Feedburner hacked! from Boris Veldhuijzen van Zanten on Vimeo.
It is hardly surprising that FeedBurner’s subscriber numbers can be faked. What is surprising is how easy it is to do so. As the video above shows, all you need is a Netvibes account. The folks at the Next Web in Amsterdam took a blog with 43 subscribers and turned that into 2,500 overnight simply by creating an OPML file with the same feed copied 2,500 times and pasting it into their Netvibes page. The result was 2,500 widgets of the blog feed, which FeedBurner counts as separate subscribers.
Why does this matter? Blogs like to tout how many RSS subscribers they have because, even if it is a smaller number than direct visitors to their site, it represents their most loyal readers. That’s why we display how many RSS readers we have in the Feedburner chicklet at the top of TechCrunch (currently 850,000). For these numbers to have any meaning, though, they cannot be as easy to game as the video shows. (And, no, we don’t game our numbers).
You’d think that Google would be smart enough not to double-count these things, or at least ask Netvibes and other widget start pages to de-duplicate the numbers for them by user. What appears to be happening here is that FeedBurner counts each widget for a particular feed on Netvibes as a separate subscriber, regardless of whether that widget is on ten thousand different user pages or repeated ten thousand times on the same page. The same thing happened a couple years ago with Pageflakes.
Update: Netvibes VP of Product Development Franck Mahon responds in comments that it is working to fix the problem of duplicates, but that there are other ways to “hack the numbers.” And he notes that it might be more useful to count active subscribers than just people who may have added a feed two years ago and never read it.
This new reply feature in the comments section is nice but I think the replys need to be a different color so that one can easily scan for new comments Vs replys ..
Ah, so this is how TC has 850K subscribers
I don’t think TechCrunch would need to do this.. their news are good enough.. for people to be interested in.
http://blabtech.blogspot.com
And I thought bloggers were decent, hardworking people…sigh!
yeah. no.
Yeah, some of them.
Michael, how many subscribers does TC really have?
My trust in feedburner chicklets count has suddenly vanished. And anyone could be doing it… Lets wait for Feedburner to fix this and see those faked counts fall!
yeah, wonder how many Feedsubscribers Techcrunch will be left with…
this is old news
Nope it is not, at least we have only found stories about using someone else’s chicklet.
Can somebody please start keeping a before and after count! The difference would be degree that somebody is a fraud. Then the blogosphere forces bloggers to post those numbers! Its all part of the credibility scale.
Yeah, we’ve been tracking daily feedburner counts on about 2,000 blogs since April.
“loyal readers”? c’mon feedburner stats are about as reliable as any other web stat.
As a default feed in a bunch of different downloadable readers, you also benefit from those “subscriptions”. That’s gotta be a few hundred thousand.
Not surprising. I came across a blog recently with like 40K subscribers and 0 comments on nearly all of their posts.
Will we ever have trustworthy feedcount numbers or will the honest people get screwed over by inflated numbers from crap sites that take advertising dollars.
This not really “hacking” FeedBurner in the narrow sense.
It just makes use of a flaw in counting mechanisms which seem to be quite simple. A very easy way would to count every IP only once (and have it expired in 24h). But still, all those numbers will be somewhat biased.
I remember Feedburner had a similar issue with Mashable. Since Mashable was one of the default blogs to be added on PageFlakes, Feedburner counted everyone that signed up with PageFlakes as an RSS subscriber. I give Cashmore credit for solving that problem.
FeedBurner relies on Netvibes and other aggregators to accurately report subscriber numbers, so this is on Netvibes, not FeedBurner. See: http://googlereader.blogspot.c.....ed-up.html
Yup.
ty… i was going to say this but seeing as you said it, and tc has very nicely added replies… i will simply say ditto.
Call me crazy, but since the little Feedburner button is just an image button I’d think all you’d need to fake a high subscriber rate is photoshop. Sure, if someone was savvy enough to check on the location of the actual image you’d be caught but who would care that much?
I was thinking the same thing, most people would never think of checking the image source. You could also take the image source of a existing popular blog but link to your own feedburner page.
You just gave me a great idea to instantly increase my level of importance… I could even create a new, slightly different image for each day.
Yea, or you can have a javascript put the number of subscribers you want. that chicklet doesn’t mean much, whatever the site it is on. Feed distribution is way too anarchic right now to consider those chicklets a valid info.
Thanx alot Techcrunch!*running to netvibes to quickly submit 500 widgets!* I guess at this point we really can’t call you a “blog”. Blogs comment on already established news and blogs. It appears YOU continue to break stories and companies in the Web 2.0 world. Hmm Maybe I should “Blog” more about you!
