Friday, July 31, 2009

Communities vs. Networks

My first job, at IBM, I knew nothing about computers. Why they hired me to design mainframe CP logic, I can't figure out. Columbia taught Electrical Engineering as a liberal art, so while I graduated able to talk in detail about semiconductor physics and hold my own in a conversation about Claude Shannon's master's thesis, I could not use the UNIX command line. I asked a lot of annoying questions my first year there, like "so, how do I turn it on?" and "what's the difference between MVS and VM?"

My first week I needed to print some 20 page design document. I sent it to the printer queue. Then I trudged over to the room where the high-speed printer was, lined up at the half-door, told the printer tech my job number and he handed me 500 pages of gobbledygook. Probably tied the printer up for an hour. Whoops.

I asked the guy down the hall, who had been very helpful, what I did wrong. He asked me to email him the file. He then emailed me back the file in a different format and told me to print that. I asked him what he had done and he told me that the original was in a markup langauge that needed to be pre-processed. I asked him how to pre-process the files, since I would undoubtedly have to do this many times. He told me to email them to him and he would do it.

No matter how much I pressed him, he would not tell me how to do this. So I asked someone else and, in 30 seconds, they showed me the program to run.

My only explanation for why he would not show me how to do this simple thing is that he must have felt that knowing something that I didn't know was valuable. That by keeping the knowledge to himself, he made himself indispensable. In reality, he had just made himself a bottleneck. And a freakin annoying one, at that.

This came to mind recently because I've been doing a lot of networking. A friend at a mid-sized ad network asked me to find some really good, smaller ad networks that might want to partner with them. While I narrowed my list of 418 ad networks down to the ones that seem like the best fit, I figured I'd ask around if anyone knows anyone they think would be right for this. I've gotten great responses and a lot of introductions (I'm still looking, BTW, so if you know anyone, drop me a line.)

A few people, though (and some I've known for more than ten years and been through some character-defining shit with), won't introduce me to people they know. They are more than happy to have me tell them what I am looking for and pass that on, then get a response and pass that back. Just like Annoying Guy at IBM, they want me to email them the file every time. They want to be the bottleneck. And just like at IBM, the only explanation I can think of is that they think this makes them important.

The Epicurean Dealmaker had an interesting post on Goldman Sachs a few days ago. He is commenting on a Slate article:

Ms Moore points out the fact that, for all its reputation as "a devastating hive mind that can control any institution it touches, including the U.S. government," and as an gathering of the smartest minds, human and machine, on the planet, Goldman Sachs employees have proved singularly inept outside of the hive.
TED's explanation for why this is is the interesting part:
Notwithstanding what they like to tell you, investment bankers... are successful to the very extent they can maintain themselves in the flow of market information. Investment banks derive their market power and importance by maintaining dense and robust information networks across the numerous markets they participate in... Take a banker with excellent network connections out of his or her supporting environment, and he or she becomes dramatically less effective.
A while ago, on the advice of a friend, I stopped thinking about the people I knew as a "network" and started thinking of them as a "community." This made me do two things very differently: I made a lot more introductions, many at my own instigation and just because I thought the two people might find some common interest; and, I consciously tried to stop mediating connections, to stop thinking the point of my relationships is in knowing the person but rather that the point is in communicating with the person, and in helping them communicate with others. Doing this has worked unbelievably well for me.

I think people who try to mediate access to their network fundamentally misunderstand the value of networks. The value is not in the number of edges connected to your node, it's in the information that flows over those edges. I think I was present at the creation of the idea that the metric for the attention economy is the number of people paying attention to you divided by the number of people you pay attention to, and I agree that it's a seductive idea. But it is completely wrong.

The value of your network, to you, is the amount of high-quality information that flows through it, and not necessarily through you. More and better relationships mean that you get more and better-filtered information; this is widely noted. Less noted is that if the people in your network are more interconnected, the information in your network will be of higher quality. This is obvious, once you stop to think about it. (If it's not, then think about the institutions that create high-quality information and how they are organized. And if that doesn't make it obvious, go read Jane Jacobs' The Economy of Cities.)

Creating communities of interconnected people is to your benefit, even if it means that you are no longer included in every communication. In the end, the information that finds its way to you will be much more useful, and there will be more of it. If your goal is to get more Twitter followers, to have your blog more widely read or to have the most LinkedIn connections, then you're not creating value at all*. Instead, create a "dense and robust information network": introduce everyone you know to someone else you know, make sure that when you learn something that you pass it on to the people who will benefit (and not necessarily everyone who follows you), and trust that the community you help build will end up making everyone in it more creative, informed and effective.

* Although you could argue, and Richard Lanham does argue in his The Economics of Attention, that the competition for attention is the best available filter, a la Hayek. I think this is probably true where cooperation is not available, either because not enough value is being created or for some other reason. These reason don't apply here.