Pete Cashmore on Blog Reader Stats Are Bullshit
Pete just caught me up on a series of posts at Mashable that lay out a strong case for how fishy the RSS readers stas are that many have been glorying in the last few days.
I won't summarize the investigative bloggerism involved, although I followed the arguments are they seem plausible.
The conclusions at one of the posts are telling:
[from Google Reader Stats Are Bullshit (With Proof) by Pete Cashmore]Some conclusions to draw:
1. Google Reader stats are bullshit because simply being the default feed in any of those bundles will increase your stats by at least 50K to 80K. The quality or content of the feed is irrelevant, and the feed doesn’t even need to exist.
2. My early tests have shown that news, science and technology feeds get the biggest benefit from this problem.
3. The case of the Footbag feed shows that Google Reader probably does not check whether subscribers are active or not (you can’t read a feed that doesn’t exist). Most of the subs on these feeds likely took Google Reader for a spin and abandoned it the same day.
4. Therefore, most of the subscribers to these feeds may not exist.
5. Because Feedburner takes its stats from Google Reader, Feedburner stats are also wildly incorrect whenever Google’s default feeds are involved.
6. In phone calls with Feedburner last week (yes, I do research!), I learned that FeedBurner doesn’t enforce any rules regarding stat counts and particularly default feeds. They are, however, extremely nice people.
7. This problem is not limited to Google Reader, but applies to many feedreaders and startpages large and small. We’ve spoken to Feedburner about problems with Webwag, Blogrovr and Pageflakes adding tens of thousands of “readers” overnight due to default feeds. We have since asked to be removed from all the defaults we know of. We may be listed on more that we don’t know about.
8. Even when feedreaders discount inactive readers on a regular basis, default feeds will still overcount because thousands of people take RSS readers for trial runs every day.
9. The easiest way to get a default feed on one of these startpages is to own it, promise to promote it on your blog or be friends with the person who runs it.
10. Some blogs are the default feeds on every feedreader on the web (BoingBoing, Techcrunch etc). Their stats may be way out.
Pete's research suggests that my recent spurt in RSS subscriptions at Blogrovr falls into this category of shadow play, since I am featured in a Tech cluster there. Hmmm. I guess I should ask them to drop me -- if only for a moment to see if Cashmore's conjecture is right.
Hi Stowe, Marc Meyer, CEO of Activeweave BlogRovR here.
I think Pete makes some valid points with regard to the shortcomings inherent to how most aggregators report statistics, but we think that BlogRovR subscribers are much more representative of actual readership than those of other aggregators he mentions.
Being bundled in something like Google Reader doesn't mean that a subscriber will ever see your content.
With BlogRovR, being in the default bundle means as people browse anything on the web about which you've blogged, they'll see your blog posts. Which in my book is excellent value to both bloggers and readers alike—contextually relevant posts.
When you write about something I care about, that’s when I see what you’ve written, even if I can’t quite get to reading all of your blog all the time. And if I did read something and forgot, I'll see it again when I need to remember.
The advantage that BogRovR provides is that today’s scenario for dealing with RSS feeds is:
Read copiously, remember encyclopedicaly, hope to recall what was interesting next time you see something relevant. Fail repeatedly (unless you're Scoble and have total recall).
With BlogRovR: subscribe copiously, browse judiciously, receive interesting blog posts selectively.
Posted by: Marc A. Meyer | October 15, 2007 at 02:42 PM
The real question is ... why do the numbers matter?
If they are wrong, if they are right, does that really matter in terms of anything?
Posted by: Nikole Gipps | October 25, 2007 at 04:49 PM