Last week it was reported that Citigroup struck a deal to sell its wholly-owned offshore IT and BPO center Citigroup Global Services, with offices in Mumbai and Chennai, India, to outsourcer Genpact for nearly $700 million.
MORE ON Outsourcing
Neither Citigroup nor Genpact (which was the wholly-owned offshore captive center for General Electric before it became an independent company in 2005) has confirmed the deal; but the news reports have other executives with captive centers in India wondering whether it’s time to offload their offshore assets, too.
The reason: a number of IT organizations set up wholly-owned centers offshore with wholly unrealistic expectations. Do it yourself instead of outsourcing to a third party and, the thought was, and you could access new markets, retain management control, and save more money to boot. Faced with the reality of rising costs, from real estate to salaries, along with greater management investment that they’d bargained for, some companies may be wondering if Citigroup, or GE before them, had the right idea selling off all or part of their captive facilities in India.
Not so fast, says Cliff Justice. Justice, head of globalization for Houston-based outsourcing consultancy EquaTerra, says that most captive centers can be fixed by partnering with local providers and resetting expections. That means CIOs and other executives need to get more realistic about their expectations associated with human resources, management overhead, and other costs. Besides, Justice says the chances of selling a captive data center for a profit are slim to none for most would-be sellers who lack the scale or scope of Citigroup’s BPO center.
Justice spoke to Senior Editor Stephanie Overby about ways to improve the performance of offshore captive data centers.
CIO.com: What percentage of U.S. companies offshoring IT and business process services are doing so via their own captive centers?
Cliff Justice, EquaTerra’s Head of Globalization: First of all, although companies are developing captives in China, in Central Europe, in Latin America, and other parts of the world, India is still the predominant location for captives.
We count about 500 captive centers in India right now. So probably half of the companies who are in India in any significant way are doing so with a captive center. It’s predominantly financial services and software companies.
CIO.com: Why have so many companies decided to go it alone offshore, rather than outsourcing to a third party in India? Lower costs? The lack of mature providers? Desire for control? Regulatory restrictions?
Justice: The market is mature now. There really is no argument that in areas like application development and IT services and transaction processing, that the market is not mature. There are even leading edge knowledge process outsourcers in areas like analytics.
The primary driver has been control. Everyone goes offshore to reduce costs. But the main reason for not going with a third-party in maintaining managerial control. But some customer still feel they can’t get their arms around how to outsource to a third party so they’re moving to that work to a wholly-owned captive model.
But you have to keep in mind that although these companies have set up captives, most of them still have relationships with third party outsourcers for staff augmentation. So they’re not capturing all the margins the service provider would make. Motorola, Texas Instruments, American Express – they all have captive centers, but they also outsource some work to Wipro and Infosys and IBM and all the big companies in India.
CIO.com: So Citigroup is getting a lot of attention putting its 10,000 person captive center in India up for sale. Is this a reflection that the captive center wasn’t working out for Citi, or was this just an economic opportunity they couldn’t pass up? A sign of things to come or an isolated incident?
Justice: Well, Genpact was the first big case of a company selling off their captive center. [Editor’s note: General Electric sold a 60 percent stake in its BPO operation, GE Capital International Services (GECIS), to private investment firms in 2004. GECIS is now known as Genpact and still serves some of GE’s IT and BPO needs. For more on Genpact’s expansion in the Western hemisphere, Genpact Makes Mexico a Passage to India.]
A [captive center] is difficult to manage. But you don’t see a whole lot of complete divestitures. Some companies may decide to outsource part of the work and maintain a piece of it.
Most captive centers aren’t going anywhere. There are a handful that may be in real trouble, not performing as well as expected, and those companies may divest themselves of the centers or get out of India altogether. And it will make news when it happens. But it’s really rare that someone will get out of India altogether or offshore location. The overwhelmingly popular solution is to make it work by outsourcing to a third party or by improving processes.
The issue of successfully managing a captive center is one that we’re really focused on. We have a team of consultants on the ground in India dedicated to addressing captive issues. But companies with captive centers that are not performing or meeting expectations will not completely divest themselves, but go to third parties and either outsource the work or take it over in some kind of hybrid or virtual captive center fashion.
CIO.com: What is a virtual captive center?
Justice: The whole idea of a virtual captive center is that a third-party outsourcer providers all the infrastructure, people and services, but the process mirrors those of the customer. Issues like security and compliance are addressed, because the security and compliance processes mirror the customer’s internal security and compliance processes. It looks more like an extension of your company than an outsourced environment.
CIO.com: So they do things your way, even if it’s not necessarily the most efficient or cost effective way?
