Suggestions for a former employer (ie: What should Microsoft do?)


Well I used to work for Microsoft. After 10 great years, I decided to move on a few weeks ago and am now at a competitor. Thats about the extent of the “full disclosure”; now onto the Monday morning quarterbacking! (although this is really more “Friday night”)

Right now, Microsoft is in a rough position. The stagnant stock is reflective of this predicament because stock price is a valuation of future potential more than it is current performance. Many don’t understand this when they express frustration over how a company with such high revenue and net profit can trade so flat.

There are a number of trends converging in both the enterprise and consumer segments that are radically reshaping the environment that Microsoft has traditionally dominated:

Consumer:
– For millenials (folks in their 20s and younger), technology is reflexive. They aren’t “geeks” in the old sense and neither are they grudgingly dragged along. They grew up on technology and rely on it the way one relies on a car, lamp or microwave. For these folks, the computer is the web browser.

– The lions share of growth in personal computing is in mobility and non-traditional form factors. PC’s still get refreshed, but people increasingly want to live on mobile devices and the vast majority of tasks people perform (80/20 rule here) can be completed on an advanced mobile device at the current state of technological evolution. For now these devices are additive (laptop PLUS smartphone PLUS tablet), but tomorrow they may not be (unless they are priced as disposables)

– Building from that point, commodity pricing is taking effect because technology has become commoditized. Margins are tighter.

– These users expect the device to essentially disappear and for the user experience to be seamless. Computing for its own sake was never genuinely valid as a use case and it certainly will never apply to a broad audience. Simple, personal, visually appealing, fast… This is how technology is measured today. It isn’t the death of the tinkering power user, it is simply that technology companies have the opportunity to sell to about 100,000x more people than will ever be in the “tinkering power user” category if they meet these challenges.

Enterprise
– Consumerization is a real trend that is exciting for CxOs. If folks are passionate and personal about how they access technology, let them do what they want. Personal responsibility on cell phones lead the way and computers are next. Maintaining core infrastructure is complex, capital intensive, heavy on OPEX (death for a publicly traded company) and has no business value. The tremendous influence of consumer technology on these users is pressuring IT (Facebook, Google, GMail, Skype, et al) to deliver capabilities beyond their ability to quickly and efficiently deliver.

– Industrialization is the companion trend to commoditization in the enterprise. Disposable, reusable, self service components that are well integrated and can be easily managed are what business wants. IT getting out of the way of the user and technology folks employed to solve business problems, not run technology. It’s business 101 really. It is only IT Pros that fail to, or refuse to, grasp the reality here.

– Platform shift to the browser is a reality. Business computing is inherently heterogenous. Vendors have long tried to lock IT into a single vendor vision, but the broader the ambition the bigger the failure. The downside to a heterogenous environment, and a driver towards possibly considering a single vendor approach, has been managing the complexity. The web has leveled this playing field. Move apps to the browser and enable all workloads for web services interaction and a huge pressure release valve is opened. Web standards have evolved quickly in response and the ability to deliver completely standards based web apps is becoming a reality. Offline capabilities are either irrelevant because computing devices are increasingly universally connected, or are a system architecture exercise in caching and state synchronization.

Where Things Stand
Microsoft has spent the past 10 years essentially running uphill. The clock keeps ticking and the strategy has continued to be doubling down on the legacy business model while publicly claiming to be “all in” on the emerging one. The drive to the cloud and more agile licensing and delivery vehicles has been hamstrung by the desire to protect the legacy base of fat client OS installations and continual OEM and annuity revenue through arcane licensing programs and high surcharges. The effect has been that Microsoft has continued to fall behind the curve and is in real danger of losing relevancy.

Doomsday?
What is the worst possible outcome? Doomsday looks like Microsoft losing all enterprise OEM and SA revenue to “Bring Your Own PC (BYOPC)” models and having only a percentage of that revenue shift into the consumer OEM business. Why? Because given the ability to buy whatever they want, a much larger chunk of users today will buy a Mac. And tomorrow maybe a Chrome or Android machine. Consumerization and commoditization mean choice and diversity; that is inevitable. Parallel to this is the increasing trend towards growth only existing in a meaningful way in non-traditional form factors like lightweight slates and smart phones. Microsoft has missed badly here to become relevant while also failing to hedge against the trend by establishing a business model that is not dependent on platform monopoly. Doomsday means failing to gain a foothold with Windows Phone and also failing to course correct and hedge.

OK Genius, so what is YOUR big idea?
So in an ideal world, what should Microsoft do? Is the tailspin inevitable? With the resources and installed base Microsoft enjoys, there is certainly time to fix the bleeding, make the most of the existing bright spots, shore up some failing but solid ideas, cut what won’t work and re-size for the future. The process would be painful but would leave the company stronger and healthy. Here are some ideas:

– Support IT in the emerging VDI world. To continue to attempt to tie VDI to the legacy OS (requiring SA, charging a full price for a VDI license, etc) will simply accelerate trends away from Windows. There should be two ways to procure a Windows OS for the enterprise. From an OEM on a new PC, or via a VDA SKU that has a low per user per year cost and is an entitlement to *only* run virtualized. The concept of an in place upgrade of old hardware is something that IT never really wanted, and certainly doesn’t want today and whether or not it ever actually mattered, consumerization will render the point moot. In a BYOPC world Windows revenue has shifted to either OEM or VDI. If the policy is to basically charge enterprises full price for Windows on a VDI blade, it will accelerate the move of their app stacks to the browser.

