The Wired cover story on “The Web is Dead? A Debate” has in fact caused an intense debate of the question, taking place – of course – on the web.  It may be seen as somewhat ironic that on just the following day we get two pieces of news from Google that would point in the other direction: a demo of the future of browser games and a Chrome OS tablet together with Verizon (aside: sure didn’t take long after their joint pronouncement on net neutrality).

Much has been made of the potential shift of attention from the totally open web to more proprietary platforms on mobile devices.  But there is another question that I have been thinking about (in fact have a draft of a long post on this question for the USV blog):  how well does completely open work on the web?  Or put differently: are there limits to decentralization on the web?

At the risk of stealing my own thunder, here is what prompted my thinking about this question.

• Almost all the knowledge that is on Wikipedia also exists on the web at large distributed across millions of sites.  So what is it that Wikipedia adds?
  
  

• Small merchants could set up their own web sites rather easily these days and sell directly.  So what is it that Amazon adds?

Clearly both Amazon and Wikipedia are not the open web.  They represent some form of centralization and some limits.  That centralization and those limits must be adding value at least for now.  Here are the benefits that I have been able to identify:

Reputation: Buyers trust Amazon’s service.  They don’t need to make up their mind on each separate merchant’s page about whether or not it will provide a good shopping experience.  Similarly, readers trust Wikipedia’s mechanisms to catch errors relatively quickly.  Certainly way more than reading some random individual’s web page.

Convenience: Having credit cards on file together with a unified check out allows people to buy from multiple merchants in one go on Amazon.  On Wikipedia all the information on say an individual is pulled into one place that might otherwise live across a dozen different sites.

Discovery: Based on its knowledge of a large number of purchases, Amazon powers a discovery process by suggesting related or alternative products.  Wikipedia doesn’t do much of this, but there is some discovery of new topics and or related items that one might not have thought of oneself.

Uniformity: Every product page on Amazon looks the same, making it easy for someone looking for something to browse quickly through a large number of items and know exactly what to look for.  Ditto for Wikipedia.  I believe that is much more important than people think.  The cognitive overhead associated from switching between pages that look vastly different is huge (in fact, I believe it played an important role in the ascent of Facebook over MySpace).

Granularity: In Wikipedia the smallest contribution item is a single letter, as in the case when fixing a typo.  On the open Web, you’d have to publish an entire page.  Not sure whether this applies to Amazon also.

Convergence: Even though each Wikipedia page has an extensive history, at any one point in time, there is only a single page.  On the open web there will easily be multiple copies with slight differences of the same content.  Again, not sure how this applies to Amazon.

What is interesting (and the subject of the longer post I am working on) is that these benefits from some centralization are difficult to overcome with better technology.  At their heart, they are not technology problems but incentive problems.  And incentives are difficult to get right.