Justice: Yes, they do it your way, even if it’s not the best way. That takes care of a lot of political issues or organizational resistance to outsourcing.
CIO.com: And what do you mean by a “hybrid” model?
Justice: Some type of hybrid with combined management structures in certain process areas, integrated with the larger organization and managed under typical service level agreements (SLAs).
CIO.com: So you may gain some efficiency by partnering with an experienced local outsourcer. What’s the trade-off?
Justice: Managerial control. The whole idea of the hybrid, however, is that you can adjust that managerial control to your comfort level. If want to move over to service provider’s platform onto their service delivery model, you are going to lose that day-to-day operational control. If you want the service provider to mirror your own company’s internal service delivery model, you’re making a trade-off. It gives you a greater degree of organizational comfort, but it costs more. With the hybrid model, you can tailor it to your needs.
Managing these captive centers requires an enormous amount of management attention. There are wages issues, currency issues, all kinds of things that come into play.
For example, the captive center is paid for by the parent organization in U.S. dollars, but it has to pay its employees in rupees. With the U.S. dollar depreciating, that’s a problem.
Attrition is also a huge issue. Managing the flow of resumes and hiring and keeping the human resources structure working – that’s a big undertaking. And a captive center rarely attracts the best talent, because it offers limited career paths. The general perception in India is that captive centers get the grunt work, even if their plans seem more grandiose. IBM, Infosys, and Wipro can offer long term careers, so captive centers have to pay more than them to attract the best and brightest.
CIO.com: As you describe all the management issues and headaches – and the fact that, according to some research the average captive center actually costs more than the outsourcing alternative. [See Comparison of Outsourced Versus Captive Solutions for Capturing Outsourcing Value.] One wonders why anyone would take the captive route.
Justice: There are regulatory issues such as data control requirements, particularly in the financial services area, that make some things difficult to outsource. Some of it is also the whole idea that outsourcing may be something that certain companies don’t feel like they can manage. When you outsource, you do have to give up a certain level of control in how the process is carried out and put more effort into relationship management. You have to have a certain level of discipline within a company to move in that direction – being able to articulate requirements, to manage business unit demand. There’s also the matter of governance.
CIO.com: The common wisdom used to be that the average company might start off working with an outsourcing provider offshore, and as they ramped up operations, move to the captive model to capture more savings and garner more control. But it sounds like the opposite is happening.
Justice: Yes, a lot of these big companies – GE and Motorola and American Express started out heavily captive and over time, outsource more and more of it.
CIO.com: So how do you measure the success of your captive center?
Justice: Understand where your captive is compared to peers and third parties. If I’m running a captive center, I would at least want to know that so I can make some informed decisions.
You would want to benchmark so you know where your costs are compared to the third parties. Most outsourcing advisors who do a lot of deals aggregate this kind of data.
But price is only one element. You also need to measure performance and customer satisfaction. One of the biggest complaints with offshore captive centers is not their performance but satisfaction levels within the business units.
Salary data is also out there. What are you paying your captive center employees compared to your peers. Service delivery cost also goes beyond salaries. The management requirements associated with a captive center can drive up costs.
A McKinsey study indicated that a captive center is 30 percent more expensive than working with a third party outsourcer. I agree with that. In fact, that’s probably on the low end. There are exceptions. There are some finely tuned captive centers out there that operate extremely well. But there are the exception.
That wasn’t the case three or four years ago, but the market has matured to the point that these service offerings are now more standardized in the open market and third parties are able to provide these services more cheaply.
CIO.com: Is part of the problem managing expectations? A lot of organizations may have underestimated how much management overhead would eat into productivity and savings in a captive offshore delivery center.
Justice: Expectations are always a problem offshore, whether you’re talking about outsourcing or captive expectation. People who outsourced to India expecting 60 percent savings and easy access to the best engineers in the world were disappointed.
One of the bigger problems is the fight for talent in the big cities. In some places, higher level managerial talent is as expensive as it is in the U.S. And real estate costs have doubled. In a market where the expectation is that everything is cheaper, that’s an unwelcome surprise.
Some companies started captive centers and didn’t understand what they were getting into. And some sent their best and brightest over to run it and do it very, very well.
CIO.com: Are there certain cases where a captive center has no chance for success:
Justice: If you have 200 or 300 people offshore. You won’t get a good return on a small captive. If you have scale, 1000 to 2000 employees, a captive center makes more sense.
CIO.com: So do you advise against selling off a captive center, the way Citigroup is doing?
Justice: Selling is what everyone wants to do. They think they have this valuable asset that they can now monetize it. But it’s very rare that they have something in their internal captive center offshore that would actually be worth a lot of money to a service provider. They are few and far between.