Moving Every Invoice in the Group to One Building
Companies consolidate transactional finance, human resources, and procurement from every business unit into a single centre. The savings are real and the failures are consistent, and both come from the same source.
The Duplication Being Removed
A group with twenty business units frequently has twenty accounts payable teams, twenty payroll functions, and twenty procurement processes. Each was built locally, uses different systems, and follows different rules.
Most of that work is transactional and identical in substance. Processing an invoice is processing an invoice, and doing it twenty different ways in twenty locations is expensive without being valuable.
A shared services centre consolidates those activities into one organisation serving the whole group, typically located somewhere with lower labour costs and adequate skills availability.
Where the Savings Come From
| Source | Mechanism |
|---|---|
| Scale | Fixed supervisory and system cost spread wider |
| Standardisation | One process instead of twenty |
| Labour arbitrage | Location with lower cost per role |
| Automation | Worth building once for a large volume |
| Specialisation | Staff doing one task rather than many |
The automation row deserves emphasis because it compounds the others. Automating an invoice matching process is not worth building for a unit handling two hundred invoices a month and is clearly worth it for a centre handling two hundred thousand. Consolidation creates the volume that justifies the investment.
The savings are frequently attributed to cheaper labour and mostly come from doing the work one way instead of twenty. Standardisation is the part that is hard and durable, and location is the part that is easy and copyable.
Why They Fail
The failure pattern is consistent enough to be predictable.
Lifting and shifting broken processes. A centre that receives twenty different processes and continues running twenty different processes in one building has captured the labour cost saving and none of the standardisation benefit, while adding coordination cost and distance from the business.
Retained complexity. Business units negotiate exceptions during the transition, and every exception preserved defeats the standardisation. A centre serving twenty units under twenty variations is running twenty processes again.
Knowledge loss. Local finance staff carried undocumented knowledge about customers, suppliers, and local requirements. Transition plans assume documentation captures it and it never fully does.
Service deterioration. The business unit that could walk down the corridor now raises a ticket, and response quality is measured by a service level agreement that captures speed rather than usefulness.
The Governance Question
The structural issue is that the centre serves internal customers who did not choose it and cannot leave.
Without a market discipline, the centre can meet its service targets while its users regard it as unhelpful, and the users can blame it for problems it did not cause. Both happen routinely.
Organisations address this with service level agreements defining response times and accuracy, chargeback mechanisms allocating the cost to business units in proportion to usage, and governance forums where users can escalate.
Chargeback is the most consequential and the most argued about. Charging units for what they consume creates a discipline on demand, since a unit paying per transaction has an incentive to reduce transactions. It also produces disputes about allocation methodology that can consume more management attention than the amounts involved justify.
What Should and Should Not Move
The dividing line that works is between transactional and judgement work.
Invoice processing, payroll execution, expense auditing, master data maintenance, and standard reporting are rule based, high volume, and benefit from consolidation.
Business partnering, commercial negotiation, pricing decisions, and anything requiring context about a specific market generally does not, because the value comes from proximity to the business rather than from process efficiency.
Organisations that moved judgement work into centres to hit a headcount target have typically reversed it, having discovered that the finance business partner who understood the operation was doing something the transaction processor cannot.
The Evolution
The model has changed in two directions.
Global business services extends the concept beyond finance to include human resources, procurement, legal operations, and information technology under one organisation, on the reasoning that the coordination benefits are larger across functions than within one.
Automation has reduced the labour arbitrage rationale substantially. Where a process can be automated, the location becomes far less important, and several organisations have relocated work back closer to the business because the headcount no longer justified the distance.
That second development is worth noting for anyone evaluating a location decision now. Choosing a site for labour cost, on work that will be automated within five years, optimises for a variable that is about to stop mattering.
The Bottom Line
Shared services centres capture scale in work that was duplicated across a group, and the durable saving comes from doing it one way rather than from doing it somewhere cheaper. They fail when exceptions are preserved during transition, which reproduces the original fragmentation in a single building, and when judgement work is moved along with transaction work. Automation has largely removed the labour arbitrage rationale, which means the standardisation argument is now the only one that survives.