or less expensive than BOT when viewed only through the initial investment. Such a comparison, however, does not capture the full cost of establishing, operating, transitioning, and ultimately sustaining the centre.
- Ignoring transition costs. In a BOT arrangement, the transfer of people, processes, systems, knowledge, governance, and ownership involves real cost and management effort. Excluding these costs can materially distort the comparison.
- Ignoring partner fees during the operate period. The partner’s fees and margin during the operating phase represent a genuine cost of the BOT model and should be incorporated into the full lifecycle assessment.
- Ignoring management and governance overhead. Both models require ongoing management, governance, and leadership attention. A captive model may also involve internal costs that are not always fully reflected in the business case, particularly where existing teams take on additional responsibilities during the build.
- Ignoring the cost of delayed capability. A slower captive build may delay the point at which the centre begins to deliver its intended capabilities and business value. The economic impact of that delay should be considered rather than treated simply as a timing issue.
- Ignoring potential stranded costs. Commitments made during the build or the BOT operating period may become stranded if business priorities, scope, location, or strategic plans change. Where such risks are material, they should be recognised in the economic assessment.
- Using inconsistent cost assumptions. One of the most significant problems is comparing Captive and BOT using different assumptions, time horizons, cost categories, or treatment of internal resources. A seemingly favourable outcome may simply reflect differences in how the two models were costed rather than a genuine economic advantage.
The remedy is not necessarily greater forecasting precision, which is difficult to achieve at the early stages of a GCC decision. What matters more is consistency, transparency, and discipline in the assumptions used. A robust assessment should consider total cost of ownership across setup, operation, transition, and steady state; apply comparable assumptions to both models; and make the potential cost of delayed capability and stranded commitments explicit wherever they can reasonably be assessed.
Exhibit 4. Where the Pitfalls Sit (Author’s Experience-Based Model)
| Pitfall | Captive | BOT | Both |
| Underestimating the build | Yes | ||
| Overestimating internal capability | Yes | ||
| Transfer treated as a formality | Yes | ||
| Culture not embedded during build | Yes | ||
| Model chosen before need understood | Yes | ||
| Weak governance | Yes | ||
| Poor location decision | Yes | ||
| Economic and commercial (TCO) pitfalls | Yes |
Interpretation: Note that the most consequential pitfalls, choosing before understanding, weak governance, poor location, and economic mis-comparison, are shared across both models. This reinforces that the model choice, while real, is not where most failures originate.
Lessons Learned
Rather than restate the observations above, I have distilled them into a practical sequence a leadership team can actually follow. The steps are ordered: the first five apply before any model is chosen, and the last two apply once the choice is made.
Before deciding
- Define the strategic objective. Be explicit about what the centre is for, what work will move, and the long-term operating model you want.
- Assess readiness. Score the organisation honestly against the five dimensions of Organisational Decision Readiness before comparing models.
- Assess the location. Test talent depth, competition, attrition, leadership availability, wage inflation, and regulatory viability, not just cost.
- Assess risk. Be honest about tolerance for early-stage risk and the capacity to manage it.
- Assess the economics. Build a total-cost-of-ownership view spanning setup, operate, transition, and steady state, using the same assumptions for both models.
After selecting Captive
- Build internal capability. Invest in the local leadership, technology environment, and management bandwidth the build demands, and plan for it to be harder and slower than expected.
After selecting BOT
- Design the transfer from day one. Treat the transfer as the main event: embed your culture, standards, data, and security controls during the operate phase, and resource the transition as seriously as the build.
Strategic Recommendations for Leaders
For leaders facing this decision, I offer a set of recommendations designed to support an informed commitment rather than favouring any particular model. I also view the decision as having three potential outcomes rather than two: Captive, BOT, and a hybrid or staged approach that combines elements of both. The hybrid option becomes particularly relevant when the organisation’s circumstances present competing signals, such as a strong need for control combined with limited internal capability or a need for rapid execution. For this reason, it is incorporated into the decision framework rather than treated as an alternative only after the primary choice has been made.
- Start with a candid readiness assessment. Before comparing models, assess the organisation’s capability to establish and operate a centre, its tolerance for early-stage risk, its need for speed, and the degree of control required for the work involved. This assessment should also consider local-market knowledge, talent availability, technology readiness, and governance capability.
- Match the model to the circumstances, not to precedent. Use the relevant trade-offs to determine which approach best fits the organisation’s circumstances. A model that worked successfully for another organisation may not be appropriate where the strategic objectives, capabilities, risk appetite, or operating environment are different.
