Getting In The Right Rooms

This is a free weekly newsletter about the art and science of building and investing in tech companies. To receive Investing 101 in your inbox each week, subscribe here:

This week, I find myself thinking about small groups. In part because last night, I sat down with small group of incredibly sharp people and listened to them articulate the future of biosecurity, open source AI, the moral implications of prediction markets, and a dozen other topics.

I’ve written before about how any solution requiring more than five people to do something, saying “if they would just” is usually way harder than it seems.

Five people. I keep coming back to that idea of just a small number of people. I mean, come on, I’m limited to having to coordinate five? Five people can’t build a supersonic plane or run a semiconductor fab of a bank or a country.

I think, instead, that idea is measuring how much a group has to agree before it can move; not a limitation on the output. We can do a lot of things, but you have to determine the weight of the friction you can bear.

In a group of three, the number of total connections is three. In a group of ten, the number is 45. Fourteen? 91. Each person you add doesn’t add one degree of complexity; it adds an additional layer of complexity for each person who is already involved.

Output scales roughly linearly. Hire one more engineer, get roughly one more engineer’s worth of work. Makes sense. But the agreeing scales quadratically. Hire one more engineer and you’ve just bought every existing engineer a new human they have to stay aligned with.

Open source software figured this out a long time ago. There’s an exceptional book called Working In Public that explains Yochai Benkler’s three conditions for a commons to actually produce anything: intrinsic motivation, modular tasks, and low coordination costs. “When the costs of coordination outpace the benefits, the commons breaks down as a useful production model.“

Not just gets slower, but full on breaks down. Stops being a way to accomplish things.

Another great example of small coordination is Skunk Works, but in a lot of ways its a cautionary tale. The story of how an “original six-man operation quickly was shunted aside by eighty-five bureaucrats and paper-pushers.” Then the Navy showed up with fifty more overseers “who stood around or sat around creating reams of unread paperwork.”

Which is crazy given some of the massive, audacious accomplishments they pulled off before the auditors showed up. Like the Blackbird, arguably the most audacious aircraft anyone had ever built; it needed “two Air Force brass, three Pentagon officials, and four key players on the Hill” to get rolling.

Certainly, there’s coordination between groups of varying sizes. I’ve written before about how IBM, in its golden age, managed to coordinate 260K pencil pushers alongside the vision of eight “Wild Ducks.” Similarly, Fairchild Semiconductor started with just eight people to get started. But it needed a lot more people than that in order to go on and lead to the creation of ~92 companies and ~$2 TRILLION in value.

The Idea Factory, a book about Bell Labs, emphasized the reality that the group, especially an interdisciplinary group, was better than the lone scientist or small team. From transistors to lasers, solar cells, information theories; that came out a full-on campus. But while the overall production may have been powered, like IBM, by hundreds of thousands of pencil pushers, the innovation always came from the few.

Over time, those at Bell Labs acknowledged that the work “seemed to have shifted decisively to big projects involving hundreds of people,” no longer “heroic research on a new amplifier, done by a few men in a hushed lab.”

Buffett put it bluntly. Berkshire is “a compact organization” that “lets all of us spend our time managing the business rather than managing each other,” and he and Munger “always found a telephone call to be more productive than a half-day committee meeting.” It reminds me of Amazon’s famous two-pizza teams. In fact, get your own pizza and just give me a quick call.

IBM, Bell Labs, Berkshire, Amazon. Notice what all these small team models share. (1) A few people holding real authority. (2) A much bigger machine behind them that doesn’t have to be consulted. And (3) something upstream pointing the two in the same direction.

Will Manidis articulated the maximalist reality of this idea: “doing anything at global scale requires at most like 25 people. Stop believing otherwise.” If the binding constraint is coordination and not capability, then the highest-leverage move available to almost anybody isn’t working harder or hiring faster. It’s getting into a smaller room with more authority in it.

As Margaret Mead put it, “never doubt that a small group of thoughtful, committed citizens can change the world; indeed, it’s the only thing that ever has.”

Stop waiting for everyone to adjust and buy-in. Find your people, get in a room, seize the authority, and change the world. It isn’t as hard as it sounds.

Thanks for reading! To receive Networked Conviction: My Investing Journal, a collection of portfolio updates, Requests for Startups, and investing ideas for paid subscribers, you can sign up below: