The business of being a cog in another company’s wheels might not be glamorous, but it can be lucrative. Third-party providers of routine operational services—such as the processing of payrolls, the movement of inventory and goods, the management of data centers, and the provision of extra manufacturing capacity—took in more than $1 trillion around the world in 2000, according to Dun & Bradstreet. What is more, the market for these services doubled in size from 1997 to 2000. Analysts estimated that this eclectic sector, made up of companies that in essence take over activities other companies choose to outsource, would grow by 25 percent in 2001 and could experience similar growth in 2002.

What is propelling the expansion? In theory, providers of nuts-and-bolts services have long had a compelling pitch to make to the large and midsize companies that make up their customer base: "Why should you directly manage the routine operational activities that make up a substantial portion of any business? Overhead tasks such as billing and human resources, as well as supply chain processes such as procurement, manufacturing, and logistics, are not an intrinsic part of your business; they are infrastructure activities. Let us manage them for you. As specialized providers of these tasks, we understand how to undertake them better than you can, and at lower cost."

Before 1997, however, customers were more likely to nod their heads in agreement than to sign up. For one thing, they were wary of becoming hostage to providers for any activity they handed over—providers that might gradually raise prices and reduce service. For another, customers and providers alike found themselves saddled with interaction costs (such as managing the relationship, monitoring the delivery of services, and coordinating the exchange of information between customer and provider) that had not been anticipated and sometimes exceeded whatever value one or the other party had hoped to gain. And some providers couldn’t deliver the promised efficiency or productivity improvement: either they were unable to exploit or even obtain economies of scale, or they couldn’t streamline or transform the processes they managed. Ultimately, they did little that a customer couldn’t do for itself.

But in the past few years, the realities of outsourcing infrastructure activities have changed rapidly. The plummeting cost of communications, the widespread use of standardized interfaces such as World Wide Web browsers, and the quickening pace at which companies are automating data have cut interaction costs sharply. Providers of infrastructure-management services, or "infraservice" providers, have entered the market in increasing numbers, and customers have become more accustomed to them. Providers and customers alike are learning how to structure deals to the benefit of both parties. Even so, we estimate that customers still manage in-house more than 90 percent of their routine operational services—an enormous opportunity for both start-up and established infraservice companies (Exhibit 1).

One company that has seized an infraservice opportunity is Johnson Controls, which has expanded from the manufacture of air-conditioning, heating, and lighting systems for buildings into the business of operating and maintaining the buildings themselves, along with their systems. In 2000, the company’s facility-management unit posted $1.5 billion in revenue (almost 10 percent of the total), and it is growing at twice the rate of the rest of Johnson Controls.

For every such company, however, several firms are languishing. In studying more than 400 third-party providers in various sectors, we found that some, such as application service providers, have struggled from the outset, unable to turn commodity offerings into sustainable businesses. Even in more attractive segments, such as contract manufacturing, certain providers are failing to flourish (Exhibit 2).

To put it simply, the providers of infraservices are successful when they manage to capture value that eludes their customer base

Quite simply, providers succeed when they capture value that their customer base can’t. They do so by carefully defining how much of a customer’s process they will manage, by opting for fixed-price contracts when possible, and by minimizing the amount of customization they do. In the long run, the providers must realize not only economies of scale but also economies of skill—by codifying the process innovations developed while serving one customer so that they can be used in work for other customers.

Creating value from scale and skill

Identifying opportunities to become an infraservice provider would seem to be fairly straightforward. Start by picking an operational activity, any operational activity. Specialists in payroll processing, for example, would typically handle this task for a number of companies, thus spreading fixed costs and achieving economies of scale. Because such specialists deliver only this service, they have the focus needed to identify areas that are susceptible to improvement and the knowledge needed to act successfully on that awareness (economies of skill). Any operational activity undertaken on a continuing basis would seem to be fair game: running call centers, managing warehouses, making meals served on airlines, processing mortgage loans or credit card transactions.

Not all opportunities to gain economies of scale and skill result in businesses. Consider the different approaches taken by two call-center providers: Sitel and Convergys. At first glance, call centers would seem to be a great infraservice business. The fixed costs of maintaining them aren’t trivial, and they experience busy periods and slack ones. A provider could lower its costs by consolidating the centers of several customers and balancing the call load.

But as Sitel discovered, it is possible to underestimate just how substantial these economies of scale must be to achieve levels of value that customers couldn’t reach on their own. New technologies (such as automatic dialers) and process improvements that allow supervisors to manage more people have raised the minimum efficient scale for call centers. Yet Sitel still fails to achieve scale economies greater than those of the big call centers (with 300 or more employees) maintained by many large companies. Other third-party call-center aggregators have achieved similar scale, and as competition among them intensified during the past two years Sitel’s service became a commodity.

