Imposing a Metcalfe Tax
November 11, 2010
Bob Metcalfe's most famous pronouncement is known these days as Metcalfe's Law, the idea that the value of a network increases with the scale of its number of members. Metcalfe formulated this concept in the context of telecommunications networks, but it is broadly applied to online social networks as well, and taken almost as an article of faith when designing new networked systems.
What happens, though, if a user on a service breaks the connection the network? A member of the network could do this either by disconnecting from the network (temporarily or permanently) or by choosing to omit or falsify some of the metadata that connects them to the network.
Why would someone make the network less pervasive or reliable? Well, there are lots of reasons. danah has admirably outlined some of the social reasons that teens disconnect, and of course I've been rambling for a while about the fact that unexpected expansions of existing networks often negatively impact those with less social privilege. But there are lots of other reasons people would want to unplug from networks that are smaller, or less focused on personal data, than services like Facebook. On GitHub, while forking is a feature, some who choose to collaborate on code there might be unwilling or legally unable to allow certain types of collaboration, and may want to sever their work's connection to the network of collaborators there while still using the robust technical capabilities of the service. On systems with tag-based navigation like Flickr and Delicious, tagging provides personal utility for information that a user wants to keep private, but the available privacy settings necessarily sever that photo or link's connection to the rest of the network that shares its tags.
In short, there are legitimate reasons users would want to contradict a service provider's desire to maximize the network effects of their community. These reasons primarily center around privacy, control, security, policy and other social considerations that provide important justifications for resisting the pull of Metcalfe's Law. But current network providers are often tempted to disable, circumvent, or modify the terms of a user's connection to the network in order to satisfy the provider's business goals, resulting in a sort of "connectivity creep" as we see today in Facebook's ever-evolving privacy policies, but as we'll inevitably see across a large number of network-focused sites. Right now network providers have a disincentive to give users the power to sever their connection to the network, even in limited ways, because it could jeopardize the growth of the network itself.
Maybe there's a way to reconcile this tension.
A Metcalfe Tax
Today, the low-level layers of our communications networks are ideally designed to see breakage as damage and route around it. But perhaps the social layers of our networks should see breakage as opportunity, and build revenue models around it.
There's a model for this already: GitHub. When a user wants to use a non-open source license, or keep their code private, they can pay for the ability to do so. This is important, because using a closed license breaks the network effects that open source licenses are designed to enable around code modification, and obviously keeping code private breaks the network effect of being able to see the history of a project or contributor.
Could other services adopt this model? Let's consider the biggest of them all: Facebook. If Facebook simply charged a nominal fee (it could be different prices in different parts of the world) to give users full, rich control over the ways they are connected to and discoverable by the network, how would things be different? First, Facebook would directly profit from users who wanted control over their presence, turning them from a thorn in Facebook's side into their best customers. Second, Facebook would have a bottom-line motivation to keep making better, more usable, more powerful privacy controls in order to serve them, instead of often treating privacy features as an afterthought. Third, the users who value their privacy or control would be more likely to advocate the service to their friends instead of warning them away, since they would have their primary concern addressed. And finally, the places where a user's information or identity is shared on the network would become much more clear and predictable if some parts of those interfaces were designated as paid-only features.
Now, "payment" in this case doesn't have to just mean dollars. In our newly gameplay-obsessed industry, maybe some networks would allow users to pay for these privacy features by completing tasks or performing functions that were important or valuable to the site. But overall, the process would be the same: The site would benefit from users who have legitimate reasons to not stay completely plugged in to the network all the time. And a precedent would be set for having the site be rewarded for honoring those desires, instead of constantly pushing against them.
So, which site's going to be the next to build a Metcalfe Tax into its business model? Seems like one of those rare delightful moments where a new site's business model could be inherently aligned with the desires of its users, instead of working in tension with them as the network grows.