Facebook, identity and throwing sheep

Way back in the dark mists of time, when Facebook first launched its platform on the desktop, one of the first hit apps was something called 'Superpoke'. Superpoke did quite a few things but the one that got all the attention was throwing sheep at people. That is, they'd open their Facebook news feed and it would say 'Benedict threw a sheep at you'. 

Of course, a website that did that would never work - you'd never get a critical mass of people to open an account on a new site just for that. But Superpoke could plug into the Facebook platform so you could do this fun little social thing right away with almost no friction. 

A lot of the social apps bubbling up now remind me of this. As I've written several times, by plugging into the smartphone address book, camera, photo library, notifications etc the frictional barriers to doing a new social app fade away: the smartphone is a social platform in the same way that Facebook is. The obvious expression of this is WhatsApp and similar things that directly address the core Facebook use cases. But it seems to me that there's at least as much potential in doing things that use the platform without trying to take over a core use case - things like throwing sheep. That is, the smartphone social platform enables a lot of experimentation with new ideas and behaviors that don't need to be your core comms channel and that would never have worked on the web, and (for a bunch of reasons) might not have been possible on the desktop Facebook platform.

Snapchat is arguably one of the biggest of these, and Secret is another. Firechat is also an interesting example - it leverages the wireless autodiscovery features in iOS7 to do hyperlocal chat. Of course this isn't quite as easy as Superpoke - you still need to install an app from the app store (for now, though that may well change) but the friction is still pretty low. With apps like Line, WeChat and Kik you can see people trying to pull this experimentation back up the stack and put it inside a social app again - that might be the right model for some things, but of course you're trading friction for flexibility. Making your own smartphone app needs that initial install but has much more power. 

I also think that (as I suggested here) retailers should be thinking about how they can leverage the social platform aspects of smartphones - shouldn't the Zappos app show you which of your friends have it and let you share shoes directly? Again, doing that well on the desktop would be really hard, but on a smartphone it's just a tap or two away. 

This takes us around to Facebook again. Perhaps the problem is not that people use WhatsApp instead of Facebook Messenger - rather it might be that they use Sephora instead of Facebook Messenger. This is partly about unbundling WhatsApp, just as WhatsApp unbundled Facebook, but it's also that the fads and gimmicks and silly little things (otherwise known as 'fun') don't happen within Facebook. The time sinks don't have to happen within Facebook. And maybe the commerce apps don't need to connect to it. 

The question implicit in all of this, of course, is identity. It's the machine-readable identify that allows all of this low-friction social experimentation. But what is the irreducible common denominator for connecting to your friends? Is it your Facebook identity? How much does it matter to Facebook if it isn't, if it still happens in something Facebook owns? Is it your PSTN phone number (which Facebook will actually let you use to find friends with the smartphone app)? Or do you change that from time to time without caring? The broader phone address book? Your email address? BBM Pin (cough)? Location? Would Apple try something within iOS (with the fingerprint scanner)? Where in the stack does the identify sit - the network, OS platform or something further up? Actually, I suspect there isn't any single common point that any company can own. 

Whatsapp and $19bn

Facebook just bought WhatsApp, paying $16bn in cash and stock ($4bn cash, $12bn stock at current prices) and $3bn in RSUs. WhatsApp has 450m active users, of which 72% are active every day. It has just 32 engineers. And its users share 500m photos a day, which is almost certainly more than Facebook. 

This is interesting in all sorts of ways - it illustrates most of the key trends in consumer tech today in one deal. 

First, it shows the continued determination of Facebook to be the 'next' Facebook. It's striking to compare the aggressive reaction to disruption shown by Google, Facebook and other leading web companies today with how some of their predecessors a decade ago stumbled and lost their way. 

Second, the winner-takes-all dynamics of social on the desktop web do not appear to apply on mobile, and if there are winner-takes-all dynamics for mobile social it's not yet clear what they are. There are four main aspects to this: 

  • Smartphone apps can access your address book, bypassing the need to rebuild your social graph on a new service
  • They can access your photo library, where uploading photos to different websites is a pain
  • They can use push notifications instead of relying on emails and on people bothering to check multiple websites
  • Crucially, they all get an icon on the home screen. 