Convergys, by contrast, offers services that go beyond routine call-center functions. For large telecom companies, to give one example, it can tally the minutes that end users spend on the telephone and generate bills accordingly. Convergys has a volume of business sufficient to give it the resources needed to deploy the highly sophisticated computer applications that do the job, and its mastery of those applications in turn attracts leading clients. This chicken-and-egg conundrum is explained by the origins of the company: in 1998 it was spun off by Cincinnati Bell (now known as Broadwing), which in the mid-1990s consolidated its call centers and thus had both the scale and the skill to offer valuable call-center and related services to other businesses.

Some aspiring infraservice providers, though, have failed to realize that unless they extend their innovations to a number of customers, their businesses will hit a wall, for eventually individual customers are likely to bring the streamlined activity back in-house and to manage it just as effectively as a specialist could. Only economies of scale provided by a large client base and the economies of skill developed by confronting a variety of situations can enable specialists to maintain their advantage over their customers.

Moreover, infraservice firms and their customers are now more closely integrated than ever and are often in constant touch. Communications between the two must be transparent. For instance, infraservice firms and their customers may need to adopt technologies such as electronic data interchange or new Internet-based software that will allow the companies to transfer business information efficiently. Underestimating the cost of doing so can defeat the very logic of outsourcing.

Value from design

How then do start-ups and incumbents identify a real opportunity to build meaningful economies of scale? In our research, we have found that successful infraservice providers design their businesses knowing the answers to the following questions:

1. How much of a customer’s business will we take on?
2. What pricing schemes will enable us to capture value from innovation and to limit interaction costs?
3. How can the customer’s demand for tailored service be reconciled with the infraservice provider’s need to standardize?

Working through these three questions helps would-be infraservice providers to assess more accurately whether a particular infraservice business has legs.

Define boundaries to achieve scale economies

One pitfall for a company trying to determine what range of activities to offer lies in the fact that the boundaries of a potential infraservice business can be set almost anywhere. Such businesses must decide whether to carry out only a particular activity or to handle an entire function. To arrive at the right decision, an aspiring provider must, among other things, sort through its existing capabilities and assess the possibility of acquiring related skills.

Many of today’s struggling infraservice providers have limited themselves to the precise task for which they had economies of scale at the outset. In doing so, these companies deprive their customers of the value that might have been gained had they offered to handle a larger portion of the function. Five years ago, for example, most providers of third-party information technology services managed only a sliver of the overall IT needs of a large business: data center management, applications development, network services, or the help desk. The customer gained some value but would have realized still more had a single innovative provider taken on the entire IT function. (And, of course, the lower the value to the customer, the lower the payments to providers.) Moreover, any customer that contracted with different companies for a number of IT services incurred the incremental cost of managing the provider of each of them, while each provider incurred the cost of coordinating its activities with those of the other providers.

Infraservice firms that have defined their boundaries too narrowly can also constrain innovation. Some logistics firms, for instance, provide nonmanagement personnel for their customers’ warehouses. But because such logistics firms have so little control over the way the warehouse as a whole runs, they cannot adequately distinguish the job that they do from the job that their customers did.

In most instances, the providers of infraservices would be better off defining their boundaries more broadly, at the level of processes

In most instances, service providers would be better off defining the boundaries more broadly, at the level of a process. The human-resources provider Exult, for example, offers customers the whole range of HR services—in contrast with most of its competitors, which focus on delivering narrow slivers, such as benefits management or payroll processing. By taking over the entire function, including the customer’s HR staff, Exult can change the process and therefore its underlying economics. The infraservice firm automates many HR services, offering a Web-based, self-service utility that reduces the need for personnel. Changing only selected HR services and leaving others in the hands of the customers’ in-house HR staffs would hardly move the needle on costs. Changing all of these services shifts it significantly.

Price to capture value

Price structures influence not only the incentives for both parties but also their interaction costs and the provider’s future negotiating position.

The two most common pricing choices—costs plus and gain sharing—have more often destroyed value than created it. With costs-plus contracts, providers lack any incentive to reduce costs.¹ Customers sometimes believe that such contracts will save them money by capping the provider’s margins. But costs-plus contracts also limit the incentive of the provider to squeeze costs, because they guarantee it a profit margin that no longer depends on the efficiencies it can realize by innovating, by exercising its purchasing power, or by hiring more productive staff.

