Chapter 8 Organize Around Customers and Processes Joint ventures Reverse holding companies Amoeba structures Pull together; split apart Enterprise units Organize around processes Domes, not pyramids
Organization is the vehicle for a businesses' growth ambitions.
Organize Around Customers and Processes
Excerpt from Rethinking the Corporation
By Robert M. Tomasko
Ideas from the world of physical design (Chapter 7) are starting to appear, at least in part, in organization structures throughout the world. Here are several of the new directions being taken, and some lessons that can be drawn from them.
Moves in the right direction
Many benefits can arise when moving from a being a segment of a planar structure to an independent, more plastic form. This occurred when the exploration and production division of Sun Oil became the Oryx Energy Co. Named after the fast African antelope, Oryx quickly became the largest U.S. independent producer of gas and oil. When part of Sun, it emphasized finding oil to supply its parent's refineries. But as an independent it could follow its own intuitions about where its industry was moving. Noticing that increasing environmental concerns were likely to favor gas as an energy source over oil, Oryx fueled its growth by giving greater emphasis to natural gas exploration.
IBM has moved a step in this direction, using the spin-off idea as a partial basis for reorganizing the $65 billion corporation into more than a dozen semi-autonomous businesses. Several of these are software businesses, which like the gas side of Oryx may be able to use the separation from what has been a hardware-driven company to achieve more of their own potential.
Several smaller companies, who may aspire to IBM-like reputations in their marketplaces, are trying to avoid the splitting IBM's is undertaking by growing in sales, not in bulk. Cypress, a Silicon Valley semiconductor maker makes a point of keeping new product lines away from its existing organization. Instead each is handed-off to a separate start-up venture along with up to 19% of the venture's stock which is owned by its employees. The rest of the ownership remains with the parent. Cypress' chief executive, T.J. Rodgers, explains: "I would rather see our billion-dollar company of the 1990s be ten $100 million companies, all strong, growing, healthy, and aggressive as hell." For Rogers the only alternative is an aging billion dollar business that "spends more time defending its turf than growing."
Other corporations have taken a less extreme course to structural reform. The successful restaurant operator, International House of Pancakes (IHOP), has modified the traditional approach taken to franchising to accommodate the need for both growth and control. This Southern California based company adds 25-35 new locations each year by selecting the site, building the restaurant and directly running each outlet for at least six months before signing-up a franchisee to take-over the operation. IHOP also keeps one of its employees temporarily based at the restaurant to train the new manager and monitor service quality during this critical transition period. Like Cypress, IHOP keeps a significant part of the ownership for itself - it usually owns the building - so it can remove franchisees who do not maintain company standards. Unlike many companies that grow by teaming-up with independent operators, IHOP uses the franchise technique as a management tool, not a way to grow using other people's money. Its restaurant operators are more entrepreneurial than the salaried managers in their competitors' outlets, but under more control than some free-wheeling franchisees.
An IHOP competitor, Burger King, made a different structural change to try to encourage its 1000 company-employee run restaurants to operate more like its 5000 franchisees. Each company-owned unit is now charged rent and royalties (on the same basis as paid by the franchisees), managers' spending limits have been increased and two levels of management that watched over the company outlets was eliminated. This idea of simulated franchising is not a fast-food exclusive. Seafirst, a Seattle bank owned by BankAmerica, has started treating its individual branches as franchises, with each bank manager adjusting hours to fit local preferences as well as hiring all branch employees.
Joint ventures are an increasingly popular way to get around the limitations of an existing corporate structure. In addition to its strategy of structural subdivision IBM has formed partnerships or alliances with a surprisingly broad array of other companies - Apple, Intel, Mitsubishi, Motorola, Siemens, Toshiba and Wang. Many of these are also strong marketplace competitors of IBM. Ford has also learned that rivals can be allies. By teaming up on focused projects with Mazda, Nissan and Volkswagen, it is able to field a much broader product line than possible if it followed a "go it alone" approach to the global marketplace. And the Corning Glass Works has stayed focused on refining its critical capabilities in optical fiber production by letting six joint ventures carry the weight of building a sales and marketing capability for these products in Europe and the Far East.
