Customizing Customization

The history of U.S. business during the past 100 years has been a story of mass production and mass distribution of standardized goods. Scholars and practitioners who examined the economic landscape have generally been drawn to large corporations that built their fortunes by transforming fragmented and heterogeneous markets into unified industries.1 At the heart of this transformation were strategies based on standardization: standardization of taste that allowed for standardized design, standardization of design that allowed for mechanized mass production, and a resulting standardization of products that allowed for mass distribution.

Recently, a growing number of economists and management scholars have declared that this era is over.2 Numerous books and articles have posited that we are witnessing the dawn of a new age of customization, an age in which new technologies, increased competition, and more assertive customers are leading firms toward customization of their products and services.3 Not surprisingly, firms that have adopted customization strategies have attracted considerable attention as models of what is expected to become commonplace in the near future.4

We begin by describing these two logics. We then argue that this conceptual polarization, which took firm root in the theory and practice of management, led management thinkers to ignore strategies that combine these logics. Put simply, this view itself represents an inappropriate standardization of management theory — or, more exactly, continuation of the standardization mentality that has long pervaded such theory. Since the days of Frederick Taylor, the notion that managers should single-mindedly pursue the “one best way” has led management writers to seize on one solution or another (or, more often, one solution and then another) as the best practice. Indeed, the enthusiasm for customization today was paralleled by an even greater enthusiasm for standardization many years ago.

What has been ignored in all this is that customization and standardization do not define alternative models of strategic action but, rather, poles of a continuum of real-world strategies. By promoting customization as the answer to what ails many organizations, we may be replacing one extreme with another. Managers need to locate their strategies along the continuum, and the role of management writers is to provide the conceptual tools to make this easier.

It is not an accident of history that customization is now promoted with the same enthusiasm with which standardization was promoted almost a century ago. Today’s customization movement is a reaction to significant economic and technological forces, much as the standardization movement was in its time. But these developments must be viewed with a clear perspective. Therefore, we begin by reviewing how the standardization and customization distinction emerged in the first place.

Although most readers may be familiar with the reasons that standardization became dominant at the turn of the century, they may not sufficiently appreciate to what extent this was an act of cognition as well as a reflection of economic and technological forces. Product standardization, mass production, and mass distribution were seen as a conceptual whole, a configuration. Industries that were transformed by this process were contrasted with old-fashioned industries that continued to customize their products. As a result, two logics emerged, a logic of aggregation in contrast to a logic of individualization, that were mistakenly treated as conceptually distinct and mutually exclusive.

We argue that this conceptual polarization, which took firm root in the theory as well as the practice of management, led management thinkers to ignore strategies that combine these logics. We thus describe a continuum of five strategies, from pure standardization to pure customization, and then apply them to various industries. We then argue that the most striking recent trend has been not toward pure customization, but toward some middle ground that we call customized standardization.

The Logic of Aggregation

Economic theory takes a bird’s-eye view of markets. It strips them of their complexity and variability, aggregating firms and individuals into two groups: buyers and sellers.5 In contrast, management theory, especially in strategy, starts with the relationships of specific firms to their environments. Sellers are therefore disaggregated. Customers, however, continue to be viewed collectively as a group (or a set of segmented groups) that shares common characteristics. This has led managers and researchers alike to emphasize the advantage of economies of scale in every part of the value chain, from development to production to distribution, which in turn has promoted a focus on industries in which customers’ shared characteristics are easily established.6 Thus, in his 1980 book, Porter draws on 196 industries to illustrate his ideas.7 By our count, 176 of these have been dominated by the logic of aggregation, while only 20 can be considered to have tilted toward that of disaggregation (despite, as we shall argue, the rather more common occurrence of the latter than has generally been thought). Not surprisingly, perhaps, the industries that Porter uses most often to illustrate his framework are in bulk production of goods such as chemicals, branded goods such as soft drinks, and other products such as metal containers that epitomize standardization. The upshot of this inclination has been to associate aggregation with the search for efficiency and so to help root the concept at the core of managerial thinking.