Any smartphone app is just two taps away - a desktop site can crush a new competitor by adding it as a feature with a new menubar icon but on mobile there isn't room to do that. Mobile tends to favor single-purpose, specialized apps.

This has led to an explosion of mobile social apps - last summer I counted over 50 with more than 1m downloads on Google Play. Some sort of consolidation is clearly inevitable but it's much less clear whether we will revert to one or two. The smartphone itself is the social platform and all sorts of different ideas can leverage that, where on the desktop web they'd have needed to leverage Facebook.

So It's quite possible mobile social will have lots of services indefinitely. This creates opportunities, but also a pretty basic challenge to Facebook. Partly in response, it paid first 1% of its market value for Instagram and now close to 10% for WhatsApp, taking not dominance but at the least two of the commanding heights of mobile social. That's the right way to think about value, I think - not 'OMG $16bn!", but "is this worth 10% of Facebook?' The deal values WhatsApp users at $35 each (very close to what Google paid for YouTube, incidentally), but the current market cap of Facebook values its MAUs at $140 or so.  

Third, the sheer scale of the numbers involved is a good illustration of what the shift to mobile means. I produced a presentation here to try to drive home this point: mobile is the next computing platform and it is several times larger than the desktop internet. There are now roughly the same number of smartphones and PCs on earth - those PCs are mostly shared and immobile or locked-down corporate boxes, while the smartphones are mobile and personal. Meanwhile, the widely-discussed collapse in the cost of creating a startup in the last decade combines with both the much larger scale of mobile and the routes to market and virality offered by mobile platforms to mean that if you're very good (and lucky) you can get to astonishing scale in a short time. This scale is at the heart of the valuations we're starting to see - WhatsApp is probably now sending more messages than the entire global SMS system.  

Finally, mobile social apps are not, really, about free SMS. Mobile discovery and acquisition is a mess - it's in a 'pre-pagerank' phase where we lack the right tools and paths to find and discover content and services efficiently. Social apps may well be a major part of this, as I discussed in detail here. These apps have the opportunity to be a third channel in parallel to Google and Facebook.

Also, note this tweet from the co-founder. 

(Note: my boss at a16z, Marc Andreessen, is on the board of Facebook - the posts linked were all written before I joined)

Who buys the iPhone 5C?

Facebook's ad platform gives a wide range of targeting information, including gender, country and, as it happens, device type. So you can see what Facebook thinks the audience for your ad would be if you aimed it at British women with iPhone 5Cs. And, as it turns out, that's an interesting answer. 

So, this chart shows FB's numbers on this basis for the 5S and 5C for the USA and UK, both markets where both iPhone and Facebook have good sample sizes. 

Screen Shot 2013-12-08 at 12.54.40 am.png

There are two obvious things in this chart: the 5S is selling better than the 5C (which we pretty much knew), but the 5C has far from flopped, and women like the 5C much more than men. You can see this more clearly if we switch to percentages and compare with the broader base. 

Screen Shot 2013-12-08 at 12.54.51 am.png

Women are a (very) slight majority in Facebook MAUs overall (which is probably within the margin of error). But they're a little more likely than men to buy iPhones, particularly in the USA, and much more likely to buy the iPhone 5C.

What's going on here? There are a couple of dynamics: 

  • All iPhones are more expensive relative to the competition in the UK than they are in the USA, where the pricing structure tends to mask the price of phones. 
  • The iPhone 5S is sold a little more on technology than the 5C and is also more expensive  (again, especially in the UK)
  • The 5C is offered in a wide range of colours

So, is this gender disparity because women are more practical and buy a cheaper phone? Or because the colours of the 5C are more appealing than they are to men? Or something else?

(All this is presuming, of course, that the Facebook data is reliable, but these gaps should be large enough to rise above sample error). 

Mobile interaction models

The interaction model for the desktop internet was pretty much settled 15 years ago. It turned out that the answer was a web browser. Stand-alone apps such as Pointcast were a mostly blind alley, and while apps persisted for email and IM, and for very specific things like music, the words ‘web’ and ‘internet’ became effectively synonymous to anyone non-technical. Over time we added Ajax and better search and better social, but everything really happened inside the browser.

