Hacker News new | past | comments | ask | show | jobs | submit login
Gawker's Traffic Numbers Are Worse Than Anyone Anticipated (theatlantic.com)
94 points by ctide on Apr 20, 2011 | hide | past | web | favorite | 52 comments

I blocked most Gawker sites from my Google search results, because all the URLs seem to be broken and it takes me a few seconds of staring at a page unrelated to my search before I remember that and hit the back button.

I simply don't understand the point of the changes they made.

Edit: from one of the comments on the article (I'm even more confused about their changes now):

"Another thing that may be contributing is that after the redesign they, Gawker, put some national redirect for Canada that actually breaks the links and just dumps you on the home page. This has become so frustrating that I don't actually go to gawker sites anymore and used to log into lifehacker and i09 nearly daily."

I get the same thing here in Japan. To get to the Gawker top page, I have to go us.lifehacker.com (for example).

When I click any link from google however, I'm automatically redirected to the JP top page, so I have to take the www out of the url and replace it with us.

it's such a PITA that I've pretty much stopped using Gawker sites all together.

I get this in the UK. I've never managed to read a gawker article.

The problem is their completely moronic geo-ip auto-redirection technique of inserting a 'uk.' subdomain prefix. Deleting the prefix will usually let you view the desired article.

The fact that they haven't figured this out by now tells me there's something very wrong in their software camp.

Same thing happens to me. I have RSS feeds of their sites and after clicking on the article it would try some redirect attempt for Spain and come up short. Seems to be better now though.

same in australia. any link redirects to .com.au home page

Gawker sites have become barely usable for me - if I click one of their hashbang URLs it will sometimes helpfully redirect me to uk.sitename.com, losing the hashbang part and so leaving me at the front page and not the article I'm trying to read.

It's really annoying and entirely unnecessary. A blog is not a web app and doesn't need to be built like one.

Their site is completely unusable for me too. I'm amazed that they haven't fixed this, and amazed that it made it through testing without being spotted.

I go to m.gizmodo.com, which makes it partially useful, and the stories are readable if clicked from there. Of course, when you try another gizmodo link it fails.

What a disaster, but will they have the balls to go back?

Hmm...yes, m.gizmodo.com seems to work, provided you manually remove the '#!' and leave the rest of the URL alone.

In theory if one had a web server running locally, one could add gizmodo.com to /etc/hosts and map it to loopback, then configure said web server to catch gizmodo.com and rewrite the URL to redirect to m.gizmodo.com instead...

Like this perhaps, using Apache's mod_rewrite:

    RewriteCond %{HTTP_HOST}     kotaku\.com$ [NC]
    RewriteCond %{QUERY_STRING}  _escaped_fragment_=(.*)    
    RewriteRule ^/(.*)           http://m.kotaku.com/%1 [L,R=permanant]
edit: hmm, might need this set after the one above, for URLs without _escaped_fragment_:

    RewriteCond %{HTTP_HOST}     kotaku\.com$ [NC]
    RewriteRule ^/(.*)           http://m.kotaku.com/$1 [L,R=permanant]

Easier solution, if you use chrome https://chrome.google.com/webstore/detail/nlknelhingldnlmpgn...

Im sure FF versions exist, and it uses ca.site, so it has the old design.

A friend of mine at Gawker told me that Nick Denton got the idea for the redesign and made them do it despite the protestations of members of the team (like, everyone selling the ad inventory). It is regarded internally as a colossal fuck up forced upon them from above.

FTA: "Nick Denton wrote in to argue that the site's internal tracker has been broken for two months."

Really? Gawker has not had web analytics for the last two months, after a major redesign? That is hard to believe.

Gawker has also had a broken 'Back' button in Firefox, where you can click on an article and it requires 3 Back clicks to get you somewhere you didn't mean to go.

Gawker sites are now also serving out giant pages -- the images are not saved for web viewing (generally huge compared to their pixel area and could be much smaller with non-lossy compression). I checked out Firebug one day after the redesign, and saw almost 2MB of JS, in separate files and not minified.

I would like to hear about the process for this. Unfortunate, as I used to enjoy Deadspin as a go-to break time site.

> Really? Gawker has not had web analytics for the last two months, after a major redesign? That is hard to believe.

That also struck me as highly doubtful. Many people here work on big web apps; the idea that something as critical as analytics could be broken for two months is ludicrous. My manager would be having some sort of embolism if it was down for two days. After 2 weeks, I'd probably be fired. After another two weeks he'd probably be fired!

Analytics are the bread and butter of large web sites. If they go down, you basically stop doing business. It's inconceivable. Denton has the motivation to lie, so it's pretty clear that's what this claim is.

"Broken" may mean something other than "totally, completely broken". Maybe there's some odd bugs that cause an occasional misreporting, and they're trying to hammer these out. Maybe you can only log in sometimes. I don't know, but I think that broken could mean something besides "we can't look at any data at all".

