Connecting the world

The computing industry has always sold to segments. Mainframes to large companies, minicomputers to medium-sized companies,  PCs first to companies and then middle-class households, and so on. Today, there are around 1.5-1.6bn PCs on earth, of which perhaps half are owned by consumers, yet there are 7.1bn people on earth and 5.25bn adults. PC buyers are still a segment, and a pretty small one, and so, by extension, was the internet. 

Mobile is different. It started, like PCs, as a toy for rich people, but in the last 20 years it has spread such that 85% of the earth's population is under cellular coverage - more than mains electricity (80%) and close to the same as access to improved water (90%). 3G networks cover around 60% and will grow to 90% by 2019. Somewhere between 3.5bn and 4bn people now have a phone, depending on your assumptions (there are several billion more SIMs but much duplication). This continues to grow fast, pushing into some very low income groups: the average monthly spend in India is $3-4. In the last decade we've discovered that the value of a phone actually increases as income falls - in the developed markets they sit at the top of Maslow's Hierarchy of needs, but for a poor farmer in rural Africa or an Indian fisherman they solve problems much further down the scale. 

Mobile was always somehow separate from the mainstream consumer technology industry - there was always a wall between the two. But since 2007 smartphones have broken that wall down, and the mobile industry is going through a massive category conversion. Of 1.7bn phones sold last year just under 1bn were smartphones, and there are now around 2bn on earth. This will grow much further. The real question will be more about affordable data, and battery life when there is no mains electricity nearby. 

For the first time, then, the consumer technology industry is selling to everyone that it is possible to sell to. There are still people who are unable to participate in the economy, but everyone else - everyone able to buy things - will probably be a customer. 

This changes lots of other things. Most obviously, as I've written before, it makes the internet opportunity not just two or three times bigger but closer to ten times bigger.

It also challenges conceptions of market size or ecosystem. Since Apple only sells phones at $400 and up (for now) most of this growth in users will go to Android, which now starts at $50 or less. So the Android ecosystem will encompass everything from Instacart customers in San Francisco to ROM hackers in Kiev to rice farmers in rural Myanmar. This makes the concept of a 'customer' on the internet look much more like, say Unilever's: there are people buying soap by the gallon and people buying it in sachets

What does mobile scale mean?

Some time in the next six months, the number of smartphones on earth will pass the number of PCs. 

Screen Shot 2013-12-18 at 17.51.51.png

This shouldn't really surprise anyone: the mobile business is much bigger than the computer industry, and there have been more mobile phones than PCs since at least the late 1990s. There are now perhaps 3.5bn to 4bn* mobile phones, replaced every two years, versus 1.7-1.8bn PCs replaced every 5 years.

But there was always a wall between the two industries - neither really saw or paid attention to or sold to the other, or if they did (like Microsoft) they didn’t get anywhere. Smartphones broke down that wall - suddenly tech companies could sell to an industry with $1.2 trillion annual revenue, and suddenly mobile operators are faced with people who think about product cycles in weeks rather than years.

Both sides are slightly shell-shocked, and I often feel that neither has quite internalised the scale of the change. Mobile operators still try to make messaging services, and tech people still get surprised that you can sell more phones in a quarter than the PC industry sells in a year. 

Screen Shot 2013-12-18 at 18.12.45.png

The driving dynamic is that smartphones are not actually a new industry - rather they are a category change sweeping through a huge industry that was already there, hiding in plain sight. The great majority of those mobile users are converting to mobile phones running Unix, and mobile phones (on average) are replaced every two years, meaning this is a very rapid cycle. 

So we will end up with somewhere over 3bn smartphones in use on earth, almost double the number of PCs. (In fact, the more meaningful question is how many people will have a data plan: when you go to low income levels in emerging markets you find people who are happy to buy a $50 Android phone but operators who cannot afford to give that person a gig of data a month for a dollar.)

In truth, none of us have really internalised what change this means. The fact that Apple makes more money than Microsoft or that smartphones outsell PCs isn’t really the point. Rather, the entire internet is being changed fundamentally - both the size and the character of the internet are going to look quite different from what we have been used to. 

First, scale. There are perhaps 900m consumer PCs on earth, and maybe 800m corporate PCs. The consumer PCs are mostly shared and the corporate PCs locked down, and neither are really mobile - at best you can take them from table to table. Those 3bn smartphones will all be personal, and all mobile. So the internet goes from a shared device at home to a screen in everyone’s pocket. And that’s without considering several hundred million tablets, which blur all of these boundaries. 

