AN OFFSHORING JOURNEY
Long dissatisfied with the lack of flexibility in outsourcing models, some CFOs are settling on a complex mix of outsourcing and shared services. Is it better than the alternatives?
By Simon Littlewood

Most CFOs who have been in big company finance for at least 10 years have taken an outsourcing journey. That’s certainly true for Sujit Mukherjee, finance head and director of Global Business Services for Invensys, a US$8 billion global industrial automation, transportation, and controls company headquartered in the United Kingdom.

Mukherjee, who is based in India, has overseen major outsourcing projects for Honeywell and Bechtel, and now he leads a team responsible for global finance transformation at Invensys—which includes improving service levels, enhancing controllership, and reducing cost.

Tapped for the job of offshoring the company’s finance processes two years ago, he weighed whether to build a captive shared services center or outsource. Neither fully suited Invensys’s needs. Putting everything into shared services would centralize processes too much—some of the company’s businesses require greater customized support from finance. On the other hand, pursuing large scale outsourcing would cede too much control to a third party.

So Mukherjee settled on what might be called a hybrid approach—using outsourcing for some pieces, shared services for others, and keeping still other parts where they are in the businesses. It’s an approach that many companies—Western and Asian—are pursuing, according to consultants. Can it be made to work?

First Steps

To make his outsourcing choice, Mukherjee looked to predecessors—most prominently in Asia. Figures like Raman Roy, generally recognized to be the pioneer of Indian business process outsourcing, introduced the captive, shared services, insourcing model for transactional processes at General Electric’s GECIS unit in India. GE eventually moved to an outsourcing model for reasons of cost and control.

Like Roy, Mukherjee was open to all approaches. The first step was reviewing existing operations. With the help of a consulting firm, he poked and prodded into finance operations globally. The company has three major units—it builds rail signaling systems, as well as process control systems for the petrochemical industry, and white goods manufacturing. These businesses are distributed in 60 countries. Operations are highly decentralized. Invensys’s finance activity, for example, is dispersed across various business lines and geographies with widely differing processes and skills. Mukherjee found that over 50 percent of finance personnel were dedicated to transaction processing activities, and there was little automation and a high cost of finance.

The reviewing exercise allowed Mukherjee to establish the scope of what he wanted to do. The standard model, featuring a high degree of centralization, didn’t appeal. For one, he says, local knowledge in the diverse businesses was too important. Another reason was that operations in some countries—such as Thailand—were low-cost enough that the savings from centralization weren’t enough to justify the change. So he threw away the idea of starting from scratch, of offshoring in “a single controlled environment without legacy baggage of any kind.”

The better approach, he decided, would be to outsource selectively, and build a system to monitor the quality of services. That would mean retaining some shared service centers, altering their character, and elevating their role. In the parlance of the business, the shared service centers would become “centers of excellence,” providing an administrative overlay to look after the outsourcing. “We could deliver value far faster if we left the processes and people out there in maybe 20 geographies,” he says, “and try to develop common processes and best practices that way.” This resembled a portfolio approach—one in which the good baggage could be kept.

Mix and Match

The move to a hybrid model is one many companies are pursuing, says Arno Franz, Asia-Pacific managing partner for TPI, an outsourcing consulting firm. The wholesale outsourcing deals of the 1990s have in many cases given way to a desire to mix and match outsourcing and shared services to suit company requirements.

But that’s not to say that it’s the right decision. “What it tells me is that companies don’t have a proper, overarching sourcing strategy that meets the needs of the organization,” says Franz. “They’ve just fallen into this [hybrid approach]. What you often find at a functional or business unit level is that people are doing what they want.”

True, there are factors pushing companies into this. One in particular: in emerging markets, the big service providers are often a step behind their clients. “Especially in many Asian countries, the clients are going to be there before the service providers are there,” says Franz. A CFO may have no choice but to outsource to a local provider or use shared services even when the preference is to have someone else perform the work.

But the hybrid approach has two big shortcomings—cost and complexity. “It’s much more expensive to administer,” says Franz. “It’s complex enough to deal with one mature service provider in a big outsourcing deal. Now you multiply that by two or three service providers, maybe a shared services facility, and perhaps a captive in India. If you weren’t good at managing one service provider, what makes you think you can manage several different back office functions being delivered by different entities?”

Mukherjee’s Choice

Mukherjee, however, believes he has found the right answer. Having decided to avoid a one-size-fits-all solution, he had to work with the resources that Invensys had. The company had shared service centers for several of its larger businesses in Massachusetts, serving its English language businesses, and Brussels, for European languages. Its businesses in Asia were decentralized. The plan was to follow the model he pioneered at Honeywell and convert the North America and Europe shared services centers to centers of excellence that would monitor various outsourcing operations. In Asia, the transition would come later.

The benefit of the center of excellence plan, says Mukherjee, was that the decision to outsource and set up a shared service center, or simply stick with the status quo, could be entirely made on the business case. In China, for example, a center of excellence could look at the cost of employees for certain accounting functions. Mukherjee could weigh whether it was more expensive to retain those employees, or to outsource the activity, and make that decision region by region within China. Currently, Invensys uses an outsourcing provider in Dalian for some transactional activities.

Invensys’s outsourcing project has a four-year timeline. To launch it, Mukherjee hired a project team that paired managers from within the company with temporary, outside experts. The team will launch the transition in two stages. First, it will identify those processes which can be easily outsourced to deliver immediate savings. Following the full transition to the center of excellence model—still two years away—Mukherjee envisions a move to business planning and analytics generated at the centers and available when local businesses require them.

The question remains, of course: will all this turn out to be as—or more—costly than the big, “center of control” outsourcing models of yore? If so, will the value of analytics make that higher cost worth paying? In looking to improve the model, CFOs will have to ensure that they’re not paying too high a price for flexibility. Or outsourcing will change again.

Simon Littlewood is a frequent contributor to CFO Asia.