After Success, Devolution

The tightly linked nature of the emerging economy makes it behave like a biological community. Wars and battles were the allegories of the industrial economy. Coevolution and infections are more apt in the new economy.

Companies are like organisms evolving in an ecosystem. Some ecosystems in nature offer few opportunities for life. In the Arctic there are only a couple of strategies for survival, and a species had better get good at one of them. Other biomes are chock-full of opportunities, which are in constant flux, appearing and disappearing as species jockey for their niches. The harmony we attribute to nature is not static perfection but a complex dance of ups, downs, trips and falls, and balance regained.

Rich, interactive, and highly flexible in shape, the network economy resembles a biome seething with action, a jungle in fast-forward motion. New niches open up constantly and vanish quickly. Competitors sprout beneath you and then gobble your spot up. One day you are king of the mountain, and the next day there is no mountain at all.

Biologists describe the struggle of an organism to adapt in this type of habitat as a long climb uphill, where uphill means greater adaptation. In this metaphor, an organism that is maximally adapted to the times is situated on a peak. Imagine a commercial organization instead of an organism. A company expends great effort to move its butt uphill, or to evolve its product so that it is sitting on top, maximally adapted to the consumer environment.

All organizations (profit and nonprofit alike) face two problems as they attempt to find their peak of optimal fit. Both problems are exacerbated by the constant turbulence of the network economy.

First, unlike the industrial era’s relatively simple environment, in which it was fairly clear what an optimal product looked like and where on the stable horizon a company should place itself, it is increasingly difficult in the network economy to discern what hills are highest and which summits are false.

In biological terms, the new economic landscape is “rugged,” disrupted by gulfs, precipices, and steep slopes. Trails are riddled with dead ends, lead to false summits, and made impassable by big-time discontinuities. Because the economic terrain is jumbled with no overall pattern, there is no certainty that a company intending to head up a slope toward a peak new market is actually climbing anything larger than a hill. In biospeak, they may succeed in getting to the top yet find themselves stuck on a suboptimal peak.



Big and small companies alike…

…have to deal with their new landscape. It’s often unclear whether a firm should strive to be on top of a mountain (for example, to be the world’s most reliable hard disk manufacturer), when the whole mountain range beneath that particular peak may sink in a few years (if everyone moves their storage onto large protein arrays). An organization can cheer itself silly on its way to becoming the world’s expert on a dead-end technology. (The nuclear power industry offers one example.)

Turbulent times mean that local success is not global success. A company may be at peak efficiency, but on the wrong mountain. The trick is to select a high-potential area to excel in.

Some of the most perfect technology was created just before its demise. Vacuum tube technology reached a zenith of complexity just before it vanished. As MIT economist James Utterback writes: “Firms are remarkably creative in defending their entrenched technologies, which often reach unimaginable heights of elegance in design and technical performance only when their demise is clearly predictable.” It’s relatively easy to arrive at a peak of perfection. The problem is that perfection can be local, or suboptimal, like being the best basketball player in your state, but unaware of national tournaments. While a firm is congratulating itself on creating the world’s fastest punch card reader–the fastest in the universe!–the rest of the economic world has moved on to the PC.



The harsh news is that “getting stuck…

…on a local peak” is a certainty in the new economy.

Instability and disequilibrium are the norms; optimization won’t last long. Sooner, rather than later, a product will be eclipsed at its prime. Indeed, an innovation at its prime increases its chances of being eclipsed. In Mastering the Dynamics of Innovation, a study of innovation in the automobile industry, Utterback concludes that “an unhappy byproduct of success in one generation of technology is a narrowing of focus and vulnerability to competitors championing the next technological generation.” The product may be perfect, but for an increasingly smaller range of uses or customers.

While one product is perfecting its peak, an outsider can move the entire mountain by changing the rules. Detroit was the peak of perfection for big cars, but suddenly the small-car mountain overshadowed it. Sears was king of the retail mountain, but then Wal-Mart and Kmart’s innovations created a whole new mountain range that towered above it. For a brief moment Nintendo owned the summits of the video-game mountain until Sega and later Sony built separate mountains even higher. Each of the displaced industries, companies, or products were stuck on a less optimal local peak.

