Tuesday, August 01, 2006

10,000 More Microsofties - What Do They Do?

So there I was, languishing sideways in a sun-drenched chair, indulging in bon bons, and getting engrossed in another summer-time read when Todd Bishop and I had a quick conversation.

"...Microsoft increased by 10,000 in FY06," the ever vigilant Mr. Bishop shared with me, keeping an eye on the recent postings to Microsoft's PressPass site.

Goodness gracious!

A fine G-rated reaction. And thank goodness I didn't have a bon bon in my mouth right then.

This was a heck of a kick-off to the 2006 Financial Analysts Meeting. Who are these people we've hired around the world? Sales and marketing? The field? MSN? Perhaps I'm reading too much summer-time dreck, but it makes me wonder if we're hiring all the folks we can to keep them away from the competition, tucking them away in pods along the The Dalles and other international locales. I shared my opinion with Mr. Bishop: they certainly are not developers, unless we've really lowered the bar and have decided it's okay to hire people who don't need to know how the JIT works, let alone those memory pointer thingies. You can spell HTML? Hired!

Unless someone has a secret to share on how to hire lots and lots of super-talented developers.

Redmond employee count grew by what... holy crap, almost four-thousand people! And our total global growth, including Redmond, was originally projected at 5,000? No wonder we have streets that turn into parking lots around campus. You know, you can only put so many rats in a box before the stress gets to them and it gets toxic and ugly. And the Redmond box is pretty much full. The infrastructure can't take much more, especially if you consider the ancillary jobs that open up based on the increased Microsoftie jobs.

I can only hope that FY07 will not be another 16% growth year. But I'm one beaten down individual, as if each one of these new folks have walked right over me as I sputter, "Mini- Microsoft- Lean- and- Mean- ooch!"

As for the FAM... lots of good coverage all over. I know, we didn't give anyone confidence about Vista's slip ship date (including a report that the press groaned to the answer about when Vista would ship... like we stuck our stock's chin out for a prime right hook). Personally, I would have loved some theater. Like BrianV being onstage giving that initial answer, "...we slip so that we don't ship shit!" and then being fired (or sent to MBS or something) and Jon DeVaan being yanked out of the audience and asked the same question, "We're shipping on time! I promise!" Now that would be refreshing.

Well of course that's not how it's shaping up in the near term. Some folks are grumbling about Sinofsky cronyism as Windows reorganizes for the future. I'm excited for any change that seems to be busting up the levels of hierarchy, clearing out the old, and bringing in the new. Some folks aren't going to be happy with any changes. But I think it's promising. What would you do differently? One positive outlook:

JonDe is a microsoft hero. He and EE team will make windows a tiger again.

It makes me wonder if the folks being given the bums rush are looking for jobs elsewhere or if this SPSA grant coming up very soon is having the positive benefit of "I finally got mine and I am out of here!":

My office is currently being swamped with senior microsoft partners looking to explore options, to consider employment with my firm.

What I am trying to get a handle on is if this is the first wave of Vista related attrition, or is this people planning on moving on after the first slug of the big $1m payday, or is this the first wave of the rif?

I am just curious. It seems odd that we would see so much activity now.

(Lord, it's me... Mini. Please, please let this be the truth and not some sadistic soul playing me for the fool. It's been a rough year, Lord, and I need something hopeful like this to make it through!)

Well, if your office is Google then I hope folks realize they might have a rush to beat... let's see, what's that URL from today's PI? Here we go: http://www.google.com/aspirejennifer - time's a-wasting!

Other goings on...

(1) FAM demo blowout: I was in too much shock from the employment growth to complain about this. In the meantime, both Larry Osterman and Rob Chambers have owned up to what went wrong and why. That's pretty damn open and honest. Some commenters wrote conspiracy theories around it being a staged demo, but somehow I think if it was staged it, ah, wouldn't have crapped out.

(2) Microsoft Lays Out Plans at Analyst Day has the following encouraging note:

It's only a matter of time before Wall Street takes notice of the strides that Microsoft has made of late to transform itself from a slow-moving software behemoth into a company that can better take advantage of new trends in the marketplace. Microsoft's renewed focus on online services is a good move for the company. Meanwhile, the company has taken full advantage of the XBox 360's head start on the other next generation consoles, and it has been doing a good job of leveraging the system's popularity to generate new revenue streams. With all of Microsoft's initiatives set to come together next year, fiscal year 2008 should see huge profits for the firm.

Well, it was encouraging until that FY08 part... better than FY-never.

