At a recent gathering of executives in charge of information technology at major U.S. companies, several panelists discussed their experiences with outsourcing. One CIO, representing a large Midwest utility holding company, told the audience that his firm's decision to outsource had been a wise one, and he quickly ticked off the usual reasons: lower costs, fewer personnel headaches, the ability to focus on core business issues, and so on. Other panelists weren't so sanguine. They described, at length, higher-than-expected costs, poor levels of service, and dissatisfaction with everything from their ability to communicate with their outsourcing vendor to the vendor's ability to deliver on promises. In fact many of them said they had demanded that their contracts be renegotiated. After the utility-industry CIO heard this, he acknowledged that his own experience with outsourcing had in fact been so bad that now, realizing he wasn't alone, he was heading back to his office at once to demand new outsourcing terms. Welcome to outsourcing 1998, when a new reality is taking hold: Handing off information systems to a third party--even a multinational firm with decades of experience and thousands of highly trained employees--isn't the cure-all it was widely believed (and marketed) to be. Without proper management, good contracts, and realistic expectations, outsourcing can become a nightmare. Costs far exceed what customers thought they'd pay, service levels are poor, responsiveness to changing business conditions is slow or nonexistent, and migration to new technologies rarely follows a sensible strategic blueprint. That would be bad enough in any situation, but what makes it worse is that in most cases the customer has to bear a good part of the blame. "There's no doubt that outsourcers have painted an overly rosy picture," says Robert B. Chapman, coauthor with Kathleen R. Andrade of Insourcing After the Outsourcing (1998, American Management Assn.). "But what makes outsourcing so appealing is that IS costs are out of control, and that's because companies don't understand their own business requirements." Other experts say that customers are also guilty of: failing to benchmark their own operations, thus robbing them of any adequate way to measure an outsourcer's claims and eventual performance; believing that IS problems will go away once an outsourcing deal is signed; and not realizing how critical it is to manage carefully the relationship with the outsourcer for the life of the contract. Whether for those reasons or others, dozens of companies from Mutual of New York to General Electric Co.'s appliance division to Cummins Engine Co. Inc. to LSI Logic Corp. have renegotiated, switched vendors, or reasserted control (i.e., "insourced") over their own IS operations. Will this widespread dissatisfaction ultimately slow the strong growth that outsourcing has enjoyed for the last decade? Probably not. Despite the growing pains, analysts say outsourcing is still a booming business. In the last six months alone there have been major deals announced by British Steel PLC, Chevron Corp., and other industrial giants. Market-research firm Dataquest Inc. predicts that outsourcing will grow from a $43 billion business in 1997 to a $90 billion business by 2001. In fact outsourcing, once limited to a company's data center(s), now may extend to desktop services, applications development, even the formulation of IT strategy. There is no aspect of IT that an outsourcing vendor won't happily take on and promise to do better than the client can do. And yet "there is absolutely no doubt that there's a backlash against outsourcing," says Alan Gonchar, president of Compass America Inc., a consulting firm in Reston, Va. "It can be traced back to the way many deals have been structured, with terms that couldn't have continued over the life of the contract." Gene Procknow, managing director of the outsourcing advisory service for consulting firm Deloitte & Touche, agrees. "Many of the contracts signed in the early '90s are being revamped." Although he says that some of that activity can be traced directly to "vendors who have oversold and overhyped what they can do," IT and senior executives at customer companies aren't simply victims of slick sales pitches. "Many thought that they could abdicate responsibility for their IT functions," Procknow says. "But you're not abdicating it; you're just delegating it, and that means you still have to manage it." For some companies, even delegation has proved impossible. Take LSI Logic, a chipmaker in Milpitas, Calif. The company signed a five-year deal with IBM Global Services in 1995, but terminated the agreement last year and moved all its IT operations back in-house. "We insourced it all in 90 days," says CIO Lam Truong. "It was a very tough challenge, especially since at the same time we were implementing a new ERP system, but we managed to pull it off." Was all this to counter a "bad" outsourcing deal? Not exactly. "Being in Silicon Valley, we have to cope with a very dynamic business environment," says Truong. "When we outsourced we became very dysfunctional--we lost the linkage between our business strategy and the systems that help make it happen. Our need to change probably outpaces other industries, and in that scenario outsourcing probably isn't an option." IBM--which along with Electronic Data Systems Corp. (EDS) and Computer Sciences Corp. constitute the "Big Three" of outsourcing--has had to restructure or abandon a number of long-term deals, but remains optimistic about the future of outsourcing. "Our perception is that there is no backlash," says Stephen Huhn, vice president of