There is only one way out. The stuck organism must devolve. In order to go from a peak of local success to another higher peak, it must first go downhill. To do that it must reverse itself and for a while become less adapted, less fit, less optimal. It must do business less efficiently, with less perfection, relative to its current niche.

This is a problem. Organizations, like living beings, are hardwired to optimize what they know–to cultivate success, not to throw it away. Companies find devolving unthinkable and impossible. There is simply no allowance in the enterprise for letting go.



And the better the company…

…the less room there is for devolution.

Everything about a modern organization is dedicated to pushing uphill. The CEO is trained, and paid well, to push the firm toward the peak. Quality circles get the entire workforce marching uphill toward optimal performance. Consultants monitor the tiniest detail, trying to eliminate anything that might keep the company from attaining the peak of perfection. Reengineering wonks zero in on computer data showing which parts of the organization are lagging behind. Even the receptionist is in search of excellence.

Where in the modern company is the permission, let alone the skill, to let go of something that is working, and trudge downhill toward chaos?

And have no doubt: It will be chaotic and dangerous down below. The definition of lower adaptivity is that it places you closer to extinction. But you have to descend and risk extinction in order to have the opportunity to rise again.

Economist Joseph Schumpeter calls the progressive act of destroying success “creative destruction.” It’s an apt term. Letting go of perfection requires a brute act of will. And it can be done badly. Management guru Tom Peters claims that corporate leaders are now being asked to do two tasks–building up and then nimbly tearing down–and that these two tasks require such diametrically opposed temperaments that the same person cannot do both. He impishly suggests that a company in the fast-moving terrain of the network economy ordain a Chief Destruction Officer.



With or without someone in charge…

… of creative destruction, there is no alternative (that we know of) to leaving behind perfectly good products, expensively developed technology, and wonderful brands, and heading down to trouble in order to ascend again with hope.

Once upon a time this march was rare. The relatively stable markets and technological environment of the industrial era were smooth, not rugged. Only a few parameters changed each year, and they changed gradually. Opportunities arrived with forewarning. Those days are over. The biological nature of the new economic order means that the sudden disintegration of established domains will be as certain as the sudden appearance of the new.



There can be no expertise in innovation…

…unless there is also expertise in demolishing the ensconced.

There is nothing wrong with perfection. To be maximally fit for a niche, to serve optimally, to seek the peak of perfection–these will always remain the goals of any firm, or individual. So why let go of perfection at the top?

The problem with the top is not too much perfection, but too little perspective. Great success in one product or service tends to block a longer, larger view of the opportunities available in the economy as a whole, and of the rapidly shifting terrain ahead. Legendary, long-lived companies are intensely outward-looking. They can spot a global peak and distinguish it from the many false peaks. They understand that an inward focus, especially a narrow focus on being “world’s best” in some matter, can work against long-term adaptation by blinding the organization from seeking new heights. Better for the long haul is an outward perspective that is always seeking alternative mountains to climb.

This outward vista is all the more critical in the new economy because perfection is no longer a solo act. Success is a highly interdependent enterprise, encompassing a network of vendors, customers, and even competitors. A firm needs to explore widely, outside of the current favored position, and at times contrarily.



Letting go at the top is not an act against…

… perfection, but against shortsightedness.

In addition to the scarcity of leaders willing to disassemble the profitable, and the natural bias of companies toward perfection, there is another reason why letting go is so hard. Economists Paul Milgrom and John Roberts studied the competencies–the winning traits–of a large number of firms in modern manufacturing and concluded that competencies of companies tended to occur in suites, or in a guilds of skills.

This natural bundling of traits makes it very difficult for contenders to challenge a successful firm. As Richard Nelson, an economist at Columbia University says, “Successful firms often are difficult to imitate effectively because to do so requires that a competitor adopt a number of different practices at once.” Companies can buy technology and human skills in a particular area. But gradually acquiring one or two competencies at a time does no good when you are attempting to displace a highly successful firm. The whole suite of mastery has to be acquired simultaneously in order for you to be competitively effective. A firm such as Disney is almost inimitable because of the difficulty of obtaining in one swift swoop its highly integrated mix of skills.

The natural bundling of traits also makes unraveling for devolution immensely difficult. To devolve demands going against all the best qualities of an organization all at once. The organic world offers a number of lessons in this regard. Biotechnology is built on the knowledge that most genes don’t code for anything themselves. Most genes regulate–turn off and on–other genes. The genetic apparatus of a cell, then, is a dense network of hyperlinked interactions. Any gene is indirectly controlled by many other genes.

