Phase 5 of 10 · Design

Scope and towers

What exactly will the centre do, and in what order?

The big picture

Scope is not a list of processes. It decides the talent you need, the technology you must deploy, the governance that will work and the value you can promise. Scope shaped by whoever objects least produces a collection of fragments nobody would have designed.

The decisions that matter

  • Which processes or process segments move
  • How much of each process the centre owns: execution, exceptions or end to end
  • Which business units, geographies and legal entities are in scope
  • The tower structure, and the phasing into waves

What good looks like at the end

  • A process heat map: ready now, ready with redesign, stays with the business
  • A tower design with its logic stated
  • An end-to-end ownership map (who executes, who approves exceptions, who fixes recurring issues)
  • A wave plan with wave one built for stability

Who owns it

The programme lead and future GCC Head, with process owners; criteria agreed by the sponsor before consultation

When

Alongside location and operating model design, finalised before transition planning.

Choosing a tower design

ApproachBest suited whenRisk to manage
Process-based (e.g. Record-to-Report, Purchase-to-Pay)Several ERP instances, early-stage centre, standardisation focusFragmented business ownership; SLAs disconnected from outcomes
Function-based (e.g. Corporate Finance, Business Finance)One large entity, distinct business unit needs, partnership mandateDuplicated expertise; harder enterprise standardisation
Outcome-based (e.g. Financial Close, Vendor Management)Mature centre (3+ years), single ERP, strong sponsor alignmentHigh design complexity; needs cross-functional data

Standardise first, then evolve the tower structure as capability grows.

What specialist advisers say

Curated from published work by Everest Group. Summarised in our own words; follow the link for the full piece.

  • Everest Group

    GCC mandates are moving beyond cost arbitrage toward more strategic, innovation-led and higher-value functions. Worth designing later waves with that direction in mind.

    From the report summary; the full report is for members.

    Global Capability Centers (GCCs) Market Outlook for 2026, 5 March 2026

In detail

The approach

  1. Apply the three filters to every process

    Standardisability (same inputs, same outputs), system independence (can it run with remote access?) and judgment intensity. Anything that fails the third filter does not belong in wave one.

  2. Produce the heat map

    Classify every process as ready now, ready after redesign, or staying with the business. Use real exception rates, not theoretical ones.

  3. Choose the tower logic deliberately

    Process, function or outcome based. Be explicit about which and why; towers formed by default are redesigned later at higher cost.

  4. Negotiate end-to-end ownership now

    Map every step: who executes, who approves exceptions, who owns recurring issues. Ownership given away during stabilisation is very hard to win back.

  5. Phase by progressive complexity

    Wave one: high volume, high standardisation, low exception rate. Wave two: moderate complexity. Wave three and beyond: judgment-intensive work, centres of excellence and analytics.

  6. Set up scope governance

    Test every later change request against suitability, capacity and capability, technology readiness and value impact. The GCC Head owns this.

Readiness signals

The signs that need attention before a centre moves on from this phase.

  • Every process in scope has passed the three filters, honestly
  • Wave one processes have known, low exception rates
  • Each tower has a stated logic and an outcome owner
  • End-to-end ownership is agreed in writing with the business
  • A scope governance owner and change criteria exist

Where organisations go wrong

  • Letting business units veto processes from scope

    What works: Agree objective criteria before consultation and apply them consistently.

  • Wave one chosen on volume alone

    What works: Volume plus standardisation plus a low exception rate.

  • The accountability cliff: the centre runs the middle, the business keeps the ends

    What works: Own the process from trigger to close; escalate only genuine judgment calls.

  • Scope creep, one “small favour” at a time

    What works: A formal scope change test owned by the GCC Head.

Questions for leadership teams

  • Have we been honest about judgment intensity, or called borderline processes transactional?
  • If a tower misses its SLA, who owns the root cause, and can they fix it?
  • What is the actual exception rate of each wave one process?

Read more