In mobile this is quite different: nothing is settled. We have the web and apps and of course app stores, and then we have many complications - voice, in-app payments, web apps, hybrid apps, widgets, push notifications, social messaging apps, Google Now and Siri. Then there’s the hardware layer - images, barcodes, NFC, bluetooth, location, motion sensors etc. Innovative and disruptive new interaction models can very often find a route to market, far more easily than they could on the desktop internet. Sometimes, they scale to a hundred million users in a year to two. And we have more and more waves of innovation coming, with things like local wireless from Apple and deep linking to within apps from Android, and a very fast-evolving social messaging space, and more things in 2014 and beyond.

So, we can actually have a pretty limited idea of what the dominant interaction models will be in 5 years. 

(There is a dream, of course, that all these nasty choices and options will go away and we can go back to the nice, simple, limited web, but that doesn’t seem very likely just yet.) 

One of the big changes here is the removal of monopolies. If the web is not the only interaction model then web search loses power as a discovery and acquisition channel. And in parallel Facebook’s desktop monopoly on social has not transferred to mobile and it seemly unlikely that it ever will (I wrote about the reasons for this here). So both of the key channels on the desktop are smaller and less crucial, and also work significantly differently, and are pretty poor at driving some key types of engagement, now that you’re not just looking for a click on a link. This changes lots of things, and creates lots of new opportunities.  

The puzzle for Google is how it brings its vast, decade-old machine learning project to bear on this new complexity of data and behaviour. The obvious problem is that data and behaviour within apps are effectively dark matter that it can’t track (hence the deep-linking initiative in Android). But this is balanced by much richer data collection. Your Android phone feeds it with data all the time - where you are, what you look at, where you go after you search and what you did the day before. The challenge is finding the right ways to collate and present that data - Google Now is one example but probably not the only one. The search box and the page of results is just one possible interface to that machine-learning project - what does Google look like after the search box?

Social faces a different set of challenges. It seems to me that on mobile Facebook will never have the near monopoly that it briefly enjoyed on the desktop - smartphones remove most of the frictional barriers that keep you on one social network. But mobile social more broadly is a vast opportunity. With web search no longer the dominant channel, social, on a far more social device than the PC, has an open door to push at. Tencent announced that the first 5 games that it launched with Wechat integration, starting in August, have had 576m registered users. Mobile social is an engagement, interaction and distribution channel, and it appears to be much richer, and probably much bigger, than social was on the desktop. 

If this is the end to near-monopolies in acquisition and discovery, it’s also interesting to think about it as the end to monocultures. If the interaction model shifts away from web search, that change makes different models and different types of behaviour possible. In turn, one might ask - what models and companies and behaviours were precluded by Google on the web? What good ideas didn’t work because of the way Google did search and the way Facebook did social? How did that monoculture shape things, and how does that change now? 

Instagram and Youtube

Instagram is looking like a great acquisition. It had 30m users when Facebook bought it in April 2012, and has now passed 150m, just 18m later. 

Screen Shot 2013-11-09 at 10.33.00 pm.png

 (The change in colour signifies the acquisition.) 

This reminds everyone, naturally, of YouTube - again, a product that has become far more popular since acquisition in late 2006 (this chart shows the only consistent data that seems to have been released).

Screen Shot 2013-11-09 at 11.14.16 pm.png

There's a difference here, though. Youtube is the dominant online video sharing platform but Instagram is not, remotely, the dominant photo sharing platform on mobile.  

Facebook's latest disclosure is that 55m photos are shared a day on Instagram, and another 350m on Facebook itself.  But 350m a day are also shared on Snapchat, and 400m on Whatsapp. And we don't know the numbers for Line, or WeChat, or the next half-dozen services to be launched that we haven't seen yet. Meanwhile Instagram has 150m monthly active users but Whatsapp has 350m and there are close to a dozen others with more than Instagram. 

So as it turns out, Facebook did not solve the unbundling problem by buying Instagram - even in photos. It bought just one of many mobile social products, and not even the biggest

All of these new services are driven by the fact that smartphones have characteristics that remove most of the defensive barriers that Facebook has on the desktop:

  • The smartphone address book is a ready-made social graph that all apps can tap into
  • The photo library is open to all apps
  • Push notifications remove the need to check multiple sites
  • Home screen icons are easier to switch between than different websites

The fluidity with which you can move between these apps seems to be breeding very fluid use cases. The original analysis was that these were unbundling Facebook in a semi-coherent way - most obviously, Instagram was taking photos, a core Facebook use case, and moving them to a different, specialised app. But it doesn't seem to be as clearly defined as that.

People aren’t using Instagram for photos, WhatsApp for text, Line for stickers... they’re using everything for everything. Instagram to tell people you're running late, WhatsApp to share holiday photos, Snapchat to make plans for the evening and so on. WhatsApp and Instagram are not in different categories - they're direct competitors for time and attention. Instagram, Snapchat, Line and all of the others are all poking away at different social behaviours and different options in the same communication space. 

This shouldn't really be a surprise: we already have three social apps on our phones - voice, SMS and email, and we don't stick to a rigid set of use cases for each one. These new apps just add more options into the spectrum. That, of course, points strongly against consolidation into a single winner.

So buying Instagram certainly looks like a good trade - it would be worth a lot more if it was selling today. But as a strategic move, it's looking increasingly irrelevant. Is FB going to buy Whatsapp, Snapchat, Line, Kakao and the next ten that emerge as well? Sure, some of those will disappear, but it doesn't look like FB will crush the competitors the way it did on the desktop. On mobile, FB will be just one of many. 

Just maybe, Facebook might have been better off rethinking the core product instead of buying what turned out to be just one of a swarm of alternative services. 

This, of course, prompts comparison with another (in)famous acquisition - Flickr's purchase by Yahoo. There was a period when AOL and Yahoo went around buying up lots of cool new web services as their portal model came under threat. They then generally mismanaged them, but that wasn't really the point. No matter how well they ran these acquisitions, they couldn't buy every great website that there was. Neither can Facebook. 

 (Update - I got the exact photo sharing numbers transposed when I wrote this - updated with links)

Dead social networks and the value of history

This Google Trends chart provides a neat shorthand illustration of the rise and fall of five different social networks. All of these had strong lock-in effects. All of them thought that there were barriers to switching away. All thought that people would not want to abandon their history of photos, interactions, tags. But when it came to it, people just walked away from all of that accumulated history and adopted a new platform. 

 I've been wondering how broadly applicable this message is. How much value do ordinary people place on their email history, for example? At least half of the subscribers to my newsletter are using Gmail, and in the tech world we tend to think of such things as extremely sticky. But real people walk away from services that they're supposedly locked into with disconcerting ease. And when a corporate person (as opposed to a VC or entrepreneur) leaves a job they leave their email behind too. When I left NBC Universal at the end of 2008, I left behind my Exchange mailboxes, including, say, my 'Launching Hulu in Europe' folder. I can't really say that I need it now. It has a certain historical interest but no professional value. 

Meanwhile a lot of my professional interactions are on Twitter, which are ephemeral and unsearchable even after a day or two. (There's a start-up idea in here, somewhere.) 

I wrote a post on the unbundling of services recently in which I suggested that a key issue for Facebook is that the smartphone phone book and photo library are resources that any app can tap into, where a competing desktop social site had to recreate them from scratch. This, obviously, makes it easier for smartphone apps to unbundle segments of the Facebook experience - text messaging, photo sharing, games etc. It also enables new experiences, such as Snapchat. All these apps sit on top of these smartphone resources in the same way that Facebook apps sat on top of Facebook. And most people use more than one and many use three or four in parallel, both with different people and the same person. 

So the underlying relationship has far more value than any record of the messages exchanged. People switch between apps and dump them and their archives on a whim, or even in a deliberate detox (it seems to me this is part of the point of Path - it's Facebook, but with only your real friends). The value is in the contact list on the smartphone - the social services and the conversations and things shared themselves are ephemeral.  

