Yes, Facebook's popularity will decline

Yes, Facebook's popularity will decline

Summary: Remember Compuserve? AOL? Yahoo? Secondlife? Facebook? Yeah, you remember Facebook now - but not for long. 'Social' services have a definite lifecycle and Facebook may be on the downside.


Remember when Princeton researchers published a paper asserting that Facebook would lose 80 percent of its users by 2017, based on analyzing Google Trends data for Facebook searches? Everyone had a good laugh over that, not least of all Facebook, whose clever comeback used similar data to find that Princeton was in trouble.

Yet a second study (Strong Regularities in Growth and Decline of Popularity of Social Media Services) by researchers Christian Bauckhage and Kristian Kersting of Germany's noted Fraunhofer Institute has reached a similar conclusion. And they looked at 175 different Internet services — including Friendster, Myspace, Secondlife, Yelp and Zynga — across 45 different countries.

Their conclusion:

"...our analysis reveals that collective attention to social media grows and subsides in a highly regular and predictable manner. Regularities persist across regions, cultures, and topics and thus hint at general mechanisms that govern the adoption of web-based services."

In other words, many social internet services are like fads. And like most fads, interest tends to peak and then decline over time.

Peak Facebook? 

The authors note that search terms do not represent absolute user interest — people may stop searching for Facebook because they have it bookmarked or use an app — but a large body of research has found Google Trends data is a good index of consumer interest.

They graphed their findings for several major services, including Facebook and Amazon. Here's the Facebook graph:

Screen Shot 2014-07-09 at 10.39.23 PM

They used several different diffusion models to fit the data and found that the shifted-Gompertz model had the best fit across the data. That curve also gives a massive falloff in Facebook's relative interest.

But this isn't true of all Internet services. For example, Amazon's curve looks very different.

Screen Shot 2014-07-09 at 10.38.59 PM


As the authors note:

Due to the simplicity of the diffusion models considered here and because of their apparent empirical validity and theoretical plausibility, the results reported in this paper therefore provide a new baseline for research on the mechanisms and long-term dynamics of collective attention on the web.

The Storage Bits take: To be clear, the research doesn't predict imminent doom for Facebook. 

But it does suggest that the Facebook fad faces user interest declines just as other services before it have.

Interest has already declined among younger teenagers — one reason Facebook paid $19B for WhatsApp — and Facebook has more challenges ahead. Concerns over information security, selling of information, and constant UI changes have soured many people I know.

But given the "social" nature of social services, is it so surprising that they are more like fads than utilities? Facebook will struggle to stay relevant to users' lives.

Comments welcome, as always. How has your use of Facebook changed in the last year?

Topics: Storage, SMBs, Social Enterprise

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • Anyone remember...

    AOL Neighborhoods,

    Fashion fades over time and is reborn in something with a fresh skin/name.

    I don't use FB, but am thinking of running a private/controlled FB-like software on my own server where I can let family and any of their friends use it. No AD's no tracking, just staying connected and sharing as desired.

    FB is good for re-connecting with people, but once you've done that...just both move to something more secure/private on a private server. That's my thought anyway.

    Time will tell.
  • i've never used any social network

    due to privacy concerns over the aggregration and selling of my information. Another factor is fear of what a 'friend' might say that reflects on me professionally. I don't need to risk being associated with someone who makes a racist, xenophobic, or homophobic comment.

    I am hoping a lot these decline due to fears over privacy and that the US will get some privacy laws with teeth.

    I don't allow my girlfriend to post very many pics that include me on her Facebook account. I don't
    • You're using one

      You're using one right now. What do you think these forum system are?
      Buster Friendly
      • i don't consider these social networks

        i'm not sharing any details of my life, i don't use the same username anywhere else, & i use disposable emails.
        • It's interesting

          Before the my spaces and facebooks (when I was young) this sort of thing is exactly what an online society was - forum, chatrooms, irc.

          I spose it comes down to you investment - if you make friends on here/invest in the comments section, it is a social network. If you just comment on the odd post or whatever it isn't.

          I would disagree that a social network automatically requires the sharing of your personal data, I see that as a modern definition from the likes of facebook. When I used chatrooms and irc, I had very good aquantancies I knew only by their username/handle.
        • But you are revealing personal details...

          In the body text of your initial post, you disassociate yourself from racist, xenophobic, or homophobic remarks by others. This categorizes you just as surely as if you had taken a poll. Then, you mention your girlfriend. Again, this removes some possible options from a profile that could be constructed about you.

          It could be that you're fabricating these qualifiers, in which case you'd be doing what others have recommended be done to obfuscate the reality of one's life by constructing a faux life for yourself online.

          Bottom line - if you're online, you're leaving footprints.
          • Facebook

            and other sites people consider social networks are quite different since people use real names and many other specifically identifiable forms of information, whereas with comments pages, etc, all people need know are your ideas, and ideas are far less individual. (Yes, I know people can use fake names, etc, on social networks, but that's besides the point).
    • Privacy Concerns...

