In 1984 managers at Hewlett-Packard’s outpost in Boise, Idaho, launched a fledgling business based on laser technology for computer printing applications. They expected initial unit sales of 10,000 a month; they achieved 100,000 shortly after launch. By the end of the decade, HP’s managers had catapulted their Boise start-up into a $5 billion growth engine that was driven not by just one printer technology but by three. Today HP’s printing business extends into digital photography, wireless-information distribution, and e-commerce imaging. It continues to drive corporate growth.

The secret? HP’s managers relied on a corporate-level process we call "patching" to create a continually shifting mix of highly focused, tightly aligned businesses that could respond to changes in the market. In the case of laser-jet printing, the executives lopped off pieces of the core business to form new business units for such products as networked printers, keeping the laser-jet managers focused on their booming operations. They launched businesses in related products such as scanners and faxes. They transferred businesses from one division to another to make better use of skills and to optimize scale. To drive new growth, they combined businesses to create critical mass and increase cash flow. Most significant, Hewlett-Packard’s managers relied on patching to develop a second printer business built around ink-jet technology and later used patching to revitalize their "Wintel" and UNIX computer businesses. Today these businesses represent roughly 80 percent of the company’s revenues.

Patching, then, is the strategic process for routinely remapping businesses to take advantage of changing market opportunities. It can take the form of adding, splitting, transferring, exiting, or combining chunks of businesses. Patching is less critical when markets are relatively unchanging, but when they are turbulent patching becomes crucial. By dynamically adjusting businesses to match changes in the marketplace, managers are more likely to focus on high-potential businesses and to create economic value for the corporation. For managers in dynamic markets, patching becomes an enormously important skill to add to the corporate repertoire. Corporations that don’t patch are burdening their employees with misaligned organizations.

At first glance, patching may seem to be just another name for reorganizing. But patchers see things differently. While managers in traditional companies associate structure with stability, managers in companies that patch believe that structure is inherently temporary. Patchers also develop corporate strategy differently. Traditional managers set corporate strategy first, whereas managers who patch keep the organization focused on the right set of business opportunities and then let strategy emerge from individual businesses.

Our understanding of patching comes from almost a decade of research into the reasons behind corporate success in high-velocity, intensely competitive industries. One phase of the research took us inside 12 successful companies in different parts of the computer industry—the prototype of this new competitive reality. We then tested the relevance of these ideas in other industries through targeted case studies.

Evolution, not revolution

Managers who patch make lots of usually small, frequent changes to their organizational structures—think evolution, not revolution. These patches can take several forms; splits and additions are the obvious ones, but there are also combinations, transfers, and exits.

Dell Computer regularly uses splits to focus more closely on target markets. In 1994 Dell split into two parts: one, for transactions, dealt with customers who bought equipment in quantities of one or two; the other, for relationships, catered to customers who bought in quantities of 50, 100, or 1,000. By 1996 those two parts had become six. Since then, Dell has announced a new split almost quarterly. Commercial-relationship accounts are now segmented into corporate and small-business segments; government accounts are split into federal, state, and local segments; other nonprofits are divided into areas such as education and medicine. As a result, Dell’s business-level managers stay tightly focused on increasingly specific market opportunities, which they can exploit in very targeted ways.

Managers at Dell’s archrival, Compaq, also patch, but they rely more on adding new product divisions, such as storage devices and workstations, and on exits, such as the company’s departure from the networking business. Compaq’s corporate executives keep business-level managers focused on taking well-defined product-market "hills" without dictating the strategy for making this happen. Managers at Cisco Systems also frequently reshape their portfolio by adding new businesses, but they do so primarily through small acquisitions—more than 20 in the past four years. As a consequence, Cisco has transformed itself from a small networking company into a broad-based telecommunications player organized around a group of very focused business units.

Bombardier is an example of a company outside the high-tech industry that relies on patching. Managers at the Canadian manufacturer of aircraft, mass-transit vehicles, and other forms of transport undertake frequent, small reorganizations—especially those involving a kind of transfer called "raising the orbit." For example, several of Bombardier’s manufacturing businesses had developed fledgling service operations. Recognizing a chance for growth that was not central to the direction of any existing business unit, corporate executives pulled services into a new business, with two benefits: services was elevated to a higher orbit of strategic prominence, and its managers became energized and focused on their opportunity. It has since become an important part of the corporation.

