Facebook may lose 80% of users by 2017, say Princeton researchers | Hacker News Reader

Facebook may lose 80% of users by 2017, say Princeton researchers

theguardian.com

170 pointsadrienm159 comments

So the argument is:

- Facebook's growth curve looks like that of an infectious disease

and

- "Ideas, like diseases, have been shown to spread infectiously between people before eventually dying out, and have been successfully described with epidemiological models,"

therefore

- Facebook will die out as per the epidemiological model.

Allow me to respond to that in a manner appropriate to the level of rigorousness of the argument:

Bullshit. Shoddy analogy at best, wilfully stupid attention whoring at worst.

The largest indicator of bullshit (as I've commented below) is that based on their model, ALL infections of this type go to zero. They could have modeled Apple, Google, Christianity, or Bitcoin and reach the same conclusion (albeit with different timeframes, I'm sure).

The rise in consumption of hamburgers follows an infectious curve-- thus we predict that humanity will be hamburger-free within 2090. /s

Agreed. Not everything is a fad.

I am rather surprised by this constraint on the model, as an extremely straightforward modification to the standard SIR model allows for diseases to reach an endemic equilibrium state.

> Bullshit. Shoddy analogy at best, wilfully stupid attention whoring at worst.

how did you get so many karma points using such coarse language? It brings down the level of discourse.

Even pg noted that top comments are largely awful. I find just skipping the first one makes reading HN a much more interesting experience.

Your observation reminded me of the Million Short search engine that skips past the first N ranked hits.

Is there a name for the phenomenon that the first entries in any contested ranking tend to be less interesting?

There's something about the ranking systems used on the web in the way they tally the nearly-effort-free actions of hordes (clicks, likes, upvotes, star ratings) that means they represent one kind of magnitude (statistical average impulse behavior) but very little of another (deeper reflection & significance). Deeper stuff tends to see less impulse attention and therefore sinks down the ranking, but is there some other way to identify it?

It's intrinsic vs perceived values in ranking. You could have some rough measure of intrinsic value of a comment by analyzing its content: length, grammars, wording sophistication, etc. It's definitely not trivial and maybe open to spams.

It's the same phenomenon that creates pop music, pulp literature, prime time TV, summer blockbusters, two-party voting systems, and AAA games. The mainstream tends to be safe, non-threatening, and appeal to a wide variety of audiences. There may be a genre/topic/idea that better suits people, but it takes work to find it and not everyone wants to put in that work.

Not everyone will like Thelonious Monk, and the same people who like him might not like Americana or electro house or black metal. But no one complains about smooth or cool jazz. No one complains about mainstream country. No one complains about techno, no one complains about hard rock on the radio. They're watered down, appeal to wide audiences, and are non-offensive in their broad market. It takes no effort to like mainstream culture, and it doesn't demand anything from your intellectually. You're going to like the Billboard Hot 100, or at the very least you're not going to hate it, because it's been manufactured to be liked by everyone.

Same thing with high rated comments in a lot of cases. While the poster might not actually be saying anything, might not have a point, might not have any intellectual depth to their comment, their comment offended no one, was easy to understand, and can be appreciated at face value without needing to think about it. If you can get people to like your ideas without making them think, you're going to be popular. If you make them think, they have the chance to decide if they like what you're saying or not, and if you make them think, you give them the chance to decide that they don't like you.

Is there a name for the phenomenon that the first entries in any contested ranking tend to be less interesting?

Sturgeon's Law, maybe.

There's also the Wadsworth Constant. (Linking to life-hacker rather than the reddit thread as it gives more background: http://lifehacker.com/5849342/skip-past-the-boring-intro-on-... )

> Is there a name for the phenomenon that the first entries in any contested ranking tend to be less interesting?

'The fanatics are picturesque; mankind prefers to see gestures rather than hear reasons.'

the topic doesn't warrant a high level of discourse

The level of discourse was brought down to below zero by the article. I'm actually raising it, if anything. This is an appallingly bad "forecast" and I'm slightly ashamed on behalf of Princeton for allowing their name to be associated with it.

Also, if you look at my comment history, I don't habitually use "such coarse language". I only use it in appropriate circumstances - like in response to this article. The coarse language has little to do with the karma.

Edit: worth noting that the title of the article on Guardian is "Facebook will lose 80% of its users by 2017", not may. That has been editorialised by the submitter. The article is even worse than you may think if all you read was the title.

Well I think ur bullshit.

What a great way to communicate.

I wasn't communicating with the authors of the article, just with the commenters here. I was not even insulting the authors (only the "forecast"). You, on the other hand, are throwing insults at people (me in particular) - and doing so in a largely inarticulate manner.

Another pretty poor analogy, I suppose. I can see why you liked the article. </sheldon-joke-mode>

it's only a bad forecast if it doesn't happen. if it does happen, then it's a freaking awesome INTUITIVE forecast. Okay, so there's no science behind it? It doesn't matter. There's no science behind a lot of things, yet they work.

When you recognize insane patterns and connections, much like an Artist as opposed to a Scientist, things can work. All that needs to matter is that the intuition has to be right. The science can come after.

Oh give me a break. He made a very good point -- the article takes the epidemiological model analogy too far and in a non-rigorous fashion. You are the one "lowering the level of discourse" by focusing on form instead of content and by writing in clichés.

we don't even know yet if it takes it too far. What happens when 2017 comes along and facebook has lost exactly 80.00% of users since 2014? Will it still have been an analogy taken too far?

Yes - 'a broken clock is still right twice a day', and scientific conclusions cannot be accurately or scientifically applied via analogy.

i guess that could be true. The other day i was trying to convince someone that you cannot use a mathematical conjecture if a proof has not been found yet. I think I succeeded.

Why do you believe this analogy will not apply to Facebook?

Social media is called 'viral' for a reason: because its spread is apparently analogous to infectious disease. This analogy certainly does seem to apply to other memes and social networks of the past - why would Facebook be an exception?

The argument isn't that it won't apply to Facebook. Rather, the assertion that the analogy will apply is unscientific. It is an assertion of the form: A and B have behaved the same so far. Thus, A and B are of the same nature, and will continue to behave the same indefinitely--an improperly drawn conclusion

"Viral" is just a word. The fact that we have repurposed it to refer to electronic media does not magically impart to electronic media the properties of biological viruses.

The conclusive headline (use of the word will) is the Guardian's editorial decision, not used by the scientists, if that's the main objection.

