Follow by Email

Friday, July 19, 2013

Observing Microsoft, Part 3

When a company chooses a strategy, it is usually important that the strategy must make sense given its existing business model. A strategy of changing the business model, however, is a much harder one to implement and takes years. And that's one of the reasons why I'm observing Microsoft.

OMG there's so much to catch up on! But it's clear the trends I was referring to in my previous installments are being realized. To start with, I looked at their Surface and Windows 8 strategy, and then I looked at their management of the Windows brand, and its subsequent performance in the crucial holiday season.

Converting themselves into a hardware company, in the Apple model, is sheer madness for a software company like Microsoft. It will kill off their business model very quickly, I think. And yet they continue to do it, company culture be damned.

Ballmer is a coach personality, and clearly business looks like a football game to him. I can imagine him saying "if a strategy is not working against our opponent, then we must change it up". But it's clear that it's much easier to do this with a football team than it is to do the same with a company of 100K employees.

So I wonder why Microsoft doesn't just focus on making business simpler? Instead, they have been making it more and more complex by the ever-expanding features of Office, their business suite.

Software, hardware, nowhere

As one of Steve Jobs' favorite artists, Bob Dylan, once said "the times they are a changin'". And Steve knew it, too. At TED in 2010, Steve said that the transition away from PCs in the post-PC era had begun and that it would be uncomfortable for a few of its players. I took this to mean Microsoft, particularly. But how has it played out so far?

Microsoft is a software company that dabbles in hardware. Most of its revenues come from software, but remember that they make keyboards and mice and also a gaming console. These are only dabbling though, because the real innovation and money is to be made in gadgets like phones, tablets, and laptops. But their OEMs make gadgets, which requires a significantly greater level of expertise and design sense. So Microsoft's entry into gadgets can only represent their desire to sell devices, not licenses. They want to be like Apple, but specifically they want to own the mobile ecosystem and sit on top of a pile of cash that comes from device revenues. And the OEMs like HP, Lenovo, Dell, Acer, and Asus are a bit left out; they must compete with their licensor. That can't be good.

So Microsoft is clearly changing its business model to sell hardware and to build custom software that lives on it. Hence Surface RT and Surface Pro. But their first quandary must be a hard one: what can they possibly do with Windows? Windows 8 is their first answer. The live tiles "Metro" style interface is unfortunately like greek to existing Windows users. The user experience, with no start menu, must seem like an alien language to them.

This entire process is beginning to look like a debacle. If it all continues to go horribly wrong, the post-PC era could happen a lot sooner than Steve thought.

Microsoft ignores their core competence as they blithely convert themselves to a hardware company. Specifically, I think that's why they are doing it badly.

They could end up nowhere fast.

Microsoft's numbers

Microsoft is a veritable revenue juggernaut and has done a fairly good job of diversifying their business.  An analysis of Q4 2012 reveals the following breakdown of their business units in revenue out of an $18.05B pie:

23% Windows and Windows Live
28% Server and Tools
35% Business
4% Online Services
10% Entertainment and Devices

This reveals that business is their strongest suit. Servers also speak to the business market. Online services also largely serve businesses. Each division, year over year, had the following increase or decrease as well:

-12.4% Windows and Windows Live
+9.7% Server and Tools
+7.3% Business
+8.1% Online Services
+19.5% Entertainment and Devices

This reveals that Xbox is their fastest-growing area. It is believed that Xbox is leaving the PowerPC and moving to AMD cores and their Radeon GPUs. This could be a bit disruptive, since old games won't work. But most games are developed on the x86/GPU environment these days.

It also shows that their Windows division revenue was down 12.4% during the quarter year over year. This involved a deferral of revenue related to Windows 8 upgrades. Umm, revenue which most likely hasn't materialized, and so you can take the 12.4% as a market contraction.

Why is the market contracting? Disruption is occurring. The tablet and phone market is moving the user experience away from the desktop. That's what the post-PC era really is: the mobile revolution. Tablet purchases are offsetting desktop and laptop PC purchases. And most of those are iPads. It gets down to this: people really like their iPads. It is a job well done. People could live without them, but they would rather not, and that is amazing given that it has only been three years since the iPad was released.

The consequence of this disruption is that PC sales are tumbling. If you dig a little deeper, you can find this IDC report that seems to be the most damning. Their analysis is that Windows 8 is actually so bad that people are avoiding upgrades and thus it is accelerating the PC market contraction. On top of the economic downturn that has people waiting an extra year or two to upgrade their PC.

Microsoft CEO Steve Ballmer stated in September 2012 that in one year, 400 million people would be running Windows 8. To date, it appears that only 80 million have upgraded (or been forced to use it because unfortunately it came installed on their new PC). That's why I said we need to ignore that deferred revenue, by the way.