In some situations the best basis for collaborative links is around information. Information partnerships can be a creative way for one company to share the scale of another. Information sharing was the basis for the recoupling of the small Italian textile companies formed when several large family-owned conglomerates were broken apart, as described in Chapter 3. Another Italian-based business, Benetton, has used information systems to build a global fashion design, manufacture and retail empire.
The world's largest consumer of wool, Benneton's name is on more than 6000 stores in 80 countries. It calls itself a "vertically disaggregated company," one that relies on others for many of its critical business functions. All but a handful of Benneton's stores are owned by franchisees. They are each served by one of 75 regional agents, who also own stores, and earn a commission based on their territory's sales. They advise the store operators, recruit new ones, and take and pass on their merchandise orders to headquarters in Italy.
Operating out of a 17th century villa in Treviso, a few miles north of Venice, Benneton's executives coordinate the three key elements of the business. Designs for new garments are developed twice a year by 20 freelancers, located throughout the world. Submitted to headquarters by floppy disc, they are finalized and made into production patterns there. Samples of the season's offerings are made and sent to the regional agents to show to the store owners. They then decide which will sell well in their localities and orders are placed.
These form the basis for the next few months production, 80% of which is done by subcontractors to Benneton using fabrics purchased, knitted and colored by the company. Finished clothing is shipped directly to the purchasing store.
Actual selling to those who will wear the clothes is left to the franchisee, although sales patterns are closely monitored so information about hot items and popular color combinations can be quickly factored into the production schedules. Bennetons' electronic information systems are so closely coupled that they only make garments already sold to their stores.
At Treviso, in the middle of a corn field, is one of the few tangible assets the company owns, a large and very modern warehouse. So strong is the idea of computerized information in Benneton, that the operations of this essentially robot run facility are modeled after the way a computer stores information. This five floor high structure can hold more than a quarter million boxes of clothes, all kept in essentially random order throughout the warehouse. As a new box is received from a subcontractor automatic cranes and conveyers place it wherever space is available. No effort is made, as in most warehouses, to place the box with others of its type or final destination. This is similar to the way electronic information is stored on a computer disc, with different parts of the same paragraph scattered all over the disc's memory locations. But in both cases, retrieval is no problem; the computer quickly remembers the exact location of each item of clothing as easily as it finds each word of text and instantly strings them together in the order specified in the paragraph. The lesson here is that ways of organizing that seem like "no-organization" from the hierarchal mind's point of view, can make a lot of sense when coupled with computer-fast information processing about the exact location of every item. The pay-off for Benneton is the most efficient use of expensive warehouse space.
Benneton's information partnership-based structure is held together by more than electronic data. The company is well known for its bold advertising campaigns that deals with global concerns such as the AIDS epidemic and racial harmony. They attract customer attention in ways much stronger than the campaigns of its more-timid competitors. They also serve to project an image, a sense of a corporate aesthetic, that connects the values of the headquarters executives with those producing and selling its products and those using them. It says we stand for something important beyond only garment selling. This kind of broad-bandwith communication can help overcome, through empathy, what might otherwise be difficult walls between parties not part of the same corporate hierarchy.
Wal-Mart, the world's largest retailer, has taken a more focused, but just as successful, approach to creating information partnerships with its various constituencies. It links itself to concerns of many of its U.S. customers with a procurement policy emphasizing "Buy American" which is highlighted in advertising and store signage. It has also pushed the use of technology to its fullest by using "electronic data interchange" to release sales figures regularly to the makers of the products sold, who are then expected to automatically replenish Wal-Mart's inventory. In addition to reducing inventory size, it eliminates the need for middlemen such as merchandise brokers and independent sales representatives who formerly were the human links in this information flow.
Visa, the world's largest credit card, is an example of the new corporate forms emerging from the electronic information network revolution. Called by its founder, Dee Hock, a reverse holding company, it is owned by its 22,000 member banks and other financial institutions. Built on innovation, it created the first system for banks to electronically transfer money to each other, it serves as a good model of an intermediary organization that can allow companies to extend their boundaries without adding a cumbersome hierarchy.
