Expert Opinion: Philip Evans

Communities: Important partners in the deconstructed business ecosystem

6 June 2016

4 min read

Thanks to modern technologies, transaction costs are falling dramatically. But transaction costs are the entire reason for vertical integration. So, in many instances, vertically integrated business definitions are breaking up. This redefines not just businesses, but entire industries.

Within any value chain, you typically have different functions with different characteristic economies of scale: intense in some functions, negligible or even negative in others. But because they were tied together in a single business definition, these effects tended to offset each other: the big company might enjoy economies of scale in its factory, but the small competitor might be advantaged in being closer to the customer, both physically and emotionally. This is why, averaged across the value chain, we have typically seen only gently increasing returns for corporations as a whole. Moreover, small operations close to the customer are characteristically most nimble and innovative; large operations with scale economies and standardized processes are most efficient. With the two kinds of activities tied together, corporations have lived with a trade-off between innovation and efficiency, a trade-off that no amount of clever internal organization can finesse.


The relentless downward pressure on transaction costs brought about by technology, however, melts the glue that holds value chains together. Increasingly, businesses can and must exit disadvantaged functions. The result is “deconstruction”: the breakup of value chains and the redefinition of businesses. When businesses deconstruct themselves, we call it outsourcing or divestment. When deconstruction is done to them, we call it disruption.

When the value chain breaks apart, each activity can operate at the scale that suits it best. If an activity is scale-intensive – running a data center, for example – it can be outsourced to a specialist with economies of scale beyond the dreams of most businesses. Hence cloud computing, Amazon Web Services and Microsoft Azure.

On the opposite end of the continuum are activities with no economies of scale, or even negative ones. Liberated from the transaction-cost-driven need to be part of a larger enterprise, these activities can now be performed by very small enterprises or by individuals. Often these congregate in large communities connected by a platform provider. The members may compete with each other or they may collaborate. Like a swarm, they massively duplicate each other's efforts, blindly try things, and pile onto successes. Their collective behavior is rarely efficient, but it is extraordinarily adaptive, experimental and innovative.


One of the best-known examples of this is the emergence of developer communities for the iPhone and Android platforms. These developers tend to be small businesses or individuals, but they are effective because the operating systems they develop for (and the distribution and billing services needed to market an app) are provided by Apple or Google. The developer's job is just to come up with an app that fills a need. They provide the innovation; Apple and Google provide the scale.
Another kind of community is populated by those sometimes called "e-lancers," individuals providing and purchasing personal services and finding their customers over such internet platforms as eBay, Etsy, Kickstarter, TaskRabbit, Airbnb and Uber. Uber, for example, enjoys far greater economies of scale than a typical taxi company when it comes to technology deployment, marketing and exploiting the "network effect" (the number of local drivers it can connect to a potential passenger, and vice versa). But the individual driver is a sole proprietor, free to fit chauffeuring flexibly into the rest of his or her schedule. That may sound like a minor difference, but it is Uber and not the taxi industry that will automate and scale ride-sharing, carpooling and, eventually, the co-ownership of autonomous vehicles.
Wikipedia and Linux, meanwhile, are examples of purely collaborative communities, where nobody "owns" the product and contributors receive no financial reward. These communities work because the collective "artifact" program) is highly modular, enabling many to contribute without the top-down planning characteristic of a corporation. This does not mean that these communities are entirely "self-organizing"; there is an informal hierarchy of text editors and code maintainers, ultimately sanctioned by the mild authority of the founders, Jimmy Wales and Linus Torwalds. In contrast, one reason the community of Bitcoin developers is running into serious problems is that its founder, "Satoshi Nakamoto," is a pseudonymous figure who retreated into silence years ago. Even collaborative communities need a light touch of traditional leadership.
Purely competitive communities also can form around innovation competitions. Allstate, the US-based insurance company, for example, posed an actuarial problem: predicting bodily injury losses from insurance application data. Allstate collaborated with innovation platform Kaggle, which orchestrated a contest. Teams of academics, industry researchers and hobbyists competed. Within 12 weeks a solution was posted that improved on Allstate's formula by a factor of nearly three. The winners received a cash prize, and Allstate got the solution for a fraction of the cost of an internal development effort.
The largest communities, of course, are the social networks, which can be thought of as publishing media where the consumers are also the journalists. Social media are replacing conventional media in commanding consumers' attention. Provided their commercial use of personal information continues to be accepted, they are destined to overwhelm conventional media in their ability to deliver precisely targeted and persuasive advertising messages.

In all of these examples, the members of a community are individuals or small enterprises, the smallest units of economic organization. But the community itself can be extraordinarily large. This is possible because of the development of global internet platforms. It also is possible because communities are not constrained by the need for complex, non-scalable hierarchical coordination, as would be true of corporate employees. A community also is advantaged because of "network effects"; the larger the community, the greater its value to each member. Thus, for certain tasks, globally scaled communities not only get things done but are economically advantaged over traditional corporate organizations.


Corporations need, therefore, to look closely at activities where their scale adds nothing, or even detracts. They need to ask whether outsiders – especially their own customers – could do stuff, especially innovative and customized stuff, better than internal staff. Often, the correct response is threefold: to use an external community platform or create one's own; to expose data and resources to exploration and tinkering by others; and to stop performing that function internally.
This is a long way from the traditional repertoire of managerial methods: planning, accountabilities, standardization and incentives. The corporation of the future will increasingly need to act as platform rather than actor, curator rather than manager, enabler rather than director. The challenge is to let go of traditional notions of hierarchy and control in order to tap the enthusiasm and imagination of others.

Related resources