A gain-sharing contract better motivates the provider to innovate and to reduce operating costs, but it also raises interaction costs.² This is the most expensive kind of contract to negotiate and monitor because the parties have to define and accept precise cost projections for every situation. Further negotiations, in which each party blames the other, are almost inevitable when the savings are lower than expected. The incentives to innovate are limited, too. Customers of one leading facility-management firm scrapped or modified gain-sharing contracts that had yet to expire, because the infraservice provider, having reduced costs "too much," was reaping too large a windfall. Both costs-plus and gain-sharing contracts extract a price from infraservice firms as well by revealing their costs and profits and thus betraying their future negotiating position.

Fixed-price contracts are a better option. When prices are fixed, providers keep the rewards from process innovation. To keep costs under control, European airlines set fixed prices for catering services. The airlines decide what they will pay per meal, competing caterers then concoct a selection of menus, and on that basis the airlines then choose the caterer.

Fixed-price contracts are also less costly to negotiate and do not require customers to be continually auditing their providers’ expenses, as they must under costs-plus and gain-sharing contracts. Exult, for example, offers a fixed-price contract to customers. Because the firm controls the entire HR process, it can charge a fixed price for each of its customers’ employees without worrying that unanticipated costs will erode its margins. But this form of pricing, too, has its hazards, chief among them the difficulty of developing a strong understanding of the actual costs.

Decisions about which pricing models to adopt are tightly linked to decisions about where to set the boundaries of the infraservice firm. Fixed-price structures succeed only if the firm has, in the first place, decided on service boundaries that are wide enough to give it the kind of control of overall costs that Exult has achieved. By contrast, the logistics firm that provides its customers with warehouse staff can’t control overall costs, so it must offer a costs-plus pricing structure for the limited services it does provide.

While providers should try to negotiate fixed-price contracts for their services, they must recognize that in all likelihood they will have to adopt different pricing schemes for different services. The shipping company APL installs logistics information systems on a costs-plus basis, offers access to its tracking system for the term of a fixed-price subscription, and charges a fixed price for every call for special information the customer makes to the call center. The choice of pricing scheme will depend on the receptiveness of the customer and the underlying economics of the offering.

Modularize customization

Customization undermines economies of scale. To win business in a competitive market, Corio, an application service provider, bowed to its customers’ demands for a high degree of customization. The further Corio went in that direction, the faster it saw its economies of scale disappear.

But providers can rarely maintain a strict no-customization policy. Exceptions include Sabre, which has standardized its airline reservation systems for travel agents, and Automated Data Processing, which has sufficient market power to resist the requests of its clients, mostly midsize companies, for a customized Web-based payroll offering. Successful infraservice firms can strike a balance by offering "standardized customization," or a modular set of services. APL, for example, has created a series of modules that enable customers to design their own feature requirements into its logistics systems, without the provider’s intervention.

From economies of scale to economies of skill

As more and more of the routine operational activities of a company become automated, the task of turning them over to third-party providers becomes easier and more economical. Meanwhile, the reduction of business processes to digital form brings the company and its providers closer together, further reducing costs.

If digitization means lower interaction costs and thus makes outsourcing more feasible economically, the plethora of firms that rush into the resulting competitive space will badly need some way to differentiate themselves once their scale-efficient services have become commodities. Indeed, some are already moving in this direction. A few contract manufacturers and logistics providers are maneuvering to control a bigger piece of their customers’ supply chains, a step that will require them to be innovative managers of inventory and information and to change their customers’ underlying supply chain processes.

Today many infraservice providers can improve a particular customer’s routine processes but face difficulties in codifying the lessons learned from serving one customer and then applying them to serve others. (Sometimes the problem is that separate management teams handle each large customer account.) In time, the kind of codification and general application that builds scale will become easier to achieve, causing more attention to be paid to economies of skill—that is, to the improvement of processes and the exploitation of insights (Exhibit 3).

Economies of skill will open up opportunities for incumbent companies to launch infraservice businesses that leverage the knowledge and information that have been acquired from one business to create value in a new and seemingly unrelated one.

• A pure-play credit card company could create an information-based direct-marketing business to take the place of the in-house marketing departments of other businesses.
• A retailer or franchiser could use its knowledge of demographics and real estate to build a site-selection and real-estate-management business that performed those functions for companies unversed in them.
• A partnership between a bank and a credit scorer could go into the business of managing credit risk.
• A leading manufacturer or construction firm could sell procurement services.
• A retailer could extend its supply chain and logistics processes to other retailers, thereby consolidating assets and coordinating orders.
• A company could turn its accounts-receivable management and collection skills into a full-service order-to-cash business.

While the new infraservice businesses are likely to be more knowledge based and less asset intensive than those that are already in operation, the infraservice businesses of the future can learn from their predecessors and avoid the snares that await them. Indeed, piloted correctly, these new companies could take their place among the great growth engines of the next decade.