A captive earns its next phase of growth by doing more of the value-added work which boosts the bottom line of onshore desks – the senior, judgment-led work – and getting deeper into headquarters as a result. That supersedes the main KPI of the old pre-AI world, which was really about headcount growth. This piece sets out how a captive (a Global Capability Centre, or centre of excellence) builds that senior layer and turns it into greater onshore penetration.
The first constraint is not that the talent is scarce or the open market too dear – India has the depth. It is that attracting and holding judgment-grade people takes a different employer model than a centre built for process work is set up for: different infrastructure, policies and culture. Neither is right or wrong; they suit different work.
The second constraint is the comfort levels and the remote team management capabilities of the onshore desks that you are targeting. Both those things are trainable.
The service which enables future growth is built from two choice parameters:
Scope – create a capability the centre does not yet have, or strengthen one it already runs.
Operating model – place the team onsite inside the existing centre, or offsite as a separate unit with its own pay, hiring and staff handbook (the environment judgment work tends to need).
Two things hold across every version. The onshore–offshore integration layer that converts the work into onshore penetration. And a destination: Frontline's people become the centre's own employees in due course, so the capability and the institutional knowledge stay inside the centre. Throughout, the work ships under the centre's name.
It is the practical companion to India captives' next strategic move, which makes the wider case for why this matters now.
What the right answer has to do
Before the parameters, the tests. Any route to senior capacity that works for the captive – as opposed to working for a vendor – has to meet four conditions.
It has to deliver now. The judgment layer is needed for the next operating-model cycle, not the one after.
It has to build the centre's own people, not merely stand in for them. Senior capacity that arrives and develops nobody has solved this year's problem and left next year's intact.
It has to end in the centre's hands. The capability – the people, the workflow, the stakeholder relationships – should finish inside your organisation, not as a contract you renew without end.
And it has to work through the centre. The work ships under the centre's name; the new desks served at headquarters are the centre's wins; the expansion story belongs to the head who commissioned it. An arrangement that routes around the captive fails this test by definition, whatever else it delivers.
These conditions hold whatever you choose. The parameters below are how you build something that meets them.
The two choice parameters which build the service you need
The service is assembled from two independent choices. You set each one to your own situation; they combine freely, and no combination is off the menu.
Parameter one – scope: a new capability, or an existing one
This is what the work is, not how it is structured. Either we create a capability you do not yet have – a new asset class, a judgment-led product, a workflow onshore desks want that would otherwise go elsewhere – or we strengthen one you already run but that lacks depth or seniority.
Creating fits when headquarters' demand is ahead of your bench. Strengthening fits when the desk exists but the judgment layer on top of it is thin. Either way the work runs under your name.
Parameter two – operating model: where the people sit, and under whose rules
Onsite, the team works inside your existing centre and infrastructure: fastest integration, and it uses what you have already built.
Offsite, the team is held deliberately outside that infrastructure – which is often the point, not a compromise. A separate unit can run different pay scales, different hiring practices and a different staff handbook; it gives you room to experiment with new ways of working, or to give higher-end judgment work the employer model it needs – which is rarely the one a process-led centre was built around. Run it as a time-boxed experiment or keep it permanently.
Neither pole is the better one by default. The right answer depends on whether you want speed and reuse, or freedom and a clean sheet.
In both options the analysts present themselves as members of the India captive at all times when they face onshore or clients offshore.
Where it leads: the people become yours
Whatever scope and operating model you set, the engagement is built to end in the same place. Frontline's people start as ours, working through the centre and under its name; in due course – the horizon is around three years – they become the centre's own employees. While the engagement runs, your own analysts develop alongside seniors they would otherwise never sit with, so the internal bench grows; and when the people transfer in, the institutional knowledge, the relationships and the ownership stay inside the centre rather than walking out at a renewal.
Setting the two together
The parameters are independent: pick a point on each. "Strengthen an existing desk, offsite." "Create a new capability, onsite." Either combination is coherent, and centres often move as trust builds – an onsite engagement that later spins out an offsite unit, say. Nothing is fixed once. The only test is that the combination meets the four conditions above.
