The list was updated and Blogoscoped bounced to 9th. But hey, it beat out CNN. :) |
That ranking is obviously inaccurate because it only includes some blogs and news sites. For example, Hack a Day – 12,479,
Scientific American – 66,934 subscribers are not included.
Also Google Reader's numbers seem off and are different from Feedburner's numbers. Blogoscoped is in one of the default collections (the Google one), so the number is higher that it would be otherwise. |
> Blogoscoped is in one of the default collections (the Google
> one), so the number is higher that it would be otherwise.
[I commented on that issue on TechCrunch, so I'll re-post here, with some slight edits.]
I second that you need to take into account the default bundles in Google Reader. We could further divide them into two types of bundles: those topic bundles immediately visible upon sign-up (the “default” default bundles), and those topic bundles which you need to expand.
There are three “default” default bundles (you still need to click “subscribe”, but they are immediately visible during sign-up):
- News (BBC, CSM, ESPN, Google News, MarketWatch, NPR)
- Sports (ESPN, BBC Sports, Bill Simmons, Off Wing Opinion, Sports Frog, True Hoop, Yanksfan vs Soxfan, Footbag WorldWide)
- Fun (Colbert Report Videos, Daily Show Videos, Google Video Top 100, Quotes of the Day, Onion, YouTube Most Viewed)
Some of the “non-default showing bundles” are:
- Technology, Thinkers, Celebrities, Geeky, Photography, Cars and so on.
But does this introduce a “skew”? That totally depends what you think these numbers express. If someone subscribed to a default bundle, even *without* specifically acknowledging the feeds contained with-it, they may *still* end up being a loyal Google Reader-reader of that content, say TechCrunch (or Google Blogoscoped, which is contained within the “Google-related” bundle). It’s not that you’ll ignore the items in GReader only because they were added as part of a bundle.
On the other hand, if you want to compare blog *quality* then no, this is not a good indicator. Maybe a Google editor made a subjective choice to select these bundles (or maybe they are based on e.g. now outdated blog stats Google had). Maybe the bundles themselves are showing outdated blogs, and there may now be better alternatives. Maybe new cool blogs for these subjects came along and they are not yet contained in the bundle (suggestion to bloggers: maybe ask Google for your blog to be included? :)).
But that’s the nature of any ranking chart, that it doesn’t show quality, but just… well, specific ranking based on specific numbers being counted. And on a side-note, these rankings in itself, the minute they are published, “skew” their own numbers in a sort of Heisenberg fashion, because people are then more likely to find new reading tips by checking the top 10 or so (this even includes Google web search result rankings!). Though I don’t believe that this means you can’t make it with a fresh blog due to older blogs blocking the view — it only means it’ll take a little time (like half a year, a year) to get into this “self-increasing feedback loop”. |
I think Google could offer personalized bundles by using data from your search history, your location, your bookmarks etc. That could be a part of the recommendation feature that will be launched soon. |