This means that the internet gets several times bigger. One could talk about time, or engagement, or use, or value, and all of these metrics are problematic, but it is clear that mobile devices will become the dominant part of the internet - we will stop talking about ‘mobile internet’ in much the same way that no-one talks about ‘new media’ anymore. 

Just as importantly, the character of that internet changes too, and not just because of mobility and screen size. On the desktop, ‘internet' has really meant ‘web’, with a few exceptions such as Spotify or Skype. Everything happened inside that browser window, and that didn’t really change at all for 15 years. On mobile, clearly, that is not the case - we have a much richer array of routes to market and models for interaction, and this complexity and innovation is growing all the time. The web is no longer the only option and Google & Facebook are no longer the only option. Some of these changes in dynamics were obvious from the start - others have taken time to appear. A good example is the way that smartphones remove the winner-takes-all benefits of a leading social network, and so obliterate Facebook’s monopoly on social. There are more changes coming - mobile is changing all the time in a way that the desktop internet did not.

In a sense, then, one could suggest that smartphones liberate the internet from the browser in the same way that the browser liberated it from the command line. 

All of this means that the operating environment looks very different, and a lot of our assumptions need to change. A lot more is up for grabs, and the scale of success looks different. When a dozen guys in a garage with a hot service get struck by lightning, that means 50m or 100m users, not 1m, and in time it might mean 1bn. A great many industries that came unscathed through the first wave of the internet - the desktop wave - are now facing disruption. Lots of things that look like the right idea are going to turn out to be blind alleys, like Yahoo, but at the same time even irrelevance can mean a big business - Facebook has lost its monopoly but the mobile opportunity may be big enough that that doesn't matter. Web search is no longer the only discovery channel but Google is doing better than anyone else at finding what comes next. 

It's now the season that people are supposed to write sets of predictions of what will happen next year. For me, the prediction that comes out of this is confusion, change and opportunity. I really don't know what will be the most important thing that will happen in 2014, but that's kind of the point - everything is on the table. 


*There are around 6.5bn mobile connections, but a great many people have more than one SIM

Smart devices versus Broadband lines

Sometimes the simplest observations are the most compelling.

According to the ITU, there are now around 650m fixed broadband lines in the world. 

In the 24 months to December 2012, around 915m iPhones, Android phones and tablets will have been sold. At the current rate of growth, which shows no signs of slowing, there will be twice as many of these smart mobile devices as there are broadband lines within a year.


Of course, real life is messier than that makes it look - you need to think about multiple users per PC and multiple PCs per connection, etc, etc. But none the less, the shift in the balance of consumption over the next few years will be profound. 

Polarisation, continued

Sometimes the simplest images are the most compelling.


This chart shows quarterly revenue for the handset units of Samsung, Apple, Nokia and the rest of the industry since July 2011. It includes some estimates for small points of detail it it is mostly just the data the companies report.

Apple (with cyclical swings) and Samsung are taking almost all of the growth - one taking the high end and the other taking all the rest. 

Platform Wars

This is a summary presentation based on a longer and much more detailed report I’ve produced for clients of Enders Analysis. I’ve presented these slides at a few conferences, so I’m making it public here as well. As should be obvious, it is the basis for a talk rather than a stand-alone document, but most of the charts should be self-explanatory. 

Mobile platform wars: Apple, Android, Samsung and Facebook

US Q2 smartphone share

As reported by the US operators, supplemented by some reasonable solid estimates on my part. The iPhone had around 47% share in Q2, roughly level with Android (RIM is now very small in the USA). 

2 years ago the iPhone was only offered by AT&T, which had 30% of the market. Today it is offered by operators with 80%. I strongly suspect that the iPhone ‘5’ will be ranged by T-Mobile (whose spectrum has been reconfigured, making this more practical): all things being equal that would take it slightly over 50% market share. 

This is especially relevant in the light of this internal Apple survey, disclosed today as part of the Apple/Samsung patent lawsuit: people choose operator first and phone second, especially in the US, given the highly variable coverage that operators provide. 

Handset revenue

Sometimes the simple charts are the best. This shows quarterly revenue for handset manufacturers (using the handset business unit for companies that do other things as well), for Q2 2012 versus Q2 2011.

In the last 12 months Apple and Samsung grew, as did Sony (from a very small base). All the other ‘traditional’ branded manufacturers shrank. In total, Samsung and Apple’s revenue grew by $10.9bn year-on-year: the other branded OEMs combined shrank by $6.3bn. 

