June 18, 2009

FriendFeed Sneaks Into My RSS Stats And Hits The Big Red Button

It's tempting to go back the age-old line of there being lies, damn lies, and statistics. On the Web, where practically everything is measured and big numbers are almost always better, counting up one's followers, friends, subscribers or authority is practically a pasttime. But with each metric comes a question of validity - how did they approach that data, and is that process consistent with the world view of what is factual?

Today, for reasons known only to their team, FriendFeed started to display subscriber counts to those FriendFeed users who are importing blog posts alongside all other subscribers, displayed in FeedBurner or any other blog analytics tool. With this change, popular FriendFeed users have seen a dramatic jump in their feed subscriber counts, even if actual traffic or readership to their sites has not changed.

BlogPerfume Shows My Stats Spiked Today

A clear beneficiary of this move, my own statistics ballooned from a possibly accurate count of just over 8,000 subscribers on this blog to more than 13,000. And in parallel, thanks to my importing my posts on my wife's blog, her count catapulted from just over 50 to more than 9,000. (For a site that gets only dozens of visits a day)

My Wife's Blog Stats Are Through the Roof

Coincidentally, my RSS subscribers had already been jumping, starting in late April, for reasons largely unbeknownst to me. In the last two months, I organically saw the subscriber counts pass the 5,000 barrier and crest to the more than 8,000, as I poked through the stats and tried to find out why - considering both Google Reader bundles and possibly a part-time inclusion on the Techmeme leaderboard as factors. But now, pointing to that growth seems silly, given FriendFeed flipped the switch and gave me a big, albeit likely false, foundation.

FriendFeed's Impact Rivals that of Google On This Site

The company's comments on this change state that "you are putting your words in front of a lot more people", so theoretically, they should be counted. But I believe it is less-intensive to follow someone on FriendFeed than it is through standard RSS, and I have no idea how this handles duplicates, though I can guess it's somewhat controlled, given my own stats jumped by a mere 5,000 when my wife went up by more than 9,000.

Rob Diana of Regular Geek clearly made his comments understood, when he said, "Subscriber Counts Now Mean Nothing".

Since you don't have admin access to my FeedBurner stats, you can see the jump by taking a look at Blog Perfume's Feed Analysis tool here.

So the question is - why? Did the FriendFeed team just want to extend the visibility of how much impact their service has with bloggers? This move makes them a clear rival to Google in my own statistics. Or did they really think this was a way to show, accurately, how many people you were exposed to? Either way, as I said on a thread in the site, what's done really can't be undone. I hate upward spikes as much as I hate downward spikes, as we've seen when FeedBurner and Google FeedFetcher miss each other in the night. But it's not accurate, especially when it comes to small blogs hiding on big accounts (like with my wife's blog on my ID). I just hope Twitter, Facebook and other sites don't choose to do the same thing, or we can call the whole tracking bit a wash.