Microsoft has just Palmed itself, ending any hope of a mobile comeback. In the usual style of mis-managerial triumphant retreats, they got the press to write it up as a stunning victory.
The defeat in question is what others have been shouting about all week, Android and iOS apps now have an ‘easy path’ to Microsoft OSes. While this may sound like a serious victory to the technically naïve, it is anything but, it only accelerates Windows’ doom in mobile. The only reason SemiAccurate can see that the entire tech press ‘missed’ this obvious fact is advertising, Microsoft does quite a bit from what we hear.
To understand why this ‘victory’ is such a stunning failure, you have to understand a bit of the history of Microsoft in mobile. Starting with Windows Phone 7, Microsoft has had failure after failure. Their entire strategy was folly, it couldn’t have worked but they did it anyway, it was gross mis-management.
Things started out by changing the entire OS paradigm to one that is both ugly and non-functional. Worse yet it required a total re-write of the software, not just a port. You had to re-architect all your apps for the new Microsoft way, a very expensive and time-consuming task. For some strange reason, very few devs ported to Phone 7, zero marketshare tends not to entice developers to spend time and money. Microsoft was reduced to paying devs to port their wares, the proverbial whack-a-mole with suitcases full of cash approach. As ABI’s Kevin Burden said, “Trust us, there will be apps at some point. If that means paying developers, so be it.”
Between this and the few brave devs who ported their wares to Windows phone, the app landscape was sparse, very sparse. Even with a marketshare of 2-3%, it was >10% pre-Phone 7, there was some non-paid development. Then Microsoft released Windows Phone 8 which was effectively not compatible with Phone 7. Anyone who wrote for 7 got screwed. Anyone who bought a 7 device got screwed, there was no upgrade path. In essence everyone got screwed but it was purportedly for good technical reasons. Burnt users somehow didn’t care.
With momentum in the Windows phone ecosystem almost stopped dead, Microsoft was back to hauling suitcases full of cash around and paying for ports. The grass-roots non-paid devs were few and far between on Phone 8 but those that were there actually made good money, especially on the ad side. Many attribute this to the sparse app market, no competition means anything decent probably sells. Microsoft attributed this to the thriving ecosystem and Bing ad model that was a gem for developers.
Then Microsoft did something that most onlookers and press didn’t realize was happening, they pulled the plug on a little known ad guarantee. That was Microsoft guaranteed devs that all ad spots would be filed by Bing ads or whatever it is called. Devs had it good because if there were no ads to place, Microsoft would effectively buy those slots for its own use. Everyone made money except Microsoft, they were hemorrhaging it to the point that it was even painful for them.
So they ended the practice rather suddenly, much to the chagrin of developers and people with apps on their platform. Many complained that revenue after this cut dropped to a single digit percentage of their pre-cut revenue, and not a high single digit percentage either. Devs were burnt but Microsoft stopped losing massive amounts of money. It also showed quite clearly how bad the Bing/ad/Microsoft mobile revenue model was, it wasn’t just awful, it was non-existent. There was no ecosystem and anyone who believed the rhetoric just got a nasty lesson in economics. Windows phone development was effectively dead, and dead for cause. (Please note that Microsoft has removed the forums where users complained about their business practices linked in the ITWorld article. How ethical.)
Just the other day Microsoft was crowing about selling 8.6 million Windows based phones last quarter, on the surface a solid number. Apple alone sold 61.2 million iPhones in the same quarter and they are about 25% of the worldwide market for phones. Microsoft still can’t hit 3% of the market. Worse yet iPhones sell for ~$500 and up unsubsidized, Windows phones sell for <$200 unsubsidized, at least unsubsidized by the carriers. Anyone want to bet that those Lumias with the massive cameras and high-end storage cost a bit more than $200 BoM? You can sell a lot of product if you are prepared to lose money on each one. Microsoft is prepared to lose money on each one, and does so quite often, but they still aren’t selling. Time to rethink that old adage, eh?
