The death of the newsfeed

When people get married, they are often quite sure that they will have a small, quiet wedding. None of these massive, extravagant parties with hundreds of people for us! We’ll just invite close family and friends. Then, you make a list of ‘close family and friends’… and realise why people have 100 or 200 people at a wedding. You know a lot more people than you think.

I was reminded of this recently by the fact that, according to Facebook, its average user is eligible to see at least 1,500 items per day in their newsfeed. Rather like the wedding with 200 people, this seems absurd. But then, it turns out, that over the course of a few years you do ‘friend’ 200 or 300 people. And if you’ve friended 300 people, and each of them post a couple of pictures, tap like on a few news stories or comment a couple of times, then, by the inexorable law of multiplication, yes, you will have something over a thousand new items in your feed every single day.

This is a combination of two factors - Dunbar’s number (a rule of thumb that implies that you probably do know several hundred people well enough to friend them on Facebook) and ‘Zuckerberg’s law’ (the supposed tendency to share more and more on social media over time). Combine these two and you get overload.

Pushing a little further, it seems to me that Zuckerberg’s law, such as it is, is really an observation about the models for following people that social media platforms have evolved, in which sharing something onto your own feed is not the same as sending it to any particular person. You would not send 10 pictures of your child or dog to everyone in your address book very often, if ever, and most people (under 50) would not send every funny or enraging news article they see to everyone in their address book either, but the asymmetric feed makes posting at that kind of frequency normal instead of rude. Since you’re posting it to ‘your’ feed instead of sending it explicitly to someone, it’s OK to post lots and to post less important things. That, in turn, takes us to the tragedy of the commons - we’re ‘supposed’ to post stuff, but by posting stuff, we overload each other’s feeds. Facebook’s Growth team was too good at its job. 

This overload means it now makes little sense to ask for the ‘chronological feed’ back. If you have 1,500 or 3,000 items a day, then the chronological feed is actually just the items you can be bothered to scroll through before giving up, which can only be 10% or 20% of what’s actually there. This will be sorted by no logical order at all except whether your friends happened to post them within the last hour. It’s not so much chronological in any useful sense as a random sample, where the randomiser is simply whatever time you yourself happen to open the app. ’What did any of the 300 people that I friended in the last 5 years post between 16:32 and 17:03?’ Meanwhile, giving us detailed manual controls and filters makes little more sense - the entire history of the tech industry tells us that actual normal people would never use them, even if they worked. People don't file. 

This is the logic that led Facebook inexorably to the ‘algorithmic feed’, which is really just tech jargon for saying that instead of this random (i.e. 'time-based') sample of what’s been posted, the platform tries to work out which people you would most like to see things from, and what kinds of things you would most like to see. It ought to be able to work out who your close friends are, and what kinds of things you normally click on, surely? The logic seems (or at any rate seemed) unavoidable. So, instead of a purely random sample, you get a sample based on what you might actually want to see.

Unavoidable as it seems, though, this approach has two problems. First, getting that sample ‘right’ is very hard, and beset by all sorts of conceptual challenges. But second, even if it’s a sucessful sample, it’s still a sample.

Looking at the first of these, there are a bunch of problems around getting the algorithmic newsfeed sample ‘right’, most of which have been discussed at length in the last few years. There are lots of incentives for people (Russians, game developers) to try to manipulate the feed. Using signals of what people seem to want to see risks over-fitting, circularity and filter bubbles. People’s desires change, and they get bored of things, so Facebook has to keep changing the mix to try to reflect that, and this has made it an unreliable partner for everyone from Zynga to newspapers. Facebook has to make subjective judgements about what it seems that people want, and about what metrics seem to capture that, and none of this is static or even in in principle perfectible. Facebook surfs user behaviour.

It’s useful here to compare the newsfeed challenge with the Google’s search results challenge. Google has to work out the best 10 results to show, using all sorts of judgements about what signals work best and what signals matter more, just as the newsfeed does - it can’t just show you the results by some objective measure like date or file size. It can offer complex controls and filters to fiddle with, but like Facebook it has to get things right without such controls because most people won’t ever touch them. And also like Facebook, of course, it has people trying to game the system. Unlike Facebook, though, Google has explicit intent - you told it what you wanted to see. And so if Google shows me exactly what I told it I wanted, it’s succeeded, even if I ‘shouldn’t’ have searched for that. Facebook has no such direct signal. There are things it ‘shouldn’t’ show me, even if my uncle did share them. But what are those things, and who should decide, and what does the weighting look like?  

