Blog Post

Facebook Dip Seasonal, Here is Why!

Stay on Top of Enterprise Technology Trends

Get updates impacting your industry from our GigaOm Research Community
Join the Community!

[qi:011] Update: Hey Michael when you say, “Om Malik says Facebook traffic is tanking” it is not quite true. As I said in my original Facebook-traffic post from late last night, I don’t quite buy it. My exact words, “It just doesn’t make any sense! Unless of course, the unthinkable is happening. I hope to chat with FB tomorrow and get more on this.”

Well, I haven’t been able to get Facebook to respond, but I did talk to others and have gotten to the bottom of this “dip/tanking” issue. Paul Sutter, co-founder & president of Quantcast explains that the dip we are seeing is because of the panel model adopted by comScore.

ComScore has a panel that has a bias toward Internet users who log on from home. The same is true of all measurement panels – Nielsen, Hitwise or Quantcast. As kids go back to school, they vanish from the panel, even though they are still using Facebook, from school dorms. “We have a hybrid model, a combination of panel and direct measurement,” he says. Direct measurement means a website places a Quantcast pixel on their website and that is used to measure. Sutter points out that their own panel also saw a similar dip, but in reality if direct measurement was in place, the traffic should go up.”

ComScore’s Andrew Lipsman disagrees with Quantcast explaination.

Quantcast’s claim that the comScore panel has a home bias is nonsense. comScore’s panel includes home, work, and university audiences and each sample is weighted and projected according to the size of the respective Internet population.

Of course, it would help, if Facebook explained what’s the story.

Facebook spokeswoman just got in touch from London and explains….

It’s important to note that we focus on active monthly users rather than registered users. Active monthly users in the U.S. grew from 17.7 million in August to 19.4 million in September. This active monthly user base has been growing at a rate of about 3 percent per week since January 2007 which means we have been doubling about every six months. Globally, we ended September with 44 million active monthly users as compared to nearly 9 million active monthly users for the same month in 2006. Our page views have also been increasing and went from 54 billion page views for August to 57 billion page views in September.

That settles it for me!

(My original update is below the fold with information and data from ComScore and Hitwise.)

Well well well….thank God for my disbelief. The 9.3 percent decline in unique visitors and 3.8 percent decline in page views for Facebook in September are part of a historical pattern. Hitwise and comScore have both confirmed this information. Andrew Lipsman of comScore left a message in response to our previous post:

Last year, there was a similar dip in visitation at Facebook, which suggests there are some seasonal factors at play. Another important factor that has not been mentioned is that September has 1 fewer day than August. Assuming all other things are equal, there would be approximately 3 percent less traffic in September just because it has fewer days. So a fair portion of the decline can be explained by that factor. Always be careful in looking at one month trends, because seasonality can always be an issue.

The Hitwise spokesperson had this to say, in addition to the historical chart they sent our way:

There is a small dip for Facebook in terms of U.S. visits it appears for the last few weeks, but looking back over time this could be more seasonal than anything. We’ll know more in the coming weeks.

hitwise facebook stock data

We are still waiting to hear back from Facebook. More when I have that information.

Update: Andy from Marketing Pilgrim got this information before I did. Good work Andy!

22 Responses to “Facebook Dip Seasonal, Here is Why!”

  1. Gotta love the guy who offers the greek week explanation. Dude there are not enough greeks in Greece to account for that drop.

    As I mentioned on your previous post. Why are you comparing a chart of ‘total uniques’ with a chart of ‘US market share’???

    These graphs are two completely different animals. Social sites don’t take as much of a summer hit as the net in general so the broad based net pick up at the end of summer last year showed a drop in FB market share. Yet they still increased total uniques last year. That’s not the case this year.

    Nice try with byzantine quantcast explanation, why don’t you stop rationalizing and just accept multiple independent streams of data from other sources that agree with dip??

    So… you were wrong.

  2. Question? Do these statistics companies still only using ie to collect their data? How would many college kids going from their parents pc to their new shiny mac that runs safari/firefox effect these numbers?

  3. Also keep in mind that as students return to school, Greek rush week events occur. At UCSD at least, all of the girls in the Greek system disable their Facebook accounts and set MySpace privacy to maximum during the week due to Panhellenic regulations. I’m pretty sure this has quite an effect.

    Also, look for the drop when people disable Facebook accounts for Lent. (This is noticed internally, though the outward effect on stats may be minimal.)

  4. Andrew Lipsman, comScore

    Quantcast’s claim that the comScore panel has a home bias is nonsense. comScore’s panel includes home, work, and university audiences and each sample is weighted and projected according to the size of the respective Internet population.

  5. Thanks for researching this weird dip.
    In Israel, Facebook has become extremely popular in the past month or two. Everyone’s talking about it, and some bloggers that objected FB publicly have also joined in…

  6. Since all the measurement panels (Comscore, Netratings, Quantcast, Compete, Hitwise) share this same home-bias flaw, all of these panel measurements will show the same flaw.

    Direct measurement is really the only way to solve this problem.