Worse yet Microsoft started out asking huge sums of money for Windows phone, $35 was the figure several OEMs told SemiAccurate during the Windows Phone 7 launch time frame. It is now zero. Microsoft’s revenue is based on selling OSes and software, now they have to give it away and more to move 3% of the market in units. Apple doesn’t lose money on phones, nor does Samsung, LG, Xiaomi, and most of the rest of the OEMs..
Officially Microsoft will make it up on services and revenue but as we showed earlier with the Bing ad placement debacle, there isn’t a functional ad model to make up the revenue. Nor is there an app ecosystem to take a cut of, it is almost non-existent. That is where the biggest gun in Microsoft’s arsenal comes in, Office.
Microsoft’s rather idiotic and legally questionable bundling of Office with WART meant nothing less than doom for that platform. Sure you could argue that a slow, expensive, incompatible, and painful OS would have the same effect, we won’t argue against that position. When a vendor is charged $80-90 for a forced OS+Office bundle, it may work if your devices have a high enough ASP. Forcing that load on a low-end tablet is crazy especially when your competition, more capable and well liked by consumers, sells for $99 retail at the low-end for hardware and software. It was once again folly. It also failed, WART died a well deserved death an any consumer who bought it, well, they got about what we expected.
To change topics a little here, lets move to Office on mobile. Microsoft’s strategy to leverage their way into mobile was simple, use Office. If you wanted Office on your mobile device, you had to use Windows. To force the issue even farther, they made the desktop Windows 8 into a mobile-like OS to familiarize people with the Microsoft mobile offerings. SemiAccurate said it was a debacle that would ruin Microsoft’s desktop sales. Those taking Microsoft money had a distinctly different view. (1, 2, 3, 4 and thousands more if you look.)
In some ways the plan worked spectacularly well, Windows 8 users were almost immediately familiarized with the Windows Phone/mobile OS offerings. Unfortunately they hated it on a level that was obvious to anyone but Microsoft executives. Windows 8 dropped sales of PCs worldwide by about 15%, a stunning number, it was an abject failure. As SemiAccurate said at the time, trying to force a terrible OS on users by using Office as a lever had no chance to work, and it didn’t.
The first problem was obviously the impact on Windows desktop sales, it cratered. The second problem was brought on by mis-management, Microsoft seems to have no backup plan for this eventuality. Instead of users flocking to Windows Phone and tablets, they flocked away from Office. You could almost hear the shock from bewildered Microsoft executives over this one, same for the tech press. If you look at it without ulterior motives, you have to wonder what color the sky is in their world, it was an obvious failure long before launch. Sure Ballmer is gone but the new crop seems to have the same blinders on. So by trying to force mobile OSes, Office, and that ecosystem on users, they just destroyed their desktop stronghold.
Luckily there was a ray of hope in the tablet space, Surface. When the Surface tablet was announced, only SemiAccurate said it would doom Microsoft, everyone else took the money and ran with glowing reviews. If you read what SemiAccurate wrote, we said that Surface fundamentally destroyed the ‘partnership’ Microsoft had with OEMs. Before Surface they wouldn’t dare step out of line by offering alternative OSes like Android or Chrome, after they see them as their only chance. Once they all launched ‘alternative’ OS devices, almost every OEM noticed they get better margins, far better, on non-Windows devices. The PC market is over for Microsoft, it is just a matter of how long it takes to bleed out.
If you look at the numbers however, they aren’t that grim. Sure PC sales are in the toilet but Chromebooks et al are not eating into Microsoft’s sales, take a look at the latest Gartner numbers, don’t they look good? Better yet Gartner is predicting that even though PC sales dropped by ~5% this quarter, they will pick back up soon. How long have they been shovelling that same BS for? (Note: They are far from alone, take a look at the chart here for another shining example of industry ethics.) If it wasn’t for the XP EOL corporate sales bump, they would have quite the losing streak going for predictions.
