Is the web really dead?

Wired uses this graph to illustrate Chris Anderson and Michael Wolff's claim that the world wide web is "dead." ff_webrip_chart2.jpg Their feature, The Web is Dead. Long Live the Internet, is live at Wired's own website. Without commenting on the article's argument, I nonetheless found this graph immediately suspect, because it doesn't account for the increase in internet traffic over the same period. The use of proportion of the total as the vertical axis instead of the actual total is a interesting editorial choice.
You can probably guess that total use increases so rapidly that the web is not declining at all. Perhaps you have something like this in mind: graph2.jpg In fact, between 1995 and 2006, the total amount of web traffic went from about 10 terabytes a month to 1,000,000 terabytes (or 1 exabyte). According to Cisco, the same source Wired used for its projections, total internet traffic rose then from about 1 exabyte to 7 exabytes between 2005 and 2010. So with actual total traffic as the vertical axis, the graph would look more like this. 3.jpg Clearly on its last legs! Assuming that this crudely renormalized graph is at all accurate, it doesn't even seem to be the case that the web's ongoing growth has slowed. It's rather been joined by even more explosive growth in file-sharing and video, which is often embedded in the web in any case. Update: It's also worth adding that bandwidth, though an interesting measure of the internet's growth, isn't so good for measuring consumption. It doesn't map to time spent, work done, money invested, wealth yielded... Does 50MB of YouTube kitteh represent more meaningful growth than a 5MB Wired feature? And, as others point out in the comments, many of the new trends are still reliant on the web to work, especially social networking.


  1. You make Tufte and Few proud! I thought the people at Wired worshiped at the altar of Tufte (as should everyone!). Boo, Wired. They know… should know… better than that.

      1. Well of course DNS is obsolete. We all use host tables now, downloaded from, don’t you?

    1. Alan, Email is definitely not extinct. Even with the increase of IM, email in the corporate world is anything but booming. Email marketing is on the increase, just as the above graphs represent.

      I would love to see any type of data you have to back up your claims.

  2. I don’t know about other people, but I usually use the web to access and find the streaming video listed in their graph. So yes, the type of data is different, but I don’t know if it should be broken out like that. I mean, if you load a YouTube video, shouldn’t all the data received because of that webpage be considered part of “the web”?

  3. I find it interesting that DNS appears at all on the graph. DNS is required to operate pretty much all of these types of traffic, but it uses very little bandwidth. That’s why it appears early on in the first graph, then vaporizes almost immediately as total traffic grows. Total web traffic in the early 90s must have been very small for it to show up at all.

    1. The web was entirely text based in the early 90’s, Mosaic released in 1993.. which was the first browser to support images.

      I can’t guess what the average size of a web page was back then, but pages over 100k in size were probably rare. These days anything under 100k is unimaginable.

      1. No, Gopher supported image formats quite nicely, thank you. I was there, I even ran a gopher server myself back when the Mama Gopher was still a desktop mac. I think it was an SE-30.

        The web didn’t bring images. The web brought a simpler interface, so less technically skilled people could get online. Sadly the interface is SO simple that the stupid people got online too… I kid, I kid! We’re all stupid.

    2. DNS didn’t disappear from the graph because total traffic is growing,it disappeared because the size of transactions increased. When you use 100 bytes of DNS traffic to sent a 500 byte email, DNS is a noticeable percentage. When you use 100 bytes of DNS traffic to transfer a 1GB movie, the DNS traffic is inconsequential

  4. So… are they claiming that because video and peer-to-peer traffic is increasing, then the “web” is dead?

    Because if the “web” is Telnet, FTP, and Newsgroups, I’m ok with that.

    1. Yeah, ooh, Telnet use has fallen to zero. However, if they’d called out ssh (perhaps that’s lumped in w/ “other”?), my use alone would show a spike higher than P2P’s.

  5. I think if anything, the phrase “World Wide Web” is fading a bit from common use, even though it obviously retains its technical meaning.

  6. Big fan of Chris Anderson. His research and writing is a reminder that there remain some really good journalists out there.

  7. Yep, looks like another case of misrepresenting statistics to advance a false argument. Proportions are always dangerous, and bandwidth is totally irrelevant to importance or the “life” of a medium. An email, text, tweet, etc. all put together may take up less than one one thousandth the bandwidth of a youtube video — that doesn’t mean that video is one thousand times as important or alive.