- Plan the full lifecycle from the outset. For a captive model, the planning should extend from initial setup through operational maturity and steady state. For BOT, the build, operate, transfer, and steady-state phases should be designed as one connected lifecycle rather than treated as separate stages. The eventual ownership model should influence decisions from the beginning.
- Establish governance early. Define decision rights, accountability, performance expectations, governance forums, and ownership responsibilities before the centre begins operations. Effective governance should remain a management discipline throughout the lifecycle rather than being treated as a feature of the chosen setup model.
- Test the decision against the evidence. Where the assessment points strongly toward speed, local expertise, and reduced early-stage risk, these factors should be given appropriate weight rather than being overridden by a preference for direct ownership. Similarly, where the organisation has the capability, readiness, and need for direct control, the appeal of speed should not by itself determine the decision. The objective is to allow the evidence from the readiness assessment to guide the choice.
Exhibit 6. Model-Selection Decision Framework (Author’s Experience-Based Model)
| If the organisation… | Captive | BOT | Hybrid / staged |
| Needs to move quickly | Yes | Yes | |
| Has deep local knowledge and capability | Yes | ||
| Places the highest value on day-one control | Yes | Partial | |
| Wants to limit early risk exposure | Yes | Yes | |
| Has patience and appetite to own the build | Yes | ||
| Is uncertain about the location or approach | Yes | Yes | |
| Requires culture embedded from the start | Yes | Partial | |
| Has mixed or conflicting signals across the above | Yes |
Future Outlook and Emerging Trends
Several shifts are reshaping the GCC setup decision, and it is important to consider them without favouring either model.
First, the work performed by Global Capability Centres continues to move up the value chain. As GCCs take on more strategic responsibilities, the importance of control, organisational culture, capability development, and integration increases under both setup models. Industry analysis describes this evolution as a shift from traditional cost arbitrage toward capability and innovation arbitrage, with centres increasingly taking ownership of end-to-end products, platforms, and business capabilities (EY, 2025; Everest Group, 2025). As the strategic importance of GCCs increases, the advantages of seamless integration under a well-established captive model and the challenges of transferring culture and capabilities under a BOT model become more significant. The implications, however, will differ depending on the organisation’s circumstances and priorities.
Second, the range of setup options is expanding. Hybrid and staged approaches are becoming more relevant, and the traditional captive-versus-BOT binary is increasingly being complemented by arrangements that combine elements of both models. Analysts have begun to recognise hybrid delivery as a distinct category rather than simply an exception to the two primary models (Market Research Future, 2025; Business of GCC, 2026). From my experience, this development can be valuable because it allows organisations to tailor the setup approach to their level of readiness rather than forcing every situation into a binary choice. At the same time, hybrid approaches introduce additional complexity and require greater clarity around ownership, governance, responsibilities, and the point at which control changes hands.
Third, location strategy is becoming increasingly complex. India continues to be a major GCC destination, while organisations are increasingly considering distributed, multi-location portfolios rather than relying on a single offshore location (ISG, 2026; Business of GCC, 2026). Eastern European hubs such as Warsaw and Krakow have developed strengths in areas including engineering, cybersecurity, and multilingual finance. Latin American markets, including Mexico and Colombia, provide nearshore options with time-zone alignment for North American organisations, while Southeast Asian markets such as the Philippines, Malaysia, and Vietnam continue to develop capabilities in customer operations and digital services (Hexaware, 2026; GCC Pulse, 2026). The drivers behind these decisions increasingly extend beyond cost and include access to specialised talent, time-zone coverage, geopolitical diversification, and regulatory considerations.
This evolution has direct implications for the captive-versus-BOT decision. Location is no longer necessarily a single decision that follows the choice of operating model. An organisation may, for example, use a BOT approach to enter an unfamiliar market while maintaining captive operations in locations where it already has strong knowledge and capabilities. A multi-location strategy can therefore result in different setup models being used within the same GCC portfolio. This reinforces a central argument of this paper: location and operating-model decisions should be assessed together, with neither treated as an afterthought.
AI and the Changing GCC Setup Equation
No discussion of the future of GCC setup would be complete without considering artificial intelligence, and I believe it is important to address its implications directly rather than treat it as a passing trend. Industry evidence suggests that AI is beginning to reshape some of the economic assumptions that contributed to the growth of GCCs. As automation and agentic AI take on an increasing share of routine, rules-based work, the traditional labour-arbitrage rationale for establishing centres may become less significant, while the emphasis shifts increasingly toward capability and innovation arbitrage (EY, 2025; ORF, 2026).
