In practice · 02 of 09
The board approved the centre. Nobody has said what it is for.
A location decision usually arrives as a headcount number and a cost target. What is missing is the design.
What a headcount number does not tell you
Which capabilities move, and in what sequence. What the receiving organisation actually looks like. Who owns what on day one. What has to be true before anything is handed over. None of that is answered by a target of a thousand people in a named city, and yet the target is usually the only thing that has been agreed when the programme starts.
In the absence of a design, the centre gets designed around supplying engineering capacity, because capacity is the easiest thing to specify and the easiest thing to count. That was a reasonable default for twenty years. It is a poor one now, at precisely the moment engineering capacity stops being the scarce input.
The uncomfortable version. If a centre exists to supply a skill, and that skill is the one being commoditised fastest, then the centre's business case is exposed to the same force that is meant to justify it. That is worth confronting during the design, not during the second-wave review.
Classify the capabilities before anything moves
The first piece of work is a classification, so that where work goes is the output of a judgement rather than a policy. Capabilities differ on things that matter and that a headcount plan cannot see: how much of the knowledge is written down, how much regulatory load they carry, how tightly they are coupled to a specific individual or a specific customer, and how much of the work is judgement rather than execution.
Some capabilities should move first because they are self-contained and evidenced. Some should move late because the receiving organisation has to earn them. Some should not move at all, and saying so early is worth more than any sequencing plan, because the expensive failures are the capabilities that move and then quietly move back.
Design the receiving model around what does not commoditise
The receiving organisation is a design problem in its own right and it is usually treated as a recruitment problem. The difference shows up in three places.
- Ownership, not staffing. A team that owns an outcome end to end behaves differently from a team assembled by skill and pointed at a backlog. The second is cheaper to recruit and more expensive to run.
- Seniority and escalation. Between a headquarters and a delivery centre sits a gap on escalation, seniority and what "done" actually means. It is a recurring failure mode and it is almost never on the programme plan.
- What the centre is allowed to decide. A centre that can execute but not decide will send every judgement back across the time zone, and the cost of that appears as delay rather than as headcount.
Set the evidence gates before the saving is recognised
The most common structural error is recognising the saving on a date. The transfer is scheduled, the date arrives, the saving is booked and the transition is reported as complete. Then incidents arrive at month three to six, when fixes cost most, and the saving reverses in a line nobody connects back to the decision.
The alternative is to recognise the saving against a condition. The transfer passes a gate, the gate is evidenced by someone who is not delivering the transfer, and the evidence is dated and kept. This is the single most useful thing an independent party contributes to a capability centre, and it is the thing a delivery partner cannot credibly do for you.
Common questions
We already have a centre. Is this only for new ones?
No, and the second-wave case is often the stronger one. An existing centre that is delivering capacity rather than capability has usually reached the limit of what a headcount model can produce, and the next increment of value comes from redesigning what it owns rather than from adding people to it.
Is this an offshoring service?
No. We are independent of everyone who delivers, including the offshore vendors and the systems integrators. We do not staff the centre and we do not compete with the firms that do. The judgement about which capabilities should move, and whether some of them should move at all, is not one you want made by a party paid per seat.
Does the cost case still work?
Sometimes, and less often than the model assumes. Labour arbitrage has narrowed, and in several markets a regional location in the same country is now price-comparable with an offshore one once the true dual-running and escalation costs are carried. The case that holds up is the one built on capability, resilience and coverage, not on a unit rate.
When this comes up. Comes up when a centre is approved but not designed, at a second-wave scale-up, or when an existing centre is delivering capacity and not capability.
How it is delivered
Compass and Blueprint carry most of this, with Vault holding the transfer gates and Watch retained through the first waves. Each module is a fixed deliverable behind a go or no-go gate, and the baseline earns the design. The full set of modules is here.
Related situations
- Regulatory evidence and auditLocate where control logic actually sits against where policy says it sits.
- Independent estate assessmentBuild the estate picture independently, and separate what is locked in from what is portable.
- Technology and software due diligenceRead the estate inside the deal window instead of sampling it.
Tell us what you are trying to land.
A short conversation about your situation and whether an independent accountable role is the right instrument. If it is not, we will say so. No deck follows automatically.