The following maxims capture the basic imperative of the logic of aggregation: (1) reduce the impact of customers’ variability on internal operations, (2) do so by identifying general product and customer categories, and then (3) simplify and streamline interactions with the customer. Years ago, these maxims were reflected in the resolution of long-standing difficulties in product design, production, and distribution. Over time, the experience of these functional areas coalesced into a well-defined set of strategies that promoted the advantages of scale economies while suppressing the pull of market heterogeneity. After a long string of notable successes in such industries as shoes, automobiles, and fast food, these strategies came to seem both self-evident and compellingly logical. Aggregation strategies became the norm against which effective business behavior was judged.

Yet it is often forgotten that all this was accomplished despite considerable initial resistance — from workers, consumers, retailers, and even managers. To overcome this resistance, in 1916, the proponents of the new order launched what came to be called the “standardization movement.”8 They contrasted the virtues of aggregation with the vices of individualization, thereby creating the polarization that has shaped U.S. management thinking ever since.

The polarization between aggregation and individualization centered on the role of standardization as a prerequisite for mass production and mass distribution. The advocates of aggregation argued that the full potential of mass production could be realized only if design, sales, and delivery of the product were standardized as well.

During World War I, these advocates prevailed on the U.S. government to enforce the “simplification” of U.S. industry. After the war, legal sanctions were relaxed, but, by then, the movement had a momentum of its own. By 1929, a survey of eighty-four product classes showed a reduction in variety at times amounting to 98 percent of its 1921 level.9 For example, the number of bed blanket sizes dropped from seventy-eight to twelve; hospital beds that had come in thirty-three different sizes were all standardized to a single size by 192910; and the colors of men’s hats were reduced from 100 to nine.11 The vigor of the standardization movement was so pronounced that two French travelers of the period concluded that the U.S. businessman “has standardized the individual in order to be able to standardize manufacture,” a claim they would hardly withdraw today after a visit to McDonald’s.12 As one manager explained, “Over a period of years of experience with builders and architects, as well as homeowners, we have found that the five-foot tub is on the average an adequate size bathtub for the average size person.”13

In retrospect, it may be difficult to grasp the magnitude and speed of this transformation, but two examples may help. In 1900, according to a U.S. government survey, there were 23,560 custom shoe-making establishments. Some twenty years later, barely a few hundred remained; most had been replaced by large, integrated operations producing the now familiar standard-sized shoes.14 Or, from the opposite side, before people were standardized into those five-foot bathtubs, one leading manufacturer’s output was one customized tub per day!

The practical lessons learned during this period were subsequently incorporated into the mind-set of the newly emerging management “profession.”15 When the study of business became a scholarly pursuit at the turn of the century, the mass production and mass distribution firm was singled out as the most rational approach to gaining competitive advantage. Management thinkers counseled against the proliferation of products, and even such an influential writer as Lyndall Urwick warned that the temptation to respond to customers’ demands could be ruinous: “To allow the individual idiosyncrasies of a wide range of customers to drive administration away from the principles on which it can manufacture most economically is suicidal — the kind of good intention with which the road to hell or bankruptcy is proverbially paved.”16

In the mid-1950s, there began a gradual move away from the extreme forms of aggregation, under the label of “market segmentation.”17 Encouraged by automation in manufacturing as well as the shift from rail to road in transportation and changes in the mass media, companies began to target specific groups of consumers. Nevertheless, the logic remained the same, for this was not a movement toward serious customization so much as toward the aggregation of the submarket or the market class.18 Indeed, the mass market in many industries had reached such a large scale that segmentation posed few real dangers to efficiency in production if not strictly in distribution.

The Logic of Individualization

Although the logic of aggregation became dominant in many industries, there remained areas of economic activity where it failed to take over. Obvious examples are certain traditional crafts, such as personal tailoring, fine jewelry making, fine restaurant cooking, and grinding prescription eyeglass lenses. More significant, perhaps, is the capital goods sector, where products continue to be designed to customer specifications and manufactured in job-shop facilities. In industries such as pulp and paper machinery, steam turbines, commercial aircraft, flight simulators, and construction, the individual customer can be deeply involved in every aspect of the transaction and expects key product decisions to be negotiated jointly. Resistance to aggregation has also been common in transportation, leisure, and automotive products, where firms often respond to the needs of individual customers despite the ceaseless drive toward greater economies of scale.

