Doing business on the Web is doing nothing less than deconstructing the organization. And like any phenomenon or challenge, deconstruction must be managed.

The value chains that define a business, the supply chains that define an industry, the customer relationships and brands that define a franchise, and the organization charts that define hierarchy and power are all premised on the “glue” of information. Rich information, flowing through limited-access channels, bonds business structures together. But the salient fact of our professional lives is that this glue is melting. Information flows are being separated from physical ones. Rich information is being freed from its traditional limitations of reach. Markets are therefore inter-penetrating where organization had previously prevailed. Businesses are being redefined. So are the universes of possible suppliers, customers and competitors. So too are the scope and power structure of the corporation.

Information channels that glue business together are being blown apart by the advent of universal connectivity and standards. These connectivity and standards enable rich communication unconstrained by the need to establish a shared vocabulary beforehand. They enable a massive expansion of reach with no loss of richness in the information shared. They enable markets to replace managerial coordination. They enable no-cost searching and switching. They support symmetry of information. They enable fluid teaming within organizations. Just as easily, they enable fluid teaming across organizational boundaries. If “relationship” means a warm and fuzzy inertia, they undermine relationships. If “brand” means an uninformed preference for the tried and true, they undermine brands.

The penetration of the Internet and the explosion of e-commerce are only the beginning of his process. So far, the Internet has blown up the traditional trade-off between richness and reach only for the delivery of static, largely textual information. But the underlying technologies will continue to improve processing speed and memory by factors of 10, and bandwidth by a factor of 100 every five years. Rich connectivity will spread to every kind of intelligent device. Switched, interactive sight, sound and motion will become ubiquitous. Entertainment and information will merge. Intelligence and standards will penetrate ever deeper into content, transforming the Web into something approximating a vast, neural network.

This is a more far-reaching phenomenon than the advent of a new channel for advertising and transactions. It challenges specifics of marketing, sourcing, distribution, collaboration, employment and risk bearing. It challenges the scope and rationale of the organization. It challenges assumptions about the major sources of competitive advantage: scale, scope, franchise, share of market, share of wallet, relationships and core competencies. It “deconstructs” the very definitions of a business and an industry.

THE CHALLENGE TO STRATEGY

To understand the nature and implications of this deconstruction, just consider some central tenets of conventional strategic thinking.

Conventional strategic thinking assumes a value chain, the sequence of discrete steps that define a business. Such steps belong to the same value chain because each one is customized to fit the others and could not be operated separately without incurring the risks that arise from mutual dependence. The elements of a value chain generally may have different drivers of competitive advantage, but because they are tied together, it is only the average advantage that matters.

The competitive advantage of a business is defined in the context of competitor businesses, which collectively constitute an industry. Because the competitors in an industry are similar, and because competitive advantage is averaged across multiple value-added steps, there is not much difference between the most advantaged and least advantaged. Which is why one competitor rarely seems to drive the others out of business.

This industry stability guarantees that most competitors have things roughly right. But if things are roughly right, then deficient pieces of the business formula can be fine-tuned, taking the other pieces as more-or-less given. Therefore the incremental pursuit of the incremental improvement is a principle that works. Focused change programs achieve results. Hierarchical systems of control and accountability add value. Strategy can be partitioned and compartmentalized, and with it the organization that executes that strategy.

Conventional thinking also recognizes the supply chain, a set of vertical links among suppliers and customers tied together again by the pattern of coordinating mechanisms they have built. Indeed, supply-chain relationships have narrowed and deepened as companies have sought tighter coordination of logistics or design. The retail customer is part of this chain, committed to predictable patterns of behaviour by his real costs of searching and switching. Strategies of cross-selling, brand extension, relationship building and maximizing lifetime value are all premised on the existence of asymmetries of information between the seller and the buyer and the consumer’s substantial costs of comparing alternatives.

