By many measures, outsourcing is a great success story. Service providers may have taken a severe beating from politicians and the press this election season, but they seem to have won the battle that counts most — persuading executives that outsiders can often do nonstrategic work cheaper and better. The result is a worldwide IT and process outsourcing market that Gartner predicts will swell from $293 billion in 2003 to $429 billion by 2008. As one executive said in a CFO magazine survey last summer, "Outsourcing cannot be stopped."
But you don't need to look far for signs of trouble. First, there are the high-profile flops, including JPMorgan Chase & Co.'s recent decision to pull the plug on a seven-year, $5 billion deal with IBM; the cancellation of a big contract between The Dow Chemical Co. and EDS; and EDS's ongoing trials with the U.S. Navy, which have cost the outsourcer $1.1 billion and contributed to a recent credit downgrade. Then there are the failure rates for outsourcing in general, which range from 25 percent to well over 50 percent, depending on the study.
Granted, failed transactions are nothing new. But the size of these failed deals alone demands a closer look at what went wrong — shredding a $5 billion contract is jarring for both client and provider. Moreover, it appears that better project management may not be the answer: most outsourcing problems derive from the way executives conceive their deals and set them up.
Last year, the board of a New England-based technology company decided that the firm should emulate its peers and outsource its software development to India. It seemed like a good idea — the average hourly cost of a software developer in India is $6, compared with $60 in the United States. But the project has gone poorly: the U.S.-based employees are struggling to manage programmers sitting halfway around the world, and much of the work coming back from India has been subpar. "The board said, 'We want you to outsource,' " reports a finance manager at the company. "That command flowed down to the executive team and they forced it onto the guys managing the project. We just rushed into it."
This illustrates two problems that undermine many outsourcing efforts. One is that companies simply sign up for projects without enough preparation. "It can be like an M&A activity," says Mark Lutchen, partner and leader of PricewaterhouseCoopers's IT business risk management practice. "Once the CEO has decided he wants to do the deal, it's up to the CFO and everybody else to make it work."
The second problem is more worrisome: executives often strike big deals for the wrong reasons. They may, for instance, blindly imitate competitors when in fact outsourcing is not appropriate. Or they may turn to outsourcing as a way of papering over deeper problems.
Consider a situation that was common in the early 1990s — outsourcing as a backdoor way for financially strapped companies to raise cash. "The dark side of outsourcing has always been that it can really just be financial engineering," says Bruce Caldwell, an analyst at Gartner. "The outsourcer might say, 'We'll take over your assets to help you balance your books. You'll get some cash to the bottom line for the next quarter, and in exchange we'll have revenue from you for the next 10 years.' It's basically a loan."
Such deals are less common today, not least because they pose big risks for outsourcers themselves. But analysts and consultants say that companies continue to make another error — outsourcing as a way of shifting responsibility for a bad process. "Sometimes with large outsourcing projects, companies think that they're transferring responsibility," says Brian Keane, CEO of Keane Inc., an IT outsourcer and consulting firm.
This rarely works. One reason is that complexity actually increases with outsourcing. What was formerly a technical job when the process was in-house becomes a management challenge complete with detailed performance measures, outside relationships, and the need to adjust the outsourcing project to changes in the business.
Furthermore, when a company hands off a defective process with no plan for a fix, the process just gets worse. "I've seen companies say, 'We hate our process and our systems,' but their fundamental proposition is to do it cheaper," comments Bob Pryor, CEO of Capgemini Energy, a joint venture between Capgemini and TXU, a Texas-based energy concern. "Employees think they've got someone to fix the problems. But when they don't see a fix — in fact, the process may be worse in some areas — you take real hits to consumer satisfaction."
As with nearly any business concern, help is available — for a fee. In recent years, a subindustry has sprung up around outsourcing. Former executives, lawyers, and other consultants are all hiring themselves out to help decide whether and how to outsource.
An obvious choice for many companies is to bring in consultants from the big service providers to tell them how to repair their processes. But caution is in order here. According to analysts and former employees of the service providers, because the consultants have a strong incentive to win more outsourcing business for their firms, the client may not receive unbiased advice.