More recently, of course, there has been a move to greater customization in a wide variety of industries, including services. In some cases, the change has been highly visible. The standard telephone service of the past has given way to a varied menu of features from which customers may select their own preferred combination. The same transformation has taken place in banking, insurance, and clinical laboratory testing. In many other instances, specific companies are pursuing customization, while the rest of the industry continues to follow an aggregation logic. For example, Ultra Pac, a company that makes plastic food containers, can customize 400 to 500 different kinds of packaging and ship them within three days of order. Its competitors can ship an item within three days as well, but only if the order conforms to standard designs.19 Textile/Clothing Technology Corporation, an industry research group, has launched a more revolutionary initiative. It is developing three-dimensional scanning systems to produce custom-made clothes. It shapes a figure on a computer using the information on individual measurement and then relays the data to agile manufacturing facilities where the finished product is shipped within days.20

Of course, some of this can go too far. In the mid-eighties, Toyota, which had been in the forefront of the manufacturing revolution, launched an all-out effort to offer buyers individually customized automobiles. Initially, Toyota seemed well on its way to accomplishing its goals. However, by the mid-nineties, production costs began to soar, forcing Toyota to abandon the effort. Other Japanese automobile companies that followed Toyota’s example ran into similar problems. Nissan, for example, which at one time boasted that customization meant it could produce “any volume, anywhere, anytime, of anything for anybody,” retreated from customization when it became clear that buyers did not want eighty-seven different varieties of steering wheels.21

In industries where the logic of individualization is pervasive, different forms of marketing, production, and product development dominate. In marketing, firms seek to develop a direct relationship with the individual customer. In production, products can be “made to order” or “tailor made.” And since products can be designed for particular customers, research and design can lose much of their isolation from the marketplace. In effect, the orientation is toward the management of each transaction. But, as we shall see, there are industries where product, process, and transaction vary markedly in their degree of customization.

Between Aggregation and Individualization: A Continuum of Strategies

Although pure aggregation and pure individualization are perceived as opposing logics, this influence has not led to the emergence of two distinct groups of strategies. Instead, we find a continuum of strategies, depending on which functions lean to standardization and which to customization. In the manufacturing firm, to take a common example, production managers often see aggregation as the best way to increase efficiency, whereas sales managers often consider individualization as the surest approach to increase sales. It is perhaps not a coincidence that the term customer looms large in sales managers’ vocabularies, while production managers prefer to speak of outputs and schedules. The differences in vocabulary reflect clashing viewpoints that remain unresolved even after working compromises have been achieved.

But the best solution is not necessarily a compromise. In just the operating processes, some firms tilt one way or the other because of the needs of the customers they choose to serve, while others favor intermediate positions. The latter reflect an organization’s ability to customize partway back in its value chain, while retaining standardization for the rest. Since the cost of customization tends to increase in proportion to the number of product changes, it makes sense to customize the downstream functions first. Firms may offer customers special delivery services or individualized financing, while refusing to allow changes in production. Or, beyond this, they may be prepared to assemble on demand, according to customers’ requests, while refusing to modify a product’s core design and the standardized fabrication of its parts.

Thus value chain customization begins with the downstream activities, closest to the marketplace, and may then spread upstream. Standardization, in contrast, begins upstream, with fundamental design, and then progressively embraces fabrication, assembly, and distribution. These two approaches give rise to the continuum of strategies based on standardization and customization that we introduce here.