The article has a more measured statement: "Extrapolating the best fit model into the future suggests that Facebook will undergo a rapid decline in the coming years, losing 80% of its peak user base between 2015 and 2017."

Observing common network effects and using it to make predictions seems like valid scienfitic research to me. The hypothesis may or may not be correct, but it's a (very) testable and falsifiable hypothesis. The science is good.

http://arxiv.org/pdf/1401.4208v1.pdf

Edited to add - jarrett: I accept your point that the research could be more thorough; it's a shame only Myspace data appears to have been used. However, I think this is a valid thread of research and the word 'bullshit' was unwarranted.

Yes, it is a testable and falsifiable hypothesis. So is "George Washington will be elected President in 2016." Nonetheless, we can fairly argue that both hypotheses lack a firm grounding in the evidence.

> Observing common network effects and using it to make predictions seems like valid scienfitic research to me.

It's valid I suppose, but purely speculative at this point. It would be much more meaningful if a study demonstrated that social networks almost always behave like viruses in both their growth and decline, than to simply observe a similarity and growth and extrapolate the decline.

>Nonetheless, we can fairly argue that both hypotheses lack a firm grounding in the evidence.

Then make that argument. Because just saying "I don't like it" over and over is not making any such argument.

Which of the above posts is "just saying 'I don't like it' over and over?" This thread is a discussion of the structure of a certain argument. To recap:

swombat3 explains why he thinks Princeton's method of reasoning by analogy is flawed. ronaldx disagrees, asking why the analogy wouldn't be applicable. I reply that, in my view, ronaldx has disagreed with a different point than the one swombat3 made. ronaldx replies in defense of Princeton's claims, citing the testability and falsifiability of the hypothesis. I reply that there is little evidence for the hypothesis, despite its testability and falsifiability.

So, all participants have advanced specific, well-articulated arguments throughout. I don't see anything similar to "I don't like it."

Humorously, this discussion has become rather meta by now. We're now arguing about the form of argumentation used in a discussion that was itself about forms of argumentation. It's like we're living out an LSAT question. I find that sort of conversation hard to resist, so I suppose I'm just rising to the bait.

>swombat3 explains why he thinks Princeton's method of reasoning by analogy is flawed. ronaldx disagrees, asking why the analogy wouldn't be applicable. I reply that, in my view, ronaldx has disagreed with a different point than the one swombat3 made. ronaldx replies in defense of Princeton's claims, citing the testability and falsifiability of the hypothesis. I reply that there is little evidence for the hypothesis, despite its testability and falsifiability.

There's not an argument in this entire paragraph. Is it meant to be a list of pointers to arguments?

Neither "X is(n't) flawed," "X is(n't) applicable," nor "X has (very little/lots of) evidence" is an argument.

An argument for falsifiability is an argument that the statement of a particular belief has content and is not just a trick of language, but can't be said to be an argument for or against that particular belief.

> There's not an argument in this entire paragraph. Is it meant to be a list of pointers to arguments?

As stated, it's a recap, or summary, of arguments previously made. So no, the arguments themselves are not contained or fully expressed in the summary. In this sense you're correct that it's a list of pointers to arguments--arguments which can be readily located earlier in the same thread.

> Neither "X is(n't) flawed," "X is(n't) applicable," nor "X has (very little/lots of) evidence" is an argument.

Correct. Those are conclusions. The arguments presented in support of those conclusions are available earlier in the thread.

> An argument for falsifiability is an argument that the statement of a particular belief has content and is not just a trick of language, but can't be said to be an argument for or against that particular belief.

Agreed. That's why I said the falsifiability of Princeton's hypothesis was not an argument for the likelihood of that hypothesis. I think we're on the same page about this.

Follow-up reply to ronaldx's edit: I personally would not have (and didnt't) use the word "bullshit" in this context, so I agree with you there. I still agree with swombat3's overall argument, though.

Where does this "swombat3" come from? My username is swombat, not swombat3.

I believe "3" comes from "3 hours ago", which was displayed next to your username at the time when user jarrett mentioned you. So I presume it's just a slight copy-paste typo.

Wow, that was a really silly goof on my part. Too late to edit I'm afraid. My apologies.

No worries, it just puzzled me, particularly since it was repeated across multiple comments!

Not everything needs to be scientific. Even a lot of mathematics is figured out first using intuition. The proof may come much much later.

Some people do have intuition that is superior to other people. They can see patterns and connections in things and intuit the correct answer, or predict the right path. Much like Steve Jobs. It's an artistic, intuitive way of looking at things rather than a purely scientific way. Sometimes it doesn't work; sometimes it does.

Fair point. Indeed, not everything needs to be scientific. Though for empirical claims that science is capable of addressing (as opposed to, e.g., what is morally right), claims that are based on sound scientific reasoning are more worthy of attention than claims that lack such a basis. I consider Princeton's claim to be one which fits this criterion: It is empirical, and it concerns a topic which is amenable to scientific inquiry. Therefore I find it appropriate to critique the claim in scientific terms.

Why do I believe it will? Seems like a pretty extraordinary claim. There are plenty of other tech businesses that saw meteoritic growth, and they didn't all peter out like diseases.

Many forces that are at play with diseases (immunity for one) are absent with tech products, and a big one from tech products (the fact that the tech product is built by a company filled with intelligent people who want it to succeed) is absent from diseases.

Infectious diseases seems like a pretty tenuous analogy for a 10-year old tech business with thousands of employees.

As someone partial to the concept of memetics and Universal Darwinism, I happen to like when biological/Darwinian principles are applied to higher-level concepts. That being said, there may be ways to differ with the article's analysis even within the biological analogy: it's possible that Facebook and its users have a symbiotic relationship (social features in exchange for traffic/attention/ad impressions) in which case the user base may continue to stay with the site until an even more advantageous website comes along.

Their model is based on search query volume. Is it possible people simply don't need to Google for Facebook any more?

An analogy can an have predictive power when there is no evidence?

It depends on the virus. Most viruses you can only get once, then you fight off the infection (according to the model, anyway). Facebook may be a little more like herpes.

Perhaps the main reason why we are so strongly inclined to talk of the head as the locality of our thoughts is this: the existence of the words "thinking" and "thought" alongside of the words denoting (bodily) activities, such as writing, speaking, etc., makes us look for an activity, different from these but analogous to them, corresponding to the word "thinking". When words in our ordinary language have prima facie analogous grammars we are inclined to try to interpret them analogously; i.e. we try to make the analogy hold throughout. -- We say, "The thought is not the same as the sentence; for an English and a French sentence, which are utterly different, can express the same thought". And now, as the sentences are somewhere, we look for a place for the thought. (It is as though we looked for the place of the king of which the rules of chess treat, as opposed to the places of the various bits of wood, the kings of the various sets.) -- We say, "surely the thought is something; it is not nothing"; and all one can answer to this is, that the word "thought" has its use, which is of a totally different kind from the use of the word "sentence".

^ that's Wittgenstein and not me writing there, but I believe that responds to your claim adequately.

Right but we can do science here. Myspace, Friendster, Google+... We know a lot about how social networking plays out. There's enough of a sample size there to start doing some observations and making some hypothesis.

No, there's not. In what universe is (optimistically) 5 data points a sufficiently large sample?

If you have 6 things, then 5 is as good as you can get to extrapolate about the 6th.

Yes, when you have only 6 things to look at you might take an educated guess based on what might have happened to the first 5. It's the sort of contextual, gut-driven judgement people call a hunch.

That's fine for a conversation about facebook after a few beers... it's pretty silly to take that and write a flippin' paper about it and present it as "research".

It's not five data points.

It's five trials with many, many data points - what you're fitting is a time series.

Hell, for some epidemics (epidemic modeling being what I do for a living) I'd kill for five example cases to fit a model against.

So there's 'we know the way social media plays out' and there's actual science and the two phenomena have nothing in common beyond 'we drew conclusions based on what we saw'. The difference is that the latter is rigorously controlled, tested and peer-reviewed for according to a vetted methodology, and the former is no different from saying 'basically all asian people I have met are good at math, thus this new asian person I have just now met must also be good at math'

"So the argument is:"

It's an interesting way of looking at it, but there doesn't need to be a more complicated explanation for the eventual fall of Facebook than the fact that it's always been the pattern for social networks. There's fundamentally more novelty than real value to it so people will lose interest and it will be supplanted by whatever else has the property of being novel. (And lucky, and in the right place at the right time.)

You just don't need something like Facebook to stay connected to your close friends, and the interactions you have with your larger "network" are inherently shallow and capable of being disrupted in the move from one social network to another or disposed of completely.

it's always been the pattern for social networks.

For individual social networks, yes. But what if we lumped all social networks together, and looked at aggregate usage over time? It would constantly increase. I think that instead of dying out, it will plateau until we reach saturation. Other communication technologies follow this model: radio, television, telephones, cell phones, etc.

Rather than being a novelty, I think that the past decade has demonstrated that online social networks are something most people want and will continue to use.

To be clear, you and I are making contradicting factual predictions about the future. Neither of us knows which prediction will turn out to be true. At the moment, the available data fits both of our predictions. We'll know in about 20 years.

There are so many people that are saying this can't happen for a variety of reasons, at least that is the common reaction I am seeing. I think this type of denial is misplaced.

That search traffic trend (assuming it is accurate and not some type of bug) is very worrisome, incredibly worrisome.

I did a comparison of Facebook, Yahoo and Google search trends in Google Trends. Neither Yahoo or Google saw the same type of decreases, and Google has tons of mobile accesses too, especially since it is built into the UI of Android phones, no need to search for Google.

http://www.google.ca/trends/explore?hl=en-US&q=/m/02y1vz,+/m...

I don't think one needs to use a fancy disease model parallel in order to predict doom for Facebook, you just need to look at those search traffic trends.

I think this article is generally right in the broad strokes if something doesn't change -- can there be a Facebook game changer here?

Navigational searches on Google from users going to Facebook are being displaced by direct access to the FB app on their smartphones, which is not measured in Google Trends.

That, and the assumption that FB is equivalent to MySpace circa 2006, make me completely dismiss their claim.

That is why I did the comparison with Google, Google is built into the UI of Android and it is also integrated into the Omnibar of Chrome (or whatever it is called) so there is also a displacement of search traffic too on Google, but it didn't see those declines. Now this isn't a perfect control, but it is something...

FB habits are very different from Google or Yahoo. Last earnings call, FB said that half of their daily users are mobile only. About 25% of Google searches are mobile.

http://techcrunch.com/2013/10/30/nearly-half-48-of-daily-use... http://www.businessinsider.com/mobile-sends-a-quarter-of-sea...

Here's a hint: people don't Google for Google.

iirc, at least a few years ago `google` was actually one of the most popular Google searches.

I'm sure some people do search for 'google' but it's definitely not as popular as Facebook.

Put another way, if we're going to follow the thesis that Google and Facebook search traffic are equivalent, the graph linked above would mean that Facebook is a much larger business than Google.

But that was probably because people just typed "google" in the Firefox's address bar a lot.

>That search traffic trend (assuming it is accurate and not some type of bug) is very worrisome, incredibly worrisome.

Why would anyone be Googling "facebook" at this point?

For the same reason people google for google or gmail: they don't know the url.

Yes, there really are people like that. Lots of them.

Didn't introduction of omnibars in browsers killed that trend? The second time such an user tries to type in "gmail" into their browser, they'll get a gmail.com completion, not a search query.

The thing about infectious diseases (the model they are using) are that the virii just die and that's that. What is not clear is where 80% of Facebook's users would be in 2017. Google+ ? Somehow I don't think so. This would be much more compelling for me if there was some Facebook alternative that was growing rapidly in the 14 - 24 demographic but the Princeton guys don't mention one, and I've not come across it yet. Its hard for me to believe that 80% of Facebook's users will stop using it and go back to no equivalent.

> Its hard for me to believe that 80% of Facebook's users will stop using it and go back to no equivalent.

Really? Why? Who's to say that "social media" (at least in the sense of a unified platform with a real name) wasn't a fad?

Plenty of people are moving to a more fragmented internet life, with personal messages on Snapchat, photos on Pinterest, discussion on Reddit, etc.

I agree, things that I used to do on Facebook, discuss topics and articles, say whats on my mind, play games, and post photos can all be done better else were. Where I see a shift in my Facebook friends list is most people have inactive accounts but are active on Twitter, Redit etc.

Case in point: in the 1990s, portals were big, and every popular website wanted to be a portal. Likewise, between 2006 and now, most popular websites integrate social features.