global business development. "The business is far from mature, but outsourcers have already delivered great value, with spectacular cuts in unit costs, enhanced service levels, and other benefits." One of those benefits, Huhn says, is "greater flexibility for customers," thanks to contracts that take into account mistakes made in the past, including the critical issue of price. There are at least two downsides to the financial terms in most outsourcing contracts: It's impossible to compare them to other deals, and some contracts make no provision for price cuts as computing becomes cheaper. "There is no price list put out by vendors, no real way to compare," says consultant Gonchar. "We see a 300% price difference between companies who know how to drive a good bargain and those who go into an outsourcing deal uninformed about the terms they may be able to strike." Outsourcing vendors say they can't supply a price list because every deal is different. Not only does the scope of services vary, but many deals have provisions for the outsourcer to acquire a company's computing facilities or other assets, or absorb its staff. "But each deal doesn't start in a vacuum," counters IBM's Huhn. "There are a number of standard computing environments, whether a data center, desktop support, etc., that provide some basis for comparison." Good outsourcing contracts avoid a flat-price structure in favor of one that periodically assesses typical costs for providing computing services and adjusts the price when appropriate. "No matter how good a vendor is," Gonchar says, "there will be technological developments that no one can foresee that will drive computing costs down. If you're locked into a flat-price contract this is going to make you unhappy." But even lawyers who help write such contracts caution that customers should not assume that better contracts will solve all their problems. "Whether or not the relationship between the customer and the outsourcer is successful," says Richard L. Fogel, a partner with Gordon & Glickson PC in Chicago, "depends less on the contract than on the compatibility of the two organizations and their willingness to work together to solve problems." Executives who have recently entered into outsourcing arrangements say that's a lesson that's not lost on them. When British Steel announced, at the close of 1997, that it had signed a 10-year deal worth 400 million (US$650 million) with Cap Gemini S.A., the huge European IT services and consultancy company, forming a true partnership was very much on the company's mind. "We wanted a company that could not only take over our existing systems on day one for the same cost," says Roger Thackery, director of IS technology and services for British Steel, "but could also help us transform the company to meet our business goals." Thackery adds that while a well-drawn contract provides a good start to an outsourcing arrangement, "the relationship doesn't come in a contract. When the inevitable tough times come, you have to be able to get issues on the table and talk about them, not sweep them under the carpet." He echoes the sentiment that management of expectations is vital, and one way to do that is to "talk to as many reference sites and third-party sources as possible and find out both what has worked and what hasn't." Odds are that a good deal of those opinions will be negative. A 1997 survey by Deloitte & Touche found that outsourcing was not meeting expectations in several key areas, ranging from price and service levels to strategic business issues such as the ability to concentrate more fully on core competencies or transition to new technologies. "While we do see vendors make outrageous claims, such as promising 'legendary' service that is, at best, only equal to what existed prior to the deal," Deloitte & Touche's Procknow says, "buyers are also to blame for what amounts to an expectations gap." And the buyers agree. "When we first outsourced," says LSI's Truong, "we had no understanding of what to outsource. We had no road map, no metrics. We just had a hodgepodge of systems that we turned over to the vendor, hoping for the best." Companies that are more deliberate fare much better. Five years into its 10-year, $500 million outsourcing deal with EDS, Bethlehem Steel Corp. is happy with the arrangement. "We put together a list of 12 or 13 goals for our outsourcer," says IT director Tom Conarty. "We wanted to cut costs, have access to the latest technologies, improve our skills base, and more." That road map not only helped the steel giant make its vendor selection, but keep the relationship on track once a deal was struck. "We structured ours as an incremental contract that allows us to buy more or fewer services as our needs dictate," Conarty says. "That's worked out well, because over time we've seen that we may need less in one area, more in another." The concept of benchmarks--quantitative measures of a company's performance levels versus its costs--is a key not just to striking the proper terms with an outsourcing vendor, but in deciding whether to outsource at all. For example, while Bethlehem Steel is midway through its 10-year deal and British Steel has just entered its own long-term outsourcing arrangement, Inland Steel Industries Inc. took a hard look at outsourcing last year and decided it was doing fine on its own. "We've benchmarked for years," says CIO Bill Howard. "And [through] that, downsizing, and upgrading our technology recently, we sensed we were doing pretty well." Still, if outsourcing would have made more business sense, the company was prepared to do it. "We put together a very extensive RFP," Howard says, "and gave