      ...are definitely valid with Facebook, but I'm not sure how having others post xenophobic etc comments would cause harm so long as you set the security filter not to share everything on your page with the general public.
  • i used to love Compuserve back in the day

    also BBSes. But they weren't aggregrating & selling your info, either.
    • Wasn't it like

      Wasn't it like $24/hour for the premium stuff?
      Buster Friendly
      • it was

        if you used 9600 baud. It was much cheaper for 2400, but I don't recall the exact pricing.

        i used a program called TapCIS that would connect, download forum subjects, & disconnect. Then i could mark the threads I wanted to download, & it would connect & get only those. It was extremely efficient & severely limited how long you were online.
  • I was on FB for a couple of months

    until I realized what a PITA it was keeping up their UI and security setting changes. When I have to read articles to keep up on something that is supposed to be "fun" then they've lost me.

    I like GotThumbs idea of a private FB-style server. When you strip FB to its essentials it is simply a very easy way for non-techies to build a web presence. Why hasn't the open source community done that?

    R Harris
    • "Why hasn't the open source community done that?"

      We have. Or at least we've tried. There was this thing called Diaspora a while back.
  • Facebook dying under weight of it's arrogance.

    Facebook dying under weight of it's arrogance.

    The political bias of Facebook has been evident when pages created that are anti Obama are quickly shutdown as hate speech, but pages calling to the hanging of Herman Cain for being a republican house slave and an uncle Tom were left up for over a year until after he withdrew from the race for the presidency.

    As recently as this week Facebook pages referring to Hillery Clinton as the Killer of Benghazi and supporting a criminal negligence investigation were being removed as hate speech while pages calling for the murder of a high school cheerleader have been left up for months.

    If Facebook cannot remain unbiased in its policy enforcement than it deserves to die when the public walks away from it.
    Hates Idiots
    • True - People will leave like they left MSNBC, CNN, the lapdog media etc

      When you take a political stand you automatically offend at least 50% of your audience. Johnny Carson knew this. He made fun of Democrats and Republicans and that is why he was so popular. Compare him to David Letterman - nothing more than a hate filled frustrated old man who is leaving with low ratings.
  • The first rule

    The first rule of analysis is you never assume past events predict future trends. The scale difference between those other sites makes them completely incomparable too. Facebook would be better compared to Google for scale.
    Buster Friendly
    • To The Contrary

      Matching similar items, one in the present to ones in the past, is an effective analytical technique.

      Baseball statistical analysis uses this technique. It makes us think that a 21 year old who hits .300 in the major leagues is expected to have a better career than one who first hits .300 at age 25. The future may differ from the past, a ball player may have bad luck and get injured, but in terms of allocating resources, the 21 year old is more valuable in terms of money to be spent and players to be traded for.

      Whenever we discuss market saturation or early adopters, we are indirectly referencing this way of thinking about events, in that it comes from past studies and recognition of the same patterns of product adoption and market peaks and maturations of automobiles, telephones, phonograph records, washing machines, and so on.

      One has to make an effort to find historical analogs that approximate the current situation. It is a bit presumptive of you to say that the use of past portals and social sites as analogs was thoughtless.

      I can see why these were chosen. Social media has a network effect, the competition is with entertainment, and the battle is over time and attention. AOL, MySpace, and 1970s CB radio all were things people did because people they knew did it. Scaling, to my mind, only relates to the potential number of network nodes, but all were large enough to have significance.

      Returning to your assertions. A lot of reasoning is inductive. Normal distributions and statistical plots become more similar when sample populations are larger and there are correlative factors.

      I'd say that evidence indicates Facebook has been keenly aware of the AOL/MySpace pattern and its applicability to them and they are doing what they can to find what their faded predecessors missed. I'm doubtful of their ability to break the pattern, because I think it's a fundamental given that people change how they use their time and attention, whether it's in one's lifetime or as a population. Yep, I'm assuming that people stay people, daring to defy your first "rule."
      • I can see why too

        I can see why too because on the surface they seem similar but not when you actually look at the details especially scale. Would you also Google search popularity will drop because of Yahoo, webcrawler, lycos, AltaVista, excite, hotbot, jeeves, dogpile etc all did? It's just bad analyses all around.
        Buster Friendly
        • That's why I put the FB and Amazon charts in the post.

          I see sites such as Google and Amazon as utilities - search and shopping - whose appeal transcends fads such as MySpace and FB. Until something better comes along - as happened with the search engines you mention - they'll continue to garner the traffic.

          R Harris
          • I Think That's Utterly Nonsensical

            The Yahoo search engine declined because Google was better. Yahoo Groups declined because a) they kept altering it into something useless, and b) better services came along. Myspace declined because people found a better alternative. If a better online retailer than Amazon comes along, it too will have to change or decline. We haven't seen people stop using services along the lines of anything you've mentioned just because of the passage of time. They replaced those services with something better. All services peak and plateau, but true decline will only occur once a service has become superceded by something else, or otherwise becomes irrelevant, and humans will always find the need to socialise, brag, share information, etc.