LinkedIn has a similar issue. Once you've accumulated a few hundred LinkedIn connections, every time you look at the news feed you see people you have no recollection of ever meeting. They're people you met, once or twice, a few years ago, and it seemed worth connecting to them, but now they're just names. That is, many LinkedIn connections are like the business cards you find in a drawer and struggle to place. You file them away safely for another year or two, and then, the next time you find them, you shrug and throw them away, not because the information is out of date but because the relationship that's referenced is long gone. How many LinkedIn 'connections' are equally worthless? Rather like the tags and shared photos on Facebook, you care about them at a moment in time, and then, when you really think about it, you're happy to abandon them. Maybe, like Path and Facebook, there should be a LinkedIn clone that's just for the people you actually know. 

Snapchat is the purest expression of this view. Any interaction, whether professional or purely social, is a conversation that fades over time if it isn't continued. After a certain point the archive of that interaction has no value (except perhaps to your company's lawyers). And the division is between the relationship itself versus the many different media though which you might contact a conversation - phone calls, SMS, coffee, Instagram, beer, twitter, Snapchat, WhatsApp, Facebook. And both the relationship and the messages have half-lives. An app that reminds you that you met this person has strictly limited value - if you can't remember them, are they really in your network?

Lots of companies have big user bases and big accumulations of user data. And they think that this gives them a lock-in. But maybe the only stickiness comes from the mere presence of users - more like a nightclub than a bank. If your friends move, you'll move in a second, and the dynamics of smartphones mean there are no barriers at all to moving. Owning the address book, and perhaps the photos, are the only real levers of control, and it's very hard to dislodge the underlying platform owners from that. 

That of course begs the question - what is the irreducible, underlying, unchanging point of identity? Is there one? An email address? A PSTN number? A Facebook/twitter account? Or is it ultimately a personal, real-world connection?

Unbundling: AOL, Facebook and LinkedIn

One of the recurring themes of the consumer internet is the cycle from aggregation to disaggregation - bundling to unbundling. There is a lot of value in services that pull everything together in one place, but over time that value starts to recede, the lock-ins keeping people there weaken and the appeal of having separate, specialised products grows. And then, after a while, the appeal of aggregation starts to grow again. We saw this in the past with AOL, and now with Facebook on mobile.

A big part of the dynamic of Facebook unbundling is the different mechanics of apps versus the desktop web. Any smartphone app can use the phone book and PSTN numbering system to access a ready-made social graph, removing a lot of the friction Facebook benefits from on the web. A core Facebook use case is photo sharing, and a smartphone's built-in photo library offers much more fluid ways to share photos across multiple services. In effect, the phone book and photo library are reservoirs that any app can tap into, where on the desktop, creating that resource is a major pain and a major moat for Facebook. Meanwhile, push notifications to apps mean you don't have to check different web sites for updates.  And it seems that pressing 'Home' and jumping to another app to share something else in another way with someone else is much easier than loading a different web page. 

We can also see this unbundling happening to Craigslist, or at least that’s what many people hope. Once a charming reminder of the web’s uncommercial roots, Craigslist is now, to many, a fat complacent monopolist with a nasty litigious streak, whose founder claims no responsibility for the company he owns. It appears to have taken over the newspapers’ business while taking on all of their attitudes. There are a swarm of services, often mobile first or mobile only, trying to peel off parts of the Craigslist offer, or do things Craigslist should have been doing. AirBnB is only the most obvious. Chris Dixon has a good note about this here, and Andrew Parker produced this great graphic back in 2010. 


LinkedIn is another interesting example of a product waiting to be unbundled, and it looks especially vulnerable because of the consistently poor execution of so many core features.  

To give just one example of this, a few weeks ago an old colleague of mine took an important new job at Microsoft. I found out because he posted a photo of his badge to Facebook and I saw it, by chance, on my fortnightly visit. But when I went to LinkedIn I saw this, filling the top half of my news feed.

Screen Shot 2013-08-22 at 11.29.07 pm.png

I spent 10 minutes searching without seeing his new job listed anywhere other than his profile itself, and indeed I couldn't actually find a way for LinkedIn to tell me about it - there is no view of 'job changes in my network', though that would seem like something that would have been implemented at launch. However, it's eager to tell me Deepak Chopra’s ten tips for how successful people wipe their arses. 

To re-iterate the point - LinkedIn has my entire career history. It knows I've spent 14 years in mobile, tech and finance. And this is the news it recommends me. 