    Many people are just determined to do and say stupid things. It is not easy to stop them because they often know what they are saying is stupid, they just don’t care…

  8. If they are talking about traffic in MBs, then video of course takes up more traffic space… yes?

    Is this “traffic” in what value? Bandwidth?

  9. Please don’t show the Wired chart to Edward Tufte. He’s one of my favorite people and it would be a loss to society if his head exploded.

  10. That’s another issue… bandwidth. What is the ‘effectiveness’ of video vs email? Granted, video uses WAY more bandwidth. ‘web’ lies in between. Perhaps ‘hits’ would be a better stat to use. One video = one web site = one email… needless to say, with a fair bit of tweaking, as it’s not always tru that one video = on wesite.

    1. They do have a fantastic track record on predicting the future of the Internet at Wired.

      I’ve often thought that one could have made a lot of money throughout the Internet age by consistently betting against their predictions in the stock market. They’ve been wrong about everything, much more profoundly than could be attributed to chance alone.

    2. Well if you replace ActiveX/Java with “flash” the predictions in that article aren’t *that* bad.

  11. Tim Berners-Lee here.

    I’m sorry. We lost. Now everyone get out. They need to build a big Gopher machine here. Sorry.

  12. I’m sorry, but this post is just flat-out-wrong, and is an example of what the original poster is complaining about. Of course web traffic is increasing, usenet traffic is increasing as well – ALL NETWORK TRAFFIC IS INCREASING-. However, if you want to actual show whether it’s becoming more or less relevant, you have to compare it to other data sources, which are also increasing. So yes, relative to all other types of data on the traffic, port 80 traffic is going down.

    Now, a more germane question is whether that matters. Web traffic is text and graphics, and is relatively low-bandwidth compared to video. Streaming video takes up a lot of bandwidth, while even a heavy-text page will clock in at a hundred kB max. So there is a mistake, but the representation of the graph isn’t the problem.

  13. Also, a general tip for reading journamalism like this… *Any* time an author makes a “bold” claim that such-and-such is “dead,” they are wrong. They are *always* wrong about this. I’ve read about more supposed Google-Killers than I can remember. When an institution does enter a decline (like newspapers), people know it, and articles about it don’t have the same sensationalist appeal. These articles read like a mad-lib:

    “[Popular, useful institution] is dead because of [fad no one actually uses OR new institution that is perfectly capable of coexisting with old institution]. What will it mean for [unrelated phenomenon]? How will you live in the [soon-to-be dated buzzword for the new times we are supposedly entering]? Here is the story of [marginally topical anecdote] to prove our point…”

  14. Am I correct when I say that most networked mobile phone apps use the Web to connect those sweet APIs?

  15. Very disappointing from Wired. Just because they cant sell their magazine printed on dead trees anymore, they try and twist facts?

    As others point out above, 95% of the video traffic mentioned above is found on Web pages.

  16. I agree, that’s awfully misleading, even if technically true. A 10min youtube clip takes up more bandwidth than a large library of books, which could take a lifetime to read. I can spend 12 solid hours a day reading text on the web, or I could download a couple 3 minute songs, and the traffic would end up the same in the end. I spend far, far, far more time reading text on the web than I do downloading video, I watch maybe a half hour worth of video per week, but I spend at least 8 hours a day every day surfing the web. I suspect though, that half hour of video consumes most of my bandwidth though.

  17. so is this proportion of total data or proportion of requests? the latter would be a bigger deal than the former as video is just inherently much, much bigger. but as a proportion of requests it is a bigger deal.

    also, while the statistics do seem misrepresented, and the headline is stupidly sensational as headlines are wont – the underyling idea that a lot of internet traffic is moving away from PC+Browser to a variety of other kinds of devices and presentation modes is totally accurate.

  18. I’d like to point out that the basic point of Wired’s article — that non-web internet use is seriously hot stuff now that mobile apps have matured — is very timely and sound. The problem is in trying to illustrate this as a battle between web vs. not-web, where it’s tempting to illustrate the relative loss of ‘market share’ as a dramatic decline.

