Why small teams score big

Socialism works! So wrote management guru Tom Peters, in Forbes ASAP, in the 1990s. You scoff? Okay, you’re half right. Socialism works, but only in very small numbers. It doesn’t scale.

The same is true of teams in business. Think of any typical ones: Research, a product invention, a marketing launch or a sales campaign. A team of people happily committed to the project and to one another will outperform a brilliant individual nearly every time. But the team must be small—really small—or it will get in its own way.

This isn’t a new idea. In Roman times, a squad was made up of soldiers who could effectively hear their commander’s orders during the clash of battle. The squad’s size—eight—was defined by the number of soldiers who could share a standard tent. Two thousand years later, Amazon’s founder and CEO, Jeff Bezos, speaks of the “two-pizza rule” as the guidepost for team size. “If it takes more than two pizzas to feed the team, the team is too big,” Bezos likes to say.

Is it a coincidence that Roman legion squads and Amazon’s strike teams are of similar size? No. The very nature of the human brain—in particular, short-term memory—revolves around what psychologist George Miller famously called “the magical number seven, plus or minus two”. That is, human short-term memory is capable of capturing and briefly holding between five and nine items of information. Think of zip codes. The brain has a limited repertoire of tricks to enhance that power. (One of them is “chunking” small clusters of data, such as the way we see telephone numbers as a three-digit area code, a three-digit local prefix and a four-digit direct number at the end.)

This means the optimal size of small teams is the same as the effective range of short-term memory in our brains. Our minds seem to work best in that zone of seven, plus or minus two. Below that, the team often devolves into pairs or trios; above that, team communication begins to break down. But why, exactly, does it break down beyond the two-pizza-size team?

Hard math of impossible connectivity
The answer lies in the mathematics of networks. To understand the mag- nitude of this effect, let’s look at the smallest number of connections in a team and progress forward, showing the total number of connections:
2 members = 1 connection
3 members = 3 connections
4 members = 6 connections
5 members = 10 connections
6 members = 15 connections
16 members = 256 connections
32 members = 1,024 connections.

Notice that after the low-digit numbers, the equation settles down into N2, with N representing the number of team members. The complexity of the network grows far faster than the number of team members does. At 1,500 team members—the typical size of a $500 million (revenue) company or division—the number of interconnections reaches 2.25 million. And that creates an obvious problem. Human beings can only handle, much less maintain, much smaller numbers of connections. That’s why, as the number of team members grows, relationships degrade quickly.

Most of us are good at staying in contact with five or six other people on a constant basis. But doing so becomes a whole lot tougher with a dozen or more. At 50? Fat chance. Not even those rare individuals who have photographic memories for faces and names are able to stay in touch with a team that size in the same way they can with one made up of only a half-dozen others. Despite the tools of social networks, texting and videoconfer- encing, we still don’t have the time or the bandwidth to continually maintain hundreds or thousands of close, personal connections.

That’s why bigger teams almost never correlate with a greater chance of success. That’s why socialism fails beyond a dozen or so people. It has nothing do to with ideology and everything to do with the hard math of impossible connectivity.

Rich Karlgaard is the publisher at Forbes