What people seek in Facebook, they'll start finding in more directed services. Facebook was made for a world where smartphones and affordable or free fast mobile internet access was rare. A shrunk version of the regular web version can be enough for adults who are already used to Facebook, but teenagers and children will flock to user experiences that explore the advantages and idiosyncrasies of smartphones.

Maybe Facebook the company will own some of these new services and interface ideas, but Facebook the old timeline/friends/profile service will get obscured by the new, shiny, more intimate new services.

Facebook definitely seems worried about smartphones. They tried to make their own phone interface, and (despite some good reviews) it was a commercial disaster: http://news.cnet.com/8301-1035_3-57583557-94/heres-why-the-f...

Agree! Indeed, the comparaison with portals springs to the mind. While the amount of items (links, people, services, cultural niches) available increases, the need for more specialised and fragmented avenues of consumption increases. It's not difficult to imagine a scenario when one can pick and mix completely their online experience and still find their their social connections across different platforms, with the use of some meta tool, like an OpenId that also exposes the patterns of activities of one's friends.

As a purely empirical, nonrepresentative sample, I can see that most of my friends on Facebook are fairly quiet nowadays and my own account is virtually dead. It's still quite useful to have a FB a/c in order to authenticate on other apps; maybe FB will transforms itself into the aforementioned meta-tool...? Until it's superseded by a more specialised one...

Snapchat and Pinterest is hardly fragmented internet life and it has every bit as big a world domination wish as FB.

For me, its a scale thing. The 800M users is 80% of Facebook's user base give or take. Seems like more than a fad, definitely an itch was being scratched. For example the migration from MySpace to Facebook was pretty clear, but I don't see the bulk of Facebook's audience heading to some other service.

That said, its entirely possible that "social" is a fad, and certainly that internet social is way to wide a network than old fashion neighborhood social (invest in NextDoor I guess :-) but I'm still convinced that the value proposition of amplifying your ability to maintain a larger web of relationships is still a thing. Not many people are natural super connectors and Facebook takes more ordinary people a lot closer to that.

At some point people will just get sick of having a separate social network (with separate, even though maybe synced, friend lists) for every single thing they do. I for one would love if Facebook ate Instagram, Foursquare and Yelp - I'm already sick of using separate apps for something that could benefit from being tightly integrated with FB.

It seems that that 14-24 demographic is flocking to less general "social networks" and more image-centric or one-to-one social networking apps, like Instagram and Snapchat. Essentially networks that allow them to not see what their parents/aunts/other families are sharing or taking pictures of.

I think the social usage is becoming much more fragmented. I tend to keep up with my family via WhatsApp and Instagram, business contacts via Linkedin and the rest is Twitter.

Hasbro doesn't die just because the current crop of 5-12 year olds outgrow their products. Instead, they simply target the next wave.

On the other hand, social networks seem to adhere very strongly to a particular generation or crowd. I don't find it too difficult to believe that different crowds use different social networks (due to network effects bootstrapping at differing rates in differing groups) and that consequently "the Facebook" of the next generation will be Snapchat or something else.

Facebook's challenge, then, is to figure out how to become Hasbro or else it will become Myspace.

Yes, and as Hasbro does so, they destroy one of my favorite games as a kid (Risk). Sorry for the tangential comment, but I was floored I discovered what they did to the game, as I was searching for Risk (the classic) at local stores to buy for my kids.

First, it's viruses, not virii. And viruses are not the sum totality of infectious diseases.

Second, as I mentioned in another comment, a very easy extension of this model allows infectious diseases to reach a state where they exist at equilibrium within the population. I have no idea why the authors didn't use them.

For such a bold claim, I'm not really impressed with the model. Basically, the authors take the well-known SIR model [1] (susceptible -> infectious -> removed) for simple modeling of infectious diseases and modify its removal term to be infectious as well (rather than the infected population just decaying exponentially).

The big problem I have with their model is that the population flow is still one-directional, so ANY infection modeled with these equations will eventually go to zero. There is no term in the equations to account for social adhesion that would prevent people from leaving, and there is no pressure for people who have left Facebook to join again. The authors have validated their model by curve-fitting it against MySpace's decline after the fact, so there's no real prediction here.

It's an interesting model, but take any predictions based on it with a grain of salt.

For anyone interested in reading on similar subjects, here's a paper I co-wrote in which we incorporate a reinfection term into the SIR model and use it to model viral memes [2,3].

[1] http://en.wikipedia.org/wiki/Compartmental_models_in_epidemi...

[2] http://www.sciencedirect.com/science/article/pii/S0307904X11...

[3] http://stash.synchroverge.com/files/viral_memetic_model.pdf

Looks like this thread was killed by the flamewar detector and then resurrected? Or was it flagged? It troubles me when conversations on HN are just "disappeared" like that.

For those who didn't notice: this article disappeared from the front page, and another related submission (https://news.ycombinator.com/item?id=7105451) eventually rose to the front page, where I commented that this was strange and linked to my post on the matter (https://news.ycombinator.com/item?id=7105414). The newer submission was then killed and this article was resurrected. My post was also briefly killed, though it was brought back with a reduced rank. Not sure what happened, but it's all a bit confusing.

Previous discussion from a different report of the same new: https://news.ycombinator.com/item?id=7104904 (19 points, 3 hours ago, 18 comments)

The first comment is long and interesting: “Um, folks, this is a non-peer reviewed paper written by some Princeton students. [...]” https://news.ycombinator.com/item?id=7103987

Facebook is way more integrated into everything than MySpace ever was. This is not just a post pictures and sell music kind of place (although it certainly is a big part), this is everything from hardcore socialresearch platform to businesses making their entire living off the platform to entertainment and so on.

The thing with Facebook is that it's not for the youngest people anymore it's for those who used to be young (+25)

And so naturally kids aged 17 doesn't want to be were their parents and older siblings are hanging out. That does not mean that FB those on FB right now will drop out if anything I think it will grow.

I know that most lawsuits against Facebook for a litany of privacy violations end up going nowhere.[1]

But there just might be a tipping point, soon, when one of these class action lawsuits actually becomes a significant drain on Facebook's resources.

[1] Campbell V Facebook http://www.scribd.com/doc/195351255/Campbell-v-Facebook

I think it's dead man walking, especially with an unsustainable revenue base.

Care to elaborate?

their price implying they'll have 150% of all advertising dollars spent in any form. Or something roughly like that.

Doesn't mean its dropping to zero, just astoundingly overvalued.