Contrast Visa's success as a focused federation with the "go-it-alone" attempt of United airlines to build one organization that included the key services needed by a business traveler. Its Allegis Corporation attempted to put under one hierarchical roof the operations of the airline, Hertz car rentals, Westin's hotels, and an automated reservation system linking all three. This conglomeration failed, partially because of pressure from Wall Street which was skeptical that managerial synergies would emerge. The financial markets doubted the an executive from a successful car rental agency would be able to contribute much operational wisdom that would help the airline grow its business.
The Visa network only took off when it was spun free from one of the network's nodes, its parent Bank of America. Groupings that seem logical because of information-commonness, can suffer when placed in a relationship based on hierarchical ownership. Flow happens best horizontally, not up and over. Two Harvard Business School professors, Ben Konsynski and Warren McFarlan, say this well when they observe: "Managers from companies in reciprocal industries should now be plotting common approaches to customers through relational databases, not plotting how to take each other over."
Audrey Freeman, the Conference Board's labor economist and an astute observer of changes in corporate form, likens many emerging corporate structures to the basic biological organism, the amoeba. This unicellular microscopic animal is a prototype of the plastic structure, it is almost constantly changing its shape. It is impossible for an amoeba to move without reorganizing itself. It ingests food not through a designated mouth, but by extending an armlike portion of its body toward whatever has stimulated its interest. The food is then surrounded, covered with some cellular material, engulfed into the organism and circulated around the portions of it needing nourishment.
The amoeba grows by mitosis, the division of one amply nourished
cell into two. This is not unlike Illinois Tool Works' growth
strategy. The company actively extends feelers to find market
niches where it can be the dominant player. Then it stays so
close (engulfs?) to the customer that its engineers or marketers
discover a new need and invent a solution. Then comes the corporate
mitosis; the new product, and associated employees, are spun-off
as a new entity. Its current divisions include one that is the
world's largest maker of plastic buckles and another that invented
the plastic loops that hold six-packs of beer and soda together.
Both entities resulted from this hiving-off process. Health care
products giant Johnson & Johnson follows the same logic.
When the endoscopic (small-incision surgery) portion of one it
its 160 odd companies showed significantly greater growth potential
than its surgical products parent, it was quickly granted independence.
Pull together or split apart?
No single plan is appropriate for every company wanting to move its structure in this direction. Downsizing alone certainly will not work; frequently all it creates is an organization looking like a warm wedge of Swiss cheese. What a company needs to do to create a horizontally-oriented structure depends on where it is starting from.
For some companies, the starting point will involve a pulling together of inter-related activities that may have strayed over-time. Siemens "discovered" that the employees who designed railway switches had no easy horizontal organizational linkage with those making train's electrical systems. After pulling these, and other railroad-related activities together in one division with a clear group of common customers many ways to share and leverage capabilities and market intelligence emerged. In less than four years, sales almost doubled in what is frequently written off as a low growth, mature industry.
Pulling together activities to achieve sharper customer focus is equally important in service industries. Swissair, Swiss Airlines predecessor, used the routes in its schedule - the traveling of which is how customers experience the airline - as the basis for creating its organizational clusters. As part of its revitalization efforts, Sears, Roebuck & Co., also realized that most customers tend to come to one of their stores to purchase from a specific department, not to scan the array of all Sears' merchandise. But its organization structure was built around the store as its basic unit - a reflection of Sears once successful strategy of emphasizing its own brands and unique approach to customer service. Sears changed this by identifying seven of what it calls "power formats" (such as "Brand Central," its store-with-in-the-store for consumer appliances and electronics). Each is a separate organization, pulled together from what were quasi independent store operations and procurement groups. Each is responsible for its own buying, pricing, selling and profitability. No longer will the apparel department hold a sale in hope of maximizing its store's profits by hoping busy shoppers will have their car serviced by Sears on the same visit. Nor will Sears be bound to keep each format in each store. Some formats, such as its Home Life furniture operation, have already been spun-off into focused stores of their own.