You may not game your numbers for TC, but they are likely very inflated by a function of time and default subscription deals TC had. I could be completely wrong, but you could post your “reach” numbers anytime you talk about TC RSS subscribers, and I notice this never happens.
I’m not in any way suggesting that it’s the kind of BS that goes on via Netvibes, or that there is any attempt on TC’s part to game the total number. I am suggesting that the number is probably extremely inflated anyway.
Interesting conclusion.
Are users really more likely to subscribe to a feed if it has a higher number of readers? I always thought that was just a feature to make other bloggers jealous (and yes, I use it too
Why is this a surprise? All the analytics services can be gamed at some level. Yes, all of them.
Gaming the numbers is only part of the story here. Follow up this article with what people can do (or are doing) with these gamed/inflated numbers.
CG
I wish personally there is some other feed counter similar to feedburner yet more stable, I f*** lost over 450+ subscribers and 182 email subscribers back in March. I never received email from feedburner or google.
Excellent! Signing up for Netvibes right now. Feel free to add my feed to your netvibes account, too
This is a real bad thing, quite some time back there was a competition between Shoemoney and John Chow, and somehow Shoemoney managed to get his feed count high enough on the last day, wondering if he used this trick.
The same thing was also repeated by John Chow in recent times, at least now we can know how this king of magic happens.
Right, the services report the number of subscribers in the HTTP header when they ping feedburner, like “143 subscribers”. So they can send any number they want, and Feedburner really just relies on them being honest about it. There’s no way for Feedburner to count the actual subscribers, except if they start creating code and rules for every single site that reports to them.
Worse than faking your own blogs numbers is the ability to screw with anyone elses numbers to the point of it being useless.
Anyone could come along tomorrow and make any other blog look like it has 50 million subscribers, ruining their ability to track their own number of subscribers.
The whole system will turn to crap, and nobody will display their numbers anymore.
I wouldn’t recommend trying this. Google is certainly aware of the issue and could come down very hard on sites that have obviously jumped to several thousand users from just a handful. That said my Feedburner count goes up and down so much because of reporting problems with the system, I’m not even sure how they’d be able to tell. I wrote about the Feedburner Myth a while back on Sciencetext - http://www.sciencetext.com/the.....-myth.html
If everyone turns the other way and no one asks questions everone makes more money….until advertisers realize that they aren’t getting their money’s worth!
a smart advertiser would do research
and not go by rss counts
rss feeds # are easy to game
i know of several others ways to do it
For me this is about whether you should than whether you could. I could do this easily by embedding my feed page in an iframe on some high-volume site that allows me to (I won’t mention who). But I want to know real numbers so that I can see the effect of a campaign or series of posts. I am interested in conversions, not a number. By gaming this system I would be doing myself more harm than good.
Now if only it could tell you if the people that read blogs were worth counting in the first place.
it’s damn stupid to do a cheat like this. totally meaningless.
stupid only if you get caught lol
but many only subscribe to rss feeds if they see a high number
This is hardly a feedburner ‘hack’; it’s just a case of netvibes being crap. Beyond that there are much simpler ways to inflate your feed numbers.
It’s not hard to fake the feedburner counter…
Netvibes has an awsome tech team, I’m sure they’ll sort this out
We are working on a fix to filter out in the reporting the duplicates while still allowing people to add several instances of the widget to their startpage.
While doing some tests this morning I found some other aggregators with the same bug (which is in no way an excuse to not say that we did suck) and there’s also another easy way to hack the numbers: all you have to do is spoof the request the aggregators send to FeedBurner with wrong numbers. You don’t even have to use those aggregators (even if we love when people do use Netvibes)
In top of this it makes me think about what really is a subscriber. We, and competitors, do report as subscribers users who even didn’t came since a long time just because they did subscribe a long time ago (by the way we do only report users who did manually add the feed to their page and not because it was on a default startpage).
Wouldn’t it be smarter to only report unique active users for the last 30/60 days for instance?
Of course lots of blogs would loose subscribers if all online and desktop aggregators would do this but it’d give a better overview of what are the hottest blogs currently.
What do you think?
I agree with Franck Mohan,
Showing 888K subscibers doesn’t mean that TechCrunch gets that many visitors daily.
If aggregators start showing only active users for last 30/60 day, then we would find the real juice…
How about that , if all bloggers now this hacks than even for 2 days blogs old can have thousand subscriber and cheating their advertisers
I read about a recent contest, Blogging Idol, somewhere and it was pointed out that people are gaming the contest.
They ramped their subscribers so fast but they forgot all about traffic and comments which did not experience the same explosive jump.
Ultimately, readers are mature enough to realize if a blog is worthy of such high numbers. I will raise eyebrows if I don’t see great content and web design.