"When the recession hit, our management asked, 'What's recession-proof?' The answer: 5-to-10-year contracts," says one former Accenture consultant. "It was clear to the consultants that we had to try to develop any engagement into a services contract."
But outsourcing is not always the best answer for a client. For instance, the right choice for a company hoping to spend less on its call center might be to do without one. "What if process engineering eliminates the need for the call center?" asks Tom Rodenhauser, president of Keene, New Hampshire-based Consulting Information Services. "The issue [for the IT- services firm] is that the consulting solution has to point to outsourcing."
Michael Murphy, a partner in the technology practice of law firm Shaw Pittman LLP, argues that while this may be true in some cases, it's not a reason to avoid using the consulting arm of a service provider. "Their recommendation could be a valuable contribution," he says. "The trick is to be alert to the conflict of interest and manage it." This could mean seeking a second opinion from an independent firm.
How Big Is Too Big?
Another reason many outsourcing megadeals founder is that the model itself may not be right for all projects. In theory, managing one vendor rather than several should be easier. The value of the project should drive the outsourcer to do everything possible to make it succeed. But the very size of a megadeal can be a liability. "A deal can get into trouble when it's a total transformation: 'Over the next 10 years, we expect you to replatform our systems and take us into the modern age,' " says Keane. "It's so large and ill-defined that there's a lot of room for disappointed expectations."
Using one provider for a large, many-sided project may mean trading expertise for the convenience of one-stop shopping. While this may seem like a fair trade, Pryor of Capgemini contends that it can cause trouble down the road — when a large deal stumbles, the reasons very often stem from the provider's weak points. "As a service provider, you have to ask whether you have the capabilities to deliver on all of the client's expectations. In some of these megadeals, they don't." With its TXU project, Capgemini is addressing this problem by subcontracting the human-resources component of the work to Hewitt Associates, a model Pryor expects the firm to apply elsewhere.
Single-provider megadeals can work, say experts, as long as companies take steps such as breaking the contract into discrete pieces with clear metrics, and establishing a governance structure that gives the outsourcer continual feedback from the client. But increasingly, companies are turning to another model: smaller projects with more providers, an approach known as "multisourcing." This has the advantage of providing more specialization and flexibility.
Farming the work out to multiple providers is no panacea, though. JPMorgan tried this in the late 1990s, corralling several big firms into the much-hyped Pinnacle Alliance. But the service providers had difficulty coordinating their work, causing the deal to collapse under its own weight. (Ironically, JPMorgan's solution was the single-provider project with IBM that it canceled in September.) "You can use multiple providers, assuming you have someone deciding who does what when," says Rick Nathanson, CEO of outsourcing advisory firm Nathanson & Co.
One deal worth watching is Procter & Gamble's $4.2 billion effort to outsource most of its shared-services operation, which spans IT, human resources, and facilities management. P&G has parceled out the work to three providers: Hewlett-Packard, IBM, and Jones Lang LaSalle. The company initially intended to give the whole package to one company, but had second thoughts when the outsourcing sector declined a few years ago.
Filippo Passerini, P&G's chief information and global services officer, argues that the company's approach will help it avoid the fate that has befallen both single- and multiprovider megadeals. First, the deal involves enough providers without adding too much complexity. "Having three providers strikes the best possible balance between managing multiple relationships and getting experts in each area," says Passerini. The company has assigned to each service provider a senior executive and a team of employees who will oversee the relationship, monitor hundreds of performance measures, and coordinate the work.
P&G has also avoided the temptation to outsource the entire shared-services operation. It will keep about 40 percent of the work in-house, including sensitive areas such as IT systems architecture and financial reporting.
Despite the problems with outsourcing, there's little doubt that — done right — it can improve things for companies. "I know of no industry where a company can be a leader without a high-performance IT organization, and the fastest way to get there is by smartly and selectively outsourcing," says Keane. That may be true. But as the urge to outsource spreads to companies new to such projects, there is a real risk that outsourcing will continue to be a good idea that often turns out badly.
Don Durfee is research editor at CFO.