Other businesses will give more attention to splitting-apart activities or functions. Hewlett-Packard's profits rebounded and number of new product introductions soared when a single computer division became two: one selling minicomputers and workstations directly to large corporations; the other, printers and personal computers through dealers to the mass market. Each requires different sales strategies; each now has its own sales force.
This idea can be applied to an entire corporation, or to improving just one of its functions. Niagara Mohawk Power, a large upstate New York utility abandoned its industry's historical pattern of tight vertical integration, and broke its operations into four entities, each responsible for its own growth and profit. One is focused on providing gas service to it customers, another to delivering electricity. A third unit generates electricity from traditional sources, and the other produces it from nuclear reactors. At Illinois Tool Works, threatening competition from the Far East in the early 1980s forced them to reshape the approach they took to manufacturing. Compared to many Japanese factories their facility in Elgin, Illinois was best characterized not "so much an assembly line as a loose confederation of city-states that periodically engaged in mutual commerce." Several advanced manufacturing techniques were applied to improve the situation, including the drastic change of isolating the equipment and people who produced nine of the plant's high volume components, and setting them up and small, stand alone operations in Wisconsin farmland. None of these plants has more than 25 employees; at each a much stronger sense of identification of worker with product occurs than was ever possible in the centralized, but balkanized Elgin facility.
Both Merck and USAA used the composite team as the basic building material to form a structure based on splitting-apart larger, more monolithic operations. USAA, a Texas-based insurance and financial services company several decades ago eliminated the large functional groups that handled insurance underwriting and servicing. Instead five smaller organization were created, each with all the specialists needed to handle all of a customer's needs. Each was assigned a fifth of USAA's customers - each grouping a similar geographic and actuarial cross section of the entire customer pool. Then the five groups were put into competition with each other to provide motivation and identify the company's next generation of management superstars. Employees who only knew and cared about their little segment of the company's operations quickly became insurance generalists, able to see how they could help all the pieces best fit together.
This splitting apart technique works just as well outside the U.S. In Italy, headquarters for multinational pharmaceutical giant Merck's research into viral diseases, lab operations have been subdivided into teams. Each focuses on specific diseases or promising chemical compounds. But they do not just discover something new and pass it on to somewhere else in Merck for testing, regulatory approvals, and selling. Each team stays with the product through the marketing phase, giving the researchers a stronger stake in their work's commercial success.
For a handful of corporations, the best starting point may be to do a little of both, almost simultaneously take apart and put together. ABB Asea Brown Boveri is a master at this. Created through the merger of the largest electrical equipment makers in Sweden and Switzerland, ABB has organized in a way that puts substance behind the hackneyed phrase: "Think global, act local." Its quarter of a million employees are located in over one hundred countries. Its founding chief executive, Percy Barnevik, calls it a multi domestic organization - a federation of national companies. Other ABB executives maintain it is not a global business, only a collection of local businesses with intense global coordination. Its operations are divided into 1200 individual companies, each averaging about 200 employees, with most located in only one nation.
They are "intensely coordinated" through two structures. One is a traditional geographic-based line management structure. The heads of each of the 1200 companies report to the head of their "country company." This person serves as ABB's "national CEO" and ensures that the operations there are following all local laws and making the best use of the locally available talent.
The company heads also report to one of 50 "business area" leaders. These business areas include specialities such as power transformers, instrumentation, electric metering, etc. The business area leader, with no more than five functional specialists to provide assistance, is responsible for maximizing the results of that particular area globally. As many as possible of the areas are represented by individual companies in each country.
Overseeing these parallel structures is a twelve-person executive committee and the chief executive. Back-up support for the global management of this over $25 billion in sales corporation is provided by a 100 person headquarters staff unit in Zurich.
The objective of this very lean, matrixed structure to allow the many local ABB companies to operate in as much of a stand alone manner as possible, while at the same time spreading new ideas, products and factory capacity across as much of the world as practical. ABB's structure is the largest attempt of its kind to mix big with small. To date it has been an effective way to assimilate rapid growth through acquisition and operational rationalization. It is well worth watching to see if it can also be a conduit for growth from within.