    But when that market is exploding so rapidly, all it means is that the web is growing appropriately according to the needs of an ecosystem it largely facilitated. The web is often the access point for these new goodies anyway.

  19. Can someone explain clearly the difference between the “web” and the “internet”? I thought they were/are the same thing….

    1. I’ll field this question if nobody else will.

      The internet is a massive network that supports a wide variety of traffic. This includes instant message traffic, peer to peer traffic, bittorrent traffic, and of course the world wide web, among many others. These different types of traffic follow different “protocols”, which are essentially complex rule sets dictating how programs will communicate over the internet. The HTTP protocol is protocol that supports the web.

      The Web, on the other hand, is the webpages you view in your browser; a single protocol. Instant messages and file transfers and video are not part of the web.

      One good analogy i’ve heard is that the internet is like a railroad, and the web is one of the many types of train cars that run on it.

  20. Ridiculous claims like this aren’t isolated to tech journalism. It seems to be a case of the “everyone is like me” fallacy that runs rampant through society.

    I’ve met people who have sincerely stated in conversation, “yeah, but everyone has an iPhone.”

    Just because some people are the bleeding edge of finding new ways to use technology doesn’t mean that the greater, slow-moving populace isn’t still using the older conventions.

    Yeah, the web is dead. Tell that to all the cubicle monkeys in Corporate America reading these articles using Internet Explorer 6 on their CRT monitors. Their out-of-touch managers are still printing emails and have only now started playing Farmville.

  21. Something else that completely invalidates their argument is that you could spend 3 minutes watching a hi-def movie trailer and then 30 minutes reading material on the web and yet the trailer consumed more bandwidth. Also in the past 10 years online video has increased in bandwidth size as technology has improved and bandwidth increased. We used to watch video the size of postage stamps that was horribly pixelated and now we are watching hi-def video that fills the screen and looks incredibly good. A site like, might be using more images these days, but hasn’t increased in size of bandwidth over the years compared to video.

    So basically bandwidth is just a horrible measurement for how the web is used and whether it’s alive or dead.

  22. Another thing to bear in mind is that bandwidth is a poor measure of consumption. It gets cited because its easily quantifiable. A feature like Wired’s, for example, represents a few MB of traffic for each ‘view’. Some random youtube clip might represent dozens.

    But it’s not the case that the youtube movie’s ‘view’ represents more growth than the feature, even though it uses far more bandwidth. 1MB of the web might represent a greater input of work, resources and time than 25MB of video, and likewise may yield more value. It might take much longer to consume, too.

    A good illustration might be to compare the bandwidth of a movie to a home rental. We wouldn’t think of doing this normally, because the media aren’t (historically) measured in quantifiable digital terms.

    It’s natural to think of seeing a movie in a theater as a better experience than at home, but fundamentally they’re the same ‘unit’ of cultural consumption and all that. But whereas a movie projection might represent several terabytes of digital-equivalent ‘bandwidth’ the same movie on VHS tape would represent less than a single gigabyte.

    Unique visits/requests and so on are a better indicator, but even then–these metrics are the ones relevant to web advertising–it remains just a single data point in an entire book of black magic.

  23. I’m wondering what some of those small disappearing services look like in actual bandwidth. Especially telnet. I was just using telnet the other day, but I’ll bet that one has actually gone down.

  24. Also, they’ve made some serious category errors. Video is a media format, while web (presumably http) and peer-to-peer are transport mechanisms. You could draw a graph of video vs. photos vs. text, or you could draw a graph of http vs. p2p vs. ftp etc. But you can’t mix and match across categories.

    How many ways is it possible to fail in a single chart?

    1. This sounds like something that should be published in The Journal Of Irreproducible Results. (obscure science humor magazine)

  25. My head exploded in the first paragraph, right at “During breakfast you browse Facebook, Twitter, and The New York Times — three more apps.”

    If you’re going to characterize a bunch of stuff that rides on top of HTTP as other than web traffic, then you’re clearly aiming for some kind of No True Scotsman argument.

  26. The web’s power comes from the humble link. The fact that you use Tweetdeck to read Twitter doesn’t change the fact that you follow all those links in tweets to plain old web pages.

    Maybe if apps started launching other apps I might believe the web was at risk, but apps really just supplement the web, not replace it.