At the corporate level, conventional thinking also recognizes core competencies, know-how that may not be specific to one business but can be an advantaged asset for the corporation as a whole. Lines of business are selected, and so the overall corporation is shaped by the desire to leverage these competencies. The role of the corporate centre—indeed the identity of the corporation—is defined around the transmittal of core competencies to all business units.

All these premises of strategy—the value chain, the stabilizing effect of averaging, and the definitions of the industry, supply chain and core competencies—are taken as givens in the process of making and then executing business strategy. But they all presume something about information. They presume that information channels bind activities together, and that corporate organization is the advantaged information channel. It is precisely these assumptions that are becoming obsolete.

The first step toward obsolescence is the advent of open, universal communication standards—which provide a common infrastructure for the exchange of rich information unconstrained by the traditional necessity to flow along pre-established channels. Connection to a dedicated information channel therefore ceases to bind activities together.

The second step is the advent of content standards for the definition, storage and presentation of particular kinds of information. This eliminates much of the mutual dependence among activities connected by that information. Syntax and vocabulary cease to be a challenge, since everyone speaks the same language. But once that happens (and it is happening very fast), once the same infrastructure and the same content standards prevail, there no longer is a trade-off between richness and reach in the sharing of information. Layers in a value chain are no longer tied to each other; supply chain relationships, including those with the final customer, are opened up.

When the Internet, extranets and intranets respectively support rich information exchange with the customer, suppliers and within the corporation, and when the Internet, extranet and intranet are exactly the same thing, there is no longer any necessary difference between information flows within the corporation and across its boundaries. The unique function of organization, hierarchy, brands, relationships and intermediaries to channel rich information among a privileged few is called into question. Compound this with the increasing fluidity and transparency of capital and labour, and it becomes apparent that internal hard-wired organizational structures, the boundaries of such an organization and the core competencies incubated in it are no longer necessary or stable. The glue melts. Industry structures, the premises of strategy, are “deconstructed.”

MANAGING DECONSTRUCTION

Much of management is about specificity and tactics—stepwise solutions to prioritized, practical problems. Deconstruction, in contrast, presents a dizzying degree of fluidity, indeterminacy and instability. It is the opposite of what most managers are trained to deal with. That makes it discomforting. It is mentally exhausting. One craves a simple forecast, straightforward prescriptions, a specific set of “to dos” for Monday morning.

That is really difficult. Deconstruction is not a new set of rules about strategy. It is an argument that says that all the old rules of strategy apply, but at a much finer level of granularity. With the partial exception of pure information businesses, the strategies themselves are essentially the same: scale, market share, cost, innovation, capabilities, competencies and the rest. But those strategies have different objects. The task is therefore one of identifying these new objects and rethinking and reapplying the familiar principles of competitive advantage.

There are no shortcuts for doing this, no simple, across-the-board substitutes for hard thinking about business specifics. However, we would like to describe 12 guiding principles that may help to rethink a strategy in an era of deconstruction.

No business leader today can presume that the definitions in his or her business will be valid a few years from now.

Deconstruction means that traditional business definitions can no longer be taken for granted. Suppliers, customers, competitors, industry, value chain, supply chain, consumer franchise, and relations with employees and owners all become variables. They are reshaped by the shifting economics of information and by the strategies pursued by players who exploit those economics. Moreover, the process of deconstruction is continual. Driven by connectivity and standards, progressive advances in richness and reach will challenge successive business definitions with the possibilities of further deconstruction. There is no end-game.

Consider books and music. CDNow was the first company to dominate the sale of music on the Internet. But when Amazon entered the category, it took only a few months to overtake CDNow. It is quite apparent that in cyberspace, music retailing is not really a category separate from books. Perhaps the very idea of “music retailing” is merely a mental throwback to the physical world, where the peculiarities of inventory and merchandising make it a distinct business. Indeed, after a phase of frantic warehouse construction, Amazon is rapidly redefining its business as a retailer of a wide variety of products, many of which are not physically handled by Amazon at all. Its ultimate business definition may reflect a certain kind of customer relationship and a certain kind of shopping experience, rather than any traditional product-specific category. Meanwhile, the Rocketbook format for books, and MP3 (and other digital formats) for music, threaten to convert distribution into e-mail, rendering all those warehouses obsolete. Business models come and go and the survivors transform themselves continually. There is no endgame.