(The Apple number is slightly misleading because it catches the company in a cyclical dip in iPhone sales - run-rate growth is higher)

Not shown: Huawei and ZTE plus (literally) thousands of other smaller manufacturers in China and now India, which is increasingly difficult to estimate with a straight face: many of these are effectively Mediatek vendors. In effect, the market is polarising between ultra-low-cost manufacturers at the low end and high-volume high-efficiency  manufacturers at the high end.

Another (slightly flippant) way of putting it: the market is polarising between quad-core phones and quad-SIM phones. 

Nokia and Apple

Nokia reminds me a lot of Apple in the pre-Job days. 

It doesn’t have the loss-making, inept manufacturing, but everything else is there. The OS strategy is in chaos, the product lineup is bloated and confusing and the internal politics are out of control. Hardware teams and software teams do whatever they want with no reference to each other and no-one has any idea what OS will be on what device, what OS will come next, what device to buy or what internal fiefdom they should be dealing with. 

Meanwhile both companies struggled to get past a sense of ‘but we invented that!’. Apple invented (sort of) the GUI, while Nokia invented the idea of a consumer-friendly, easy to use mobile phone, yet both watched helplessly as newer upstarts zoomed past them. Both the Apple OS and the Nokia OS (more Series 40 than Series 60, admittedly) began as industry-leading products with great UX, both had far too much bolted onto them and both eventually collapsed under their own weight. Both went through a long painful period of confusion as first one then another future path was picked up and then dropped, and both went through and extended period of denial as they pretended to themselves that the success of Microsoft/Apple was due to something other than the fact that they had BETTER PRODUCTS. 

Of course, Apple achieved salvation by buying NeXT, which both reset the OS strategy and brought back Steve Jobs, the only man to create a billion-dollar company three times (Apple v1, Pixar and Apple v2). A lot of the comment around the new Nokia CEO centres on the fact that he clearly isn’t the same kind of product visionary. Does it matter?

Well, Nokia has lots of problems that don’t feature on Techcrunch. It has a massive business selling phones for $10 in emerging markets that is now under attack from cheap Chinese manufacturers using MediaTek chipsets. It has a solid mid-range business in phones running Series 40 that is about to get kicked in the teeth by Android (and the iPhone Nano?). And it has a high-end business in ruins. 

But running across all that is a deeply dysfunctional company that is incapable of executing - even if you agree with Nokia’s strategies and actually like the handsets they announce, the company doesn’t deliver them. The N97, released 2 years ago as ‘the answer’ to the iPhone, didn’t have a reasonably bug-free firmware for more than a year after launch. The N900, this year’s iPhone killer, didn’t support paid downloads on the Ovi Store for months after launch, and still doesn’t have many of Nokia’s flagship features. Any given Nokia handset has dozens of different firmware variants floating around, all with slightly different capabilities and all needing individual testing by developers. 

This shows up in little things too: the high-profile flagship retail presence was supposed to sell the Nokia content experience, but the handsets on display had no preloaded content, no SIM cards and no working Wifi, so there was no experience to see. Who approved that? Who stood in a store painted black and lit in purple neon and said ‘yes, this is the way to sell to women’?

So Nokia needs someone who can solve a corporate culture that looks at broken products and says ‘who cares, ship it’. It needs to reset the OS (Android? Windows Phone? Who knows). It also needs to fire dozens of failed senior managers. So far, so obvious. But I think there’s another, less obvious problem. 

Apple’s greatest strategic advantage in mobile is that it only sells one phone. This means it can go to mobile operators with a ‘take it or leave’ it proposition - Apple can ignore operator demands for customisation. This protects Apple’s user experience and makes life far easier for developers. If you’re happy with 5% of the market, and aim for 10%, one handset is fine. But Nokia sells 400m phones a year, at every price from $10 to $600. It has to have lots of phones on the market - not 50 (apparently the target for 2011) but several dozen. This is exactly the same problem Android is starting to have. If there are a dozen inter-changable handsets on the market then operators can demand that you remove that free GPS and pre-load their paid GPS instead - or preload Bing. 

In other words, the very fact that Nokia (and by extension Android) is putting more than one device on the market makes it impossible to hold the line against mobile operators in the way that Apple can and does. That makes it much harder to match the Apple experience, no matter how good your strategy, OS and execution are.