Some random researches with no solid background on this fit a fairly flexible curve to data and people go mad?

Just looking at simple principles falsifies this: Facebook's older public is quite solidified, there's so much built on top of it, etc. Myspace didn't have to enormous infrastructure or the general social network outlook we currently have. Even by pure inertia it'd say it would take more than 5 years for FB to loose 80% of it's users, even if a really promising competitor emerged.

There seems to be a call here for a better-quality top comment. I may not succeed in providing one, but I will try. [Pauses to finish reading article.]

"John Cannarella and Joshua Spechler, from the US university's mechanical and aerospace engineering department, have based their prediction on the number of times Facebook is typed into Google as a search term."

That's a convenient methodology, but of course it's a methodology that cannot be applied to anything before the availability of Google search data, so it doesn't even provide a model for explaining why Google supplanted AltaVista. (Hint: even at its healthiest, I don't think most users of AltaVista found AltaVista by doing a Google search first.)

"They tested various equations against the lifespan of Myspace, before applying them to Facebook."

Myspace, of course, is the classic example of a social networking site that once looked unstoppable, until it pretty much died. But data-fitting to just one data set is probably not enough model testing for making a prediction about a different social networking site.

"But Facebook's chief financial officer David Ebersman admitted on an earnings call with analysts that during the previous three months: 'We did see a decrease in daily users, specifically among younger teens.'"

Now we're talking about some meaningful data that actually pertains directly to Facebook. If Facebook corporate officers are concerned about a trend, maybe that trend is really meaningful to Facebook's business prospects.

"Investors do not appear to be heading for the exit just yet. Facebook's share price reached record highs this month, valuing founder Mark Zuckerberg's company at $142bn."

Well, that's the nub of the issue for anyone who buys shares of Facebook stock. If it still looks like holding Facebook stock allows a profitable exit, maybe there is nothing to worry about here (at least for investors).

The article engages in further speculation about Facebook making a lot of money in the future. That could happen. My prediction about Facebook stands: "Facebook will go the way of AOL, still being a factor in the industry years from now, but also serving as an example of a company that could never monetize up to the level of the hype surrounding it. I used to see friends on AOL. I never felt an obligation to help AOL monetize just because of that. Networks are a dime a dozen. Right now, Facebook is a very convenient network, and I like it. I do not predict that Facebook will make a lot of money because of users like me."

I could be proven wrong, but that's what I think the future of Facebook holds. I admit my methodology is just about entirely drawing an analogy with AOL, which I used a lot back in the day.

I think the methodology is deeply flawed because of the growing number of mobile users who are mostly using native Facebook apps and they definitely won't be Googling for that.

I would prefer to think it means people have realized they can type facebook.com into the address bar and just go straight there. But that's probably just wishful thinking.

fb.com also works and is shorter.

Modern browsers have made it much easier to get to the page you want, I think. If you're in the address bar, probably for most people "f" or at most "fa" will pre-fill facebook.com, then you're talking 3 button-strokes... I'm curious how many one-letter chrome-autofill-suggestions I regularly rely on, but for me, muscle memory dictates t-f-enter to get to facebook (t opens a tab when using vimium, but you can substitute command-t if you like).

Then again I am a developer and I use vim, so I think about things like how many keystrokes it took me to execute a command... :/

It's even simpler, most of what I've seen they suspend their machine with the visible tab on facebook, resume and check what's new.

I agree. What that fact shows is that people are no longer searching for Facebook (either to get the site or find out what it is) but now people either know what Facebook is, or have been to the site enough for it to prepopulate when typing fa..

A better stat (which Facebook provide in one way or other) to base on is the number of active users, time on the platform and devices used to interact.

One difference between AOL and Facebook is that Facebook has hired some of the most talented CS folks of this generation, e.g. Yann LeCun. If their management has the vision to foresee the social network's slide to irrelevance, they can leverage the talent to adapt. This is how it works for companies that have adapted well and lived through and grown amid some of the greatest technological upheavels ever: IBM is over a hundred years old, Microsoft is close 40+, HP is 50+, Xerox, Sony, Intel, Honeywell, AMD, Dow, DuPont, Johnson & Johnson, P&G, 3M, etc. etc.

Hire top talent and make strategic acquisitions ---> Innovate ---> Get lucky with strategy at the executive level ---> Sail or sink.

You could also go the way of Kodak, which always hired top talent, and still lose out, but at the end of the day, predicting success or failure of such a huge entity (in terms of $ and no. of users) isn't as easy as it is made out to be in this paper. I think everyone tends to take a simplistic approach. While AOL came out of the IT boom, so did Amazon, eBay, PayPal, Akamai, and a bunch of other companies which continue to do well.

With the kind of investment that Facebook is capable of attracting, they can either become the next IBM or the next AOL.

AOL hired talented people too.

Maybe, but they had some weird management issues. When they acquired Netscape back in '98, there was a notorious all-hands meeting at which an AOL executive informed the disgruntled Netscapers that they were "perfectly free to leave", and so they did, draining that unit of some top internet engineering talent.

Every place >=1 employee has weird management issues.

The problem isn't with the code -- it's with market & business viability and end-user desirability.

"Get lucky with strategy at the executive level"

I love how the most important part is hand waved away to luck.

IBM, Amazon, Ebay, PayPal, Akamai all had/have a solid, long-lasting strategy around making things that are: 1) Desirable to users/buyers/decision makers 2) Feasible, but sufficiently advanced from a technological perspective and 3) Viable from a business perspective, over something more than a short-run.

AOL and Facebook both did 1&2 pretty well (exceptionally well, in many ways) but #3 both fell short on -- Facebook has never had a really strong monetization strategy and now they're potentially slipping in terms of user desirability (amongst some segments) and still, they're failing to find an outright win with their monetization strategies. Ultimately, I think the OP was pretty spot on.

Executive strategies change and their outcomes depend a lot on luck. And it is difficult to predict. Most of our discussions about Kodak, AOL, etc. (like this one right here) are with full benefit of hindsight. Look at how Google (or any other big company out there) has diversified. My point is just that: when you have hired the smartest people on the planet and are attracting investment, you have the opportunity to become the next great company. Whether it will happen remains to be seen, and cannot be predicted with simplistic analysis. Especially when investors (whose day job it is to watch Facebook) are still happy enough to invest in the company.