  27. The killer app on the iPad is not the NYTimes’s. It’s not Wired’s.

    It’s that darn Flipboard, an app full of mucky user generated content. Yuck.

  28. Are Wired’s overblown headlines still credible? See the July 2008 issue declaring “The End of Science”, a method which was no longer needed since we can find truths more easily by throwing more and more people at every question — using technologies like the web.

  29. I demand to know why the google wave protocol isn’t featured on this chart. This is clearly a Jobs-Wired conspiracy to downplay the success of google wave!

  30. Anderson frequently practices this kind of cargo cult scientism, presenting outrageous claims with lots of convincing-looking graphs in primary colours that appeal to autodidacts. His “Long Tail” assertion became had just enough truthiness to become an article of faith during the last decade, even as study after study using actual market/sales data refuted his assertion as to its numerics. The work by Elberse, Netessine, Tan, Page and Bud clearly shows a real-world sales volume more like a log-normal than a power law distribution. The difference between Anderson’s assertions and the actual work using real sales data is striking. It’s possible that Anderson allowed himself to be misled by some high-volume Internet retailers in the hope that their competitors would divert valuable resources into pursuing the chimerical long tail while the retailers cleaned up at the left-hand-side of the curve.

  31. The original graph would be a great critical thinking exercise for middle school kids.

    What is sad is that fundamentally flawed statistics and graphs like this one are constantly used in politics / policymaking.

  32. So web traffic and all other modern traffic is still exponentially rising? None are in any danger of “dying” anytime soon, at least not without some unforseeable event occurring in the future to cause it. I know bb is a very hi-brow crowd, but do any of us even know another person in our lives that would buy into such a ridiculously skewed & spun interpretations of statistics such as these?

  33. This type of journalism comes straight from the “insert brand or product name here”-killer school of thought.

    There seems to be this idea that everything is a zero-sum game, something must always win, something must always be killed.

    It’s simply poor journalism trolling for pageviews.

    The iPhone doesn’t need to be killed, Google doesn’t need to be killed, Facebook doesn’t need to be killed. What is needed is open, fair competition with lots of competitive choices, which isn’t the kind of story lazy journalists like to write.

  34. Let me tell you what the problem with this type of article is. Sure, you are getting some hits today. You feel that the laxity in your editorial standards was warranted by this measure of success.

    But I’ll give you my case as an example. Not an avid reader of Wired, but once in a long while I check you up. Had a neutral view of your brand and product. I remember reading some interesting things in here. If I had to push my somewhat faint impression of Wired, I’d say that your articles typically don’t strain my intelligence, but sometimes, they are mildly interesting.

    I came here through Daring Fireball today. Read a couple of articles panning this article, and another putting it into a context of past disgraces similar to this one (which I wasn’t aware).

    So, great. You got me to come to this page three times today. But, in my case at least, your brand is shot. Hard to recover from that. Really hard. Chances I’ll visit Wired in the future instantly halved at least.

    That’s a really bad deal for you.

    PS: By the way, the same self-destructive mechanism of short-term gain build upon pooping on your own brand is really prevalent in most media. Sad.

    1. Dear Anon @ #53. This is BoingBoing. Not Wired. Suggest copy and paste your views on THEIR comments page.

  35. Looking at my own bandwidth monitors, I see that email traffic is gigantic, as is virus and worm chatter. Apparently they filtered out the spam and malware before they built the chart, making it even more misleading and useless.

    My captcha is “Colleen’s theStank”. Wtf?

  36. THis is just another “everything is different” “sky is falling” non-story.

    Breaking out video as somehow seperate from the WWW is just stupid. youtube has a www at the beginning. where is this land of online video that I don’t access through a www. Netflix on my xbox?

    we were never accessing the www on our TV’s or while sitting on the toilet, so the fact that direct to TV services and mobile now have a market isn’t a loss for the interwebs, its a loss for broadcast TV and magazines.

  37. I’d like to see (hypothetically) how the graph stacks up in terms of actual time spent viewing/reading/consuming in addition to just by bandwidth – when I find something really interesting it’s usually a simple web page that I spend several minutes reading (or hours – the list of 100 greatest magazine articles is a perfect example) while the same amount of bandwidth in video is (I’m guessing here) much shorter than a few minutes.