Thus, most attributes in a biological organism usually travel in the genome as loosely coupled associations. Blue eyes and freckles, say. Or red hair and a hot temper. Two important consequences follow from this. First, to get rid of the redhead’s feisty temperament by evolution may also mean–at least at first–getting rid of the red hair. Animal breeders know this dilemma firsthand. It is difficult to breed out an unwanted trait without breeding out many desirable ones. Chicken breeders can’t get rid of a chicken’s aggressiveness without throwing out its egg-laying proficiencies.

Secondly, the interlocking guild of competencies, which give organisms and organizations their advantages, becomes a drawback during change. The increased interlinkage of the network economy heightens this dilemma. In the network economy, the skills of individual employees are more tightly connected, the activities of different departments more highly coordinated, the goals of various firms more independent. The net brings the influence of formerly unrelated forces to bear upon each potential move.



The more successfully integrated…

…a firm’s capabilities are, the harder it is to shift its expertise by changing just a little. Thus successful firms are more prone to failure during high rates of change. (Success makes it easy for the successful to deny this fact.) Indeed, the very success of successful organizations makes them conservative toward change–because they must unravel many interdependent skills–even if some are working fine.

The problem that IBM faced with the arrival of the personal computer in the early 1980s was not the problem of acquiring technological know-how. As a matter of fact, IBM already knew how to build personal computers better than anyone. But the package of proficiencies the blue suits had honed over the years to make IBM indomitable in the mainframe computer field could not be gradually adapted to fit the new faster-paced terrain of desktop-based computing. IBM was supreme in the old regime because their sales, marketing, R&D, and management skills were all optimally woven into a highly evolved machine. They couldn’t change the size of the computers they sold without also altering their management, forecasting, and research skills at the same time. Changing everything at once is difficult for anyone, anytime.



Because skill guilds constrain…

…(and defend) an organization, it is often far easier to start a new organization than to change a successful old one.

This is a major reason why the network economy is rich in start-ups. Starting new is a less risky way to assemble an appropriate new set of competencies than trying to rearrange an established firm, whose highly intertwined bundles resist unraveling.
In a rugged economic landscape, about the only hope an established company has for adapting to turbulent change is by employing the “skunk works” mode, which reflects another biological imperative. Computer simulations of evolution, particularly those run by David Ackley, a researcher at Bellcore, demonstrate how the source for mutations that eventually conquer a population start at the geographical fringes of the population pool. Then after a period of “beta testing” on the margins, the mutants overtake the center with their improvements and become the majority.

At the edges, innovations don’t have to push against the inertia of an established order; they are mostly competing against other mutants. The edges also permit more time for a novel organism to work out its bugs without having to oppose highly evolved organisms. Once the mutants are refined, however, they sweep rapidly through the old order and soon become the dominant form.

This is the logic of skunk works. Hide a team far from the corporate center, where the clever can operate in isolation, away from the suffocating inertia of success. Protect the team from performance pressures until their work has had the kinks ironed out. Then introduce the innovation into the center. Every once in a while it will take over and become the new standard.

Economist Michael Porter surveyed 100 industries in 10 countries and found that in all the industries he studied, the source of innovations were usually either “outsiders” or else relative outsiders–established leaders in one industry making an entry into a new one.



To maximize innovation, maximize the fringes.

Encourage borders, outskirts, and temporary isolation where the voltage of difference can spark the new. The principle of skunk works plays a vital role in the network economy. By definition a network is one huge edge. It has no fixed center. As the network grows it holds increasing opportunities for protected backwaters where innovations can hatch, out of view but plugged in. Once fine-tuned, the innovation can replicate wildly. The global dimensions of the network economy means that an advance can be spread quickly and completely through the globe. The World Wide Web itself was created this way. The first software for the web was written in the relative obscurity of an academic research station in Geneva, Switzerland. Once it was up and running in their own labs in 1991, it spread within six months to computers all around the world.


Archives - This site operates under a Creative Commons License.

This is a blog version of a book of mine first published in 1998. I am re-issuing it (two posts per week) unaltered on its 10th anniversary. Comments welcomed. More details here.
-- KK