Hewlett-Packard’s managers have relied on a wider repertoire of patching maneuvers than most. In one move, managers combined a new networked laser-jet printer business based on an emergent technology for an established market with another printing business based on an older, cash-cow technology. The rationale was twofold: to transfer market knowledge from the older business to the new one and to fund the new business. More recently, HP’s managers used a transfer move: they built a "parking lot" in which the established scanner business was temporarily grouped with new businesses in Internet applications and wireless appliances, thus giving critical administrative scale to these enterprises. Perhaps best known is the now legendary split of the ink-jet from the laser-jet business, though this patching move was only one of many that built HP’s printing empire, which now dominates the industry around the world. The company’s recent split into two parts, however, is more a classic reorganization than a patch.

As these examples demonstrate, managers at patching companies often make the same moves again and again, creating routines—from selecting the patch to precise details of implementation—that work whether their patching changes are small, as they usually are, or large. Therefore, although all managers find large changes more challenging than small ones, managers at patching companies usually execute large moves more effectively than their traditional competitors for one obvious reason: they are limber from frequent repatching, while their competitors are, comparatively, out of shape.

Size matters

Managers who patch well pay close attention to the size of their business units. They balance on the edge of chaos, between agility and efficiency. Small business units allow managers to focus on the specific demands of key customer segments and to exploit niche opportunities for growth. They also make it easy to pursue fragmenting markets, markets evolving at different rates, and businesses that need different operating models. However, units that are too small may result in inefficiencies: they may have excessive overhead, require too much coordination, and suffer loss of scale in economies or market power.

NovaMed Eyecare Management, a medical start-up, is a telling illustration. This highly successful venture provides top-quality eye care with cutting-edge technology at a competitive price. How? For one thing, by carefully managing the size of the practice groups in each of its markets. In a regional market like metropolitan Chicago or St. Louis, there will be enough doctors in the group to create the right scale to support the professional management, information technology, ancillary services, and media access that drive efficiencies. But the number of practitioners will be small enough for the doctors to make timely decisions and avoid cannibalizing one another’s practices. If the practices were bigger, NovaMed could not be as selective as it is in choosing its doctors. Being bigger would also create too much competition for patients, confuse relationships with referring providers such as optometrists, and threaten the local character of health care. If the practices were too small, overhead expenses would damage profits, and the number of specialist physicians would have to be limited. NovaMed executives arrive at the optimal patch size by carefully segmenting markets based on specific geographic and operational parameters. Senior management readjusts the scale as changes—for example, new surgical procedures—open up fresh market opportunities.

Managing patch size is important for new ventures such as NovaMed, but for big corporations trying to respond with agility in dynamic markets, it can be crucial. Microsoft is a good example. Although the company itself is a software giant, Microsoft executives try to keep the staffs of their applications businesses at or below 200 people so that managers can develop an in-depth understanding of their businesses and motivate people—particularly their star developers—effectively. Having bigger business units would make it possible to develop larger, more complex software programs but would lower motivation by distancing developers from their products and so ultimately hurt the business. Conversely, excessively small patches wouldn’t give Microsoft the programming scale needed to write substantial software applications.

The level of uncertainty in a market also affects optimal patch size. As a rule of thumb, turbulent markets favor focus and agility and hence small size; static markets favor economies of scale and hence large size. Because the edge between agility and efficiency shifts as markets evolve, patching managers must track the subtle shifts along that edge closely. Consider Dell again. The Internet and multimedia technologies have created extreme turbulence in computing, thus altering the balance away from efficiency toward agility. In response, Dell’s managers have not only created more patches in recent years but also cut the average patch size (as measured by revenues) despite the company’s overall growth. As a result, Dell has been able to fine-tune product offerings, sales approaches, and profit levers in customer service, thereby achieving extraordinary financial performance.

Laying the groundwork

Patching requires modularity, detailed and complete business- level metrics, and consistent, company-wide compensation

Patching won’t work unless the company’s infrastructure supports the process. That requires modularity, detailed and complete business-level metrics, and consistent, company-wide compensation.

Modularity is the most important of these elements. A patching company’s business units need to be focused and discrete so that they can be combined seamlessly. A complicated organizational structure—in particular, one with lots of shared services or cross-business committees—will slow patching down.

