This is an old story; shades of every high-end vedor that's ever been wiped out by a low-cost player. As a hardware guy, it reminds me of the response you'd hear from POWER, PA-RISC, Alpha, SPARC, etc. vendors when you mentioned x86 workstations.
At the time, my response was, sure, maybe you've got a better product now, but what fundamental advantage do you have that keeps Intel from using their large, and growing, userbase and economies of scale to crush your R&D spending and pull ahead?
Software doesn't have such high fixed costs, and MS has a ton of money to spend on R&D. But, this still sounds like a lame defense.
This might be more contentious, but I believe that people that are still using Windows Server (and Microsoft offerings other than AD) are doing so because it's easier to use/more familiar than Linux/open-source alternatives: Throw Joe-Linux on Acme-Cloud-Provider and use Weekly-Flavor-Config-Management? Sounds hard. Click some buttons and fudge your way through group policy? Now we're talking something I can do to pay off a mortgage.
Because it's so easy, there are a factor more Microsoft admins around than there are Linux admins. And they're fairly productive.
Plus, Microsoft software works really well together. Windows + Windows Server + SQL Server. Works great.
Linux is great if you have highly skilled technicians to support it. There's way more tweaking involved than with Windows; this makes it ideal for anyone running a centralised model where they're deploying thousands of identical machines (like the web). It's not so great for corporate environments where there's a high level of variance..
It's basically the Ford vs Ferrari argument: Microsoft enables companies to employ average quality labour to produce a "good enough" result. Ferrari employ top engineers to produce an extremely high quality result.
Ford scales, Ferrari doesn't.
Turns out after showing them they liked it and they are currently running the Linux servers. Windows machines are used throughout the company for the regular users, but Linux for their servers storing all the patient information files, user data, etc. After spending about 3 months as an intern I essentially taught the Microsoft certified staff enough to manage the servers well enough (it's really not all that hard) and they still have them running (roughly a year later).
Point is, you really don't need highly skilled technicians, none of them had used Linux before my internship, yet all of them still seem to run the servers fine. One REALLY nice thing is that their Linux servers go down far less, are quicker, and are actually more secure than their old Windows 2000 servers they were running.
Recently, I had to set up a development environment for ASP.Net development and was taken aback by the sheer amount of things I needed to install and knobs I had to turn to get it working. I've never had so much trouble on Linux nor OS X.
The same is true for getting a working build environment for our main product (CPython-based). I find Windows to be quite hostile to non-VS environments.
It's time to retire ye olde Linux-is-complex argument. It really does not hold up to scrutiny. I say this as someone that uses Windows 5/7 days and OS X 7/7 days, so I'm at least a little bit informed about the OSes.
It doesn't compare as well to Postgres - but the better tooling may well make it worth the $10k in licence fees. That's a month of engineer time in maintenance.
SQL Server's cheap for what you get when compared to the commercial competition. Informix is pretty much dead (and had horrible tooling). Oracle is even more expensive. And compared to Oracle, Microsoft look like angels. Ingres is a bit of a joke.
The big expensive MSSQL licences also include some pretty nifty BI tools, including decent implementations of data mining algorithms.
In server markets Linux + Unix have roughly 65% market penetration and it's growing.
Windows Server has 35% and it's shrinking.
I use Server 2008 R2 (and upwards) at work and its really not bad, it's miles ahead of the previous versions which were horrible to use. Is it not possible that people use this stuff because its useful?
And my point was perhaps the reason people are using Windows server is not because Linux is too hard and unfamiliar but because just maybe it's a better fit.
If everything in the MS ecosystem "just works" then I am either willfully overlooking its great features or like so much else, it's a function of what you know best.
There's little to gain by switching to another OS if you have enough money for licenses and you know its ecosystem well.
Microsoft gives you freebie tools that work most of the time.
Also, because Postini was in the death throes for some time before Google Vault was released, many bigger enterprises went out and bought third party archive/ediscovery solutions.
Google is maturing as an enterprise vendor, and the simplicity of the licensing model and lack of client side software is appealing to the "new economy" companies: small businesses, businesses using Macs/BYOD and businesses using lots of freelancers/temps.
For the traditional enterprise shops, Google can be a real pain. You need to buy 3rd party stuff for things like identity federation, for example.
At the end of the day, it's called competition, and it's great. Instead of Ballmer & Co. coming down off the mountain and letting us know what we're going to do for the next 5 years, we have a real marketplace and can say no.
They make almost all of their still-quite-impressive profit from the markets that they still completely dominate, like enterprise productivity software. They are probably losing money on products like Windows Phone and Surface.
So what happens when Microsoft has credible competition in every market it sells to? That will soon be the case (if for no other reason than the fact that Google is actively trying to kill them), and I'm not sure even Microsoft is confident that Microsoft can compete and stay profitable in that environment. I see a lot of belt-tightening in their future.
I don't really think a company can get--or stay--as big as Microsoft is, without the rent-seeking ability afforded by near-total monopoly of a major market.
You can ask the same question of Google. Enterprise with Google is a grown up hobby project that isn't a big earnings generator. Will google apps stay cheap if googles golden goose (advertising) declines?
IBM just laid off over 3,000 people last month. To quote pg, "Consulting is where product companies go to die. IBM is the most famous example." http://paulgraham.com/startupfunding.html
Will google apps stay cheap if googles golden goose (advertising) declines?
Nothing lasts forever. No matter how much we want to think that major corporations are permanent institutions, they have a lifecycle, and when they lose control of their market they may continue along for quite a while in a zombie-like state, but eventually they will wither away and die. Someone outside of Google will eventually figure out a better way to do online advertising, and Google will find itself in the position that Microsoft is in today.
Microsoft is trying really hard to get a new lease on life through a product pivot, but I'm just not sure it's even possible at this point. Their products are generally more expensive and not convincingly better than the competition's.
Stock chart doesn't suggest imminent death either.
The larger companies get, it just gets harder to maintain that level of quality, and the consequences show up in the output. A Google at 50K employees looks very similar to a MS or an IBM when they were at that staffing level.
Look at Intel they have been stuck at 100k for decades.
The only exception I can think of is Samsung but this is really over a much smaller time frame so doubt it will sustain.
Same goes for MSFT vs Linux or GOOG, etc. There's nothing new here.
For example, Excel vs. Google docs: you can do basic charts and graphs in Google docs, but clearly Excel has many more features, such as data analysis tools, and when you connect Excel to Microsoft SQL server you have reporting + data mart, where Google Apps offers nothing similar (not without a lot development work needed to set it up)
Google Apps is good but often you still find businesses also need Office suite. If Google were to say, directly connect Google docs to a data warehouse tool such as Dremel, the add a lot of advanced data analysis features to Google docs, then things would start to get interesting.
At all the companies where I've worked at, I've never seen anyone use anything more advanced than a few basic functions.
The _vast_ majority of users in a corporate environment use:
1> Font changes (color, size etc)
3> Hyperlinks, ToC and other forms or document organization.
It is silly to buy 1000 Desktop or Office 365 Excel licenses, when only 20 people will use a substantial amount of its features.
More importantly, people are not emailing each other Office documents. I've seen companies who do _have_ google docs do it.
If I wasn't being emailed Office documents on a daily basis I'd assume my colleagues and clients didn't like me any more!
Either way, Microsoft will see their unit margins drop. This is good business strategy for Google given their rivalry.
The inbox size vastly increased (500MB => 30GB) which makes the record keeping much easier, just never delete anything. Before we had to empty our mail into offline archives.
Everyone still uses office for word processing and spreadsheets. The online storage is nice for sharing files though.
We do pretty well with a more or less entirely Google Apps workflow.
And even if they do - where to go? A lot of non it companies don't have even real ops ... hosting your own email and servers is expensive in that case. And potentially dangerous a modest string of bad luck could wipe you.
And precisely because a company like Microsoft or Google would do this, you need to fear them, because they would do this with other governments, too, including the Romanian one. It's not just NSA you have to fear. If the laws allow it, the Romanian government could get all your secret documents from Skydrive or Gmail, just as easily.
Did you read about the Cryptocat security bug? We learned about it because it was open source. Maybe NSA found out about it even earlier, but at least we did, too, and fixed it. Now imagine the same type of bug was in Skype. Then NSA would be the only ones who know about it.
If you are able to decipher such backdoor in the code, or do you expect the code to have a comment saying /* backdoor */ ?
Microsoft's position in enterprise is extremely strong. Apple is not making any serious attempt to challenge that (and probably never will), but Google (and Linux) pose a real threat to Microsoft in the long run.
Assuming the offering is just the same, but for many it may be good enough