Socialbakers calls out The Guardian, says there’s no Facebook exodus

Look, can we all stop writing panicked articles about Facebook losing millions of users based on Socialbakers stats? It seldom turns out well.

The social analytics service has, for the second time, written a blog distancing itself from an article written by The Guardian which suggests that Facebook has lost large numbers of users in its largest markets.

Using stats from Socialbakers, the article suggests that Facebook has lost nearly nine-million US users over the past six months along with some two-million in the UK. It also claims that people are switching off the world’s largest social network in Canada, Spain, France, Germany and Japan. It largely puts these losses down to fatigue.

Socialbakers thinks this is bunk. CEO Jan Rezab notes that “the Facebook stats found on our page are not primarily intended for journalists, but rather Ad estimates for marketers”.

“We state, quite clearly, on our site that these figures are rough estimates and cannot be used to determine Facebook traffic,” he adds.

Rezab also the notes that this is not the first time his company has had to clarify the nature of its stats with The Guardian. The previous instance saw The Guardian reporting that Facebook had lost a number of UK users in December 2012.

At the time Rezab explained that the data needs further interpretation to be understood in context, and that monthly active user counts can be easily influenced by casual users.

The latest Guardian reports follows fast on the back of what the company called a “massive misunderstanding” about its use in South Africa. Reports from that country suggested that it had shed nearly a million Facebook users. Those reports had a people in the country scratching their heads, especially considering that all previous indications had suggested that Facebook use in the country was on the up.

In reality, there a re a number of reasons for fluctuations in social network user stats. While more than half of all Facebook users log in every day, there are a significant number of people who can’t be bothered and log in a few times a year — or once every 31 days, missing the 30 day period measured by the ad interface. Log ins can fluctuate, as people head to Facebook to share their thoughts on elections, major events and significant holidays. Measuring all those users also becomes problematic as the tool does not always count mobile users or those logging in from work, who could move around and be categorised in different locations or classified as ambiguous.

Ultimately when it comes to reporting on stats from companies like Socialbakers, or any stats in fact, its down to us as journalists to ask ourselves a couple of questions: who are the stats being produced for and how did the people presenting the stats come up with their numbers?

For The Guardian to make the same massive mistake twice however seems a little bizarre. It’ll be interesting to see what its explanation is, but it’s unlikely that the usual mutterings about overstretched newsrooms and stressed out editors is going to cut it this time.