The Enterprise Unit
Where the SBU's structure included all the activities necessary for it to function as a stand-alone entity, the enterprise unit has only those most vital to its competitiveness - primarily those representing critical and cutting-edge capabilities. Other needed capabilities are purchased in the marketplace or shared with other enterprise units. This is the direction ABB, Benneton and Corning Glass are taking.
Where the SBU used the "worker and boss" pair as its basic organizational building block, the enterprise unit relies more on reinforced jobs and composite teams to get things done, in the manner of a Merck and USAA. While SBU's, and the hierarchical superstructure above them, were full of part-time managers with fragmented responsibilities, enterprise units will have few managers. But all will be load bearing. And while the SBU's implicit objective was to grow the business, and the size of the organization, as much as possible, the enterprise unit is more mindful of the advantages of scale. When growth happens, it is dealt with through the kind of corporate mitosis practiced by Cypress, Illinois Tool Works and Johnson & Johnson.
SBU's are usually a reflection of the intersection of a corporations' products with the markets it serves. Organizing around these market segments provides better focus than can be obtained when the organization is based on specific products or the functions needed to produce them. The product-market matrix is a good way to plan when an industry is relatively mature and stably segmented, but it tends to produce organizational fossils when the rules of competition keep changing. Instead the enterprise unit's outputs are best defined as those meeting common customer needs, not just common customers. As these needs evolve, so do the enterprise units. Every member of an enterprise unit has what Paul Allaire, Xerox's chief executive, likes to call a direct line of sight to each customer. Small scale and sharp focus make this possible.
Organize around processes,
What from the top-down might look like a great deal of freedom and flexibility, is - from the bottom-up - an under coordinated, under leveraged patchwork of middle management fiefdoms. The well-meaning basis for most of these enclaves is the seldom-questioned assumption that the function is the best unit to structure around: "Let's put the product designers in one department here, manufacturing over there, sales out on its own, and personnel and legal far enough away to keep out of everyone's hair. And then we'll let each use their best functional judgment about how to best serve the corporation" While it is popular to talk about the necessity for plastic and planar structures, this basically skeletal model is still dominant throughout most corporations. This, of course, leads to the wall building and corporate sclerosis that are so trendy to decry.
The streamlined flowcharts emerging from process reengineering are attempts to deal with this excessive compartmentalization. But they will remain weak half-steps unless their logic is transferred to the organization chart. There is no point in resizing work along the horizontal lines of process flow if, in the end, it is only going to be awkwardly stuffed into a vertical, functional, structure. Really managing a company by means of its key processes requires more than an overlay on functional fiefdoms. To do this we have to think about creating organization units for each key process, instead of departments for each functional speciality. If the first level of the organization chart is inhabited by people in reinforced jobs or composite teams, the second tier is home to load-bearing managers of the enterprise's most important business processes.
Lawrence Bossidy, Allied-Signal's former chief executive, is a firm believer in this style of organization. He feels every enterprise unit has five or six basis processes, and that they provide the best basis to organize around. They are the engines that deliver a company's capabilities to its customers, and they will be different for each enterprise Top-down capability analysis is the key to identifying the processes worthy of driving the structure. For some firms the list will include a seamless stream of activities that obtains and fulfills orders from customers. Others will want to include the cycle from product creation to market introduction, and some companies - such as Wal-Mart - may also want to stress a well-greased logistics machine, or all the activities necessary to find, motivate and retain loyal employees in time of increasing labor shortages.
A house divided against itself cannot stand - and neither can a company. Business processes cannot, despite the optimistic hopes of matrix organizers, cohabit a structure that also is subdivided along functional lines. Too much day-to-day effort will be consumed within the company sorting out these self-imposed conflicts to do a good job of continually improving the business processes and keeping them pointed in the same directions as their customers' needs are moving. Planar and skeletal structures mix as well as oil and water; they can be only temporarily homogenized, but only if a lot of energy is spent shaking the container that holds them. If processes are chosen as the basis for the second tier of the corporate structure, then most inhabitants of the existing functional groups will need new homes, and their old departments must disappear from the organization chart. Functional walls can cease to be load bearing, because the managers who head each of the handful of key processes are. New structures depend on new building materials.
