As you might know from our posts, a lot of the research work that Tim and I do relates to the role of networks in transferring knowledge within organizations. When the right knowledge goes to the right place, the result can be a change of products or processes – in other words, innovation. When we work with larger organizations it becomes apparent that transferring knowledge becomes harder because people start to lose track of who-knows-what and who-knows-who. The result tends to be that these organizations start to face declining returns to scale as learning and innovation become harder.

In Bill’s talk he brings up a slide of declining returns to scale in the Japanese beer industry. This isn’t surprising because we know that all industries work like this. As the industry produces more product over time, it gets harder to find more efficiencies to drive the unit cost down.

Most activities incur declining returns to effort too- even simple tasks like a computer game such as Tetris. As we expend more effort, the marginal improvements we make diminish. Note this this issue of diminishing returns is different from the diminishing returns to organizational size but it relates to my next point.

But while there is a general principle of declining returns to scale and effort, Bill showed a curve generated from the computer game, World of Warcraft, which shows increasing returns to effort and scale. As players put more time into the game, their rate of progress increases rather than decreases. So what is going on here?

I haven’t played World of Warcraft but a big part of it is that it’s a virtual community of gamers. As people get more familiar with the game they become part of the network that enables the exchange of expertise. The bigger the network, the more expertise. This increasing returns phenomenon is a network effect and it is the presence of the network that means WoW players get increasing returns to effort and Tetris players get declining returns.

I think that there is a lot of value in trying to set up these virtual networks to get around the problem of large organizations, and to take advantage of untapped expertise in the population, but I also think that there are going to be some limiting conditions to the effectiveness of these networks.

WoW seems to work well because it is a very focused community with its own language. This means that everyone understands most things that are being talked about. In many organizations that we do our network research on, there are many communities of specializations, which results in a fragmentation of the network.

Another issue is that most of the knowledge involved in WoW is codifiable. We can put the information into words and numbers and the receiver will be able to make sense of it and learn. In organizations, a lot of knowledge is tacit and experiential, which limits what can be achieved in virtual space.

These caveats aside, the prospect of increasing returns to scale through the digital innovations that are being brought into organizations are tantalizing. The “rules” of learning and organization may yet be totally re-written.