Experimentation Is The New Planning

Let’s be honest: You have no idea what’s going to happen to your industry. That’s why you build your organization into an engine of possibility.

Technology is a bitch. It affects every industry, often in ways that are difficult (if not impossible) to anticipate. There’s always the possibility that a Napster or a Netflix or a Wikipedia will arrive to completely disrupt your business or industry.

So it makes sense to have some kind of system that allows you to continually develop options and explore possibilities, so that when the day of disruption does arrive, it finds you ready with a few alternatives in hand. The time to seek those alternatives is now—not later, after a crisis has already arrived.

Let’s Be Deliberate: Real Strategy Emerges

An evolving portfolio of strategic experiments gives the management team more choices, which means better odds that some of the choices will be right.

Management theorist Henry Mintzberg makes a distinction between deliberate and emergent strategy. Deliberate strategy relies on senior leaders to set goals and develop plans and strategies to achieve them. Emergent strategy is a strategy that emerges from all over the company, over time, as the environment changes and the organization shifts and adapts to apply its strengths to a changing reality. Emergent strategy is an organic approach to growth that lets companies learn and continually develop new strategies over time based on an ongoing culture of hypothesis and experimentation.

Deliberate strategy is goal-oriented. It asks, "What do we want to achieve?" Emergent strategy is means-oriented. It asks, "What is possible, with the means we have at our disposal?"

A Portfolio Of Experiments

Diversity breeds creativity—ecosystems are richest where habitats and species overlap. With more connections and diversity comes more creativity: diverse communities are more interesting, more provocative, and more stimulating.

Emergent strategy requires that the company continually generate a broad range of hypotheses, testing them in small-scale experiments, and feeding the more successful experiments while pruning the failed ones. In order to innovate in a sustainable way, a company should have ongoing bets of all sizes, at all points in the power-law curve—a thousand small, a hundred medium, and one or two large—at any given point in time.

In 2005, Google set a formula for distributing its engineering efforts: 70-20-10. Seventy percent of Google’s resources are devoted to improving search and advertising, Google’s primary source of revenue and profits. Twenty percent is allotted as free time for people to pursue projects of their own choosing. And ten percent is invested in scaling up the most promising ideas that emerge from the 20% time, the wild cards that could develop into whole new lines of business.

More Experiments Means More At-Bats

Larger companies have an advantage because they have the resources to fund more experiments. The more things you try, the better your chances of discovering something valuable. Not surprisingly, GE’s Jack Welch, Google’s Eric Schmidt, and Amazon’s Jeff Bezos have all made very similar statements regarding ongoing experimentation.

  • Jack Welch, GE: "Size either liberates or paralyzes. We tried every day to remember that the benefit of size was that it allowed us to take more swings."
  • Eric Schmidt, Google: "Our goal is to have more at-bats per unit of time and effort than anyone else in the world."
  • Jeff Bezos, Amazon: "You need to set up and organize so that you can do as many experiments per unit of time as possible."

Strategy by Discovery

Emergence is self-organization, order that bubbles up from the bottom instead of being pushed down from the top. Emergence is common in complex systems where agents have the autonomy to move around and interact to discover possibilities. For emergent strategy to be successful, there must be enough autonomy, freedom, and slack in the system for people and resources to connect in a peer-to-peer way, like they do in Silicon Valley.

When certain ants need to find a new nest, a few scouts will head out in various directions to search for a new home. When a scout finds a suitable nest, it will spend some time evaluating it. The better the nest, the shorter the time the ant will take. Once the ant has accepted the site, it returns to the main group, where it tries to recruit another ant, whom it then leads to the site. The recruited ant forms its own evaluation, and if the site is acceptable, it will then recruit others in turn. More and more ants are recruited, in an escalating commitment to the site, until the number of ants at the new site reaches a tipping point, which triggers a new behavior. The scouts stop recruiting and begin transporting other ants until the entire colony has moved.

In this manner, an ant colony, working only with local information and without any centralized decision authority, can find the best new site and move the entire colony there in a few hours.

Employees at Mailchimp, an email marketing company with about 100 employees, decide on new features and services in a similar way. If someone has an idea, they attempt to recruit another person to help them work on a prototype or to help convince others. At Mailchimp, people get excited by good ideas, and they are trusted, so they have the autonomy to follow their instincts. To be recruited, a person must consider it more interesting or useful than the things they are already working on. Like the ants, recruitment turns to escalating commitment over time as more people are recruited to the project. When enough people are recruited, a team is formed and commits to seeing the project through to completion. In this way, ideas compete for resources and the best ideas end up bearing fruit.

Computer game maker Valve allows workers to self-organize based on individuals’ interests, passions, and feelings about what’s right for the company. Employees are encouraged to manage themselves, to find the work that most interests them, and to contribute wherever and whenever they see a need. There is no management, and nobody reports to anyone else. Employees choose their own projects and are 100% self-directed.

People commit to projects, and project leaders emerge based on informal consensus. Temporary organizational structures arise based on the needs of a particular project or team, but they are disbanded when the work is done.

Valve has no formal management or hierarchy. It is a company designed to tap into—and to feed—the creative energy, passion, and imagination of its workers. In fact, working at Valve is so fun, you might call the employees players.

Valve is a privately-held company and doesn’t publicly release its revenue figures. But sales are estimated at a billion U.S. dollars per year. And according to Valve cofounder and managing director Gabe Newell, profit per employee at Valve is higher than Google, Amazon, and Microsoft.

These are examples of emergence at work. Nobody is directing people where to go and what to do. Nobody is allocating resources from the top. People and resources self-organize based on horizontal, peer-to-peer activity.

Emergent strategy is not strategy by prediction, it’s strategy by discovery.

Reprinted by permission of the publisher, O’Reilly Media, from The Connected Company by Dave Gray and Thomas Vander Wal. Copyright (c) 2012 by Dachis Group. All rights reserved.

Dave Gray is senior vice president at Dachis Group, the social marketing optimization software and solutions leader, and Founder of XPLANE, the visual thinking company. Follow him @davegray.

[Image: Flickr user Dov Harrington]

Add New Comment

1 Comments

  • jring281

    Strategy, the allocation and scheduling of resources to overcome impediments to achieving objectives, must change with every increment of change in objectives, resources and outcomes (of applying the current strategy). 
    The most obvious and dynamic example is Formula 1 automobile racing, at least by the few who win races. 
    A way of arriving at an "emerg-ible" strategy is described in the Handbook for Interactive Managment by John Warfield and A. Roxanne Cardenas, www.jnwarfield.comAn example of applying IM in pursuit of a strategy for evolving the system engineering 'profession' to a whole system modeling capability can be found at http://ieeexplore.ieee.org/xpl...
    Also illuminating is the advice by Bill Rothschild, architect of the GE Strategic Planning scheme.
    http://www.strategyleader.com/