Skip to Content

Business Lessons from Darwin

January 1, 1997

Natural selection may contain a key to helping companies develop faster, as well as more efficiently and singlemindedly. So concludes James Hines, a senior lecturer in MIT’s Sloan School of Management, after spending two years testing analogies between the behavior of organizations and the evolution of biological species.

The parallels between organizations and organisms are striking, Hines maintains. As a prerequisite for evolution, for example, an entity has to be able to pass on a basic unit of information that initiates a stream of actions. In organisms this unit is the gene. “But corporations also have something that produces a stream of actions,” Hines says, “and that’s a policy.” By policy he means any rule about how decisions are made-such as “We will raise prices when inventories are low, and lower prices when inventories are high.” Every organization has a large genome of such policies-hundreds or thousands of them, both formal and informal-that govern their success in the marketplace.

In much the same way that genes are inherited through reproduction, policies are inherited through learning. While organizations foster learning in a number of ways, says Hines, “the most powerful mechanism is hierarchy”: ideas and bits of ideas are handed down from superiors and combined with the concepts that underlings already carry in their heads. It is through this process that companies both perpetuate certain traits, such as a nurturing or a hard-hearted culture, and gradually develop and change.

The trouble with corporate evolution, however, is its haphazardness. In nature, reproduction offers organisms a simple and reliable way of passing on their genes, which are then evaluated with an unmistakable “fitness function,” or criterion for success-namely, survival. In contrast, members of organizations are often unclear about what policies lead to success and about who among them carries the winning genes. In short, says Hines, “people don’t know who to learn from.”

But Hines is working on a solution. Taking a cue from a branch of computing known as genetic algorithms-wherein software “evolves” toward a goal through constant recombination and reevaluation-he has devised a series of steps that he hopes will allow organizations to seize control of their destiny. First the managers must decide on the fitness function for their company; they pick a goal or several goals, such as higher quality or higher profits or a more pleasant work environment. Then, says Hines, “to direct evolution, a company has to direct who learns from whom, in the same sense that plant and animal breeders have to determine who breeds with whom.” A good way to do this is by promoting people strictly on the basis of achievements that directly advance the organization’s goals. Thus, no matter where you are in the hierarchy, you know that learning from your boss will help you rise up the corporate ladder-and that whatever helps your ascent also helps the company.

The next step is recombination. This happens in organisms when the chromosomes from two parent cells “cross over,” mixing and matching the traits of separate individuals. In an organization, the personnel would be periodically reshuffled; members of one division or design team would be traded with members of another, allowing the best policies to disseminate and to recombine with other ideas in beneficial ways. This practice may sound disruptive, but Hines is quick to suggest an upside: “While you lose the trust that might have built up between the five people on a team, you also lose any animosity. And in time you’ll build camaraderie among 50 people instead of just 5.”

In theory a few rounds of promoting and reshuffling, promoting and reshuffling, should speed up the organizational learning process so that good policies emerge. Hines’s computer simulations basically concur-but there is a catch: it turns out that promotions mean little without the possibility of demotion. In one simulation, for example, people were promoted according to how quickly their group managed to get a product out the door. Hines tried to be kind to the losers. “Because I’m a nice guy, I wanted everybody to get promoted; it’s just that some people would get promoted more.” To Hines’s dismay, the simulation showed no change in time to market, even after many iterations: “Handing everybody a promotion does not give people enough information about who they should pay attention to,” he says.

But, Hines discovered, if you demote people who perform worse than average and reward those who do better than average, “you get a dramatic reduction in time to market.” It’s probably no coincidence, he says, that McKinsey & Co., the management consulting firm with the highest revenue per employee, makes active use of demotions.

Does life have to be so cruel? “Believe me,” says Hines, “I personally don’t like this result. I don’t like hierarchy and I don’t like demoting people.” But although he is still fleshing out his theories-preparing to test the effects of multiple policies and goals simultaneously-Hines is ready to give his reluctant endorsement to hierarchy as a driver of corporate evolution. Even though there are bad hierarchies that stifle creativity and communication, he says, “this form of organization may be uniquely powerful because so much is combined in each step up the ladder: money, power, freedom, your ability to accomplish things. So no matter what your motivation, you have incentives to move up the hierarchy, and to learn how to do that.”

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

This baby with a head camera helped teach an AI how kids learn language

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.