Here's something to ponder: if you could spot the flaw in Facebook's business model so easily, why can't the hundreds of investors who have millions at stake see the same flaw? (Corollary: Bearish folks can make a lot of money short-selling Facebook.)

First, don't confuse stock success with long-term financial success -- they're actually not completely aligned. Ideally, they would be, but many, many investors are irrational and/or uneducated. Many more are smart enough to prey on that and speculate/bet that they can ride the stock up and sell it at/near a peak.

Second, here's something to ponder: If the masses are so good at picking investments, why didn't everyone see the housing bubble?

In fact, some people did -- and they're the ones who were beating the drums about the need to pay attention to fundamentals and long-term trends.

Similarly, many people have been pointing out that the irrational exuberance around Facebook is not supported by fundamentals -- and is, therefore, a bubble.

Now, why could I spot the flaw in their business model? I don't actually think there's all that much special to it...it's not actually that hard to point out unsound business plans if you pay attention to the fundamentals… which seem to be so easily forgotten.

I have not said that there isn't money to be made on FB as a short-term investment -- indeed, clearly the stock isn't at its low and it's safe to presume some are going to make money off of it. What I did say was that the business model is not sound. That could change, but thus far, I've not been given reason to be hopeful that they'll do so.

data-fitting to just one data set is probably not enough model testing for making a prediction about a different social networking site.

I've watched the lifecycles of a LOT of social sites, from the beginning days of BBSs. Lots of "but this one is different" sites. They have all gone thru pretty much the same lifecycle ending in stagnation & abandonment around 7 years after starting. Some of them (CompuServe, AOL, MySpace, others which were just as big and even more forgotten) took the social tech world by storm, were poised to take over the world, and garnered lots of investment because they were so darned unstoppable. Most users got bored and moved on. Like Zaphod Beeblebrox, they're dead - they just haven't stopped moving yet. ALL OF THEM.

And Facebook shows absolutely no meaningful differentiation from that pattern.

(I want to say "now get off my lawn", but many may miss the intended humor.)

Well Facebook is 10 years old next month and still growing. Thats a pretty meaningful differentiation from the pattern.

All the other social networks were growing right up until they didn't. They had a similar percentage of the online world.

Right, and Facebook will continue to grow until it doesn't, but that wasn't the point he was responding to:

> And Facebook shows absolutely no meaningful differentiation from that pattern.

The implication of this claim is that 10 years and growing is not meaningfully different from 7 years and stagnant.

You can look at Facebook as having launched independently in a number of different markets.

Initially it was Harvard (2003), then Ivys (March, 2004), then selective colleges, then all edus. It then tapped high-school students (September, 2005). It was opened to the general public in September 2006. Introduction to other countries also generally came later in the game, and FaceBook have had to compete with alternative networks, particularly in China and Russia. (Most data here from Wikipedia's Facebook article).

It's helpful to look at each of these as a somewhat independent social network.

We're hearing now about significant defections (or failures to join) in the < 25 market. Facebook's growth overall in the US peaked ~2010-2011. It's been growing outside the US, but that's a much less lucrative market for advertisers. Breaking out growth / user patterns by introduction point would be particularly interesting as I suspect this would model earlier social network growth more accurately.

Not really. By "around 7 years" I meant rough numbers with a very broad standard deviation. Facebook might be actually 10 years old from first run, but from general social awareness thereof to now is closer to 7, with a tipping point clearly having been reached between young users increasingly heading elsewhere and a growing sense by participants of "I don't want to be here but I have to". Yes, Facebook may have more staying power just because it has hit a billion users and is getting integrated into other devices, but I still see it matching the standard (if a bit elongated) pattern of social media site lifecycle. Those billion users are finding it's hard to be meaningfully social when lumped in with a billion others. Those FB-integrating devices can and will jettison FB on a moment's notice.

In 46 years I've seen a lot of "but this one is different" claims in a lot of subjects. No, it isn't in most cases. Google, yes, is different (but even they need be very careful). Facebook isn't; users are finding they can go elsewhere.

Plus they are continuing to attract a host of litigation from both individuals[1] and nations, alike.

[1] http://money.cnn.com/2014/01/03/technology/facebook-privacy-...

http://www.scribd.com/doc/195351255/Campbell-v-Facebook

I think the sheer number of people on Facebook is a meaningful differentiation.

Also the sheer level and scale of integration of Facebook with, well, basically life. Globally.

Also one has to keep in mind that the Internet itself is also constantly growing and evolving, and today it is nothing like in times of MySpace.

I get your humor (as I pull another PBR out of my personal cooler on my porch)

I use a lot of things every day that I don't google, mostly because I know about them.

It's how they manage the ecosystem that matters, and they've never understood really what they're doing. That's my opinion of course, though all we can do see is watch to see what happens to FB.

I think a big difference between MySpace and Facebook is that MySpace targeted indie bands and teenagers - people who aren't going to be thoroughly married to a platform.

Facebook started with college students that are a little less ephemeral, but has spread out to soccer-moms and grandparents. Those folks are far less fickle.

Even if Facebook loses the younger users, it will have a rock-solid userbase for decades. The company might be doomed, but this is going to be a very very long tail. Its downfall will look more like Sears than FourSquare.

Facebook was founded in February, 2004, almost exactly 10 years ago. A lot of those college students it started off with are already the soccer moms of today.

Other differences: Facebook had MySpace and other early social networks to see what they did wrong. In addition to just the general not-keeping-up-with-the-times, the excessive customization + bored teens = eyes bleeding & having to look at different places on a page to get the same info. Although it's gotten worse it's still cleaner than MySpace was during its heyday. Also, Facebook has Zuckerburg. Whatever you say about him, he's smart and focused on keeping users, so there's constant tweaks to see what will keep them.

Speaking only anecdotally, I notice that up until a few years ago, it was common among my friends and acquaintances to join Facebook. Now the trend seems to be to leave Facebook. Not in great numbers, just here and there, and more than seem to be signing up.

Well, there's obviously a saturation point - once almost everyone (in a given social circle) has joined, there are no more people left to join. At that point, those who haven't joined probably never will.

Using an SIR model seems a shady approach for this model. The model inherently assumes an amount of time where a person is "infectious." They modified the traditional SIR model, Also, they never allow migration from the Recovered class to the Susceptible class again, meaning that once a person stops using Facebook, they will never use it again. It seems to not take into account age buckets (where Facebook may be more "infectious" with different age groups).