We develop this continuum for a manufacturing firm with four stages in its value chain: design, fabrication, assembly, and distribution. These refer, respectively, to the extent to which the firm conceives the product initially with regard to a single customer’s needs, constructs and then assembles the product with regard to those needs, and distributes it individually to the single customer (as opposed to selling it generically, as in “over the counter”). Stepping customization back one notch at a time along this chain gives rise to five different strategies (see Figure 1):

• Pure Standardization. Ford Motor Company’s strategy during the era of the Model T was the quintessential example of pure standardization — any color so long as it was black. This strategy is based on a “dominant design” targeted to the broadest possible group of buyers, produced on as large a scale as possible, and then distributed commonly to all.22 Under such a strategy of pure standardization, there are no distinctions between different customers. The buyer has to adapt or else switch to another product. He or she has no direct influence over design, production, or even distribution decisions. The entire organization is geared to pushing the product from one stage to the next, beginning with design and ending in the marketplace.
• Segmented Standardization. The proliferation of cereal brands, the variety of automobiles that followed the Model T, and even bumper stickers are examples of segmented standardization. Firms respond to the needs of different clusters of buyers, but each cluster remains aggregated. Thus the products offered are standardized within a narrow range of features. A basic design is modified and multiplied to cover various product dimensions but not at the request of individual buyers. Individual choice is thus anticipated but not directly catered to. A segmented standardization strategy therefore increases the choices available to customers without increasing their direct influence over design or production decisions. At most, there may be a somewhat greater tendency to customize the distribution process, for example, in the delivery schedules of major appliances. When pushed to the limit, the segmented standardization strategy results in hyperfine distribution — as in the market for designer lamps, which offers almost limitless variety, but not at the customer’s request.

• Customized Standardization. Automobile companies that offer the buyer the option of selecting his or her own set of components engage in customized standardization, as do hamburger chains that allow customers to specify their preferences for mustard, ketchup, mayonnaise, tomatoes, and so on. In other words, products are made to order from standardized components. The assembly is thus customized, while the fabrication is not, hence our label of customized standardization, although we might also call this standardized customization, “modularization,” or “configuration.” Basic design is not customized, and the components are all mass produced for the aggregate market. Each customer thus gets his or her own configuration but constrained by the range of available components. This is sometimes constructed around a central standard core, such as a hamburger or an automobile body.
• Tailored Customization. A tailored suit, a rug woven to order, or a birthday cake with your name on it are examples of tailored customization. The company presents a product prototype to a potential buyer and then adapts or tailors it to the individual’s wishes or needs. Here customization works backward to the fabrication stage but not to the design stage. Thus the traditional men’s tailor will show the client standard fabrics and cuts that he can adapt to the client — for example, wider lapels than normal or adjustments to accommodate an unusual physique. The client will later come back for a fitting and more tailoring. A good deal of traditional business is conducted this way; for example, in home construction, the builder will modify a standard design for particular customer wishes.
• Pure Customization. Individualization reaches its logical conclusion when the customer’s wishes penetrate deeply into the design process itself, where the product is truly made to order. Artisans who do this are well known, for example, a jeweler or a residential architect who designs to customer specifications. Of greater impact, perhaps, is the pure customization of major products and services, including large-scale production machinery, industrial instrumentation, and much construction work. So-called “megaprojects,” such as NASA’s Apollo project or the Olympic Games, represent major instances of pure customization. Here, all stages — design, fabrication, assembly, and distribution — are largely customized. The traditional polarization between buyers and sellers is transformed into a genuine partnership in which both sides are deeply involved in each other’s decision making.

Industries by Customization

We have described five strategies of customization and standardization along the operating processes of manufacturing from design to delivery. Here we wish to extend this to products, which of course also vary in their degree of customization, and to transactions, by which buyers come to agreements with sellers. We also want to extend this beyond conventional manufacturing to other kinds of operating processes, such as services and mining, for which there is no design, fabrication, or assembly. If these processes vary along the customization scale from fully standardized mass production to fully customized craft production, then so too can products (and services) vary from being commodities to being unique, and transactions from being generic to being fully personalized.

It may seem evident that wherever a particular process is located on our continuum of standardization and customization, so too must be the associated products and transactions. For example, a fully customized process should produce a fully customized product via a fully customized transaction. That, however, proves true only some of the time — and, interestingly enough, largely in the industries popularly cited in the literature. But there are other important industries that are more varied and more nuanced, so they reveal more of the real meaning of customization. These variations also carry an important message about the very notion of classifying industries.

