More on counting subscribers
October 16th, 2007 by marc
Here’s the response I left for Pete Cashmore over at Mashable regarding two of his articles, Google Reader Stats are Bullshit (With Proof), and ActiveWeave Blogrovr: Screwy FeedBurner Stats?.
Hi Pete, Marc Meyer CEO of Activeweave BlogRovR here.
Congratulations on an excellent piece on the problems with subscriber statistics from blog aggregators.
Our stats show you to have dipped by only 4k BlogRovR subscribers, not 35k!
If you show all our users disappeared it may be due to a reporting glitch. We’ve had a few days in which technical difficulties have prevented us from pinging Feedburner, causing all our subscriber base to disappear for the day. We’re working on resolving that.
At the time you asked me to remove Mashable from the default Technology recommended blogs you had 35k BlogRovR subscribers. You have about 31k today.
As part of our Technology recommended blogs, Mashable was being shown to new BlogRovR subscribers who were counted as subscribers when they chose to sign-up for the Tech bundle.
When Mashable transitioned to a non-bundled blog, we began to count explicit actions only: an explicit subscription to Mashable or opening a Mashable post summary from the BlogRovR tray. Not all of the subscribers who’d subscribed via the bundle would have yet pulled content from Mashable, accounting for the drop. Over time, these users are still shown your post summaries and as they open them will be subscribed.
BlogRovR is an everywhere-on-the-web medium for blogs to get exposed to real readers; readers who actually see your blog post summaries whenever they browse something on the web related to what you’ve written. Any BlogRovR user reading Techcrunch is likely to see a Mashable post treating the same subject right there. Some blogs we liked from the get-go and bundled early (like Mashable) have been exposed to 10’s of thousands of new readers, as you indicated.
For the record, we believe that BlogRovR subscriber counts are more reflective of real readership than the conventional counts from other blog aggregators, though we recognize that we share vulnerabilities regarding counting inactive users.
Thanks Pete for shedding light on this issue!
Yours,
Marc Meyer