Relevant update in the original post:

"Nick Denton wrote in to argue that the site's internal tracker has been broken for two months. "Those numbers are total pageviews from all sources for all sites," he said, referencing Quantcast data. "We were doing about 100m a week. At the low-point, we dipped to 75m. We're now back at 93m."

Denton is definitely a spin artist but I don't see a huge incentive for him to flat out lie given his advertisers can instantly see their campaign display metrics etc and pay on a CPM basis . . .

You'd be surprised. I bought some ads on their network and despite promised metrics, I've had nothing from them other than vague quotes. Luckily I didn't drop too much cash.

Anyone can check themselves, gawker's quantcast stats are publicly visible, in fact its the example url quantcast gives so you can just click it.

Go here: http://www.quantcast.com/ click the gawker url click the 'gawker media network' parent organization chose 'impressions' from the pulldown click 'week'

voila, the exact numbers he's quoted.

As much as the general perceived reaction to the gawker redesign has been screechingly negative, the actual traffic impact was modest. Denton claims this was a deliberate one step back two steps forward kind of thing and traffic will return in a couple quarters.

not necessarily...they can give advertisers all the ads they want...the low traffic numbers would only affect their remnant ad inventory revenue

I was never a daily visitor to Gawker or Gizmodo, but I used to end up there fairly frequently via links. What I've noticed over the past few months is that they seem to have dropped entirely out of the online social media "conversation" - you don't see people linking to them, you don't see people twittering their articles, etc. I don't have any technical explanation for that, but maybe the redesign was especially off-putting to heavy blog readers most likely to reblog/retweet things?

In my case, none of the links to their site go to the advertised page any more. If I click on one of their search results, the respective home page loads but never the article.. That bug's gotta be decimating their traffic.

Not that I'm a big fan of Gawker as a whole, but I want to know who the hell thought it was important for gawker to have an ajaxy interface with ajax page loading of posts.

Its a news site, literally the entire purpose of the site is to show text and (some)pictures, why try to make it fancy?

I used to work on a high profile old media site, and based on my experience there, I believe that the motivation is that it's easier to game ad metrics with stupid gimmicks this way. Approximately at the time I had lost hope and was looking for an exit, they were working on a new photo gallery that would count each "next" click as a new page impression. In short, it allows you to muddy the water around what exactly constitutes an "impression."

To be fair, an "impression" isn't really a great metric on a page that people sit on for >1 minute.

The guidelines we worked under at my previous job basically said "if a user action changes a substantial portion of the page, it qualifies as an impression". It seemed pretty fair to me at the time.

Advertisers (especially with rich media) get ridiculous interaction metrics from their creative, they know what people are doing with their ads. Lots of them are happy if an ad sticks on a page for half a minute and is then replaced.

Ever since the redesign their sites are almost unusable on my machine. When will sites realize that floating part of the page so that only part of it scrolls not only leads to lag but also makes for a visually confusing user experience. Slashdot did this exact thing with their header when they redesigned their site several months ago, luckily they have since switched to a fixed header.

Not only that, but I seem to get a 2px gray line that fills the width of the page and scrolls with the content. Ever since the redesign, every time I stumble upon one of their pages I click the back button.

But if that was not enough, they traded a system where each site had its own look and feel for one where every site looks exactly the same (except for a small logo in the header).

We should all be thankful for both Gawker and Digg for being such spectacular counter-examples. Sometimes you learn from your mistakes, sometimes you get to learn from other peoples' mistakes.

It's just odd watching large sites repeat the same mistakes. First digg, and now the entire Gawker sites. It's as if they paid no attention to how badly a major change can have on their viewer base from past experiences.

Seriously, if you are going to make major changes to your website, do what many other sites have done. First you put up a discussion with features and looks that you want, then wait and see the reaction. If it's positive move to the next phase, create a beta page and randomly divert a viewer to the beta site. Then have an option to input opinion so you can see reactions.

Shocking your viewers with change is a horrible way to go about updating your website.

"do what many other sites have done"

What sites have done this?

What sites Have done iterative and incremental changes, initially deployed to subsets of their users? I can immediately think of: Facebook, Amazon, twitter, zappos, and flickr.

It's generally not newsworthy because it generally doesn't lead to totally disruptive fiascos.

It's hard to know the exact drop in traffic. It's probably more than Nick Denton's claimed 7% but probably also less than the ~85% drop this article claims. The truth is probably somewhere in the middle. In any case, even 7% is a pretty massive drop following a redesign.

It's interesting because the content is the same, only the packaging and user experience has changed. It seems that reader-loyalty on news sites like the Gawker sites, Techcrunch etc, and community sites like Digg, is actually very low. Probably much lower than the sites themselves believe.

I dropped Gizmodo before the redesign, partly due to the stolen iPhone incident, but mostly due to every other article being Apple related, with the ensuing mindless and juvenile MS / Apple bashing in the comments.

Lifehacker I dropped after the redesign, as it just didn't work very well any longer.