Here's the direct link to the Princeton study cited in the article:

http://arxiv.org/pdf/1401.4208v1.pdf

One thing I don't see mentioned but once or twice is that one of the reasons the model fits MySpace is BECAUSE Facebook came around and offered a better model. Where is the better model that people are going to be leaving Facebook for? Google+? Maybe, but unlikely.

I don't buy that people are going to quit using social media, or that it is a fad. I am looking for that thing that will displace Facebook, or disrupt their network effect, but right now, I don't see anything that would work for the 2-5 year projection they make in the paper. Still possible for something to pop up, of course, but haven't seen the potential replacement for Facebook yet.

> One thing I don't see mentioned but once or twice is that one of the reasons the model fits MySpace is BECAUSE Facebook came around and offered a better model.

Salient point. I can't name another popular one among older-than-teens in the same vein (twitter I feel is a different sort of thing), and while I sometimes don't really understand why I continue to use facebook, I definitely continue to use it.

Separately from the disease analogy, Facebook could exhibit reverse network effects on the way down. Brad Burnham & Fred Wilson have also talked about reverse network effects.

Here is a link from Fred talking about Single User Utility in a Network: http://www.avc.com/a_vc/2012/12/single-user-utility-in-a-soc...

---

And, as I have said before, network effects help on the way up and hurt on the way down. If I get great single person utility from your service, it is less likely that I will follow my friends out of it when your service ends its stay on the hype cycle and the iTunes leaderboard

---

Here are some background reading from the Social Network Researcher Danah Boyd:

http://www.businessweek.com/magazine/content/11_27/b42350539...

--- Social networks appear to be a very peculiar business—one in which companies might serially rise, fall, and disappear.

Danah Boyd, a senior researcher who studies social networks at Microsoft Research, attributes their instability to the way users can bind themselves by race and class, taste and aesthetics. Influential peers pull others in on the climb up—and signal to flee when it’s time to get out. “The thing about user adoption and user departure is that it’s not a steady flow,” says Boyd. “Think of it as, you’re knitting a beautiful scarf, and you’re knitting and knitting, and you get a bigger and bigger scarf. Then someone pulls a loose thread at the bottom. And it all unravels….

“Facebook did a fantastic job of hiding behind the panic around Myspace and basically saying, ‘We’re totally safe,’ ” says Boyd. Myspace’s inability to build an effective spam filter exacerbated the public impression that it was seedy. And that, says Boyd, contributed to an exodus of white, middle-class kids to the supposedly safer haven of Facebook—a movement she compares to the “white flight” from American cities in the second half of the 20th century. Myspace was becoming Detroit. ---

I think another close analogy is fashion.

"The 870 million people using Facebook via their smartphones each month could explain the drop in Google searches – those looking to log on are no longer doing so by typing the word Facebook into Google."

You think???

I believe it. Facebook doesn't have anything going for it at this point other than the fact that everyone is on it. The problem is that it's becoming increasingly less interesting, to the point where it will be no more interesting than the Yellow Pages - everyone is in there, but nobody cares about opening it up anymore, except for once a year perhaps. At that point Facebook is dead, and it doesn't matter if people actually press that non-delete Delete button, or they stop logging in to it.

> Facebook doesn't have anything going for it at this point other than the fact that everyone is on it.

Which is pretty huge. Even if I never shared anything and never casually browsed facebook, I still would not stop using it quite regularly, say a few times per week, for things like event planning, direct messaging people (Who keeps a list of friends' email addresses?).

So the question then becomes: if facebook is slowly turned into a contact book and event planner, does that mean the end of facebook? Does facebook require billions of people hitting millions of like buttons and farming animals on their walls in order to live?

> So the question then becomes: if facebook is slowly turned into a contact book and event planner, does that mean the end of facebook? Does facebook require billions of people hitting millions of like buttons and farming animals on their walls in order to live?

It depends what you mean by 'end' but if this is all people do once there in the future, absolutely. There won't be enough daily active users to support advertising, which leads to a downward spiral. People may keep their accounts but Facebook's data (like any social network) is highly temporal - once you stop liking stuff they become more and more out of touch with ad targeting.

Exactly, but the reduced revenue is likely proportional to reduced operating costs. It will be the white-dwarf end of Facebook, but it would never die until someone creates a feasible replacement. Even at a fraction of its current size, being "just" everyone's contact book for one billion users is still quite a big deal.

So it's a definition of when Facebook is "dead" or "gone". It will never go away entirely.

People are still logging in everyday. Maybe not you, but I know plenty of people who belong to groups that log in everyday to see what the group is up to. As long as they can monetize that, it will always be around.

People with children will log on for many many more years to upload pictures of their children to share with their parents. Those parents aren't going to move to another social network. In their eyes Facebook works great why sign up for a Twitter account, and Instagram account, and a Vine account etc.

Are they monetizing it adequately, though.

FB being decimated by 2017 is ridiculous. However, it could be sidelined as Zuck and crowd can't find a way to make increasing dollars out of their userbase without driving them off.

> FB being decimated by 2017 is ridiculous.

Not really, but FB losing 80% of its users by then -- which is 8 times greater reduction than decimation would be -- is.

Doesn't "decimation" mean FB would be left with only 10% of its present user base? 90% versus 80%, not that big a difference, really. In any event, I don't believe it will happen; there's no credible alternative on the horizon, at least not yet. Google Plus doesn't seem to be taking over, and who else is there?

> Doesn't "decimation" mean FB would be left with only 10% of its present user base?

No, the literal meaning is removal of one-tenth. [1]

[1] http://en.wikipedia.org/wiki/Decimation_(Roman_army)

True, but I think most people would interpret the word today as "destroy a large number of" (as per Merriam-Webster).

Thinking about it, Facebook has a few things going for it. It's become an important communication channel outside of immediate family and Best friends. Collecting email addresses across multiple web mail accounts, that's a chore. Facebook has made messaging people easy. That convenience has got staying power, especially as in regards to mobile communication.

Teens aren't interested in Facebook anymore? Doesn't matter, they have no money to spend.

The sweet spot is advertising to older people with jobs.

The balance they need to find is the amount of directed advertising they can force on a user, without making them abandon FB... or just use it strictly as a messaging tool.

Scare off users from reading/liking/posting items in their feed, lose the other revenue stream: mining and selling social trend data.

   Teens aren't interested in Facebook anymore? Doesn't matter, they have no money to spend.