Replace the pyramid with a
What overall structure is best at containing all this high speed activity? To whom will the leaders of the various enterprise units report? How will the corporation avoid loosing the beneficial side of functional expertise? What kind of oversight will maximize capability-sharing and minimize the growth of bottom-up duplication of effort. From where will come the company's common aesthetic, its values? These are important concerns. If they are not addressed, centrifugal forces will dominate as each hard charging enterprise unit moves outward following the call of its customer's most immediate needs.
For a clue about the structural form most able to gets its arms around this energy and diversity, consider the experience of Microsoft. Its founder Bill Gates found it necessary, even with arguably the world's most effective E-mail system, to ensure Microsoft had at least an annual event where all 7,000 employees could come under one roof and, at least potentially, see each other face-to-face. What structure did Gates find to accommodate this? A dome. In his case, Seattle's large sports stadium, the Kingdome, of course.
The most efficient encloser of space
Domes are half spheres, or some close variation of this shape. They are very efficient configurations because they are very thin compared to the distances they can span. If an arch-shaped structure, or pyramid, was to cover the same space, it would have to be ten times as thick to support itself. Perhaps it would even need a supplemental external skeleton, like the Gothic cathedral builders' flying buttress, to do the job.
What is it about the structure of a dome that makes it so strong? Look up at a domed roof from inside the building. You will most likely see a series of raised ribs radiating along the dome's inside downward from its center. These vertical sections are like the meridians on a globe. You will also likely see a series of parallel hoop-like ribs starting at the bottom of the dome and continuing upward. They, in circles of decreasing radii, are the latitudinal lines of the dome.
Together, these meridians and parallels (which in some buildings are covered for decorative purposes) form a strong inner grid, a semi-circular cobweb. The meridians distribute the dome's weight downward, and the horizontal hoops keep this pressure from putting a bulge the dome's top, or spreading apart its bottom. The net result, a very stable structure, one of the stiffest and strongest ever devised.
As with all structures, the materials used to construct the dome are very important. Many Roman and Renaissance eventually developed cracks at their base. Their stone and concrete could not resist the tensions they were put under as they resisted the dome's natural tendency to expand at its base. This problem was finally remedied in the nineteenth century when steel rings were used to circle the bottom of the dome, just as steel hoops hold together a wooden barrel.
These tensions and pressures acting on architectural domes are not unlike those affecting many corporate structures.
If strong lateral linkages are not present at several levels of a corporate structure, the forces they are needed to transmit are likely to exert pressure on both the top and bottom of the company. At the top a bulge often occurs when these horizontal coordinating mechanisms are weak or nonexistent. Senior executives are overworked and find their attention is frequently distracted when they attempt to solve problems or mesh operations that should have been dealt with at lower levels. To cope with all this they may hire more headquarters staff, or put expensive electronic information systems in place, both of which tend to expand the bulge.
At the bottom, the lack of discipline and focus that needs to be provided by the mid-level horizontal linkages can cause an organization to excessively extend its boundaries. Companies without good mechanisms to maintain their focus often become victims of unrelated diversification and tendencies to provide all support services internally. The brittleness caused by narrow jobs and fragmented management responsibilities also exacerbates these problems.
The economy and strength of dome-like structures makes them a better metaphor for the shape of the new corporation than provided by the traditional pyramid. Think about the sense of freedom, and even wonder, that can be felt when entering the Pantheon, or another world renowned domed structure. Contrast this with the sense of confinement - or oppression - experienced within a pyramid.
Domes are not completely flat structures, though. Their purpose
is to contain and focus, not smother. Putting a dome-like structure
over a horizontally oriented group of enterprise units, work
processes and expanded jobs and teams may also serve as a way
to deal with some of the problems that linger after many restructurings.
© Robert M. Tomasko 2002