If you look at the fine print however, you may notice that it says, “Data includes desk-based PCs, notebook PCs and premium ultramobiles. It excludes Chromebooks, Windows-based tablets with smaller than a 10-inch display, and other non-Windows-based tablets.” Hmm, so the sales figures directly exclude Microsoft’s competition for some reason. Any guesses who funded this study, it isn’t disclosed for some reason. Any guesses as to whether the questions asked were crafted to paint a picture that was unusually positive for some vendors, OS vendors that is? Your marketshare numbers are always rock solid and unassailable if you don’t allow the competition to be counted.
If you look at the traditional PC market it was been hovering around 350M units in the pre-Surface world. According to the Gartner numbers, Q1 2015 was about 72M for an annual run rate of 280-290M. That means in the post Surface, post Windows 8 world, the PC market has lost 20% of its unit share. Remember those non-counted devices? SemiAccurate’s sources have been saying that Chromebooks represent about 20% of unit sales lately, probably a coincidence but also bang on to equal the percentage Microsoft lost. Any guesses as to why they might want to skew the surveys to exclude things eating their lunch? Wall Street doesn’t like bad news, and analyst firms never seem to give it to them, they know who writes the checks.
So OEMs are pushing Chromebooks, Android, and anything else they can in lieu of Microsoft products. Why? Because they have to, Microsoft is their worst competition now. This isn’t to say Microsoft is competent, sells much, or anything like that, just that Microsoft has an unfair advantage. By this we mean the $35 or so that Microsoft charges OEMs for Windows, Surface doesn’t have to pay this. That is more than the margins on most PCs, ~10% of the retail price for most devices. Microsoft also gets subsidies from any ads or app sales, not that this is likely to be a big number, OEMs don’t.
If OEMs don’t run from Microsoft at full speed, they are dead. All signs point to them understanding the position they are in and acting on it. Surface ended the lock on OEMs and it energized them to attack Redmond. They are. 20% marketshare loss in 2-3 years is a stunning loss that somehow never shows up in the public numbers. If SemiAccurate was cynical, we would point to the pay-to-play nature of the ‘analysis’ firms’ business model, something they vehemently deny. Make up your own mind though, start by seeing if you can figure out who commissioned which studies and why.
Back to Office we have the grand plan to turn Microsoft into a software and services company. Google mostly gives away their apps which are Word and Excel equivalents. They are easy to use, tied in nicely to their OS and GMail offerings, and work well if not as fully featured as the Microsoft Office equivalents. For the overwhelming majority they are good enough and better yet, effectively free. Tempting a teenager to buy Office 365 to get a few bells and whistles over Google Apps is going to be a really tough sell. For those that need more, LibreOffice is a damn good value for the price.
To acclimate users to the licensing change, Microsoft is giving away Office 365 to get users on board. If you buy a PC you often get a year of 365 free, a $99 value or so. This also translates into having it on your mobile device, the subscription is for you, not per device although there is a limit. With PCs selling for far sub-$200 and many come with a free year of Office 365 (Note: At the moment there are lots of examples like the HP Stream 11 for $199 which comes with a free year of 365). Better yet if you sign up for a free MS email account like Hotmail, Outlook, or whatever they renamed it to this quarter, you effectively have a free online-only version of Office 365. In short you almost assuredly don’t have to pay for Office unless you feel a moral obligation to fatten Microsoft’s coffers.
That brings us back to phones, Microsoft wants to subsidize phone hardware losses with revenue from ads and services. There is no ad revenue to be had and likely won’t be for a long time. There is no app revenue, or at least not enough to cover the hardware losses much less make net money. That leaves services of which Office 365 is about it. Leaving out the painful reality that an Office 365 subscription is a direct cannibalization of licensed Office as a product revenue, there is one more sticking point. That would be that for the time being, Microsoft is giving away far more 365 subscriptions than it is selling, I don’t know of anyone who has actually paid for it, but that is anecdotal, lots of people probably do for some reason.
The idea is to convert those free customers to paying ones when the free subscription runs out but do you think that is realistic? A teen-ager might use 365 if it is free, but are they going to shell out $99 when the year is up or will they go back to the Google apps or iOS variants they grew up with on their phone? Even if they do pay, the services revenue is not an actual net win, it is shuffling revenue from one category to another. Mobile Office has no chance of ever making Microsoft net revenue when you take that into account, but that is unlikely to stop the PR games.
