From word processing to servers to compilers, no tech company eats their own dog food like Microsoft. They even have their own search engine.
It's an under appreciated feature of their corporate culture that first hit me when I asked myself, "Why the hell does Microsoft sell project?" and realized that just about everything they have developed since DOS has been something they could use in house.
True, for the most part, but not a complete story.
One thing that MS never does is eat its own pricing or licensing dog food, which is a big part of the downside of using MS products. For example, an individual developer at MS never thinks for a second about the cost/benefit of various different OSes or servers, if they feel like installing Windows Server 2008 r2 data center edition with SQL Server on their desktop (which is around $10k in total cost to some schlub outside the company) or on an old machine they've converted into a server, they do so.
This can create a fairly substantial difference between the experience within MS and without. Also, there are tons of internal tools used throughout the development process which don't exist outside of MS or for which MS doesn't consider using the equivalent 3rd party tool which result in very different experiences. A lot of times this is justified as servicing workflows which are only relevant at MS's scale but often times this is a pretty bogus rationalization.
And our UX folks routinely use both Photoshop and Illustrator. It's almost like we can't make sweeping statements of "how it is" at a company with 90k or so employees.
Without a site license for Adobe CS, it makes sense that some orgs wouldn't have that in their budget (its like $1000 per seat!), while UX teams would definitely budget for that. The problem in essence is "design by PM," which I'm sure is not unique to Microsoft at all.
Not true, with a $1200 ($800 renewal) msdn subscription "Visual Studio Professional with MSDN" you should have access to 5 copies of every operating system and server software + professional versions of all Visual studio packages.
I'm not sure why anyone would think MS uses only their own OS for everything. Microsoft has been known to use other OS's in the past (the website used to run on Solaris right?), and we do have designers with macs (beyond just the team that does MacOffice, of course). Eat the dog food but don't drink the kool aid, anyways.
DreamSpark is again, for non-commercial use. You're probably referring to Bizspark / Webspark.
And it's far from ideal to go this route.
To get accepted into the program, you first have to start a company. As a lone developer that just wants to try stuff out and see what sticks, dealing with such bureaucracy is far from ideal - of course, in the US you can start a company probably in a couple of hours, but many in the rest of the world are not so lucky.
Then you need to apply by describing what you intend to do with it. If you have an idea for a project you have to describe it in detail, including the business plan, otherwise you can get rejected. And people have been rejected.
And then, assuming you did the above and got accepted, the *Spark programs are a ticking time bomb. When it's over, it's over. You can keep whatever software you have on your desktop already installed, but then you do have to license whatever software you're using in production (SQL Server, Windows Server).
Again, I'm mainly speaking from the point of view of someone that has a day job and that would like to try stuff out. Skipping over the fact that if you tell them you have a day job, you probably won't get accepted, it can take a lot of time to start earning money from what you're doing and thus afford those licenses at the end of the program. And it's not like successful businesses haven't been built this way.
So I just don't get why would somebody be willingly getting into this, when many companies and individuals succeeded by using the free and open-source alternatives, which personally I find technically superior.
This is an amusing digression but it doesn't change the point. At any other company the size of Microsoft the cost of installing a high-end Windows OS or a server runs into the many thousands of dollars and would normally be a very convoluted process to install. Within MS it takes all of a matter of seconds, everyone there is protected from one of the biggest pain points of using MS software, which I think is a pretty significant problem.
If this is really a problem, it applies to other companies commonly praised on HN, such as 37Signals. I suspect that the alternative of charging the street price of software off against internal budgets would be a more significant problem and garner pecks of criticism on HN.
>no tech company eats their own dog food like Microsoft
See, I've always assumed the exact opposite based on how rough the user experience is on so many Microsoft products.
I've often wondered if it was possible for a web developer at MS to sit down with IE for 30 minutes and not think "I have to fix this.... now!" When I realized how awesome Chrome developer tools were I never wanted to go back. But of course, as a web developer, you have to go back.
Or maybe that is the problem. Maybe too much dogfooding blinds you to how great the competitor is.
Just because something is used internally doesn't mean the product manager is going to prioritize fixes for internal bug reports. Testing internally does not magically add more programmers to the project.
Sure. My first instinct is to say that's a small concern, but they do have 100k employees.
Sorry, I was introduced to "dogfooding" in Dreaming in Code. I believe there was even a chapter named after it and it was a pretty big deal about how you dogfood to maintain quality. So when I hear the term that's what I assume is implied.
While true, "the PM may not prioritize fixes", a big aspect you are missing is, someone internally with a problem has access to a direct communication line with said PM, his peers, superiors, etc.
For most day to day usage issues, dogfooding is a great way to hammer out issues. Will they find issues with every use case? No, but they will find some. Also, generally, the more diverse and large the organization, the more use cases will be dogfooded (assuming the bulk of the org is encouraged/required to do so).
When you are talking a company of 90k+, encouraged dogfooding will hit a lot of different use cases.
The article states that company money cannot be used to purchase Apple products, not that Apple products are banned entirely. I know plenty of people that use iPhones on campus.
In some orgs it is actually relatively easy for a PM to get a MacBook Air, for instance. You don't even need a bulletproof justification. Although with the new crop of ultrabooks, the MacBook Air is no longer so unique, so it may be more difficult.
The new IE is actually pretty good. Microsoft knows that IE6 is an abomination and even made a website dedicated to its demise: http://www.ie6countdown.com/
My point regarding 9 is that the developer tools are buggy and painful to use compared to chrome. As a web developer, you would want them fixed as soon as possible.
MS has tons of development tools available internally that they don't release to the world. It's possible that what you're wishing you had is sitting out on Toolbox, invisible to anyone outside the firewall.
There's something of a culture of "our internal developers are smarter than our external developers" at Microsoft, so dev tools are often where the dogfooding has been the weakest. It comes from having internal developers who are hacking on things like OS kernels while the external developers have a massive population of coders churning out enterprise CRUD forms (along with a smaller population of more advanced developers), but this internal/external tool gap sometimes really bites them.
Perhaps I'm misinterpreting. But I got the impression that dev tools that qualify to be sent out into the world need to have a primary emphasis on friendliness, automation and autogeneration, and general polish. By contrast, internal tools need to be industrial grade workhorses, and if you can improve them by putting a better interface on them, that's merely nice. One of the clearest examples I can think of is the difference between their internal workhorse Source Depot and their public tool TFS (nevermind the horror that was VSS).
Like I said, I might be misreading things. And certainly the blogging community around their dev tools has improved the level of detail that goes out to developers committed enough to keep up with them.
I just came from a .NET developer meeting hosted on the local Microsoft campus. The presenter was trying to hype development of new Office 360, Sharepoint and other "code less" applications. Made the mistake of telling a room full of experienced developers they could be replaced with HTML jockeys at 1/6 their price. I decided to spend the rest of the presentation working on some stored procedures.
That's because the tooling does well enough versus your $50k dev project.
I'm not a fan of LightSwitch myself, but this is the right step forward for business software and fits in with the Office VBA ecosystem well (i.e. customise it yourself).
Sure, that's not what they/I mean though. They made something to register hours for their employees (perfect case for Access/Lightswitch) and asked us for a quote for a complex public facing SAAS solution. And the 'case' they want to make is to ask why there is so much programming involved even EVEN HE can make 'a complete application' in a few days.
Lightswitch and other systems like it have their domains and work well in those domains, but less tech clients cannot see the difference between one domain and the other. A web project we did in another 'codeless' tool (which is just not true as you are coding, just in another way) was delivered and the client said; we saw on the tool vendor's site that you just press one button to make the web site into mobile apps for all devices. So why are you charging us for that? In short; people believe what vendors say on the golf course or on their sites. Even if those cases are completely out of context; people don't KNOW that they are.