    Furthermore, if you had to rank or value every single bit (say, if you were still on dialup), do you think videos, per byte, would end up anywhere close to the top of the rankings? Sure, some would, but I’d like to think I’m correct in assuming that the vast majority of video on the internet is crap – although that doesn’t mean that a similar proportion of the video *consumed* from the internet is also crap (Though I’m sure 4chan has skewed that statistic forever for numerous cases – then again, someone’s getting a kick out of rickrolling people – there’s value in that too, I think..).

  38. An article like this angers me because they could have at least made an interesting, though unsensational point. The fact that we prefer app-style sites to the “traditional web page” should be a wake up call to developers and businesses that we enjoy having our content with minimal distractions, bloat, and irrelevance over that of the poorly designed, lowest common denominator, advertising-dominated selection that dominated the internet so far.

    I for one am sick of what the Wired-style web has turned into so far — a sinking ship whose monetary concerns have influenced both the display as well as content of their stories. Sensationalist headiness, top ten lists, “borrowed” stories now join the ranks of popover ads, inline thumbnails, and tynt-style scripts in the disrespect they show for the reader. No adblocker can dissolve them, but our own finely tuned judgement can.

    The point Wired is afraid to make is that the app-Internet has no place for a site like their own. They depend on eyeballs, they try and give us a revolution in every issue, whether it exists or not. It’s all flow and no stock.

    Apps have become services that help us have a better experience. That’s what technology has become: something that makes all of our lives incrementally better. And that’s why Wired has nothing to write about anymore.

  39. More publishers wete dreams, utter nonsense. It’s a bit like the oil companies saying, Oil is the only fuel. Tosh. Waste of life reading the damned thing (Wired Article).

  40. I think the coming browser generation products like Internet Explorer 9 and Firefox 4 which support native SVG graphics and native animation with addition of 64bit Operating systems like Windows 7 and DDR3 memory are going to bring a revolution of the Web browsing experience.
    HTML 5 is much more than a native H.264 playing, if the enterprises really invest in migrating applications in this language, there will be a profound perception of end user experiences while using the Web. More and more, the web programmers will become professional web designers, something that the Flash designer did 10 years ago.

  41. “Well of course DNS is obsolete. We all use host tables now, downloaded from, don’t you?”

    Dunno about you, but I’ve transitioned to typing IP addresses directly. Why use host tables when typing in is so easy…

  42. Wired is (dead) tired. Long live Wired. (Remember when it was a must-have magazine?) Did that drama-queen headline earn them any attention?

    Yes some orgs are making Apps, hoping to create bordellos where they can pleasure their clientele while exposing them to typical bordello shenanigans. HOWEVER, the Real Stuff – thought-provoking content of all kinds, as opposed to Virtually Contentless dreck for mental diets – is multiplying all over the web.

    I commend Evan Hansen for the “hell no!” article at Wired for spelling that out.

  43. Identity crisis period for Wired; that’s what I will say. Unless you make theses kind of highly controversial statement, people don’t talk about you much.

  44. Arbitrarily putting Web at odds with video does not account for where most of that video is accessed: the Web. Roll those video stats into the Web and everything changes.

    It’s like saying bread is replacing ovens because more people are eating sandwiches.

  45. Bandwidth numbers, whether proportional or absolute, are meaningless indicators of value. Ninety-nine percent of bandwidth-hogging video is rubbish. There’s nothing new about landfill taking up a lot of space. Old news.

  46. This article makes as much sense as comparing the length of all highways to the total distance travelled on them and using that to conclude that highways are a dead mode of transportation.

  47. I gave up on WIRED’s bullshit years ago. Remember how much they hyped CueCat? I laughed the very first time I read about it. I mean, c’mon… They’re imaginaries, not visionaries. Zero credibility.

  48. I’m left with the same feeling I had when I learned “Free” cribbed from Wikipedia without citation. Pure hubris. Pseudojournalism at it’s finest.

  49. The web provides the major means of SEARCH, which facilitates the consumption of all the other crap. LONG LIVE THE WEB.

