Rethinking Reed's Law

I posted last month on Reed's Law, the notion that Group Forming Networks (GFNs) create value more quickly than regular networks.  It makes sense to me that a network whose nodes themselves are networks (think MySpace, Facebook, social networking in general) would increase in value more quickly than traditional networks (think AT&T).

My thinking on this has been greatly influenced by Umair at Bubblegeneration and Nivi, both of whom are younger and smarter than me and thinking hard about this stuff.

But it's not just the young guys in the business who are thinking hard about Reed's Law.  Tim Oren has been busy rethinking Reed's Law on his blog.

And yesterday, Tom Evslin came out and said that he thinks Reed's math is wrong.

So, to summarize, Reed is right that GFNs set up the potential to create value much more rapidly than traditional networks.  How "much more quickly" this happens is the subject of some healthy debate.  And whether there is a natural limit to how far this goes is another subject of some discussion.

So last night, after a nice event with the Tacoda engineering and operations team, we got talking about Reed's Law. Joe Wilson, who has been working in the Internet arena for a long time, suggested to me that the value creation equation needs to also take into account the "monetization ability" of each of these GFNs.

Tacoda specializes in helping people who operate online properties "monetize" them. So it wasn't surprising to hear Joe talk that way.

I guess the bottom line for me on this stuff is that its important to understand the underlying fundamentals of online businesses, but getting the math exactly right may not be the most productive use of time. I've met plenty of great entrepreneurs over the years who know that revenues minus costs equals profits, but have no idea what a differential equation is.  And it never got in their way.

Comments

[Charlie Crystle]

In Reed's Law, or Evslin's take on it, value does not equal $ or potential revenue. So in Joe Wilson's Monetized Reed's law you would add revenue, cost, etc, and that revenue would be have to be reflected per node over time, and perhaps lines between nodes would reflect economic activity of those connections including cost, etc etc, all stuff I bet eBay has modeled sufficiently beyond the single commerce strand of auction transactions. If monetary value is the basis of the model, then perhaps the $ value is represented by the size or color of the node. The math itself is beyond my English major mind, but I can picture a 3-d, not 2-d model of this whter the third dimension is...what?

[Charlie Crystle]

Tim Oren's great post on this.

[David Gibbons]

Good Points Charlie;

Obviously, the "other dimensions" yield some function of "adoption" that limits any network actually reaching Reed's potential.

I think that Reed is more important now though because NOW that users are producing online, those limiting factors are far less relevant. In online production, the limiting factor of "competitive channels" which limits potential demand online, has far less impact.

The reason is that we've never had to consume "efficiently" - but that doen't apply to our production - when we produce, we have to use the best possible solution, or be beat out by competition - so, a "good solution" for producers is far more likely to obey Reed than a good solution for consumers will. Simplistically, this is why "all" small-sellers are on eBay, but consumers don't buy everything on eBay.

Web2 is about production. It will more closely obey Reed than Web1 did. You need to know what type of site you're dealing with - and apply Reed carefully only to the supply-side of a viral business model. The demand-side, as we learnt in Web1.0, is just as fickle and unpredictable as it always was.

[DeusExLibris (Joe Wilson)]

My comments to Fred were intended to stay a level of detail above what Charlie is talking about. While a model for a given business is clearly very useful to that business, it is often too specific to allow you to draw generalizations about a particular business sector. What I am suggesting is that the "ability to monetize" a given node is as important a component of a network's utility/value as the number of nodes.

To borrow from Tim's post (thanks for the link, Charlie!), I was trying to get at a utility function (i.e., Reed's Law, Metcalfe's Law, Sarnoff's Law, etc) that was not monotonic. For example, how useful would aFriendster, Tribe.net or LinkedIn be if all you knew about other users was their username? Not very. It is the data that you share among nodes that adds the incremental value (interests, employment history, demographic data), not simply the ability to form groups. In other words, it isn't just who you know, it's what you know about them that determines the utility of a network.

I agree with Tom's post that there is something intuitively wrong with Reed's Law. Unless you have a counterbalancing negative power function (2^-M, M approx equal to N), the exponential rapidly overwhelms all other terms and leads to ridiculous valuation excess. No linear or quadratic dampening terms have impact beyond a few iterations of N.

[Frank Ruscica]

GFN-wise, the big money opportunity lies in orchestrating a process network for customized education and career services. Such a network (a.k.a. workflow market) will maximize the aggregate productivity of all GFNs, by minimizing the lag between 1) the emergence of an opportunity for intra-/inter-GFN wealth creation, and 2) the "emergence" of sufficient numbers of GFNers with the requisite skill sets to fully "mine" the opportunity ASAP.

[Robert Young]

Fred,

FWIW, at least to me, the key importance of evaluating any venture in the context of such "laws" relates ultimately to analyzing barriers of entry. In other words, with switching costs becoming less of a competitive feature, the ability to attain networked-math growth to critical mass of users becomes a more reliable form of barrier to entry... thus impacting the value equation. Skype and MySpace are solid examples.

[David Armstrong]

Fred...my reaction to the archived post is way late....I hesitate. But you continue to fuel our fire. So here goes...

I feel like I've been close, a keen observer, to viral marketing, the real kind, ever since Hotmail was the "golden child" of it. I've paid attention to the effects of it, not necessarily the "equations". I've done some modeling on a concept of peer utility, which incorporate some of the math related to Metcalf's law (skype) and reed's law (youtube). We have built a real business around the socializing of content and products. I'd love to talk to you and commit to not wasting your time.