Several implications follow for the GCC setup decision. I present these as considerations rather than predictions.
First, AI may change initial workforce assumptions. A business case based on a large transactional workforce may become outdated as automation changes the nature and volume of work. Organisations therefore need to consider not only the work that exists today, but also how the work is likely to evolve when determining the appropriate size and capability mix of a new centre.
Second, AI capability may influence location strategy. As the demand for digital, engineering, data, and AI skills increases, organisations may place greater emphasis on access to specialised talent and technology ecosystems rather than focusing primarily on labour cost.
Third, the purpose of the GCC itself may evolve. New centres may increasingly be established around digital, engineering, data, and AI capabilities rather than primarily around labour arbitrage. This can change not only the size of the centre, but also its mandate, talent profile, operating model, and relationship with the wider enterprise.
Fourth, governance and data considerations become increasingly important. The use of AI at scale raises questions around data access, cybersecurity, governance, intellectual property, and control. These considerations are relevant to both captive and BOT models, and they become particularly important in a BOT arrangement where technology, data, systems, and associated controls ultimately need to be transferred to the client.
Fifth, BOT partners may play a greater role in building AI and digital capabilities. A partner with access to specialised technology and AI talent may provide additional value during the initial build and operate phase. At the same time, this increases the importance of defining from the outset how those capabilities, knowledge, systems, processes, and associated intellectual property will be transferred to the organisation at the end of the BOT period.
It is important, however, not to overstate the certainty of these developments. The direction of travel is becoming clearer, but the pace and precise shape of change remain uncertain. The same principle that underpins this paper therefore applies to AI as well: technology should inform the design of the operating model, but it should not substitute for disciplined assessment and decision-making.
Whatever the specific trends, I expect the underlying principle to remain relevant. As the range of setup options expands, understanding the organisation’s own readiness, objectives, capabilities, and constraints before making a commitment becomes more important, not less. The organisations best positioned to navigate this changing environment will be those that treat the expanding range of choices as a reason for deeper assessment rather than simply a reason to make a faster decision.
Conclusion
This paper has sought to give the captive and Build-Operate-Transfer models an even and balanced consideration. Both represent viable approaches to establishing a Global Capability Centre, with each offering distinct advantages and challenges. The captive model provides greater control and ownership from the outset, but requires a more demanding build and greater organisational involvement. The BOT model can provide faster entry and reduce early-stage execution risk, but introduces greater complexity around transition and the eventual transfer of ownership. The suitability of either approach depends on the circumstances in which it is applied rather than on any inherent superiority of one model over the other.
The central contribution of this paper is to reframe the captive-versus-BOT decision from a model-selection exercise into an organisational-readiness decision. Rather than treating the operating model as the starting point, the framework presented here places strategic intent, organisational capability, risk, economics, location, and governance assessment before model selection. The Organisational Decision Readiness construct and the Informed Commitment Process provide practical ways for leaders to apply this approach. This is a practitioner’s reframing rather than an empirical claim, but it reflects a recurring observation from my experience: the quality of the decision process often matters as much as the model ultimately selected.
The broader lesson is that successful GCC setup depends not simply on choosing between Captive and BOT, but on understanding what the organisation is prepared to build, operate, govern, and ultimately sustain. Organisations that assess their readiness honestly, align the model with their circumstances, plan the full lifecycle, and establish effective governance are better positioned to make the chosen model work. Conversely, a decision driven primarily by precedent, assumption, or an incomplete understanding of organisational readiness can create challenges regardless of the model selected.
If there is one principle I would leave with leaders considering a new GCC, it is this: understand your organisation and its requirements first, and let the model follow. The most meaningful setup decision is not simply the selection of Captive, BOT, or a hybrid approach. It is the commitment made with a clear understanding of what the organisation needs, what it is prepared to take on, and how it intends to create value over the full lifecycle of the centre.
- Dossani, R., & Kenney, M. (2003). Lift and Shift: Moving the Back Office to India. Information Technologies and International Development, 1(2), 21-37.
- Jha, S. K., & Seth, A. (2025). Global Capability Centres: Emerging Opportunities and Challenges. IIMB Management Review, 37(3), 100595.
- (2025). GCC Pulse Report 2025. Ernst & Young.
- Everest Group. (2025). Exploring Global Capability Centres’ Transformation Capabilities: A Renewed Provider Approach.
