Infrastructure Outsourcing in the Age of Convergence

Once upon a time - as recently as five years ago, in fact - the typical infrastructure outsourcing deal involved an IBM or EDS taking over the whole shooting match (data centre, machines, people) with the stated purpose of "transformation". Service providers could manage IT operations about 30 percent more cheaply because the cost of delivery was spread out among its customers and amortized over as much as a decade. Sure, customers ceded some control. But many viewed it as handing over a headache, with a cost benefit to boot. At the time, few customers would consider sending such work offshore, Iannone says. "The cost and reliability of telecom services offshore prohibited it." What CIO would have been comfortable outsourcing work to a country where blackouts were the norm and redundant systems and networks weren't?

Over the years, however, those telecom roadblocks were removed, redundancies were built into systems and tools for remote infrastructure management matured. "Customers started to look at the application development being done for 30 cents in the dollar in India, saying, If we could do that with operations, wouldn't that be great?" says Ross Tisnovsky, vice president of research for the Everest Research Institute. India's IT services providers, looking at network operation centres (NOCs), which supported application work and then sat idle 12 hours a day, saw an opportunity, too. Their logic, says Iannone: "I can continue to run this through the night, do infrastructure work and make money while I sleep." The offshore vendors had little appetite for customers' expensive assets and staff that the legacy providers gobbled up as part of their deals. But no matter. They had an infrastructure management sales pitch of their own. They called it "asset-light". "They could match that 30 percent savings customers could get from the transformational deal with a legacy provider through labour arbitrage," explains Tisnovsky, without taking control of assets, rebadging any employees or really changing a thing.

Today, offshore providers are starting to compete head-to-head with legacy providers, says Paul Roehrig, principal analyst for Forrester. In their hunger for this new market segment, some India-based providers have shown more of a taste for asset acquisition than in the past - making some of their offerings more like those of legacy providers. Last year, for example, Wipro acquired US-based infrastructure services provider InfoCrossing, and TCS is promising to build out its delivery infrastructure in Ohio. Meanwhile, IBM and EDS are supporting their established infrastructure deals with lower-cost, offshore labour. That data centre may be in Plano, Texas, or Boulder, Colorado, but the people aren't.

"If the contract doesn't spell out restrictions for offshore infrastructure support," says Strichman of Nautilus Advisors, "it's going there." (Both the India-based providers and legacy multinationals have thus far built the majority of their infrastructure management systems and staff in India, though they are expanding to other offshore locations, including Latin America and Eastern Europe.) Notes Rob Finkel, a partner in the global sourcing group at the law firm Milbank, Tweed, Hadley & McCloy: "It's hard to find a pure, domestic deal any more." (For more, see "It's 2008. Do You Know Where Your Infrastructure Is Managed?", end of story)

Still, differences remain. "You can't get away from the issue of scope and scale," says Forrester's Roehrig. Legacy infrastructure providers "have years of expertise taking on massive deals, large amounts of people and large amounts of assets. The Indian providers have good profitability, cash flow and a pretty decent growth strategy. The real question is: Can they grow fast enough to get the scale that they need to compete?"

Winners and losers will be determined at a later date. But the movement of infrastructure management services offshore creates new options - and new challenges - for CIOs.