If you look at OS platforms, Microsoft's future is clearly going to be on mobile devices. Yet they are not doing so well in mobile. In fact, they are becoming increasingly irrelevant, with about 80% of their Windows Phone models on only one manufacturer, Nokia. Soon, I think they may simply have to buy Nokia to prevent them from going to Android.

In the end, you can't argue with the numbers. The PC market is contracting, as evidenced by Windows revenue declining year-over-year. Tablets are not a fad. As the PC market contracts there are several companies that stand to lose a lot.


What is the Microsoft reorganization about? There are three things that I single out.

The first and most noticeable is the that the organization puts each division across devices so the software development is not device-compartmentalized, and so that Windows for the desktop is written by the same people who write Windows for the devices. At least in principle.

And, of course, games are now running on mobile devices, dominating the console market. And undercutting the prices.

This closely mirrors what Apple has been doing for years. And this clearly points out that Microsoft is envious of the Apple model and its huge profitability.

Second, in reorganizing, Microsoft is able to adjust the reporting of their financial data, to temporarily obfuscate the otherwise embarrassing results of market contraction. This is because if each division reports across devices then the success of a new device will hide the contraction of the old ones. At least, in theory.

But Microsoft made a huge bet in the Surface with Windows RT. And it's not panning out. They have just reported that they had to write off $900M of Surface RT inventory in the channel. The translation is this: it's not selling. They have instituted a price drop for Surface RT. I bet they won't be able to give them away. But when they finally are forced to, they will be the laughing stock of the mobile market.

Today, Microsoft is down 11%. That's represents a correction. A re-realization of the capitalization of Microsoft. This represents a widely-help perception that the consumer market is lost to them.

Third, Ballmer wants the culture of Microsoft to change. They have been having problems between competing divisions. Coach, get your team on the same page! Wait: they should have been on the same page all along. After all, the iPhone came out in 2007, right? Ballmer didn't think too much of it at the time. That's why coaches hire strategy consultants.

A reorg can be even more traumatic than a merger. It's all about culture, which is the life blood of a company. It's what keeps people around in a job market that includes Google and Apple.

Monkey business

I have to give it to Microsoft: they really want to give their tablet market a chance. But they are doing it at the expense of their business market. They are reportedly holding off on their Office for Mac and iOS until 2014. A deeper analysis is here.

This is a big mistake. They need to build that revenue now because BYOD (bring your own device) is on the rise and they need to be firmly in the workplace, not made irrelevant by other technology. If they lag, then other software developers that are a lot more nimble will supplant them in the mobile space. Apple, for instance, offers Pages and Numbers as part of their iWork suite. And those applications read Word and Excel files. And they can also be used for editing and general work.

Microsoft should be focusing on making business simpler. Cut down on the complexity and teach it to the young people. Reinvent business. This entails making business work in the meeting room with tablets and phones. Making business work in virtual meetings.

They certainly had better make their software simpler and easier to use. They must concentrate on honing their main area of expertise: software.

If they don't do it, then somebody else will. Microsoft should stop all this monkey business, trim the fat, and concentrate on what adds the most value. They simply have to stop boiling the ocean to come up with the gold.

The moral

There are some morals to this story. First, don't ever let "coach" run a technology company. Second, focus on your core competence. Third, and most important, create the disruption rather than react to it.


  1. PC sales are not tanking because people are switching to mobile devices. PC sales are tanking because there's no real reason to replace old PCs with new ones.

    Once machines with 64 bit chips became available in the mid 2000's, the upgrade treadmill ground to a crawl. Clock speeds no longer continually rise. 250G is usually enough disk space, and 4G or so of RAM is plenty for most casual PC use. Only hard core gamers and graphics users need to get the latest GPU.

    In the same talk you mention (I think it was a D conference, not TED), Jobs compared mobile devices to "cars", and PCs to "trucks". People still need their PC trucks to get their work done, but the one they bought in 2008 is still humming along just fine.

    1. Good insight. It's true that people still need to get work done. But more so, people are getting work done on tablets and with smartphones. You see them more in meetings, though I have found that laptops still prevail.

      But there is plenty of evidence that PC sales are slowing down because of replacement by mobile devices. Perhaps in the enterprise this is occurring slower than in consumer. Mobile devices are more convenient to keep current on the go. When I'm at my desk, though I still use a desktop machine. And that's not likely to change soon.

      Windows 8 did more to slow the PC market than any other single thing that Microsoft did. But PC market contraction in certainly not all due to Microsoft's poor judgement.

      The end of Moore's law has been heralded for quite a while, but I still see multi-core machines and GPUs making things faster for a while. The GUI experience is made smoother by GPUs, which is desirable. Perhaps not to Windows users just yet, but this is certainly occurring in the Mac market, which is a predictor of future PC experience. Because Microsoft has copied the Mac GUI for years and continues to do so.