The logic of aggregation encouraged classifications of industries that were rather stable and standardized, based largely on groupings of products and services around a set of common characteristics (e.g., plate glass, electric lamps, and cement). These groupings were manifest particularly in the Standard Industrial Classification (SIC codes), which has given economists, management people, and government officials a handy frame of reference to collect and organize data. The frame may have been too handy, however, because it has proved rather static. While economists were reifying the categories, managers were splitting, combining, and destroying them for their own competitive advantages. For some purposes at least, classifying industries along the standardization and customization continuum may, therefore, be more useful. Here it helps to illustrate the various ways in which companies adopt these five strategies in practice (see Table 1).

• Mass Industries. Mass industries, which make products such as disposable diapers and gasoline, apply the logic of aggregation fairly consistently. Production takes place in highly mechanized, inflexible facilities; the products are standardized, and, indeed, often regarded as commodities; and mass advertising is combined with mass distribution to target customers through retail or wholesale organizations, so that transactions tend to be generic. Pure standardization thus tends to be the most common, consistent strategy, although segmented standardization may become more common as the industry matures.
• Thin Industries. At the opposite extreme are what we can call “thin” industries, where customization is pervasive, as in the production of supercomputers and steam turbines.23 Products in these industries are unique and are produced in craft settings. Transactions are intermittent and highly individualized; they can be very large, complex, and protracted, necessitating considerable cooperative buyer and seller efforts. Buyers are closely involved in the design of the product and generally expect a high commitment to after-sales service.
• Catalog Industries. In some industries, the proliferation of designs favors the strategies of segmented standardization. Accordingly, firms tend to organize their products and distribution on the basis of catalogs — as in the record, book, toy, and pharmaceuticals industries. The buyer has a wide array of choices, although the transaction tends to be generic — confined to making a selection. Sometimes the short production runs can render the production process somewhat craftlike in nature, but often this is standardized mass production.24 The products themselves, while certainly not unique, cannot be considered commodities either. Product strategies, therefore, tend to be segmented standardization.
• Menu Industries. In what we call menu industries, which produce products such as printed circuit boards and financial services, buyers have a menu of choices from which to select features of the final product. Thus customized standardization tends to be the preferred strategy. Transactions involve negotiations and reciprocal relationships between buyers and sellers. Once the configuration has been decided, the production function assembles prefabricated components into finished products. Thus, while transactions in menu industries achieve a degree of customization, fabrication, if not assembly, takes place in dedicated, relatively inflexible facilities.
• Tailoring Industries. Individualization is an important factor in tailoring industries, such as residential housing and mainframe computers, where a relatively standardized core design is adapted to individual customer needs. Sellers personalize transactions to allow for considerable customer input, mostly about peripheral design changes and perhaps price, delivery conditions, and after-sale services. A craft assembly system modifies the standard core design into a dedicated product. As its name indicates, the dominant strategy in these industries is tailored customization, although the transaction itself is customized. Shifts toward pure customization can, however, take place during periods of technological discontinuity or when sellers deal with exceptionally powerful buyers.

• Routing Industries. Routing industries, such as data transmission and delivery services, offer an intriguing mixture of standardization and customization. They accept their customers’ orders in purely generic ways but then route them completely individually, in processes that are fixed and inflexible! Thus a customer of the post office writes an address on a letter and simply drops it in a mailbox. The transaction is totally impersonal. But this standardized interface produces a rather customized service; no two letters take the same route on the same day (or else they would logically be combined in a single envelope). The post office delivers 100 million letters on a particular day that have followed 100 million different routes. But while the service is customized, the process can be described as customized standardization, because the post office configures standardized processes for each order. Put differently, the letters are always handled in bulk even though each gets batched and rebatched individually. In some routing industries, however, tailored customization can also be widespread. Powerful customers of various courier businesses, for example, can demand special services in addition to those normally offered.
• Agent Industries. Aggregation and individualization also combine in professional services such as health care and auditing. We call these agent industries because sellers act as agents on behalf of buyers — custodians of their financial, technical, or medical interests. The transaction tends to be rather generic or standardized, governed by standard contracts, and stipulated by professional or technical codes of conduct. (We do not generally bargain over price with our doctors.) The seller is ordinarily far more knowledgeable than the buyer, which limits the latter’s input into the operation’s processes. Yet these professional activities tend to be craftlike in nature, tailoring highly developed (meaning standardized) sets of professional skills to specific customers’ requirements. Processes as well as the services themselves are therefore best described as tailored customization. In health care, for example, each intervention (such as an appendix operation or a drug prescription for a nervous disorder) is based on a standardized procedure adapted to a particular patient’s condition. Of course, complex interventions (such as the treatment of a kidney transplant patient with a potential stroke condition) can require such complicated combinations of procedures that they begin to look like pure customization. But, by the same token, highly routine interventions, such as certain cataract operations, can seem like pure standardization.
• Bulk Industries. Bulk industries like nonferrous metals and coal produce large volumes of standardized products that are sold (in bulk) to customers, usually on long-term contracts. Thus, while the products are commodities and the production facilities are automated and relatively inflexible, the transactions tend to be personalized. Sellers and buyers negotiate size of order, delivery conditions, and price. Hence the strategies are essentially ones of pure standardization in product and process, yet pure customization in transaction.