Deconstruction is most likely to strike in those parts of the business where incumbents have the most to lose and are the least willing to recognize it.

It is easy to deny the relevance of deconstruction by pointing to some central and possibly large part of the business that will not be deconstructed. Two examples are the editorial page of a newspaper or the investor’s need for human reassurance in brokerage. The flaw in this reasoning is that deconstruction will not attack the crown jewels, it will attack the soft underbelly.

There may be no Internet equivalent (yet) for the authoritative editorial voice of our great newspapers, but there most certainly is there for the mundane provision of classifieds. Classifieds account for 40 percent of the revenue but only 10 percent of the costs of the average newspaper. Since overall margins are about 20 percent, classifieds therefore account for 150 percent of average profitability. If classifieds migrate to the Internet, that massive subsidy to the remainder of the business will end, even if the newspaper company succeeds in dominating the electronic classifieds business. Display advertising revenues cannot be boosted to offset this loss (if they could, the newspaper would have done so already). To recoup its losses, the newspaper will have to double or triple its circulation prices. This will cause a substantial drop in circulation. But as readership falls, display advertising revenues will fall proportionately, necessitating further price increases or cost reductions. Emerging from this vicious circle, a smaller-circulation, lower-quality business model could doubtless be established, at least by some newspapers. But many—if not most leading writers and investigative reporters—will, quite predictably, lose their jobs in the process. So much for the crown jewels.

Waiting for someone else to demonstrate the feasibility of deconstruction hands over the biggest advantage a competitor could possibly wish for—time.

The Internet is the apotheosis of time-based competition: Everything is a sprint. The strategy of relying on others to make the early mistakes is very dangerous. Barnes and Noble thought that its branding and merchandising skills would enable it to easily overtake whichever company would be the first to open up electronic book selling. Despite its massive response to Amazon, there is little prospect that it will ever catch up. A six-month lag means that the game is over. “Fast follower” strategies may work in marathons, but not in sprints.

There are some exceptions: Microsoft caught up with Netscape, and Schwab with e*Trade. But these are young companies, used to running very hard and still led by their founders. For the slower-moving, incumbent organization, this is simply not an option. The experiences of too many businesses suggest that brand image and a customer list are not enough. Barnes and Noble will probably never catch up to Amazon. Toys “R” Us will probably never catch up to Schwab.

It is even worse to presume that there is no serious threat to established ways of doing business because every attack mounted thus far has failed. Some day, somehow, someone will break through, if only because the economics of information continue to improve by an order of magnitude every five years. It is only a matter of time. And when that breakthrough happens, if the established competitor is not in the vanguard of blowing up his own business, playing catch-up will be a very tough challenge.

Leaders need to wrestle with the full range of the possible patterns of deconstruction.

Businesses can be deconstructed in four possible ways. Some will break into discrete components of their value chain, unlinked by the melting of the informational glue that bonded them together. Others will deconstruct along their vertical links with suppliers, customers and consumers, when mutual relationships, stable franchises and intermediary roles are undermined by reach. Others will see the segregation of information flows into businesses in their own right. Still others will deconstruct in their organizational relations, as employees, investors, and entrepreneurs renegotiate roles, risks and rewards.

These deconstructions may be partial or they may be comprehensive. They are not mutually exclusive. They are often mutually interdependent. No successful strategy in the face of deconstruction can ignore any of these patterns.

