Windows struggles and it’s not about “app gap”
In recent years people have been complaining about “lack of apps” (or “app gap”) for Windows Phone and Windows 8. I can’t deny it — Windows for years has struggled with drawing attention of developers, who have chosen more profitable mobile platforms (follow Steven Sinofsky for more) such as iOS and Android with +95% of global mobile market share.
Unified Windows 10 came to fix that problem. Currently both desktop (x86) and mobile (ARM) versions of Windows share same kernel, runtime and app store — NT kernel + Universal Windows Platform (UWP). Of course in x86 SKUs there is also Win32 runtime for legacy desktop apps, but Microsoft sees its future built around mobile, secure and touch-friendly UWP apps acquired from Windows Store. The goal is clear. Company wants to have 1 billion active users of Windows 10 in 2018. It should help a lot with Windows Store utilization, app download numbers and increasing developers’ interest.
But will it be enough? It may be. People still (think they) need PCs with Windows so the sheer sales of W10 devices will be enough to reach the goal of one billion active users. OS adoption is paramount, but it’s only the first step. Second step requires users’ engagement and Store utilization. If majority of Windows 10 consumers will use mainly desktop programs without interest in UWP apps, the objective to encourage devs to invest in Windows will fail. But there are two strange relationships. Little quality apps in the Store => little interest in the Store. Little interest in the Store => little quality apps in the Store. How to break this vicious circle? One thing is to start with great own, first party apps (from Microsoft).
First party apps
Windows 10 is still relatively young OS with immature app platform and imprecise user interface guidelines. It took months to fix UI inconsistencies, bugs and lack of important features — even in first party apps such as Outlook, Maps, People/Contacts, Photos, Music (Groove) or Messaging. And they still lack polish, comparing to iOS or even stock Android.
The train for Windows 10 Mobile in consumer space is long gone, but MS still fights for business users and enthusiasts. With shared core and runtime, maintaining Mobile SKU is rather inexpensive and “effortless” comparing to Windows Phone 7/8. It will let W10M to live in “zombie” mode with less than 1% market share, without shutting it down completely. The plan is to gain a lot of active users of desktop Windows 10 to force devs to invest in UWP apps. UWP apps can run (with some UI/view modifications) on both desktop and mobile, so success of Windows 10 store is a success of Windows 10 Mobile store, because… there’s actually only one store. Then there might be a time for Windows 10 Mobile to resurrect and raise from ashes.
Microsoft can’t make half-efforts if they want good results. And unpolished first party apps that lack features is what I call half-effort. These are basics. Absolute minimum. Outlook, Office, OneDrive, Skype, OneNote, Sway — all these apps have to be THE BEST on Windows if MS want their platform to succeed. For now, Android and iOS users get MS apps earlier and with better quality. That’s a path to nowhere.
Services and platform
Microsoft with its famous research team and huge achievements in machine learning (ML) or natural language processing (NLP) should lead the tech industry with virtual assistants and natural (voice) interfaces. Yes, we have Cortana with great Bing (Satori engine) backend, but it is still not proactive and user-friendly enough. That’s often the difference between Microsoft and Google/Apple — attention to details. Seamless experiences, deep ecosystem integration and consumer-focused features are what distinguishes great UX from mediocre UX.
Windows 10 is still in its infancy, while iOS and Android are in the phase of gradual addition of features, deepening the app integration with the OS and making use of ML/deep neural networks for making interaction with OS more natural. What Google presented during recent IO 2016 blown my mind. I get it was only demo and most current Android devices will never get Android N, but that’s not the point. The point is, Microsoft seems standing behind in areas where it should be a leader. And it’s not about poor algorithms or lack of research/backend tools. On the contrary. One can argue that MS has the best research team and ML tools on the market. The problem starts when we get end user product that lacks polish. Lacks OS and external sources/apps integration.
Google with its huge share in search/web business and rich app ecosystem can provide much better experience with smooth user interfaces beyond app-as-separate-entity paradigm. More and more features can be done within OS interfaces without even launching an app. It is Mobile 2.0. It is the future. It is the area where Windows struggles. UWP, being relatively new, still isn’t equipped with all APIs and features available on competiting platforms. Situation is getting better, but MS is in a disadvantageous position, having to chase Google and Apple. Microsoft is always late. And that’s not a good position to compete.
x86 trap
On mobile, Microsoft is an underdog, but the company is still the king of the desktop operating systems. It’s a strange situation where MS can’t do any radical moves because of legacy and huge Windows user base, and should make such moves to be back in the mobile market. You can’t do both. A radical move would be to leave x86 and focus on ARM (maybe even own custom chips like Apple, Samsung or Huawei). Just few years ago the distinction seemed clear: “Intel = computing power, high end”, “ARM = mobile, power efficiency, connectivity”.
In 2015 ARM based A9X chip from iPad Pro changed everything (forever?). Benchmarks indicate it’s more powerful than Intel Core M lineup in some scenarios, especially in regard to GPU performance. Both Apple and ARM have made huge performance strides in the last five years, but those improvements have required higher TDPs, more sophisticated power management, and better software support. This view argues that Apple and ARM have made such huge advances because they’ve been adopting techniques and technologies that companies like Intel and AMD started using a long time ago.
Mobile computing will be (already is?) all about connectivity, fast resume and efficient sleep states — aspects that Intel struggles with. ARM doesn’t have any mysterious architectural advantage over x86 so it’s up to Intel if they fix issues and bring more SoC integration to their next wave of chips. This issues plagued Surface Pro 4 and Surface Book — Windows’ showcase devices. Microsoft doesn’t have any control on Intel’s schedule, chip architecture or bug fixing (drivers). This dependency hurts both Windows hardware and software. It is yet to be found what’s Intel’s mobile strategy, but it seems that the company already gets that the mobile war with ARM is lost (cancellation of Atom x3 — SoFIA chips). And that’s a very bad news for Microsoft whose main operating system is x86 based. With no ARM alternative due to Win32 apps dependency.
What’s next?
Optimistic scenario? Intel keeps its game strong with new power-efficient chips, Windows 10 reaches 1B of active users, Windows Store and apps are popular, developers are back and Windows 10 Mobile resurrects due to a lot of UWP apps. Windows platform is stil an underdog in mobile, but it’s far from dead. Windows desktop is still popular for productivity.
In a more pessimistic world, Windows 10 adoption is not as fast as expected, Store utilization is low so there’s little apps and little of devs’ interest. Desktop OS usage declining fast, losing significance in enterprise and consumer space. Intel lost its battle in mobile and concentrate on making money on dying PC market (milking it). Windows 10 Mobile is officially cancelled. Microsoft concentrates on cloud solutions and cross platform enterprise products.
In which version of the future do you want to live?