And that brings us back to the original point, porting apps from Android and iOS. Apps are critical to Microsoft and they have almost none because there is no real incentive for developers to write for Windows mobile devices. To fix this there are two main thrusts, compatibility with Windows proper and emulation. Once again both are folly that will end up hurting Microsoft’s already badly bruised image, something management has to understand.
Compatibility is fine, the idea is that all versions of Windows regardless of device type will run the same apps. This means you have to code in the new universal managed code paradigm, something that no one is doing. If you write that kind of code, you get access to the scant few million Windows mobile device but you lose compatibility with the billions of Windows devices that aren’t mobile, which market would you prefer to sell into?
More important is that when you sell through the Windows store you get 30% taken off the top by Microsoft. They are quick to counter that 30% is the same number that Google and Apple take for their stores, so why is this a problem? In that light it isn’t problematic but there is a thriving non-Microsoft Store way to purchase and install Windows apps, and it works. If you sell direct to users as has been the norm for the last decade or so, you get 100% of the revenue. Those channels are well established and work well now, customers use them, and coders sell through them. Which would you rather have, 70% or 100%? Think Adobe will jump on board, or Autodesk, or Valve? Yeah right. If you want to be pedantic, you can point out that Microsoft won’t sell Office on Android or iOS because of the haircut, the app is ‘free’ but useless unless you have a subscription, you can’t buy Office through either store. Microsoft, thy name is hypocrite, but that isn’t news is it?
On the app porting side there are two variants on Windows 10, emulation for Android apps and recompilation for iOS apps. Both are doomed for the same reasons, and that is where Palm comes in. If you recall Palm had a thriving business, had software that worked nicely, and syncing was trouble-free and nearly faultless in the Palm I-III days. Then they decided to hook into Microsoft apps, specifically Outlook, for mail and contacts, and suddenly things weren’t smooth. Actually that is being unfair, they pretty much never worked after that point. For some reason the syncing kept breaking even when you didn’t upgrade or change the Palm software, it was almost like Microsoft was breaking their rival behind the scenes. If I didn’t know better, and they weren’t convicted in court for such predatory monopoly behaviors, well I would still believe it. Palm made themselves dependent on Microsoft behaving ethically, now they are dead.
Now Microsoft is desperate and is allowing Android apps to run emulated on Windows. iOS apps will need a full recompile, but Microsoft promises both will be a no-brainers. I keep hearing the echos of “write once run anywhere”, and J++ but that is probably just the birds outside my window. Microsoft is putting themselves at the mercy of the good will of Apple and Google for their very survival. If the apps ported over break, it will be Microsoft who fields the support calls, not Google or Apple. As long as Google and Apple play not just ethically but go out of their way to include Microsoft, all will be well. Anyone want to bet this will happen?
So Microsoft’s entire app ecosystem is at the mercy of it’s two greatest competitors, the ones that can show you ample scars from Microsoft’s ill behavior. All they need to do is change an API to something copyrighted, like Microsoft does all the time, and all the apps break on Windows. Like Microsoft’s opening of their .NET ecosystem to other platforms that strangely don’t support Direct_X calls, Google and Apple can do the same thing. They will.
Microsoft will be chasing compatibility and catching spears from angry users for the rest of time if they go through with this idiocy. Paid reviewers love it though, they think it is great. Unfortunately if the Android or iOS versions of a program work on Windows, sort of, what is the incentive for a developer to write real native code for Windows anymore?
Worse yet is the result for the end-user. Microsoft went out of their way to make Windows Phone7/8 and Windows 8/8.1 very different from Android and iOS. User hatred and sheer OS awfulness aside, they succeeded and it is very different in every way. Paradigms that are familiar on iOS and Android just don’t work on Windows 8 variants, it is a different world. Under the hood it is also equally different but that is beyond the scope of this article.