Of course, the amount of effort going into a mediocre Huffpo clone wouldn’t matter if the core functionality of the site worked, but it doesn’t. There’s no way to see which of your contacts have changed jobs. No way to post a message to your network that your boss won’t see. The entire user experience is a mess of overlapping technologies, black UX aimed at upselling paid services and outright poor work. Pretty much any basic task to do with managing or engaging with your network is harder and more painful than it should be. It's even hard for LinkedIn, it seems: it knows I don't have an MBA (I TOLD THEM!) but recommends I join an MBA alumni group because I did my BA at the same university. 

Screen Shot 2013-07-23 at 12.15.48 am.png

But how does LinkedIn get unbundled? What's the mechanic? The dynamics of smartphones discussed above make it easier to unbundle Facebook. Craigslist's main strength is volume of traffic, and if you can gain an edge in a particular segment you can overcome that. But neither of those dynamics quite apply to LinkedIn. 

The core of LinkedIn is that it's the universal CV database. You need to have your CV there for people to find it, to look you up - in some circles not  being on LinkedIn is almost a professional failing. For specific verticals (programming, design) this isn't the case, but for white collar professionals it is. For some people an online presence per se (your blog, twitter etc) is a de facto CV and is how you build your reputation (and Klout is poking away at this), but most careers are not built in public, and most people cannot blog about their job, even if they had the time and inclination. 

So it seems to me that the low-hanging fruit is the stuff that LinkedIn just isn't doing while it pursues the Huffpo dream. Connection and meeting management (Evernote Hello). Introductions (Emissary.io). Pre-meeting stalking (Refresh). Bit by bit a graphic just like the one for Craigslist gets built up. 

The puzzle is at what point that matters to LinkedIn. The core business is about selling CVs to recruiters. So long as it remains the CV registrar, does it matter if the users only visit the site once a month, or once every six months? Indeed, if all of these apps use LinkedIn as a data source, arguably they actually reinforce its position. 

Perhaps the answer is that tech history is full of dead companies that had mediocre product but great lock-ins. Eventually, the lock-in always goes away - we have Blackberry this week to remind us of that. LinkedIn has a great lock-in and product that's mediocre for the users who enter their CVs, but pretty good for the recruiters who pay the bills. How will that play out?


The two strongest trends in Internet content are atomisation/unbundling on one hand and sealed silos within smartphone and tablets apps on the other. This is contradictory. 

As we all know, many sites see the majority of their traffic going to individual pages rather than the home page, Tumblr and Pinterest disaggregate, reaggregate and remix individual pieces to content far away from where they started, and of course social sharing on Facebook or Twitter remixes and redistributes everything. Twitter cards and the trend for social messaging services to embed content within messages take this another step. In a sense, there is no home page for any site. 

Every piece of content becomes a packet that can be routed anywhere across any service layer - but the service layer is Twitter, Kik, Line or AirDrop, not TCP/IP or HTTP. 

But at the same time, after the end of the HTML5 head fake (as Bill Gurley put it), it seems clear that apps will also be a major component of content consumption. To hope that this will not be the case is to wish to turn the clock back to 2007, and of course to ignore a pretty clear demonstration of what customers want. The last 6 years were not a temporary aberration. 

So apps are a new, permanent part of the content landscape, just like social or search before, and apps are silos. Yes, you can deep link, up to a point (including with things like AirDrop), and share back to the unsiloed web version of many content properties from within an app, but the app experience itself is essentially exclusive. You go in, you engage, and then, often after much longer than you'd spend on a website, you leave and do something else. This is quite different from the promiscuous flitting from site to site and tab to tab that's the general model for the web. And, of course, it's the complete opposite of the atomisation that's happening in parallel. 

This is a real challenge for content owners. How do you think about editorial on the premise that it will both be shared everywhere and read in the course of a half hour session with your brand's app? Do you have to pick one model or the other? If you're in the long-form business already (the New Yorker, say) this is an easy conversation, but if you run a typical magazine with a mix of content of all different types and lengths, what does your 'digital' proposition look like? How many different types of engagement do you need to think about? What's your social messaging app sharing strategy? And, of course, how do you address this across 30 titles in three cities, most of which are only run by half-a-dozen to a dozen people who're only just keeping up with updating a Facebook page?