THE QUESTION is if the smartphone, as we know it, is at the end of its innovation cycle. It’s an uncomfortable question, but a hard one to avoid. The new shiny things don’t shine as bright as they used to. Also, the market leaders, Apple and Android, are not helping with the perception that they are just cranking out the same old stuff with different sweet names.
So, many industry veterans, like myself, are wondering whether is it time to start hanging the memorabilia – like various lanyards I’ve collected from various congresses – to the wall. And is now the right time to start re-tooling the personal skill set, because knowing about mobile device construction is about as fashionable as knowing how to launch and drive space shuttles, when everyone else is spacejumping. In this megapost, I explain my view.
THE SHORT ANSWER: The party is not over yet, but the industry dynamics are such that the current duopoly is very hard to break. Windows Phone has, in theory, what it takes to break through but its highly proprietary model cripples its ability to compete with Android. So, in practice, its best shot is to beat Apple in its own game. That’s a mountain-size of a task, so it’s no wonder there’s some skepticism out there. Hence, it may well be that if the status quo is shaken, the new ideas will come from new entrants embracing fully & improving further the open innovation model that currently explains much of the Android traction.
That’s the conclusion I’ve drawn, having been few months off the industry hamster wheel. Now, my prediction is neither that novel nor controversial. It is close to the position many acclaimed industry analysts have taken, and correlates with the consensus forecast of the Wall Street.
For how I got to my conclusion, you actually have to read the rest of this very long megapost 🙂
THE LONG ANSWER: It is true that winning and losing always includes an important element of leadership and execution. But not everything can be explained by the hero qualities of the winners and the deficiencies of the losers. There is always the element of strategy – and especially in the technology businesses – the element of innovation strategy.
The theoretical view: I personally have found the theories of Clayton Christensen highly useful to the smartphone market. Obviously his research and books are a fundamental part of the industry vocabulary, but additionally I have had the benefit of following the work of one the leading experts Horace Dediu already when he was at Nokia and also beyond. His asymco blog is such a treasury of information of knowledge that it would be a waste of bits for me to replicate it here.
Instead I post here a link to the Sept 28th, 2012 podcast by Horace and James Allworth (a Fellow at the Forum for Growth and Innovation at Harvard Business School) that gives you a solid foundation in about 70 minutes. Next I’ll just summarize the key points, in my own words, in order to drive the logic forward:
- There are two fundamental ways to make a product –an integrated model and a modular model. In the former, the company retains the control of all parts of the system design that matter for the consumers whereas in the latter the architecture is “modularized” i.e. different companies specialize on the different parts of the system, focusing to make each component better, or at least cheaper.
- Integrated model tends to rule when the products are not “good enough” vis-à-vis the consumer expectations. The total control and the ability to push the envelope across the system design enable the company to solve the hard problems. Apple is a prime example of a company operating this way.
- Now, when the product is good enough, any company operating with the integration-based model faces the risk of overserving, or overshooting. People will no longer be ready to pay extra for the new improvements, and the market position can be disrupted from the below. This is usually also when the ASP:s (Average Selling Prices) start dropping
- PC industry looked to follow the pattern of succumbing to the modularity. In fact, it was locked to the stage of “no-profit Windows OEM: s and the rich Microsoft” stage for so long that people thought it was the end game. But it wasn’t. The consumer pull for the computing mobility changed everything. Suddenly, people didn’t like to put the non-branded, ugly, bulk manufactured chassis of the computer on the table of their third place, the Starbucks café. So, suddenly, the proprietary design innovation and the related hardware-software-miniaturization skill set Apple had been quietly cultivating became a proprietary slow-to-copy innovation.
- There’s inherent gravity towards modularity, but user interface innovation keeps on re-setting the cycle because getting it right requires touching so many places of the system design. That’s a task more suitable for those operating with the integrated model.
- iPhone is already giving signals of people thinking it is good enough. As Horace highlights, the sales mix between iPhone4S and iPhone 5 is a proxy whether the new (hardware) innovation provides enough of meaningful value. This makes sense. It’s an analogous situation I saw up and close with N95 and N96 – the numbers would just tell that the slider camera phone had reached its innovation apex. Adding more features or megapixels wouldn’t matter because the user sore point was the slow input mechanisms and the screen size. So consequently, the category of cameraphones needed to be re-engineered to touch and touch&qwerty smartphones.
Mobile peculiarities: Competitive people have a tendency to view the world via playoff brackets, so it is natural to expect a clear winner – a clear superiority of either the integrated or the block-building modular model. Maybe one day it will be so, but right now, the elimination battle doesn’t seem to happen. Instead, the both extremes are getting stronger, and everyone who is in the in-between land needs to think hard.