Complete and detailed business-level metrics that are comparable across businesses—on revenue, gross margins, customer preferences, product development time, and costs—are also essential for effective patching. Such metrics and the trends they reveal give corporate managers in-depth knowledge of businesses and can help make predictions about how and when to repatch. And fine-grained metrics can be particularly useful for helping managers pull apart misleading aggregate numbers. A European hospitality company, for example, was organized into one large, vertically integrated business that included brewing, pub, and hotel operations. Managers had no way to consider these very different operations as separate businesses. Only when new metrics allowed managers to understand the different business models of each operation could they repatch effectively.

Successful patching requires the movement of people, and inconsistent compensation can erect barriers to that movement

Ensuring parity in compensation across the whole company is the final infrastructure component of patching companies. After all, patching requires the movement of people, and inconsistent compensation can erect barriers to that movement.

Principles for patching

Successful patching relies on a combination of speed, strategic decision making, and follow-through. The five guidelines that follow can make the difference between a captured—or lost—market opportunity.

Do it fast

Patching decisions are best made quickly—in two or three months at most. Fast choices reduce anxiety and politicking. Drawn-out restructuring decisions are lightning rods for negative emotions. In fact, one reason reorganizations in traditional companies are politically charged is that managers take a long time to decide things because restructuring is rare. In companies that patch, however, the level of politicking is reduced, since no one expects any particular patch to last forever. The minimization of politics further contributes to the speed and efficiency of the patching process.

Develop options and then make a roughly right choice

When a patching occasion arises, managers instinctively focus on one response and analyze it thoroughly—but that process is too slow. Effective patchers develop three or four alternatives. Managers can then analyze them quickly for two reasons: it is cognitively easier to compare several alternatives than to analyze one in depth, and the crucial factors, such as the similarity of business models and profit-and-loss viability, are usually clear.

Take an organizational test-drive

Rather than painstakingly planning out the details in advance, savvy patchers come up with prototype patches before they make a final decision. This test-drive speeds up analysis and lowers the chances of major errors, just as it would in new-product development.

One common tactic—creating temporary "shadow organizations" within the existing business infrastructure—lets managers test how well various aspects of a patch will work. An enterprise software company we will call SavvySystems is a good illustration. The company, whose electronics business unit served telecommunications, semiconductor, and computer customers, was growing rapidly. Preoccupied with this high growth and with several demanding computer customers, management was neglecting opportunities in the telecommunications and semiconductor markets. The company tentatively decided to split the electronics division into three distinct business units corresponding to the company’s three customer groups. But before finalizing this choice, managers created a shadow organization to test-drive the proposed patching solution.

First, they checked the financial viability of the new units by installing profit-and-loss metrics for each of them and letting the units run against the new metrics. Each turned out to be financially sound. Next, managers tested the proposed patches from a product development point of view: they split product developers into three groups that mirrored the proposed new patches but kept those discrete groups within the electronics business unit. This shadow organization revealed that for developing products, it made sense to combine groups catering to the company’s telecommunications and computing hardware customers, which had similar enterprise software requirements, but to split computing software and semiconductors into separate businesses. SavvySystems therefore realigned the product development staff and the P&Ls, but the changes still hadn’t been made formally. In a third test, managers divided customer support into three groups that mirrored the latest patching option. Several client companies were then shuffled between business buckets because of the unexpected demands on customer support that the shadow organization revealed. SavvySystems’s managers then formally repatched the electronics business—with few postsplit headaches. They ended up with three new business units: telecommunications and computing hardware, computing software, and semiconductors.

Get the general manager right

Selecting an appropriate general manager for a business unit is essential for effective patching. The wrong manager can instantly stall patching. When HP’s managers transferred a large-format printer business from San Diego to Barcelona—to take advantage of the Barcelona unit’s strength in improving efficiency and to allow the San Diego unit to start up a new business—the move might have created a great deal of emotional upheaval. However, by also moving the Barcelona general manager, who could understand all the players, to San Diego, and placing an American with high general-management potential in the smaller Barcelona operation, managers helped bridge the cultural gap between the two divisions and gave the new general manager valuable experience.