If a CEO of tool-vendor X tells CEO Y on the golf course that their product 'fixes all IT problems because anyone can make solutions for web and mobile with it', then CEO Y will tell buy this product. Although neither CEO X or Y know actually what the product does. This is really normal. So then the product gets pushed down to the production managers and they get the 'command' to 'use it for everything, because we are shifting to this product as it will save us millions in IT costs'.
I have seen this many times (in big and small companies) and I see it with LightSwitch as well; with Sharepoint too; I know a few companies who make everything with SP, much to the dismay of the IT dep.
Edit: and I don't really know if it's the right way for businesses to go. I remember Access well; I remember institutions, for instance the main Dutch court, creating MONSTERS in that themselves because it's easy and (almost) no programming. Then when the guy/girl who did this leaves, someone needs to take it over. We are talking about a complete mess which is business critical. So I believe it has it's place, but only for small departmental apps; if they grow they should be replaced by something solid and this should be considered the prototype. IMHO.
1. They should just use something like replicon ( http://www.replicon.com/ ) for that. Don't build if something already exists.
2. Lightswitch solutions are not designed to be sold by consultancies of external developers. They are designed to be in house built and maintained as per VBA apps.
Access is still doing fine for a number of cases. I still have to maintain an Access 7.0 database for inventory tracking/management on NT4 which I built in 1997. It still does the job so there is no reason to replace it. I will be early next year moving it to Access 2010 on Windows 7 as their hardware is now dying miserably and that is the only motivation to upgrade it. Note: this is running on SQL6.5 with 50 users.
The part where people internally write horrible software on these 'code less' platforms? Company critical software which then somehow needs to be maintained by some poor programmer?
I just tell what I see in real life and that's not pretty.
Maybe your Access db wasn't such a mess. If you have over 50.000 lines of VBA written by people who never programmed but learned it organically by changing the application, then i'm not sure if it shouldn't be replaced. Any change to that system is a huge danger.
I never worked at MS and know no-one who does, so I don't know this for a fact, but I read they don't use Sourcesafe internally. I have used it and hated every second of it, so I can imagine, however, it is, literally, their dog food.
The last time I had a close friend at microsoft (and so heard the scuttlebutt) was a year or so ago, and at least from that perspective, it sounded like everybody there had an iphone -- and used google to search for stuff...
In other words, having a product, and actually getting your employees to use it when it isn't strictly necessary (e.g., they are gonna use MS compilers to compile stuff because it's pretty much necessary), aren't the same thing.
[... and a significant amount of MS dev work was done on iphone projects (e.g. in bing)!]
Like others have already pointed out, the experience of one person in a 100k+ employee company is not broadly applicable. Many employees have iPhones, yes, but many also have Windows Phones as well. I agree with you that since there is really no mandate to use MS products, it is a good sign if employees do use them, since it can signal that they truly are better than the competition.
>From word processing to servers to compilers, no tech company eats their own dog food like Microsoft. They even have their own search engine.
Wouldn't Apple be in the same position ? (except for the search engine) They have the Hardware, the OS, the compiler, their own office suite, mail program, calendar and server versions etc. the phone hardware, the phone OS.
It's a small gesture when you look at the big picture (total compensation, etc), but this means a lot to me as an employee, and really shows Microsoft's commitment to the new era of Windows 8 on beautiful devices.
It will be interesting to see how the Windows Phone 8 devices are distributed, as the previous deal for 7 was a free device only if you sign up for a 2-year contract.
Unfortunately I just moved teams, so I was in the process of ordering a new machine anyway; but still, a new Surface tablet will be awesome.
Response to snark (which I don't know why you think was obligatory):
PC World reports 5.4 million Windows Phone 7s sold this quarter alone (http://www.pcworld.com/article/262287/windows_phone_fails_to...). With 94,000 employees, this would increase Windows Phone sales by 1.7% in a single quarter, and be a much smaller percentage of installed base.
Do you believe that was a misleading statistic to use to give a general idea of the scale of the inaccuracy of the original comment? Sorry I'm being short in all these replies, I'm tired of and choosing to call people out on unconstructive snark and pedantry.
Did he announce that all employees are getting WP8 devices or just American employees (~45k)? Given that subsidies are not common outside of the US, I'm thinking only 1%. We only just recently got our WP7 devices here in China (I work for Microsoft China).
I still have not seen a Windows Phone device in the wild; I hope our market share is much better in the states!
It will take them awhile to figure this out. I just meant to say that when they announce these, it might not affect 90K worldwide employees right away, its more like the 40K in USA/Redmond + figure out how to handle in each other country based on unique circumstances (availability, taxes, etc...).
There's a lot to this, for the company, other than employees getting bonuses in hardware instead of cash. They're putting MS products in the hands of tens of thousands of people most invested in evangelizing and evaluating them - every MS software product benefits from dogfooding, why not hardware? It will presumably give MS a chance to debug enterprise bring-your-own-device that they are trying to support by implementing it on an enormous scale themselves. These are developers who may be now inclined to write apps. And most MS employees don't have laptops, so now they won't be stuck with pen and paper in meetings.
This mass upgrade will have the negative side-effect of moving all 94,000 employees to the same spot in their three-year upgrade cycle. If this proves true, it's interesting, headline-grabbing, and ultimately short-sighted, as it will result in a great deal of hurt feelings in the long run as the replacement cycle is stretched out again (for logistics and accounting purposes, if nothing else.)
Oh, and the many, many contractors look like they'll be hosed. Again.
I wonder how come MS is able to afford giving free hardware to so many of their employees. Earlier they gave the Windows Phone and now the Tablet.
Is it a good marketing strategy. Do non-MS people buy their products, based on word-of-mouth spread by MS employees. Does this strategy really work ???
It's definitely a win-win scenario (maybe win-win-win):
1. Every employee will implicitly/explicitly become an evangelist for Microsoft products. I'm not talking only about the Windows 8 and Windows Phone 8 but all the services and apps that support the platforms and ecosystem. The marketing power and reach won't be as wide as Microsoft's actual marketing effort (there is a reason why Microsoft has a whole division of marketing), but this will certainly do a good job of sparking the excitement for the community.
2. Employee morale will definitely go up. Employees will feel that they are being cared. Employees will feel that they are getting treated well. This would essentially lower the retention rate of the company. If less people are leaving the company because they are unhappy, the less money Microsoft has to throw out to replace those people.
3. (maybe) People outside of Microsoft would start to think that Microsoft is a good place to be. People would start thinking that working for Microsoft is not a bad idea, after all. This kind of decision shows the cult of the company. I think Microsoft has a lot of wrong perceptions. They might have been correct a few years ago, but things have started to change since Windows 7. Then, I keep hearing more changes and more good stuff from Microsoft lately. Definitely, this is going to help making the Microsoft brand better.
Well i believe; that for a Software company like MS; it should give software for free to its employees instead of hardware - if it really wants to show that it cares. Most employees will belong the software field and will most definitely feel cared; if software is given to them for free - inorder for them to explore their own development/testing potential along with exploring the product potential. Also by giving the software for free; when an employee leaves the company; due to his/her interest in what the software product can deliver; they become lifelong customers - buy product upgrades and related software.
as far as attrition goes; well I guess it is bound to happen in case of experienced employees, no matter how great the company is; 'cause as one moves up the ladder; number of higher positions(across various teams) start to decrease. Talented employees hungry to explore their capabilities will leave for sure due to lack of opportunity.
As far as people outside MS wanting to join - guess will automatically happen if MS meets their requirements(job satisfaction + money + environment). The case most generally is that a company will find a replacement for a vacant position on its terms(there may be exceptions). But, It need not give hardware for free to its employees - inorder to send a message to the outside world - but rather give Software for free. Software given free to its employees means a opportunity to grow; and guess that would definitely attract outsiders to explore.