  50. Surely anyone who has done a basic statistics class would know that this is poor implementation and analysis of data. Shame on you WIRED (again)

  51. It is not disingenuous – especially if measuring pageviews per Compete – to say that the www is dying, but that the internet is living. Look at Netflix’ and Hulu’s streaming growth and look at how we use email, google docs, and other services, but not necessarily the www. We read feeds first, then find the unique stories that we click through; even apps for the iPad bypass the www for news services that aren’t www.

    When ChromeOS comes along, this will become even more evident, as people will use apps for their tools, not the www.

    Yes, the growth of www still exists, but look at it this way: we still haven’t received our digital, HD content full time. When the gigabit internet arrives, people will be streaming HD video and often. Watch in 10 years from now, as HD video will suck down 20x or more bandwidth than the www. If you really don’t believe this, then you’re someone that is probably placing all your trust in that blu-ray collection that you’re building.

  52. Wait…. Where exactly does all this video reside if not on the Web? Because I was told most of it sits on sites like Youtube and Vimeo…. which were websites last I checked (although I suppose the classification of these apps on mobile devices could change that but even then it does get transferred through HTTP doesn’t it?)? And how exactly do they define Web if not as data transferred over http protocols?

    Either way this is the really the laziest sensationalist sh*t graph ever made and whoever made it should have their jobs and any journalism degree they might have taken from them…. although it seems you don’t need one to work for Wired, obviously.

  53. Did anyone wlse notice the bit that says “proportion of US internet traffic”?

    Perhaps the web is not dead in the rest of the world…

  54. Telnet, FTP, Archie, Gopher…ah, yes, I remember it well. Along with Lynx & Pine.

    Where is that rocking chair?

  55. My perception is that, to sell magazines, Wired built a controversial headline and had to play with their definition of “Web” in order to support it. Or, rather, not play with the definition — they’re avoiding letting the definition morph, which is what it has been doing for many years.

  56. I’m sorry, I couldn’t read this article as it was posted on a dead medium. Could you Twitter it directly to my Bluetooth earpiece? KTHXBAI

  57. Thanks for a new viewpoint Rob.

    I do think there is a shift happening. Apps are part of it. Increasing sophistication of PR/marketing agencies are a part of it. Stupid draconian laws are a part of it. All of it points towards more control (or influence at least) over digital communications by the big money players.

    The WWW isn’t going away though. As long as there is net neutrality, it will stay a place were people can innovate.

  58. rob, your article and the graphics you prepared make the wired story disappear.

    you are so mean. poor wired. their artwork was nice, too. and it had less whitespace!


  59. Dang, late to this party.

    So – what about making the vertical scale a “per web user”? Even while bandwidth has been exploding, it isn’t just driven by the change in usage, but the change in the number of users.

    Looking at the shift in both volume and use on a per-user basis would be interesting. I think we might see that email creeps *up* slowly, while video expands.

    I’m also curious how we measure “email”, when a huge chunk of Hotmail/Yahoo/Gmail is “web”.

  60. This is one of two things on Wired’s part:

    1) Accidental stupidity
    2) Intentional trolling

    Personally I think it’s the latter.

    If you go by the graph, nobody’s using DNS at all now and therefor in reality nobody’s doing anything else at all[1]. ;)

    [1] No smart arse comments about remembering IP addresses or host files, please. ;)

  61. IN the future, the world will devide between “LAZY” internet users and “SMART” web users. There will, of course, be some Vin Diagram overlap. Of course. Too Many Sh*theads are on internet (FB). Screw those sell outs. ALL APPS are for losers.

  62. Applications are the worst thing to happen to the internet in years. Instead of Websites improving their sites mobile capabilities, they make an annoying application? I call that internet regression, not the death of the internet.

  63. Oh good… my decision not to bother reading Wired, which I make in newsagents frequently, continues to be justified. As you rightly pojnt out, “50MB of YouTube kitteh” represents absolutely nothing other than the wide band needed for video. It doesn’t mean people are actually spending more time on it.

  64. The graph makes no sense. Why is he measuring bandwidth. He should be measuring the number of people and the number of interactions. 1 video might equal the same bandwidth as reading 26 articles and doing 248 searches.

  65. Hold on, I’m downloading this 7GB game right now while I try to push my meager 5kB comment through the internet.

    Yeah, videos are really dominating the web. ;)

  66. Soap is dead!!

    Soap is increasingly being sold in concentrated form. Sales of bottled water are exploding! Soap as a percentage of bottled liquids sold in grocery stores is plummeting!