Toward the Customization of Perspectives

A flood of recent publications attests to the widespread belief that we are in the midst of a fundamental technological change in manufacturing, communications, distribution, and retailing — a virtual renaissance of customization.25 Our own survey indicates that from 1971 to 1980, an average of only twenty articles on customization appeared annually; from 1981 to 1990, 234 articles; and after 1990, 2,324.26

The initial impetus is generally considered to have come from new information, engineering, and manufacturing technologies that have blurred the traditional barriers between batch and mass production, have opened the way to the restoration of individualization in business areas hitherto dominated by the logic of aggregation, and have served to uncouple formerly tight relationships between manufacturing and design.27 In our opinion, however, these new technologies have not had the effects accorded them, specifically the dramatic shift to customization. Rather, as we have tried to show in both our continuum of strategies and our discussion of industries, all sorts of situations continue to exist. Indeed, if there is one dominant trend — which, we repeat, must coexist with all kinds of countertrends — it is from both ends of our strategy continuum toward the middle, namely toward the strategy of customized standardization, as a number of our examples show.

The shift from pure or segmented standardization to customized standardization has certainly been most obvious. The effects of computer-aided design and manufacturing have produced striking examples of previously standardized products that can now be customized.28 Even the college textbook, the epitome of standardized knowledge, can now be customized to suit an individual professor who selects various sources and can sometimes take delivery of the published version within forty-eight hours.29

Yet the shift from the other end of the continuum — from pure or tailored customization to that same middle position of customized standardization —while less publicized, may in fact be equally noteworthy and, in fact, driven by the same technologies. The high-rise building that was once uniquely designed may today look like a combination of rather standardized components, just as the professor who once used his or her own draft materials or privately reproduced readings may today opt for one of those new reading packages.

An important consequence of this trend is that, as consumers, we lose flexibility in one area while gaining it in another. And, ironically, we lose individuality as we settle in the middle, on customized standardization. The right to specify mayonnaise on a hamburger, a bigger engine in an automobile, or the readings in a course package may give some individuals more choice, to be sure, but can hardly be described as constituting great freedom, in the marketplace or the body politic. In a sense, we lose choice as we all become categories of generic consumers in an assembly operation.

The management field is a child of necessity and experience. The challenge of understanding the large industrial enterprise gave birth to its key concepts. Experience tested and transformed those concepts into basic assumptions. Over time, what had been a loosely related network of concepts became a highly institutionalized framework. Recent developments in technology have certainly brought that framework into question. But that does not justify a shift toward the other extreme, because, ironically, even if the content of the new theory is about customization, our thinking remains standardized.

The temptation to standardize concepts is ultimately rooted in the wish to simplify the world and make our frameworks as general as possible. Managers, however, often have to ignore the theorists’ generalities because their own work brings them into specific contexts that are nuanced and, therefore, unusual — at least compared with standardized theory. For this reason, one secret of successful management today, as at the turn of the century, is to customize standard concepts to fit specific applications. After a century in which management writing has been dominated by the search for the standardized solution, isn’t it time to customize our concepts too?