      If it weren't for Metro (umm, Modern Windows UI now...) I would have written off Microsoft as a gross underinnovator. But now, I think there's hope for them.

      Multicore makes things like video processing, Photoshop, and multi-processing much faster, and there is plenty of work I do from day to day that profits from multicore.

      Will it make spreadsheets faster? Will this speed up Word?

      Maybe not, but it sure does speed up display of web pages! And that's needed by everybody.

      I work with photographs, and 250G is not enough. Hell, my laptop has a 768GB SSD on it and it doesn't seem big enough. However, to your point, photo work is *not* typical to the enterprise, and so perhaps the majority of PC users can make do with yesterday's machines.

      But make no mistake about it, the consumer market is much larger than the enterprise. This is what Microsoft is poised to lose.

    2. Robotics, automation, speech recognition, i.e. the next wave of smarts, will require multicore. Those old PCs won't suffice for long. Your computer will be your smart assistant.

    3. Please consult my post on multicore and the increased specialization of processors in computing platforms

      Really getting these new technologies right often requires some special work, like photography requires ISPs.

    4. We'll need multicore interfacing with any new custom chips, and the point remains even if we only add custom chips running with a Pentium 4 (not likely though) that the old PCs won't suffice.

    5. Multicore is only a panacea to the naive. Most specialized kinds of processing do not fit efficiently into the standard computing models. This includes most media processing, recognition processing, and security processing.

      Where multicore shines is in handling the presence of multiple apps - which require general programming. Though any sufficiently interesting app requires the specialized processors as well. And this generally can create resource issues which multicore does not solve in and of itself.

      Here I'm referring to the ISP (image sensor processor), a built-in chip for handling face recognition cascades, a built-in crypto chip, a fast resampling chip, a GPU (which may feature many shaders), a chip for media standard encoding (like i.264 and other mpeg standard), and finally the old workhorse the multicore CPU.

      Devices also feature a panoply of sensors such as typically multiple image sensors, a proximity sensor, MEMs sensors for gyroscope, accelerometer, magnetic field, or other inertial parameters, etc. Things like quartz crystals are being replaced by MEMs devices which are much more accurate. MEMs microphones are becoming common now as well. Temperature and shock sensors are especially important.

      Then there is the comm part of devices. This includes Wi-fi, bluetooth, and various forms of telephony (there are many standards GSM, 3G, 4G, LTE, and they vary by the locality). The infrared aspect of comm is becoming important as well, though wi-fi and bluetooth are excellent even for P2P comm.

      I expect that PCs will move the way of the device when it comes to the peripheral (custom) processors. If it helps to cut down on power consumption, then it will certainly be adopted. And the faster the better. Laptops run on batteries, and so that will force it.

      But even servers, which must become more and more energy efficient, must adopt this power-saving technique.

      Gee, I hope Intel is all over this. ;-) Because it'll be their hide if they aren't.

    6. So I should buy Qualcomm (Snapdragon, etc) stock? ;-)

      Nice summary.

      Our apps have to integrate with those specialized processors, and there is more than one thing happening at the same time. Multicore is more efficient (speed and energy) than trying to make one really fast core and using context-switching (is Intel's hyperthreading superior or just because most apps are not optimized for N cores?). Or I guess the salient point is the single-core just won't be fast and energy-efficient enough. Am I wrong?

  2. From my perspective, the big picture is that hardware is mass produced, but software is increasingly tailored, i.e. app markets. Thus the revenue for large companies is not in software (except the common denominator OS which can't be sold more than once per year or so). The revenue is in the mass produced hardware. Eventually with 3D printing, the hardware will also succumb to tailoring and that will be the end of the large corporation. But for now, hardware is how large corporations must survive.

    This shift in the software was caused by the internet, as I wrote in the prior blog. Microsoft is doomed and they have no where to turn. We should be talking about Apple and the threat it faces, as Apple still has window of opportunity to correct their mistakes, but it is closing fast.

    Google's (Motorola's) new MotoX is Google's realization of the above fact.

    Btw, I wrote in the prior blog that Apple needs to integrate with other devices more, and now I read MotoX auto-integrates for example with the display in your car. And Google released a $35 adapter from the smartphone to the TV.

    It is not looking good for Apple. They only have 9% share in China and Android takes the rest. Now they are falling behind Android on integration, global market share, web share, big screen smartphones, etc..

    Tablets are not the core market, I would advise Apple to focus only on the enterprise tablet size screen (or at least not expend disportionate resources on tablets), and put all their other resources pronto into:

    1. Integration (make sure Android can't lock them out, if necessary interopt with benefits)

    2. Large screen smartphones (4.7 and 5.5+)

    3. Low-priced smartphone to compete in the developing world (where the future is).

    4. Allow unapproved apps section in their app store, as long as they make it very clear they don't control what these apps do. The tailored app market is too much the future to play around with censorship.