– A Windows App Store is so long overdue its ridiculous. Steam is ancient, XBox Live is ancient, yet Apple will beat Microsoft to the punch delivering client apps in a modern and agile way to a legacy desktop OS. Microsoft should see the Apple app store, and raise it. Why not deliver OS upgrades this way as well? A user should have the ability to either subscribe to “Windows for life” which would combine the Bing/Live subscription services with an “OS subscription” all priced at maybe $20/month *or* buy an upgrade digitally from the App Store. So when Windows 8 launches, it is a front page priority “System App” that is either “FREE FOR SUBSCRIBERS!!!” or some normal upgrade charge ($129 seems reasonable). Purchase of this should kick off a deployment event that is all automated and online. A migration assistant should run. It should inventory hardware and Windows System Index and suggest trips to Geek Squad to shore up bottlenecking subsystems “you’ll be ready to go with a faster hard drive! consider buying the Intel X25-M from Best Buy and having Geek Squad install it!” (GREAT vendor tie in and partnership opportunity here). If everything checks out, the machine should be backed up to the cloud, state, personalization and transient data saved, and then upgraded and restored. Upgrade should be copying down a WIM and a bootloader, and then triggering a reboot into the loader that wipes and reimages the PC and on first run connects back up to the cloud to do a restore. All of this is done by enterprises today and more (including triggering this process from a cold boot state on a powered off machine) with tools Microsoft *currently sells*. This is something Apple *could not* match today and would be a phenomenal differentiator *and* “cloud” story.

– Make the XBox a first class citizen. The XBox is a fantastic alternate platform. It is also a great media extender and it is the living room presence Microsoft always dreamed of. It seems they dont realize this. What other explanation can there be for allowing Google to start to define the conversation around converged general computing and content aggregation in the living room with Google TV? Microsoft has had *all* of the pieces here via MS IPTV, XBox, XBox Live, Media Center and Bing in a far more functional state than Google for a good long time. Despite this, a wall remains between the XBox and the PC. That wall is the Windows group and legacy fiefdom mentalities. Why can’t “Office Live” be used on a keyboard equipped XBox by college students? Who knows…

-Get phone right or abandon it. Windows Phone should have been *all* of the things Windows Mobile had right (S/MIME, RMS, encryption, full management capability, multi-tasking, clipboard, to name a few) while also not sucking and being better than iPhone or Android at everything current in the consumer space. Somehow it came out interesting, but essentially neither. There is about a 10 minute window of space to fix this before the door slams shut potentially forever. Android has already filled the massive vacuum Microsoft left in the handset business and gave every handset manufacturer that isn’t Apple what they needed to stay alive. These guys have some passing interest in a Microsoft alternative, but it is lukewarm at best and can cool very fast. Same goes for the carriers (especially as the AT&T exclusivity expires)

-Mobilize Windows. Windows 7 needs to be reengineered at the presentation layer to run on a variety of form factors and a variety of interface paradigms. Jamming legacy Windows onto a slate and allowing OEMs to try to figure out how to make it usable is the exact same strategy that failed with Origami and it will fail again. Failure to create a robust ecosystem and make the mobile experience first class by forcing app groups to mobilize their products as well (yes you, Office team) will ensure a stillborn effort as well. It is ironic that after spending 10 years taking a run at this, Microsoft must now use Apple as a model, but Apple got it right. They mobilized OSX into IOS, built a broad ecosystem that made sense, mobilized their app stack (iWork) and now are flowing lessons learned back into the core platform (OSX app store and Lion). Increasingly, Microsoft is defining a vision (mobile computing, software plus services, etc) that others (Apple, Google, etc) are delivering on and as a result, Microsoft is losing mindshare badly.

– Either “do” or “don’t” with HyperV. Microsoft has decided that VMWare is a battle it cannot lose. Personally, I don’t agree with this, but that is a different discussion. If this is the case, then the entire thought process around the OS must be redefined. This can’t be old school mentality leading the way with virtualization as an after thought. Microsoft can’t continue to play catch up, can’t continue to have virtualization span multiple silos (especially since Microsoft business silos interoperate poorly) and *should* attempt to leapfrog the competition. Being in the unique position of controlling the development of both the hypervisor *and* the guest OS, there should be radical changes in terms of what “OS” even means in a virtual world coming soon. It seems crazy that every server guest OS, in this age of virtualization, has almost no design principals implemented that were driven by the needs of a virtualized environment, but it is *especially* crazy for Windows since a hypervisor based environment is supposedly a core strategic direction.

There are more, but this seems like enough of a start. Possibly more to come if I can find the inspiration to complete it!

Advertisements