Lucky for me, and everyone else except the blog networks, is that other content more suited for my tastes, is only a URL away. I have no commitment, no subscription or any other reason to ever look back.

Thus I think the point to take home, is that you should listen to your users / readers, no matter how obnoxious or "trollish" they might chose to express themselves. It's easy to just dismiss and ignore a point of view, if it's expressed with profanity. But below the profanity, there is usually a reason or argument, which you must find and listen to. At least if it's a recurring theme (i.e. "the new design sucks").

If you don't listen, you risk loosing your "loyal" users, probably much faster than you ever thought possible.

The redesign is a disaster but honestly Gizmodo lost me as a visitor a while ago.

Maybe I'm in the minority but personally I still find it a bit stunning that a site like this would pay for stolen property and then display it online. Even if it doesn't turn out to be considered criminal activity, it's ethically reprehensible.

If anyone wants to view Gawker sites as they were before the redesign, just append ca. as the subdomain (like http://ca.io9.com). Even knowing this, I still visit io9 1/10th as often as I did before their redesign.

Thank you! Very useful hack.

I have something to confess. I've been reading about this Gawker site for a long time already (probably ever since it became popular) - but I don't know what it is, I don't remember reading any article from it, only articles about it. Is it another us-centric thing, like the Super Bowl? Or maybe it is just not geeky enough for my liking? Well - to be honest I am not that interested in the answer to this question. I am already assuming that it is not an interesting site.

It is actually a bunch of several blogs, not just one site. They were quite popular before the re-design. It's not an us-centric thing like the Super Bowl.

What's broken traffic for me to the Gawker-hosted Kotaku is that Steam doesn't correctly redirect links from its feed of game-related articles to the hashbang targets, so I wind up at the Gawker homepage and wonder WTF is going on.

That was probably the only reason I ever visited Kotaku, and now I know better than to even try clicking.

It'll be interesting to see how this plays out, especially with the October deadline for Denton's bet with Rex Sorgatz http://blogs.citypages.com/blotter/2011/02/nick_denton_gawke...

(By October, Denton promises to pay $10 for every 1 million pageviews below 510 million, and Sorgatz has to pay $10 for every 1 million above 510 million.)

Also, a minor correction with this article:

Gawker backtracked rather quickly on the redesign, adding a button to switch the site back to a traditional blog format, which allows readers to scroll through post headlines and excerpts in reverse-chronological order. But by then it may have been too late.

AFAIK, this button has been available since the beginning of the redesign.

No.. the switch to traditional format option was added after the site were live for 2 or 3 weeks. (At least on Jalopnik) It is the only reason I went back to reading Jalopnik.

That's odd. I remember switching to the classical display on Lifehacker as soon as I saw the redesign.

I used to check SomethingAwful.com daily, until they changed their home page layout from a scrolling log of new articles (so you always saw the new stuff at the top) to a 'portal' style layout where you had different little zones promoting content from different writers. I can't quite put my finger on why it put me off so much. I've literally barely been back to the site at all. I guess the extra cognitive work being asked of me - to decide which updates I wanted to look at - was too much.

I used to be a daily consumer of gawker and gizmodo, I just can't handle it post redesign.

I really liked the old design, it wasn't broken and sure as hell didn't need fixing like this.

I used to frequently scan through gizmodo and lifehacker quickly every day or two and open any article that looked even a little interesting in a new tab. I remember being really disappointed when the redesign happened at first and the way I interacted with and used the site was no longer as painless.

I know I'm at a Gawker site when it comes up with an empty frame (I normally have Javascript disabled). So I just close the tab and move along.

What a bunch of retards.

That reminds me, I'd been meaning to unfollow Lifehacker on Twitter.

2005 called, it wants noscript back

Your privacy and security called, it wants noscript back.

Your mom called, she wants her over-applied metasyntactic template back.

well, looking on what is indexed in google http://www.google.com/search?q=site:lifehacker.com/&hl=e... there was at least a partial SEO f*ck up involved (and yes, google referred traffic is important for tabloids and other technology blogs)

Quite bearable if you view the sites on a mobile browser or just change your browser's user agent.

i heard that the reason for the re-design was partly because there are more pages viewed. meaning more advertising. so, i guess they thought it would bring in some more money....dont think that worked out to well.....

People keep telling me that the Gawker redesign was for tablet computers.


Someone at Gawker decided This Was The Way To Do It, ignored everyone else, and in the process flew their business into the ground.

I doubt they'll ever go back. The execs would probably rather lay off 10 writers than admit they were wrong. Pisses me off to no end.

When I first saw the redesign, I thought it was kinda weird but would probably be great on an ipad. But when I checked it out, the navigation was broken.

I checked it again just now on my iPad to see if they had fixed the problem. This time my iPad was redirected to the "mobile" version of the site - which is just a long list of headlines without any pictues.

Is it possible that the Google Panda Update may have contributed to it as well? If people want good content, I can't imagine a bad design would deter 50% of users... even Craigslist is horribly designed

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact