Tuesday, 14 June 2011

It's a good Facebook story. Run it and to hell with the data.

Pretty much all of the newspapers this morning are running a story on a possible decline in Facebook's UK user-base. It's fairly predictably been re-tweeted all over the place too.

The story actually came out yesterday and is - no-doubt - a cracking piece of PR for the services of Inside Facebook, who provided the stats.

Now, I've been predicting the decline of Facebook for a while and I'd quite like this story to be true. Our newspapers have done what they usually do with an interesting press release and just run the piece without checking any background to the numbers. "Inside Facebook say traffic is falling" runs the story. "We're not lying, we didn't say it was true, Inside Facebook did."

So where have these decline numbers come from?

From a subscription service called Inside Facebook Gold.

If you trawl Inside Facebook's site (and if you do better than me, then I'd love to hear in the comments) there's very, very little on the methodology that Inside Facebook Gold uses to measure active users. We've got a suggestion that it's tied to running ads on the Facebook platform and using Facebook's own tracking metrics and, er, that's it.

That coupled with the fact that so far, this decline in users is a 1-2 month blip and Inside Facebook admit they've seen bigger single month blips before, has my bullshit alert flashing amber. I'm not saying it isn't true, but how are we to know with no idea of where the numbers come from? It certainly doesn't merit blanket national press coverage.

It doesn't merit a multitude of retweets either, but they were at least as predictable as the copy-paste of a press release journalism.

As a quick example of why we need to know... I don't use the web interface to Facebook nearly as much as I used to, since I got an Android phone. The Android app doesn't deliver ads like the web interface does. For all I know, I've dropped off Inside Facebook's tracking even though my use of Facebook has undoubtedly increased since I got the phone.

That really could be one of many causes for a measured decline, depending on how the methodology works. UK Android handset sales have exploded.

Stats with no hint of the methodology, from a company you've never heard of before, that tell a story you were hoping to tell and that make a cracking PR fluff piece for the papers...

Bullshit Alert. Every time.

No comments: