OK, ex-Microsoftie with 6.5 years experience in the Services (i.e. field engineering / consulting) side of the business; 2.5 years working on Windows before that. I quit Microsoft to chase my fortune at a pre-IPO company this past January, although I still have a lot of friends working for MS.
First things first, you’re conflating two different beasts. Microsoft isn’t just Windows, and Windows isn’t all Microsoft has. Speaking of the motivations, for example co-existence with OSX, will be different from Microsoft as a whole (who would support say Office working with OSX) and Windows (who want to see OSX die a bloody death and take Apple with it). So in light of that:
Some points stated here are IMO valid. Microsoft has never specialized in innovation, more like taking another’s idea (usually legally but not always, hence the number of lawsuits) and evolving it. The list of success and failure projects isn’t wrong; it’s just overblown in that everyone screams that Microsoft is failing and dying as a company because of one significant failure, which says more about their agenda than Microsoft’s chances. Sure, Vista isn’t a raging success, but it’s far from a huge flopping failure and isn’t even close to being the worst failure in the OS arena - anyone remember WinME? MS-Bob? Total up the successes vs failures of new product launches; I would hazard a guess that Microsoft’s doing pretty damn well, in spite of a few wobbles, in the overall stakes of success vs failure of new products compared to any other business in any other sector.
Zune - yep, kinda failed to kill iPod but going from zero it’s got a 15% market share already which ain’t bad in a crowded market
XBox - can you deny Xbox is kicking some serious ass?
Vista - slow adoption for sure, and a decent amount of bad press, but Vista is still being adopted. most of the problems people are seeing are application or driver incompatibility issues - not Microsoft’s fault although Microsoft as usual is catching the blame
MSN Music - sure, it ain’t killed iTunes, but it ain’t that bad either and it’s still cranking along and making money. And integrated with the Zune it’s got some serious fans.
Tablet PC - not selling all that well compared to standard laptops, but definitely an option and have gone from 1 offering to every major manufacturer offering at least 1 tablet PC
Lots of other stuff is given away for free (Internet Explorer, etc…) so I don’t really count those as a success vs failure, but the bundled software you get for free (media players, etc…) is at least as good as, if not better than, any other free media players on the market (Real, QuickTime, etc…) although they won’t compete with pay applications.
Second, a misconception seems to be that NT was a complete re-write of the Windows code base. Not true; it was never part of the Windows code base, they started in very separate teams a long long time ago. Some (very few) code elements were shared between them, but XP was the first time the NTx and Windows 9x code bases were merged into a single product with different SKUs for Home User / Office users. This was first attempted with Windows 2000 but didn’t happen completely - Win2k was originally called NT5 and originally was planned to be the retirement of the 9x code base, and XP was basically NT6 with bits from the 9x code base they wanted to keep.
On the business side, my opinion is that Microsoft does learn from it’s mistakes but slowly and has become such a large ship that it takes quite a while to change directions as a whole or even as a whole division - over 86,000 people work for Microsoft worldwide when I left in January; that’s up from 25,000 when I started there in 1998 as a permatemp. 4 years ago Microsoft Services hired a guy named Mike Sinneck from IBM to run our global Services team at VP level. Sinneck came in, and in addition to pissing off a whole host of people, completely re-organized the field Services teams. This re-org was widely seen as a huge mistake, pushing the organization away from a model that worked (managers were subject-matter experts running a team of consultants with sales people helping place the consultants into customers) into a model that didn’t work too well (all consultants now report into a pool of managers who are not SMEs and might have no field experience at all; sales people report into others who place the consultants on site - effectively 4 levels of management for a Services engagement when we had 2 before the re-org) but the problem is this: Sinneck was canned after just over a year, but no-one changed his model because it was showing increased revenue and this cannot be altered. Profit has become the driving force, rather than efficiency, and the two do not always go hand in hand.
Another thing to keep in mind is that Windows is only one part (admittedly a large part) of a huge global business. Microsoft’s other divisions combined make far more money than Windows alone (although Windows makes more than any 2 other groups combined); so whilst there is a huge focus on Windows and it is absolutely seen as core business, it’s not the only focus anymore. Hell, X-box wowed so many people with it’s runaway success; Office is kicking ass left right and center, and the Enterprise products are capturing more and more market share every year (stuff like System Center products, Exchange, and SQL). Visual Studio is a huge tool in the developer market and growing market share all the time. Silverlight is hugely popular with web developers and it’s not even out of the door yet. Sharepoint is selling so fast Microsoft Services didn’t even have to “sell” it, just take orders.
My final thought is that in my opinion one failing of Microsoft is lack of a long-term strategic mindset. This has been evident for a long time, but Microsoft really only thinks at an Exec level 12-18 months ahead, and does a major re-organization of the company traditionally every 6-12 months. I think this is why it’s so damn tough to develop new products anymore, because if a new product takes more than 12 months to develop, there will be at least 1 if not 3 re-orgs and there’s every chance that the team you started with ain’t gonna be the one you finish with. One re-org, I ended up working in a team where I didn’t even know who my ultimate boss (boss’s boss’s boss) was for over 3 months.
To answer the OP’s specific suggestions.
Rebuild the OS from scratch:
Reverse compatibility is a main draw to any Windows OS. If Microsoft started from scratch, in addition to taking years to release a new product, it would lose a lot of that. And don’t forget - every time Microsoft end-of-lifes a product, they get jumped on from on high. When Microsoft finally killed free support for NT4 two years ago, I got no end of shit from customers for killing this ancient product. I finally started asking them - how much longer would you expect us to support a product for free? Another 15 years? 20? Yet you think Microsoft should just completely re-write the code every time there’s a new release? Never, ever gonna happen. You’re also talking about roughly doubling the size of the development org by doing this - you’d need to keep the people writing old code for supportability and service packs, and you’d have to hire a whole bunch of new people to write the new code as well. So this certainly wouldn’t help with efficiency of new development projects.
Find a reason to get people excited:
Windows doesn’t really care about OSX; anything with less than a 5% market share doesn’t get much notice in war team planning meetings, I can tell you. Windows’ big competitors right now are Google and Open Source; Microsoft knows from experience that giving away something for free that people think is of value is a difficult proposition to fight, so that’s where the next battle is going to come from. And even then, is less to do with Windows features and functionality and more to do with total cost of ownership, which according to lots of studies Windows still wins in the enterprise space (i.e. it cost less over 3 years to buy, install and maintain Windows than it does to buy, install, and maintain an open-source alternative). So this means the big bet is going to be management products for enterprises (i.e. System Center tools) rather than worrying about the Vista feature set competing against OSX.
Finally, people don’t give damn about the OS, they care about features of that OS. Vista’s features are very important to a lot of people, but they represent change to the ‘known’ way of doing things (which people don’t like) and were poorly marketed as well - does anyone outside of MS really understand what User Access Control was trying to achieve or do they just see a bunch of stupid prompts that piss them off? Combined with an industry that sucked ass getting those products to work with hardware (yes, Toshiba, I am talking about your pathetic attempts at writing BIOS code for your “Vista-ready” laptops and tablet PCs you tossers!) and application developers who were way, way behind the power curve in adoption, and you’ve got applications / features / hardware that doesn’t work properly with Vista.
Stop designing by committee:
I quite agree with you here; Microsoft needs a new Bill Gates. Even when Gates stepped down from the CEO slot, he was still ‘the man’ who was pulling all the strings and drove the interoperability strategy between groups (i.e. Office and Windows, Enterprise and Windows). Now they’ve got different people running different divisions and Ballmer at the top and none of them have the same level of respect. Ray Ozzie was hired to do Gates’ job when he leaves for real but it ain’t happening yet and he doesn’t have the power to create cross-division working yet as there are too many independent fiefdoms with too much power.
Get rid of the crapware:
*This is technically not Microsoft’s fault, but it has the power to keep this stuff from happening in the future. - *
Actually, Microsoft doesn’t have this power. Microsoft sells an OEM license to a customer (Dell, HP, Sony, Toshiba, whoever) who then can do whatever they want to that code within the license and can plug whatever bloatware they want on top. On my Lenovo thinkpad, everything from login to home screen is branded Thinkpad - I don’t see a Windows logo unless I look for it. Windows can’t stop this bloatware because once it’s left the shop, people can do whatever they want with it and because Microsoft is required by law to publish APIs that allow software vendors to write code that uses things like systray icons. You have a problem with too many systray icons? Talk to Sony, talk to Real, talk to IBM, talk to QuickTime. Don’t bitch to Windows because there is nothing Windows can do about it.
Get a handle on viruses:
Oh fer f’s sake… if people manage their PC correctly, viruses are much much less of a problem. Microsoft publishes, for free, patches once a month. Microsoft releases, for free, the Malicious Software Removal Tool, on the Microsoft web site. Microsoft published a firewall product which is given away free with XP and Vista.
Want to be fully secured against 99.9% of the threats out there? Then do this:
- Turn on Windows Firewall.
- Turn on automatic updates and install all Microsoft patches that it downloads for you
- Download and install Forefront or install another AV client.
- Do not run any attachment you get in email or click-through any links you get in email
- Turn on maximum spam filter in Hotmail (or other free email clients) and / or install a spam filter with your AV client
Guess what? You won’t get viruses anymore. Or in other words, **RTFM! **
Figure out a niche for the switch to cloud computing:
Yeah, that niche, for now, is anyone connected to a LAN. In other words, businesses. They are tearing off salespeople’s hands for this. The problem is, most of them don’t know how to do IT right, or securely, so they screw it up then blame Microsoft for their failures.
Cloud isn’t for everyone, just like Systems Management Server isn’t for everyone - sometimes legal requirements (i.e. Data Protection Act / HIPAA) will prevent it’s use in certain occasions. But don’t forget as well - there are two ‘clouds’ where the data can reside; inside of a firewall (as in a cloud that is your house’s data, or your company’s data) and then outside of a firewall (i.e. everything else - a cloud of the whole internet). Restricting access and authorizing approved access and auditing that access is merely setting the right security for access to the right cloud, it doesn’t mean opening the kimono for everyone to see everything (like that would ever be something Microsoft, of all organizations, would encourage)
Figure out some kind of mobile strategy.
The iPhone struggled because it was limited by the provider, exactly the same as any other device. You have a problem with your mobile device, talk to your mobile provider, don’t try to blame the phone manufacturer.
But nothing, and I mean NOTHING, the iPhone does hasn’t been done better, faster, and earlier on a Windows Mobile device. Microsoft owns the smartphone landscape, or at the very most shares it with Blackberry, and is actually a better choice for businesses than Blackberry because the Windows-based smartphone integrates cheaper (i.e. for free) with existing Exchange messaging systems and allows providers more control over the handsets issued. The strategy is there, and the iPhone is at best a johnny-come-lately vanity device which lacks major functionality (3g, native integration with Exchange, mobile certificate services, secure web browsing) that business users (and especially business IT managers) require.