    Soap is dead!!

  67. RayKurzweil on his visit to Dresden on June 10th beautifully talked about the impact of exponential growth of information technology.

    More on

    At first you believe the numbers, then the graph and then you question, “What is different to 20 years back?”

    Surely the amount of traffic overall, only the %s have switched dramatically!

    So be aware of numbers someone gives you, it always boils down to the basis to which you compare the change.

    Cheers, Ralf (Dresden, Germany; @LockSchuppen)

  68. The web is dead? This is news to those of us who lived outside the US. Also thank you wired for selling your soul to Apple. Apps is going to kill the web!. We are to blame! I think Wired believe that all the whole world owns an IPAD or worship Apple (the latter maybe true ). Do you know how much it cost to buy an IPAD and an Unlock IPhone here in the non US world. I rather go with the Samsung or a Nokia or dare I say a blackberry. These are not closed garden like the Apple stuff. ALL of those things are fine and dandy if you live in the States or some European countries. The web is not dead it just moved out of the US. I’m scared of the googlezon monster. Those of us outside the US will get the short end of the stick, which we already get anyways.

  69. Classic Chris Anderson – use slanted data to feebly support an obnoxious hyperbole. Seriously, does this guy just sit at his desk and think of controversial one-liners?

  70. What the Wired article strategically leaves out is the role of regulation and policy in creating the media environment. Citizens paid for the development of the Internet and we citizens must impress on our representatives the importance of maintaining the Open Internet. Our founders struggled to create an open society in which people could express their ideas in public space and through the press (which in their day had more in common with blogs than today’s newspapers). If we don’t maintain an open internet, the new public square will become an enclosed commercial space where only commercial speech is spoken. Do you want Steve Jobs deciding what messages and applications you can have access to? The political assumptions and implications of this Wired article need to be teased out and discussed as much as the technical and business issues.

  71. I think some of the issue here is that with the smarter and more efficent layouts, better browsers, modern server technologies, globally distributed dns, caching, esi, cdn and persistence has a lot of impact on these stats. Unless fixed up I don’t see how http?s traffic could be declining. They just don’t know how to measure it most likely.

  72. Conversely, the chart showing work productivity of the typical cubicle worker with access to facebook or other social networks over the last 10 years would look like this:
    50% productivity – 2000
    \ 5% productivity – 2010

  73. The claim that “The Web is Dead” reminds me of Mark Twain’s comment,

    “There are three kinds of lies: lies, damned lies and statistics.”

  74. The web is dead? Ummm, okay…that’s funny. Living in a wireless world, it would be safe to say Wired is dead.

  75. This would be interesting, if one could account for protocol use for digital software/music/video-transfer. Because that makes up the big parts of “Other” in the beginning (IRC, DC++, darknets), “FTP” afterwards (remember those ratio-servers?), websites full of free stuff (yeah, that’s where i got my mp3s back then), P2P with the clearest use-case (all those p2p-clients and torrent sites) and in the end with live-streaming of the content.

    This should account for more than 70% of internet-traffic at any time.

  76. Why are people slagging Wired here? Wired has a right to its story angle…why attack someone’s opinion? What do you read then if you don’t read magazines…crappy user generated opinion???

  77. Thanks for the brilliant post Rob. What you describe as an ‘interesting editorial choice’, i.e. ‘[t]he use of proportion of the total as the vertical axis instead of the actual total’ is pretty dominant in most stories of technology diffusion aided by graphs. Attempts at countering the imagery that results from it have been rare. Good luck with yours.

    I have a request though. Maybe I am lazy or just ignorant, but after browsing for some minutes I couldn’t find the numbers. Could you provide a direct link to the source data from which you (and Wired) draw the graphs? I like to check (and play with) the numbers behind graphs on my own. Many thanks.

  78. Excellent article, and excellent comments as well. I like it when people point out the flaws in any statistical analysis.

    Long live the number 42!

  79. Congrats once again! I was suspicious about their statements from the moment I saw the graph. This post should be attached in the comments section of the Wired article. Good work and good comments too. Keep it coming!

Comments are closed.