Mentions of "Web 3.0" in the blogosphere reached an all-time high in October 2007, and has since then not recovered. This could mean that the Web 3.0 hype is over (and suggests that maybe I should change the tagline of this blog), or perhaps are we only witnessing a temporary backlash in the interest of next Web.
Since this blog officially is about "tracking Web 3.0", I thought I should write a couple of posts on this theme. Instead of making my own predictions of what the next Web would be like, I will look at what others have said about the future.
In this post, I present data on the number of blog posts per month that mention the term "Web 3.0", from the first mention I found in October 2004 to present time. In a forthcoming post I will look more closely at the actual definitions of Web 3.0 that people have used in the past.
The tool I used to gather data about the number of posts per month was Google Blog Search, which has an option to limit the search between given dates. Notably, I also turned off SafeSearch filtering (I only came across a handful of posts with suspicious content), and I didn't count posts that were reported by Google Blog Search as "similar to the already displayed". The number of posts reported as similar were generally around 25 %, but could amount up to 50 % for selected date ranges. Not counting similar posts accounts to some degree for posts by spam blogs, which just repost original content.
October 2004 to June 2006
The first chart covers the period October 2004 to June 2006, and shows a single post mentioning Web 3.0 in October 2004, which at least according to Google Blog Search is the first time this term is used. Incidentally, the post in question was a transcript of an interview with VC John Doerr, made at the very first Web 2.0 conference in October 2004.
After the first mention of Web 3.0 in October 2004, there is a quiet period of nearly half a year, until a few mentions emerge in April/May 2005. A notable post from that period is by Dan Gillmor, who made one of the first predictions of Web 3.0. Gillmor mentions APIs, web services and the web as an operating system.
In the fall of 2005, there is an increased interest in Web 3.0. An important post contributing to this interest is from August 2005 by Jason Kottke. Jason talks about a WebOS, and suggests "Web 3.0" as an alternative title for the post. Other notable posts from this period include Nova Spivack, who in October 2005 mentions Web 3.0 in a post about a "World Wide Database", and Phil Wainewright, who in November 2005 maps out a three-layered topology of Web 3.0.
The adoption of Web 3.0 in the long tail of the blogosphere, roughly a year after the first mention, clearly increases the "noise level" in the chart. A contributing factor to this increased noise is, apart from posts by prominent bloggers, most likely also the second annual Web 2.0 conference, which was held in October 2005.
A clear spike in the blog post mentions of Web 3.0 happens in January 2006. This is due to (mostly long-tail) comments, on a post entitled "Web 3.0" by web design guru Jeffrey Zeldman, who concludes his post with the words: "As for me, I’m cutting out the middleman and jumping right to Web 3.0. Why wait?" The middleman being Web 2.0, supposedly.
July 2006 to March 2008
The second chart covers the period from July 2006 to March 2008, and takes off in the relatively quiet period between January and November 2006. This period is notably absent from A-list bloggers talking about Web 3.0. This changed drastically in November 2006, when an article in The New York Times by John Markoff started the first major Web 3.0 debate in the blogosphere. Markoff refers to Web 3.0 as the "semantic Web" and frequently also refers to artificial intelligence when describing the next Web. There were lot of high profile negative blog critique in response to Markoff's article, e.g. by Dave Winer and Robert Scoble, whose main objection was the use of the expression "Web 3.0" as something established, though the expression had not very frequently been used in the past. More constructive commentary was given by Dan Farber, who agreed to the association of Web 3.0 with the Semantic Web.
After the spike in November 2006, the mentions of Web 3.0 gradually increase and remain at a fairly high level throughout 2007. Compared to previous years, Web 3.0 was a much more common subject among the top-tier blogs in 2007, and most A-listers had some post mentioning Web 3.0. I will have a more detailed look at some of these in a companion post. A subject mentioned in the top tier of blogs will naturally percolate and amplify through the long tail, and shows up as an overall increased frequency in the chart.
There is a small spike in August 2007, with just over 800 mentions of Web 3.0. This can be attributed to the delayed coverage of Google CEO Eric Schmidt's definition of Web 3.0, which he gave at the Seoul Digital Forum conference in May 2007. Schmidt actually talks about the future applications, which are "pieced together". Further, to quote RWW, Schmidt says: "The apps are relatively small, the data is in the cloud, the apps can run on any device (PC or mobile), the apps are very fast and very customizable, and are distributed virally (social networks, email, etc)."
The All-Time High
The second major Web 3.0 debate in the blogosphere was initiated beginning of October 2007 by a post by Jason Calacanis entitled "Web 3.0, the 'official' definition". This post contributed to an all-time high of over 1000 mentions of Web 3.0 this month. Calacanis definition is quite unique, e.g. he writes: "Web 3.0 is defined as the creation of high-quality content and services produced by gifted individuals." Fred Wilson responds and calls it "Web 3.0 Nonsense" and writes: "Jason Calacanis defines web 3.0 as Mahalo.com, predictably" (Mahalo being Calacanis startup), and most people seems to agree with Fred.
The Backlash and the Future
Following the peak in October 2007 there is a backlash, and no major Web 3.0 controversy has since then followed. Despite this relatively low interest in Web 3.0 among the leading blogs, there have been a gradual increase of Web 3.0 mentions in the long tail in the first three months of 2008. This could be explained by the fact that the expression has matured and reached a critical mass, where it is able to sustain its own growth without injection from the top tier of blogs. Other reasons could of course be a general increase in the number of blog posts, and unfortunately also an increased number of spamming blogs, aka splogs. To account for the latter two phenomena, one could measure the frequency of a neutral or "time-invariant" word like "web", and normalize the current numbers with that frequency. In any case, it will be interesting to follow the evolution of Web 3.0 in the future. For example, will there be a new peak the coming fall?