When you emulate Android so Android apps will run on Windows, all of a sudden you have a different set of UI paradigms. Will a down-swipe do the Windows function or the Android one? If you override one, how is the user to know what function that is replaced by? Will Android apps update the Live Tiles correctly? I think not. Will they look like crap with generic placeholders and sort of functional features, mostly? Yup. Is there any incentive for a dev to re-code the app to work properly with Windows as well? If that incentive existed, they would have ported long ago. Now there is far less demand because it is Microsoft’s problem, not theirs, right?
Same for iOS apps but to a larger degree, the apps have to be ‘recompiled’ from the beginning. SemiAccurate thinks that recompilation will result in a long list of fatal errors that need to be recoded and reworked rather than a one-click port, but again, place your bets. This emulation and recompilation could work but it will be vastly slower, look far worse, and function questionably at best for anything other than trivial apps.
There really isn’t an upside for users. Worse yet Google and Apple can break things on a whim, introduce a new OS call that MS has problems emulating and for the next six months all the apps break. You can pretty much count on this becoming as regular an occurrence is it was when Microsoft was doing it to their competitors. Any user running emulated apps is asking for trouble.
Then there is the problem of payments and microtransactions. Google and Apple have robust and secure transaction methods that have been proven for years. Microsoft can’t secure a wet paper bag, there are more compute cycles on Windows boxes going to malware than user wares. Would you trust Microsoft’s secure payments? They don’t even try to safeguard your data now so the odds aren’t good but do with your money and credit card numbers what you will.
So if you run an emulated Android app, everything will be fine as long as it doesn’t need to call a Google service. Or do any in-app purchases or transactions. “Upgrade this app to the pro version for $1.99” won’t work, and all the little things will get annoying in large number. Will it sync your data to the cloud? Which cloud? If you have a history of using an app with a few years of data in Google, what are the odds that it will sync when emulated on Windows? Will your movies, music, and everything else move over? Not a chance in hell. Will your apps call Bing maps as seamlessly as Google maps? Shall I go on? Do you think your bank will let their app run emulated in a questionably secure environment?
In short the emulated apps will do the same thing they did on Micro-Nokia’s not-Android phone, only worse. That would be a sparsely populated store of partially working, jarringly standout apps that keep breaking for no reason. Developers can fix all of that but, well, there is no payback in doing so. The best Microsoft can achieve with these hosted/emulated/ported apps is a black eye. That said they can fix the problems one by one by reverse engineering the errors and fixing them, but that would be more costly than running around to the devs one by one and paying them to port. Need we remind you that strategy has already failed?
Worse yet there is a killer problem that Microsoft can’t get around. All Google and Apple need to do is encrypt their packages with a key that they don’t give Microsoft. It doesn’t actually need to be secure or anything more than ROT13, if Microsoft works around it, it is a DMCA violation. You might recall that this is exactly what they did to their competition with the new Office formats, trivial encryption so that when used, and it was hard not to, you couldn’t make anything compatible under the law. Emulation goes bye-bye with this trivial security upgrade but SemiAccurate is betting that Google and Apple use far more subtle and costly to Microsoft dirty tricks. After all, those two learned from the master, and they did learn.
In the end the whole emulation and recompilation strategy is doomed for the same reason as working with Microsoft doomed Palm. Microsoft is now laying their very existence on the good will of their competitors who they have badly mistreated with every dirty trick in the book over the past few decades. Now unless those two competitors go out of their way to support and nurture Microsoft’s mobile ambitions, it will likely cost the Redmond giant even more desktop market share. For Microsoft management to get themselves into this position, and do it in a public way, is once again gross mis-management. I for one can’t believe they still are this incompetent.S|A
Latest posts by Charlie Demerjian (see all)
- Another Epyc Rome TCO data point comes out - Nov 12, 2019
- What is the name of Intel’s Cascade Lake +5 server? - Nov 9, 2019
- AMD launches Threadripper 3, 3950X, and Athlon 3000G - Nov 7, 2019
- Intel messaging hits a new low - Nov 5, 2019
- Intel releases the world’s largest FPGA - Nov 5, 2019