    1. Spot on that Balmer is a coach. Microsoft lost their strategist, Bill Gates, and Apple lost theirs, Steve Jobs. Tim Cook said he didn't understand why tablet sales missed. He is an organizer, not a visionary nor strategist.

      Someone needs to lite a fire under Tim Cook yesterday. Apple's iteration cycles are much slower than Androids, and thus they can't afford to make mistakes such as being too late on big screen smartphones, being too late on using sensors to auto-integrate with the displays in the environment, etc..

      <rant>I would love to code for both iOS and Android, and would release my new language for both, but I am not going to code for any OS where the company censors what code I can offer to the customers. They don't own the customer, the customer owns him/herself. I understand wanting to build a quality ecosystem (prevent malware), but there is no logic that supports the risk the developer takes off losing an investment by suddenly being banned (e.g. when Apple disagrees with the morals of the app). There is a simple solution, which is put the unapproved apps is clearly market area and actually list the reasons the app wasn't approved, so the customer can decide for him/herself.</rant>

      Nevertheless I want to see Apple prosper, because I don't want a one-horse race. I want competition in this space. But sadly it looks like Apple is failing.

    2. Ballmer has a coach mentality. And as such, he is not a next-gen strategist. But neither is Bill Gates, who has Microsoft buying content rather than proffering it. But as we can see, neither road is completely clear.

      Please, no rants, otherwise I will have to curate your messages. Actually, that's a joke. ;-)

    3. I need to learn how to spell "light" ;-)

  3. First a macro-economic point w.r.t the relative trajectories of Microsoft, Apple, and Google. Some details were in the prior blog.

    The western empire is dying (empires move periodically from Europe to America to Asia to Europe...), but empires never collapse with hyperinflation, rather always deflation and onerous taxation (police states devolution into abandonment of the cities).

    The future (after 2016 when the USA begins its big economic collapse) after China has its 1929 handoff depression in 2016, is consumer economic volume and growth will be focused in the developing world predominantly Asia. The stock valuation of these companies should be based on their market share in Asia. The collapse starting in 2016 (maybe ending 2020 for Asia, 2033 for West) will do that.

    I hope Tim Cook (Ballmer is hopeless) is aware. You may not agree, yet I have studied the evidence that this repeats every 78 and 224 years. The Google people may not understand this, but they understand that global market share is the future, so they arrive at the same optimum long-term strategy.


    Off topic: this should probably go in the prior blog article, or preferably in one of your blogs about more abstract concepts. I tossed it here because I am not sure if you still read new comments in the older blogs?

    I wanted to share with you the shock I felt when my just turned 14 year old girl (half-pinay) emailed me after reading my blog The Universe, "Would darkness be considered faster than light since however fast the speed of light is, darkness will always be there waiting.". I will blog (on my blog) my answers (to that and her other questions) when I have time. In my opinion, she is already thinking at the abstractions of a high physicist even her math is not yet there (which is partially my fault for screwing up the primary education in PI not realizing her IQ is higher than 120s according to raven matrices). Last summer she absorbed half a semester of algebra in a week, and current middle school is too easy for her. Trying to figure out where to send her.

    1. And even further off-topic to finish off our discussion on health from the prior blog and I I will not mention it again.

      I have made two major discoveries (iodine & monolaurin/medium-chain saturated fat) and seem to have a significant improvement in my health since increasing uptake of Kelp tablets and white coconut meat over the past 3-4 days. Probably I don't have head or neck cancer (yet), because the effects cleared, but that is where the HPV likes to attack on men (and was apparently attacking me):

      I already suspected that most pathologies (including cancer) are auto-immune reactions to toxins and viruses. 95% of us in USA are iodine deficient and we need 12mg (12000 mcg) per day, not the 150mcg official RDA. Every cell of our body stores iodine, not just the thyroid and this is essential for proper metabolism and functioning of the immune system. Also the lauric acid in the mature coconut meat (becomes monolaurin in the body) is only found otherwise in human breast milk and is a powerful anti-viral and immune system support (infants need immune support bcz they don't have yet their own). The halides in our food, water, and sodas (e.g. Flouride in our water supply and Bromide in our sodas and bread) bind to where the iodine would, thus blocking uptake of iodine and causing pathologies. Even we have to be careful with seaweed (the highest source of iodine by a factor 10 or 100) because it can be polluted with bromide. There are too many details to mention in the following links.

      (ignore the text which is wrong, and listen to the shocking video)

      (cancer cure)
      (cured MS with cruciferous vegetables)
      (almonds can replace statin drugs)