As far as I know, every employee gets a MSDN subscription, which comes with tons of Microsoft software for FREE. Plus, they dogfood their own products. They dogfood Windows, Office, Hotmail (or Outlook.com), SkyDrive, etc, and dogfooding is "free". If you hadn't known, now you know.
When I talked about the retention rate, I was focused on employees who've been with the company for 3 yrs or less. In tech industry, especially in software engineering field, people switch jobs fast. really fast. All my colleagues switch jobs every 2-3 years. For companies like Microsoft, Google, and Apple, this is not good. They now have to spend money to find potential candidates, spend more money to interview those potential candidates, make full-time employees to interview potential candidates (5-8 interviewer per candidate), etc.
Its probably more about morale than marketing. Getting employees to experience and enjoy the company's products is very important. Also, they might build more apps for WP and Win8 in their spare time. But 90K people...isn't really enough to make a huge impact, perhaps the ~40K in Redmond/Bellevue, but the effect would dilute quickly as you approached downtown Seattle from the eastside.
I totally agree with employees being allowed to enjoy and explore company products; and that such experiences should be absolutely free for all employees; but wonder if this experience should be limited within the company premises.
btw, does the hardware belong to the employee or is a property of MS. If the hardware is property of the employee, how does that make sense for MS, when employees switch jobs for their betterment (job + monetary reasons) .
AFAIK, MS at Redmond does not give Windows OS(software) for free; but gives it at a discount to its employees. Why not use the same strategy with hardware – give it with some discount – say 50% discount.
Win Apps - Do you really need to give hardware to build Apps for WP and Win8 or software alone would do ???
You need hardware to test and also motivate your work. You wouldn't build an iPad app without an iPad neh?
Hardware discounts are hard when you aren't producing the hardware. You are basically subsidizing cost, and...you could imagine how much trouble that is! Better to just give it away; besides, if it was Microsoft campus store only, it would be only good for the folks in Redmond.
Attrition happens, but honestly losing an employee costs more than a phone and a tablet!
well by "software alone" - what i implicitly meant was that most employees would already have some hardware – latptop or desktop – in which case software given for free would motivate and encourage them to develop their ideas and even more apps.
If one has some hardware machine then one can install several OS’s in Virtual Machines for development and testing. one need not have separate hardware to develop apps for different OS’s.
If MS has to give hardware for free for whatever reasons, wouldn't it be a great idea to give the hardware to customers (non-MS) for free. This way if the product is really great; it will cause some buzz in the market and help the company market their product as well.
I just wonder if it’s really worth giving free hardware; instead of software. Once a software is given; and an employee discovers the potential in it; no matter whether the employee remains with the company(MS) or not; the person will become a lifelong investment and asset for the company - as he may/will buy product upgrades and related software – which may not be the case with hardware .
I would never use my company provided laptop for non-work (moonlighting) purposes. All of my personal computers are macs running OS X; I'm afraid of buying a PC in my market (China), while the local Apple Store in contrast is a very dependable and safe (Apple is also better about servicing hardware bought out of country). If I was back in the states, I would feel better about buying a personal PC.
The problem with giving free hardware to customers is that you can't really identify who would add value with it, while employees are a better (but probably still not very good) heuristic.
You will always need to test your app on a real device it is intended for; virtual machines and simulators don't really account for the real device experience. For mobile/tablet apps, this means you'll need a decent device in that category.
Hint: Users didn't ask for a desktop OS that looks and works more like their phones... or rather, like the phones that Ballmer hopes to sell them at a later date.
Fair enough. Let's say their cost for a phone, computer, and tablet is $2000. With 94,000 employees, that's 188 $M. I suppose that's pretty trivial when your market cap is 259 $B.
More like $200 for a tablet, and $400-$600 for a phone. Those are retail prices, and I'm sure MS gets a pretty good discount on those too. You tend to get things a lot cheaper if you buy a lot of them.
It's an under appreciated feature of their corporate culture that first hit me when I asked myself, "Why the hell does Microsoft sell project?" and realized that just about everything they have developed since DOS has been something they could use in house.