Merrill Lynch, for example, believed that it was safe from the competitive threat of cheaper discount brokers because of its proprietary research and the highly personal relationship between its brokers and their clients. No telemarketing operation, still less a Web page, could substitute for those capabilities. This was wrong. The uniqueness of its proprietary research is being undermined by the decision of many research houses (who used to sell their work to Merrill) to publish directly, separating the information flow from the execution of transactions. And the personal relationship between broker and client can flourish without Merrill Lynch. Independent financial advisers use the entire resources of the Net for research and analysis, and can use Schwab or e*Trade to execute transactions on behalf of their customers. Not surprisingly, it is the best brokers, tempted to quit by the prospect of putting the entire value of their client book in their own pockets, who are deconstructing the Merrill Lynch business model from the inside. None of these threats was compelling by themselves; together they are deadly.

Strategy really matters.

In a world of stability, strategy is applied economics—refining segmentation, disaggregating costs, optimizing features and the like. It is a response to supply-and demand-side realities that exist quite independently of the business leader’s ability to figure them out. But in a deconstructing world, strategy creates economic realities. Whether a standard does or does not reach critical mass, who pre-empts whom and who allies with whom determines not just the path of competitive jockeying, but the end result itself. The interplay of strategies among competitors has an autonomous impact on shaping the outcome.

This takes strategy out of the realm of applied economic research and out of the hands of staff in a department. Indeed it takes business strategy back to its military roots. Strategy has to be adaptive, not deterministic. It has to be continually reformulated. It requires alliances. It needs timely, comprehensive intelligence. It has an improvisational quality and may require ceding a high measure of autonomy to the commander in the field. Above all, it necessitates (as it does in warfare for similar reasons) the single-minded, continuous attention of the general—the CEO.

The value of winning will escalate, as will the cost of losing.

As a result of unbundling, the economics of businesses after deconstruction will often be simpler and therefore more powerful. Competitive advantage and the distribution of rewards will become much more skewed. Information flows, in particular, will tend to become either valueless or monopolies, and it really matters to understand which one. If, within a given business, there is room for only one winner, getting strategy right becomes really important. Getting it right becomes more important than not getting it wrong. And getting it roughly right before anybody else does matters more than getting it exactly right after the opportunity has been pre-empted.

Spreadsheets and cash flow projections count for little when the primary determinant of the value of a business is its terminal value, or its value as an option in some unforeseen future development. Current profitability matters not at all in a context where foreseeable growth rates exceed the cost of capital. A small probability of resounding success is worth far more than the high probability of affordable failure. The ablest managers, if they are aggressive enough, ought to have many more failures than successes. They should be rewarded for such a track record. But this fits ill with corporate norms. Corporate planning, control systems, personnel policies and cultures are designed to minimize the probability of error, not maximize the probability of success. Therefore, do some very “uncorporate” things: experiment, pre-empt or pursue contrary strategies simultaneously.

The reconstructed business definitions will rarely correspond to the old.

New businesses will emerge and agglomerate in accordance with their own competitive logic. Emerging information businesses will spill across the boundaries of the physical businesses from which they originate. In particular, many emerging “navigator” businesses will move forward and away from their origins in marketing or retailing; competitors unwilling to follow the emerging logic to wherever it leads may end up forfeiting the most valuable business position they own. Successful competitors will therefore have to be prepared to build or acquire fundamentally new capabilities, transform the business model repeatedly, make alliances with competitors or companies in previously unrelated businesses, and merge aggressively for pre-emptive scale and scope. Underestimating the requirement for acquiring new capabilities, overestimating the value of existing capabilities and sticking rigidly to one business definition are all common traps.

The hardest step for an incumbent organization is the mental one of seeing the business through a different, deconstructed lens and then acting on this insight.

Mentally deconstructing the business sounds easy in theory, but as soon as the practical implications become clear, the reaction of every organization is to resist. A “navigator” strategy sounds terrific, up to the point where being the best navigator requires navigating to competitors’ products. A “pre-emptive” strategy sounds heroic, until a succession of premature failures and the dilution of corporate earnings hit the income statement. A “disintermediating” strategy sounds great, until the dealers revolt. Companies get caught in a “cannibalization calculus,” as they try to build a new business and simultaneously preserve the value of the old. It hardly ever works. Cannibalization calculus begets half-hearted, timid, slow and compromised decisions, in an environment where single-minded speed is what matters. That is when the incumbent blinks and the insurgent steals an unmerited but devastating advantage.

