I’ve been notified by the Association of Statistics and Society* that, as a web metrics analyst, I’m required to comment on Chris Anderson and Michael Wolff’s article in Wired, “The Web is Dead. Long Live the Internet” (also covered by NPR). So, my 2kb:
Is the Web Really Dead? Uh, no. As any Communications 101 student knows, it’s very fashionable to herald in a new medium by declaring all existing media “dead”. When YouTube happened, they declared television dead. When television happened, they declared cinema and radio dead. Is radio dead? I look around this office, I look around my house, I look at the stereo in every car on the road and I say “no.” Now, has the content of radio changed since television happened? With the exception of 7-11pm on Sunday nights on WAMU, yes. Has the device on which I listen to the radio changed? Yes. Has the pleasure of listening to a velvet-voiced stranger tell me fascinating stories changed even a little? No.
So, Anderson is claiming that users are shifting from desktop Web browsers to mobile phone applications and, therefore, the Web is “dead.” The former High School Debate State Champion** in me wants to take this point by point, but we don’t have time for that. Instead, I’m just going to pounce on the largest flaw in their argument, mostly because:
Anderson and Wolff make the Number One Most Common Rookie Analyst Mistake. They confuse relative growth with absolute growth. Anderson and Wolff have a pretty compelling graph at the top of their article. It shows proportional Web traffic peaking in 2000 and dropping steeply after that. The problem with this graph? It’s showing RELATIVE traffic and thus obscuring the fact that ALL Internet-based channels have grown dramatically since 1990 and continue to grow, including the Web.
Look at our own graph of unique visitors by platform below. The green parts are, in Anderson and Wolff’s terms, “the Web.” See how they’re going up and to the right? Now, if I had made my graph the way they made theirs, you would see the green parts take a nosedive around July of 2009 simply because that’s when we started adding new platforms.
This conflation between relative and absolute growth ends up implying that Web traffic is actually decreasing in absolute terms, an implication that even snuck into our own headline: “The Web Is A 'Shrinking Minority' Of Internet Traffic.” Yes, Anderson’s addition of the word “minority” technically gets us back into relative-land, but I still walked away with the impression that Web traffic is actually shrinking – definitely not true.
Is the Web dead for NPR? Again, my exhibit A. In the last year, unique visitor audience to our main Web site have grown by 30%. If you throw in traffic to the mobile Web site, it’s grown 37%. By comparison, our audience for our mobile apps have grown from nothing in July of 2009 to more than a million in July 2010. So, the mobile apps audience isn’t peanuts, and it’s growing much faster than our main site, but it’s not going to be the majority of our audience overnight. Also, the fact that both Web and non-Web traffic is still growing shows us that people aren’t SHIFTING from Web to non-Web usage in droves. Rather, they seem to be using both.
And another thing… So, I’m pretty passionate about refuting Anderson and Wolff’s claims because I fear that, in their efforts to sensationalize these already fascinating technological developments, they actually end up confusing and intimidating you into doubting your own common sense. Do YOU think the Web is dead? Are you using your browser on your desktop less now than you were ten years ago? I’m not. And I’d bet my battered old MacBook Pro that the vast majority of people who read their article read it on...wait for it…the Web.***
* Not a real association.
** Still strangely proud of that. Sorry.
*** I’d also bet my September issue of Vogue that their article’s second biggest audience was from that other dead medium, the print magazine.