Sure they do. They spend their parents' money quite well. Being only half-way facetious. It's what they spend it on. Newest gadgets. Clothes. Entertainment (iTunes or streaming music/tv shows/movies).

Facebook's appeal is largely casual entertainment. I say largely because there are some very practical features in the platform, like photo sharing and event planning/invitation. But it seems to me the bulk of its use is simply sharing links and anecdotes. Once the novelty of the medium wears off, people will look for the next entertaining fix.

I'm not sure how different it is from massively popular TV series of decades past, such as MASH or Dallas, where millions of people tuned in every week to see what would happen next. Eventually, all series die out. In the case of network TV, the medium itself, "what happened next" was cable TV, and then the internet.

I don't doubt it will be the same for Facebook (the product, not the company).

> Facebook's appeal is largely casual entertainment.

I don't know if that's truly accurate, at least among my crowd (a few years out of undergrad, who mostly joined Facebook in late high school).

The use of facebook for casual entertainment has largely faded. It was new and novel once, and now there are so many other forms of entertainment out there that for that purpose we could drop facebook in a second.

We (specifically my friends, I don't want to generalize to everyone in my demographic) use it because its a centralized way to stay connected. For the most part, my high school friends scattered all over the country when we graduated high school, and then my college friends did the same thing. Facebook's prime utility at this point is that we want to keep up to date on what those people are doing, and so far, everyone is still there. I'm not sure if there is something that can come along that take over that same role for us. Now we may grow up and stop caring about that sort of stuff, but I think it will take more than a few years.

> a centralized way to stay connected

This is what FB has pushed all along iirc, or at least it's the argument I've always heard, and only recently (I'm 31) have I noticed that this is really what I use it for. It's fun to see pictures from a friend's vacation in Vietnam, or to read that so-and-so-who-I-like-but-not-enough-to-regularly-call got a new job or got married or something. It's bizarre, but

> Now we may grow up and stop caring about that sort of stuff, but I think it will take more than a few years.

...I only seem to care more over time, which is weird. I shouldn't discount making-more-interesting-and-smart-friends and the 'unfollow'-or-'hide all from foo on my newsfeed' features though.

That paper basically extrapolates Facebook's future from one data point (the rise and fall of MySpace). I'd be more inclined to believe them if they analysed the life cycles of more than other social network, but this way their results are more than questionable.

When something better comes along to replace or supplement it, Facebook use will undoubtedly level off or decline. Until then, I doubt it will decline much if at all.

It's gradually replacing all the independent communities with convenient Facebook communities and user groups, it's a convenient one-stop site for communicating with friends, sharing pics and generally relaxing at the water cooler.

Teens of course will always look for some alternative (WhatsApp, Reddit, etc) so as not to appear to conform to their parents' generation, but in reality there is no particularly great alternative that offers everything FB offers, especially in terms of sheer numbers of participants.

As someone who joined the Internet bandwagon relatively late (circa 2000), whenever I see such reports on /viral growth/ of social website I wonder if email and usenet had a similar /viral growth/ in the 90s. Although the number of Internet users back then were small compared to now, they were still large enough in absolute numbers for some of these statistical analogies to apply. Usenet did eventually die out (for all practical purposes), but email is still going strong (perhaps not as much as at its peak, but it is nowhere close to being dead either).

I love using facebook. It's great for connecting with friends and discovering my network. I post most of what I do there and twitter.

that being said I'm constantly looking for something new. It seems facebook has hit a wall in its cutting edge delivery of products and services. Is this a product of their IPO? I'm not sure how internally they've changed, but whatever happened, it seems to have changed how quickly the product has evolved externally.

Good luck to them though, staying #1 is difficult and requires an agile thought process.

Facebook is used by different people for many different purposes.

In Ukraine right now it is the default news channel for the opposition with the mass protests going on now, and has served a similar function in other countries.

For some people it's a way to stay in touch with close friends.

For others it's a way to occasionally keep in touch with more distant friends.

Others use it as a contact/ address book.

Others use it for games.

I don't think it's going to die out, instead it's going to evolve.

I also do think it's possible it might die out, but way too early to know what will happen.

After reading several articles about FB loosing teenagers because of "grandma knows FB" and "I prefer chats now", none of the grandma-age folks at my church, the population majority there, believe the news that FB is less of the future for communicating with the "younger generations" [because recruitment for sustainability is important].

I've stopped minding though; I no longer have a FB account, and at this point I'd prefer people learn by their mistakes.

What?

The link is to a UK google trends page and doesn't seem to work. Basically it peaks right at the IPO time and has tailed back to the pre-ipo trend. Wouldn't call it a decline.

http://www.google.com/trends/explore#q=facebook&cmpt=q

That appears to be the only referenced proof in the article (a google trends chart).

The pre-ipo trend at best looks flat, but really looks like it was declining. Facebook's emergence as an investment might actually be masking a decline in searches from people who are not investors.

Click the "forecast" button. Google disagrees.

Google, in this case, is a blind algorithm guessing at the continuation of a line. It knows nothing of IPOs.

Yes. Google does not take into account "viral" jumps in it's prediction algorithm for long term search popularity... they just simply "extend the line". Makes perfect sense. I forgot Google was run by middle school engineers.

It's easy to say now that this mass Myspace migration happened, but its downfall was much more than Facebook showing up. e.g., It had a ton of technical issues, and it got really ugly / unwieldy.

If you're interested, the book "The Facebook Effect" does a good job describing the fall of Myspace.

Facebook won't be going anywhere, just like Yahoo never did. Facebook Groups has way too many people locked in on daily basis. These are same people who are hooked on email and email lists.

I think this is more indicative that most people use Facebook mobile. I understand Googling it from a desktop browser, but most people access FB by clicking the F icon on their smart phone.

This is just the setup. They are actually modelling linkbait headlines to build a data set for a thesis on optimization of online marketing buys. You are the punchline.

That, or this is just bad science.

I'll place a $100 bet against this claim. Any takers?

$FB

Translation: short the stock

Translation: buy the stock.

(They would like to bet against the claim that Facebook will lose most of their users. That's a bet on Facebook's long-term value, and maps to buying their stock.)

Step 1: Make totally ridiculous claims about social giant

Step 2: ?

Step 3: Profit

Facebook sucks and it shall perish! It just plain sucks!

Facebook is pretty dead traffic wise already...

Then how will people log into my app?

Reply on news.ycombinator.com