The subtler pitfall is co-option and passive resistance by a skeptical, self-preserving organization.

Organizations have an uncanny ability to subvert whatever undermines their historical structures of power and reward, even when the new direction is official strategy. The IT Department takes years to build an “industrial strength” Web interface; the sales force takes charge of the “smooth transition” to the new distribution channel; the rewards for entrepreneurship are set in accordance with established corporate HR policies. These are all mistakes.

Intending to leverage its core competencies, Westinghouse assigned the new “transistor” business to its Vacuum Tube Division. Consequently, it killed in one stroke its best hope for the future of the company. That mistake is being repeated today many times over. In the majority of circumstances, there has to be a strong bias in favour of separate initiatives, reporting outside the established organization structure and protected from the conservatism of the mainstream organization. Such initiatives may require new people, different reward structures and quite possibly an independent stock market valuation. If that limits the synergies with the mainstream business, so be it. It is devastating, but in the majority of circumstances the incumbent business offers more negative synergies than positive ones.

Strategy in a deconstructing world has to be directionally right—though not specifically right—as long as the organization maintains a capacity to learn from its mistakes.

Strategy in a deconstructing world cannot be planned in the same ways as it was in the past. Planning presumes certainty, or at least bounded uncertainty. It presumes that the numbers can be analyzed, the payoffs identified. It posits milestones, budgets, cash flows, accountabilities. The entire system of corporate planning is designed to avoid error. But in conditions of high uncertainty, error is inevitable and people unwilling to make mistakes will get it right too late to claim any reward for their meticulousness. Strategizing has to be continuous; it has to be partially improvisational and flexible enough to recognize errors when they occur. The errors must then be corrected and the organization needs to move on.

The organizational, behavioural and personal baggage that they insist on bringing to the new venture too often destroys all incumbents’ best assets.

IT people struggle with a terrible headache called “legacy systems,” enormous information architectures, with layer upon layer of improvements, extensions and investments that turn out to be massively inferior to a simple, clean new box. Beyond a certain point, the problems of updating and maintaining compatibility become so severe that it makes sense to junk the entire system and start from scratch. Business systems are human software; they follow exactly the same logic. There are legacy organizations, legacy mindsets, legacy competencies. And beyond a certain point it is necessary to junk them and start again.

Current corporate planning procedures are likely to decrease value in the way they deal with deconstruction. Traditional expertise in marketing or product development may be irrelevant and dangerously misleading. Home-grown career executives may have exactly the wrong skills and attitudes. Undiluted corporate ownership may prove to be an untenable structure for distributing responsibility and rewards.

The imperative for totally new strategies implies the imperative for totally new processes.

Incumbents can be the insurgents, if they choose.

Incumbents do not have to think of themselves as incumbents; that is to presume precisely the static business and industry definitions that deconstruction denies. They can move a capability they have right into the heart of somebody else’s business and blow it up. It takes clarity of vision and consistency of purpose. It requires organizing and rewarding differently, perhaps even owning differently.

The devastating truth is that for the vast majority of deconstruction opportunities that have emerged to date, some corporation was actually the best positioned to exploit it. Almost invariably it failed to do so and a few young entrepreneurs became billionaires as a consequence. It failed to do so because it failed to think like an insurgent. It was not aggressive enough, nor was it greedy enough. It was still studying the situation.

As with military conflict, battles can be won and lost in the mind. Think like an incumbent and be obsessive about preserving old business models, and you doom your organization to reactive, half-hearted and ultimately futile gestures. Incumbents are as capable of winning on the new battlefield as anyone else is. Indeed, in some ways, they are far more capable. But they have to think like an attacker.