it to four top outsourcing firms. They had such a hard time seeing how they could do what we were doing for less that we didn't even go to full bid." Some experts believe that outsourcing can provide a marked advantage--when certain conditions are in place. "I think that when an industry is changing fast, outsourcing big-ticket items like IT makes sense," author Chapman says, "because you can maximize your ROI. Financial-services companies are good at this, at spending wisely on capital investments. Other industries are less knowledgeable about managing money that way." The other scenario in which outsourcing makes sense, Chapman maintains, is when "the IT department is out of control." That's fairly uncommon, he says, because problems are more often traceable to business units that fail to articulate or manage the implementation of business requirements. And yet disarray within the IT group certainly plays a role. In fact, British Steel acknowledges that its current reliance on mainframe computers, an assortment of homegrown applications written primarily in the '70s, and the absence of any truly integrated systems gave added impetus to outsourcing. "We are missing some important pieces," Thackery says. "For example, we have virtually nothing in the supply-chain area." British Steel hopes that outsourcing will facilitate leaving behind an IT environment in which "our projects tended to become overloaded with small enhancements, lacked strategy or discipline, took too long, and cost too much." Although British Steel has high hopes for both the technical and strategic contribution its outsourcer may provide, it has retained two important centers of expertise: a core team to manage the relationship with Cap Gemini and a group of business analysts who will act as the liaisons between British Steel's technology users and the outsourcer. Experts say that IT executives should expect to revisit the terms of an outsourcing deal several times over the life of the contract. "Since you don't know your requirements over the life of the arrangement," author Chapman says, "you can't know your costs. So the only way to be smart is to schedule fixed points for renegotiation." Increasingly, one buzz phrase entering into contract talks is "shared-risk/shared-reward," in which the customer and outsourcer frame their relationship as a partnership and describe expectations in financial terms. Although still rare (perhaps 10% of the outsourcing deals signed in 1997 were priced this way, according to the Real Decisions unit of consulting-company Gartner Group) such deals have undeniable appeal: Price is based on project performance and the outsourcer's ability to provide quantifiable deliverables. "It can be very tough to measure the contribution a vendor makes to a company's mission, not to mention embedding such terms in a contract," says Len Bergstrom, executive vice president of Real Decisions. "But when we surveyed 250 companies that had outsourced, we found that about 70% had restructured the deals at some point." Better contracts are now being written that benefit not just from the mistakes of the past, but from new players becoming involved. There are many consultants who will benchmark a company's IT performance, help it shop for vendors, and suggest alternatives. Procknow of Deloitte & Touche says that in half of all companies serviced by his firm, outsourcing emerges as a less favorable approach than maintaining IT in-house, albeit often with some changes made. Despite the attention lavished on outsourcing deals gone bad, outsourcing appears not just healthy, but here to stay. "I think that while it was pegged as a fad 10 years ago," says Barry Sullivan, vice president for Internet and electronic business at EDS, "it is now emerging as a key driver of the virtual enterprise." Allie Young, an analyst at Dataquest, agrees. "I think outsourcing deals are getting better as both parties insist on a win/win relationship and look beyond costs savings to things like strategic partnering, better communication, and greater flexibility."
How Long? |
Debate rages over length of contracts. One of the most breathtaking aspects of well-publicized outsourcing arrangements is their duration: 10 years is normal, and there have been cases in which exotic forms of joint ventures have been structured as 50-year deals. Will long contracts be the norm going forward? "While we've seen other types of outsourcing deals move toward shorter time frames," says Barry Sullivan of EDS, "when it comes to information systems, 10 years is still typical." Some CIOs believe this can't work. "My advice is never go long-term," says Lam Truong of LSI Logic. "I'd say a maximum of two years, and a one-year renewable is better." Sullivan counters, "That's like leasing a new car for a month with a $5 down payment." The complexity of transitioning systems to an outsourcer and of transitioning them again if the customer wants to reassert control or go with another vendor argues for longer-term deals, he contends. Other analysts say that multiyear deals of between five and 10 years will remain the norm, but with better flexibility on renegotiation or cancellation. "In the beginning," says Allie Young of Dataquest, "the high rate of contract renegotiations was perceived as failure. Now it's simply a fact of life and probably a necessity in long-term deals." "We signed a 10-year deal," says Roger Thackery of British Steel, "but we can terminate any time after two years, with 12 months' notice. "I can tell you it's not our intent to do that. The careful work done before signing the contract should give us an arrangement we can feel good about for its duration." |