I’ve identified at least two reasons why smartphone market behaves slightly differently than some previous markets.
- User interface adoption is not linear. It is tempting to take the extreme view and extrapolate a direct line from the ITU-T pad via touch screen and via voice assist (Siri at your service, sir) all the way to the wearable computing. If it would be so, those operating with the integrated model would have an edge for the foreseeable future. It could so happen, but I am not so sure. User interface decisions are not just rational spec decisions (e.g. 8 MPX>3.2MPX, duh) but also a conscious decision about the personal habits and the personal image portrayed. For example, it’s been clear for many years that the future is undoubtedly about touch screens. Yet so many people have pushed the inevitable switch, because they don’t want to look like idiots. Think it this way. The learning curve for PC mouse could be done safely at home, the mobile user interface you end up learning in front of your friends and colleagues. As a result, the market is always somewhat segmented against the technology adoption cycle, even though in the view of the 3-5-year view segments tend to come and go.
- Making mobile software is really hard. As the proof, the fatality rate of new operating systems is really high – so many people try but few actually get to finish. And as one coder once put it: “Moore’s law ain’t gonna rescue your sloppy code”. So if you, metaphorically, think system design modularity as a puzzle, the software piece is that ugly piece that your dog or infant daughter has chewed the corner of. It may get the job done, but it might not look pretty. The software challenge also explains why we see more software companies (Kindle, Surface etc.) moving to the integrated model with their own hardware, as opposed to hardware and manufacturing companies (Foxconn, BYD etc) adding software to their stack.
Ecosystem view: Now, some people try to outline the roadmap to the industry end game via the number of ecosystems. It is a good way to simplify the complex world for PR messages, but I’ve found it to be not sufficient enough for deeper strategic thinking.
Productization model view: In my opinion, more important than the number of ecosystems is understanding what are the most optimal productization positions in the spectrum between integration and modularity. So drawing from the frameworks above, there are two “pure” positions– the ultimate Integrator (“Absolute Best”) and the ultimate Block Builder (“Best Value”).
The ultimate integrator is the one pushing the envelope of the industry innovation on the level of the whole product. It attacks the hard problems that span across multiple system modules in a way only it can, because of its total system control. Consequently, it tends to have a high cost structure and so many constraints, but it also an opportunity to get rewarded for not just skinning but really solving the problems people have. The downside is that from the industry social graph viewpoint this type of company is high maintenance because of its high degree of proprietary solutions to the problems. These companies are like friends that make every dinner planning a nightmare because of their dietary convictions. And later they suddenly decide that the restaurant you have selected was not cool enough, and off to the replan we go again. You put up with this friend, because you love her/him and the spark she/he brings to your life. At the same time, you know that you just can’t cope with too many people like that. Obviously, Apple occupies this position in the market place today. The interesting question is, hence, how many that proprietary companies will the market place sustain.
The ultimate block builder is making things better piece by piece. The high level of modularity and the loose coupling of elements in the system design are critical so that the investment of improvement can be isolated to the system element in question, and the rest of the blocks taken ‘as is’ from the community. Being handsomely rewarded by the consumers is possible because of the community efficiency and the sharp-hence-marketable consumer problem solving. From the industry social graph point of view, these communities are hard to govern. The more structure there is, the slower the apparatus innovates and the end result can still be a mess. Or an IPR lawsuit. On the other hand, when things are done the Nike way (“Just Do It”), like things tend to be done in the technology-led businesses, you end up either with a mess or a royal mess. Android is the closest to this model. The interesting question here is not how many players the model sustains (because it is many) but who’s the Current King of the Hill and and who is the Supernanny. Currently, they are Samsung and Google, respectively, but the casting is dynamic by nature. I do agree with what ex-WP program Manager Charlie Kindel wrote already in January in his insightful blog i.e. “Google has already lost control of Android and has zero chance of regaining control. He is probably right.
Practically, most of the medicine Google has at its disposal for fixing Android are counterproductive to what has made Android a success in the first place. So the question really is whether Android can be outandroid-ed, from the within or from the outside.
Innovation moves the needle. It surely is the innovation that moves the relative positions of the companies aiming for the perfect positions. But to which direction? And how? I think the aspect of the mobile industry innovation that hasn’t been enough debated is that it really matters whether the innovation is open or proprietary. It also matters how strong the pull is for the innovation and whether it comes directly from the consumers.