Script the details

After the decision has been made on the new patch, the main concern is ensuring that the work itself doesn’t get neglected in the confusion of a complicated organizational change. That may involve keeping a development project moving, say, or making sure customers don’t fall through the cracks. Successful patchers follow a script after the patching is announced. Laying out a detailed plan—sometimes even specifying day-by-day activities for the first 30 to 60 days—helps to coordinate the many tasks and people involved in the patch. The specifics of the plan are actually somewhat arbitrary. Some patching companies have SWAT teams that swing into action to provide extra support; others use checklists of major activities. The key is to have the template ready.

An essential part of the script is to push managers to reach financial and other goals—for example, customer acquisition targets and project deadlines—quickly. Patching creates anxiety and even hard feelings, especially in difficult situations, such as combining previously distinct businesses, transferring people from one business unit to another, or adding businesses through acquisitions. If people focus quickly on clear goals, they spend more time looking toward the future and less time stewing about the past.

Another component of the template is recognizing and rewarding managers whose businesses are split, exited, or combined. Such moves usually cause emotional upheaval, so explicitly recognizing the managers’ contributions to the company is an important part of effective patching.

Common stumbling blocks

Even the best managers make patching errors now and then. "Missing the hill," the common mistake of violating the modularity of businesses, occurs when responsibility for tackling a particular product or market area is given to more than one unit. Consider the case of a computer company we studied. The general managers of two business units independently spotted the same opportunity to create software that measured the performance of computer systems. Both general managers made a compelling case for keeping the business within their own domains. With the agreement of the two division heads, the patching executive decided to split the option across the two units, giving product development to one business unit and marketing and profit responsibility to the other. He reasoned that the first business unit would benefit from developing the technical skills but that the business opportunity made a better fit with the second. With all managers in agreement, success seemed like a sure thing, but a couple of months later it became clear that no one felt responsible for the new product line. Formal responsibility had been given to an overworked lower-level manager who lacked the time and the status to make things happen. The new venture fell through the crack between two very big businesses.

Allowing one business unit to become much larger than related businesses is another common stumbling block to patching. We call it "Snow White and the Seven Dwarfs." Typically, the general manager of the large business becomes too powerful, which creates awkward and often dysfunctional decision-making dynamics and sometimes even undercuts the power of the patching executive. Business decisions tend to favor Snow White, even though that may not represent the best future opportunities of the collective businesses.

Mistakes are to be expected when managers balance on the edge between agility and efficiency. The key is quick correction. For example, the patching executive described earlier who missed the hill recognized the mistake and, within several months, repatched the new software business into a focused division.

Afterword: A new corporate strategy

In turbulent markets, businesses and opportunities constantly fall out of alignment. New technologies, novel products and services, and emerging markets create fresh opportunities. Converging markets produce more of them. And, of course, some markets fade. As a result, the clear-cut partitioning of businesses into neat rectangles on an organizational chart becomes out of date as opportunities come and go, collide and separate, grow and shrink. Predicting which competencies or strategies will be successful—and for how long—is impossible, so corporate-level strategists must continually remap their businesses to grab market opportunities.

The processes that promote this dynamic strategic repositioning are not just accelerated versions of traditional corporate processes, for the traditional ways of planning and allocating resources are too top-down and focused on control. In contrast, new corporate-level strategic processes such as patching center on change. They add economic value by helping managers mobilize and reconfigure corporate resources to capture market opportunities faster than the competition does.

Especially in dynamic markets, the payoff is worth it. Why? Business units in those markets are configured in small chunks to give those units agility. The more uncertain a market, the smaller the chunks. In fact, even entrepreneurial ventures are often structured into multiple businesses in this kind of market. As a result, the corporation has more business units to coordinate, and there is more need to offset disadvantages of scale by capturing synergies and more volatility around scope and boundary decisions. In other words, corporate strategy—reinvented to focus more on processes than on positioning—has even greater importance in dynamic markets than in static ones.

So the answer to the frequently asked question of whether the corporation can add value beyond that of the sum of the businesses is, "Yes." Managers who can quickly reconfigure resources into the right chunks at the right scale to address shifting market opportunities—in other words, managers who can patch well—can create multibusiness companies that outperform even the most efficient capital markets. Patching is a crucial, corporate-level strategic process that can build extraordinary value by dynamically stitching the quilt of businesses within the corporation.