This is the challenge to getting an algorithm right, but there is also the second problem - that it’s always still a sample. Facebook has done so much development and outreach and ’growth-hacking’ to get more stuff into the feed and to make it frictionless to share that there is now far more content available to me than I can look at. It tries to show me a sample of that, and a sample that’s better than ‘whatever was posted in the last 45 minutes’. It might or might not succeed at being better, but there remains the question - do people want a sample at all, anymore?

One basic problem here is that if the feed is focused on ‘what do I want to see?’, then it cannot be focused on ‘what do my friends want (or need) me to see?’ Sometimes this is the same thing - my friend and I both want me to see that they’re throwing a party tonight. But if every feed is a sample, then a user has no way to know who will see their post. Indeed, conceptually one might suggest that they have no way to know if anyone will see this post. Of course, Facebook’s engagement teams won’t let that happen - if I feel too much that I’m shouting into the wilderness I’ll leave (this is one of Twitter’s new user problems), and so I’ll be rationed out at least enough exposure to friends and engagement feedback to keep posting. Until you don’t. But if something was really important, why would you put it on Facebook?

I think one could suggest that this is some of what’s behind the suggestions of systemically lower engagement on Facebook newsfeeds, and behind the obvious growth of person-to person chat (most obviously WhatsApp, iMessage, FB Messenger and Instagram - three of which Facebook of course owns). The social dynamics of a 1:1 chat work much more strongly against overload, and even if one person does overshare they're in a separate box, that you can mute if you like. 

Meanwhile, you could propose that the Stories format that Snap invented, and Facebook... continued to invent, is also a way to address overload: by bundling what you want to share into one unit instead of many separate items, even if you do then still share that to many people any feed is more manageable. That in turn makes easier a key Snap thesis - that though you still share things asymmetrically, there shouldn’t be an algorithm between you and your friends. That is, maybe Stories mean you share more things, but by bundling them into one thing you place less load on your friends and reduce the need for a filter. You put the structure into the content instead of the display of the content. 

There are clearly also other trends behind any swing from newsfeeds to messaging. Messaging can be more private, have less social pressure, and be more fun. A Snapchat story isn’t a permanent record and has less pressure to show off your perfection. Stickers and filters are more fun and spontaneous than Facebook’s rigid blue boxes (and the days of throwing sheep at people are gone alone with Facebook’s platform). And some of these offer light-weight ways to interact without obligation, which was also a feature of Facebook’s model, but deliver that piece of Maslow in different ways. 

The catch is that though these systems look like they reduce sharing overload, you really want group chats. And lots of groups. And when you have 10 WhatsApp groups with 50 people in each, then people will share to them pretty freely. And then you think ‘maybe there should be a screen with a feed of the new posts in all of my groups. You could call it a ‘news feed’. And maybe it should get some intelligence, to show the posts you care about most...

A pessimist might say this looks like slash & burn agriculture, or perhaps the old joke ‘No-one goes there anymore - it’s too crowded.’ That is, for social, Metcalfe’s Law might look more like a bell curve. I don’t know what the next product here will be (I didn’t create Snap, after all). But tech like this tends to move in cycles - we swing from one kind of expression to another and back again, and we might be swinging away from the feed.

Finally, any such changes have consequences for the traffic that sharing creates. 'Like' buttons made it frictionless to post any web page you want into your feed and push it to (some arbitrarily calculated percentage of) your friends, and many hands have been wrung about how much traffic this can drive and how Facebook moves things up and down the feed ranking. But sharing links inside Stories isn't the same, today, and a link you share in a WhatsApp or iMessage group with 5 friends will only be seen by them, and Facebook has no lever to pull to make this more or less visible. On the other hand,  the 'WhatsApp forward' can take such a link and send it viral across a country, and where Facebook can ultimately kill a link or an entire source across the whole site if it really wants to, it's very different for a P2P messaging app to make that call (outside China, of course). That is, the plea from many media companies to 'up-rank' their posts in the newsfeed - to make people eat their greens - and to kill 'fake news' links is at least theoretically possible on Facebook. It's not possible in iMessage - with end-to-end encryption, Apple has no idea what you’re sharing. 

SocialBenedict Evans