The layer that runs through all of it: integration, on both sides
Whatever combination you set, the onshore–offshore interface is the part almost nobody else supplies, and the part that decides whether the rest of it lands. It has to be run as one strategy, with both sides under management.
Most integration programmes are one-sided. They coach the offshore team to communicate better and present the result to headquarters as a finished article. The half that gets skipped is onshore – and it is the half where most engagements quietly fail. Desks need framing, permissions and reasons to engage. Sceptics need converting one at a time. And some onshore stakeholders will never be fully comfortable with offshore, which is normal and workable, provided someone is managing it deliberately rather than hoping it resolves itself.
This is the work Frontline has done for twenty-one years, and the team that does it is the rare kind that has done it both in practice and from first principles – led by Dan Davies and the senior practitioners listed on our Humans page. Coaching a managing director in London and an analyst in Mumbai in the same week is a specific skill, and we staff for it deliberately. The interface work runs through every combination above as standard, and it stands alone as an engagement in its own right where a centre's analytical capability is sound but its penetration into headquarters is shallow.
Commercial shape
A fee per person per year. That's it, very simple, all included.
What to ask any partner
The diligence questions are the same whoever you talk to.
Does the work still pay its way in a year? The honest line on AI is that it is steadily absorbing the lower-judgment work, and a partner whose value sits there will be a white elephant by the next review. The work worth contracting for is the judgment layer AI does not reach – and a credible partner should be able to say which side of that line their people sit on, and mean it.
How long have they been operating senior analyst teams from India, as opposed to placing people there? Operating and placing are different trades; the first is the one that transfers.
Which institutions have trusted them with judgment work – not process work – and for how long? Multi-year engagements with bank-grade clients are the only reference class that matters here.
Will they put named individuals on the engagement, and will those names survive contract signing? A bench is not a layer.
Will they agree the transfer terms – the point at which the people become yours – at the start? A partner whose model quietly depends on the people never becoming yours has a different incentive structure than the one described here, and it will show at renewal.
Can they work both sides of the interface – coaching the onshore desks as credibly as the offshore analysts? A programme that polishes only one end produces excellent work that nobody uses. Ask what, specifically, they would do with your onshore desks, and who there has actually done it.
And will the work ship under the centre's name? Much of the market is not even built for this – the mass-market KPOs are designed to be a permanent third party, not to hand a capability over – but if a partner's own brand appears anywhere your stakeholders look, the arrangement is routing around the centre, not through it.
Frequently asked questions
Is a captive the same as a Global Capability Centre or a centre of excellence?
Broadly, yes. The terms captive, Global Capability Centre and centre of excellence all describe an offshore unit a bank or asset manager owns and staffs itself, rather than buying the work from a vendor.
Does working with Frontline mean replacing our captive?
No. Frontline works through the captive and under its name. The work, the wins and ultimately the people end up the centre's – it complements the captive rather than competing with it.
How do the analysts become our own people?
Every engagement is built to transfer Frontline's analysts into the centre's own employment in due course – around three years – so the institutional knowledge and relationships stay inside the centre.
Will this still add value as AI improves?
Yes, because it sits in the judgment layer – the work AI does not reach. Lower-judgment work is being automated; the senior, judgment-led work is where durable value and deeper onshore penetration come from.
Where Frontline stands
Frontline has been running senior offshore analyst teams from India for twenty-one years – operating, not placing – for global banks, asset managers and market-data firms, on multi-year engagements. Every combination above is one we run, on the terms described, including transfer of the people into your own employment in due course, agreed at the outset. The diligence questions in the previous section are ones we are content to be asked.
If you're weighing the senior-layer question for your centre, the next step is a working session, not a proposal: we map your situation against the two parameters, tell you which combination fits and what it would take, and say plainly where we think it wouldn't work. Start the conversation.
Read first: India captives' next strategic move
Part of: Offshore Analyst Teams – what works and what breaks