The fascinating part to me is that deciding to make an innovation open or proprietary is a clear strategic decision. It’s not a personality, leadership or execution trait – it’s calculated strategy thinking at its purest. Further, it really doesn’t matter whether it’s an explicit yes/no case within the strategy framework, or whether it is an automatic response derived from the chosen business model, it still is a position prepared by some MBA folks in the strategy department, decided/blessed/driven by some executive and (hopefully) approved by some company board. In real life, the decision can also be caused by a technical constraint that some engineer put accidentally in place by not thinking the system interfaces properly and not getting the proper guidance from the product management folks. Any which way, the open vs proprietary decisions tend to have long-lasting consequences.
Let me try to open up what I just said with an example. This is totally hypothetical and populist but please accept that in the name of landing the point. I am talking Facebook and mobile. In my 17 years of following the mobile industry, the mobile social media explosion is the second biggest whirlwind of innovation ever to come around (mobile voice call still #1, SMS tied #2, browser #3). Facebook could have chosen totally to disrupt the mobile industry by making the mobile Facebook very proprietary to the devices of their choosing. Had they done so, surely the outcry would have been tremendous, but the consumer pull they had was so huge – by any metric – that, the collateral damage to the industry relations aside, they probably could have pulled it off. They didn’t go proprietary. They chose almost the polar opposite, very open. In fact, they made making the Facebook client so easy to everyone that the technology choices almost backfired. Now, this decision probably wasn’t even that hotly debated inside Facebook. Further, it may not have been even recognized as a strategic decision as the company drive towards maximizing the user base needed the support of every platform out there. But yet it was a clear strategic choice that set up the company to a certain direction that looked totally obviously right back then, and probably slightly less obvious today as people are less bullish about Facebook’s advertising revenue prospects.
Implications: Okay, enough of theory stuff. Next I’ll walk through what this means in my view to the prospects of various industry players:
- If what is said above is true, Microsoft is not in an optimal place. Its block diagram and related plan comes across as such a Frankenstein, So much of its phone side innovation is highly proprietary, almost like a proprietary extension of the proprietary Windows computing system. They really have done some excellent work with the user interface, but one can have that only if one is ready to take the rest of the blocks too. As a result, for example, consumers end up using Zune PC software, chip vendors doing a special baseport and game devs using DirectX graphic engine, none of being the obvious choices otherwise. To complicate things further, its software-licensing based business model assumes a vibrant productization community. As a result, I tend to think that they can only be successful if they can overtake Apple – being so proprietary they are not well configured to compete with Android. I have so many friends working on Windows Phones that I sincerely wish them well, but it still is a confusing model.
- Interestingly, Android actually may be more exposed to face heated competition than Apple. The reason I say this is because I am not sure whether Android is successful because they have been uniquely good or because others have been uniquely bad. There was some competition for the “block builder” positioning from Symbian, but Symbian suffered technically (like Microsoft now) from bringing in such a large number of non-optimal technical choices. In other words, even if the business framework was deemed open enough, the technical choices were quite proprietary. Various Linux distros had all the “right” technical building blocks but, during the rise of Android, they all for whatever mystical reasons self-exploded. Don’t get me wrong in any way dissing Google’s execution. Or Samsung’s. Michael Jordan won a ton of games by just staring at the opponent and then executing fundamentals well while watching the opposition to blow themselves apart. A win is a win is a win.
- Intriguingly, up to now, Android blocks have been used to construct quite classic smartphones to cover maximum number of price points. That tends to be a scale game and not so surprisingly, has played to Samsung’s advantage. When it comes to blocks they are the closest to Legoland this industry has. In addition to whether Samsung remains as the king of the hill, the interesting question is whether the Android portfolio will actually end up being successfully segmented to smaller niches, all of which can be profitable due to the lower cost structure enabled by Android openness. The stuff that Amazon is doing is well in to that direction. You read about the rumors of luxury Vertu phones going Android (no idea if true or not but certainly logical). And if you put your scanners out, you run into a bunch of companies embracing & extending Android, some even with the help of governments. But there are also smaller new entrants. One example is Adaia, set up by some Nokia alumni, and building a rugged premium Android smartphone.
- Even more intriguing is the Linux question. Android benefits tremendously of using Linux as the base, but could there be a “more pure” instance of the block building model by throwing some of the Google middleware out and replacing that with other community-driven software for an even more capable and open alternative. And I don’t mean semi-pro consumer modding now, but a new productization block diagram for the industry players. I really don’t know. It’s not like I am hearing the Linux Savior story for the first time, but breaking through in the consumer space is hard, for whatever reason. It, however, obvious that there are people who believe in the opportunity, and that sentiment might partially explain the excitement around companies like Jolla Mobile.
Makes sense, no?