(3) Ballmer Analyzes Microsoft's 'One Big' Vista Mistake - Mr. Ballmer continues the long mea culpa around integrated innovation. Additionally, Mr. Ballmer is attending his own kind of listening tour around campus right now with Ray Ozzie, meeting with L65+ employees, group by group, and having frank conversations. Hmm!

(4) Watch out, Google, Microsoft now gets the Net - interesting that folks are writing about the Old Microsoft and the New Microsoft.

(5) MSFTextrememakeover FAM - Ante up, up the ante, or just plain bluffing - Microsoft Extreme Makeover's summarization on the FAM - very nicely done.

(6) Charles DiBona and Dylan Yolles shared some interesting insights at the post-FAM Breakfast Series event. They saw it as extremely positive that Microsoft responded with a buy-back. Next request (same as last year): a consistent significant dividend. That would increase all sorts of investment in the stock.

Why is such a dividend a bad idea?

For Dylan: Linux, not so much of a worry on the desktop given that the TCO isn't terribly different (upfront cost is negligible compared to TCO). OpenOffice: beginning to worry more about. Interesting that he brought up a worry over developers and how he sees developers gravitating towards an Eclipse / Open Source environment. What do we do to get developers excited and engaged?

For Charlie: well, he can make us feel better for suffering through a Series of Unfortunate Events: the love affair with Google, the Vista delay screwing over those looking for a Vista dividend (and they in turn screwing us over), and the unknown payoff of the unexpected $2,000,000,000 investment. He saw big changes over the past three months: communication being open, being receptive to feedback, and to act on that feedback. Kudos to those change-makers.

And the reward for, well, the ballsiest question during the Q&A goes to the gentleman who asked about the prospects of taking Microsoft private. That question resulted in Dylan sharing an interesting nugget from five years ago: Bill Gates told Dylan that if the stock got low enough Bill Gates would take the company private.

How much lower is low enough?

Updated: fixed a couple of typos, especially one where I said "slip" when I truly meant "ship." Oy.


209 comments:

1 – 200 of 209   Newer›   Newest»
Anonymous said...

(1) FAM demo blowout: I was in too much shock...

Why do you think this is a surprise? Microsoft's speech team lost all of its stars in the past years. I dont't think its speech founder XD Huang is still with the team. Kai-Fu Lee also left MS. Without leaders like XD or Kai-Fu, the current team's collective IQ is probably 30% of its competitors. Do you expect there will be a great demo?

Anonymous said...

Today's the day I decided to leave Windows

I knew it was coming for some time, but I still can't get over KJ ousting BrianV. Brian was one of the few people who had the guts to call a spade a spade and the ability to come clean when he made a mistake. I think his only fault was having too many process monkeys (e.g. basics, feature love, etc.)

Brian has both led and personally contributed to the (recent) success of Vista. His mark can be seen everywhere.

This sucks, because I was really hoping that Windows would have a chance to improve, combining the org and process contributions from SteveSi with the old school engineering leadership of BrianV et al.

I was already planning to leave my group after review #s came out. Now I guess I need to find a different division altogether (and I don't even work under COSD).

-MiniD

Anonymous said...

The first comment is crazy. On the one side we say that 70% of the competitiveness was coming due to a leader on the other hand we complain against them.

As many proposes on this blog, suppose SteveB is fired (in fact, firing SteveB seems to be a bigger goal than minimizing Microsoft of this blog. Most comments are about that leaders are bozo at Microsoft)
Suppose if SteveB is fired. Then for an independent reason something at Microsoft fails. I would say, without leaders like SteveB, Microsoft collective IQ is probably 30%. I could still say that about BillG.

You know why that demo failed, because without the presence of a leader like BillG, probably our demos are in good mood only 30% of the time. How's about the following reason, you know the demo needed the sweet voice of a leader like BillG. Other, aah.., problems have with their voices.

The fact is that Kai-Fu behavior is questioned. If the questioneer is right then even if Kai-Fu was here he would be busy passing not only his IQ but others IQ to Google in the form of confidential documents.

Anonymous said...

Excuse me if I do not understand the area of finance. Can somebody please tell me what it means to go private?

What BillG meant when he told Dylan: Bill Gates told Dylan that if the stock got low enough Bill Gates would take the company private.

Does it mean a small group of person would buy stocks from all others?

I see two problems in it. If Microsoft stock gets low enough then so is BillG's wealth. How would then he take Microsoft private?

Another problem is that, suppose Microsoft stock gets low enough and somebody tries to take Microsoft private. Then he has to buy all the stocks from the market. Just starting this process will increase the stock price again and therefore the person would not be able to afford to take the company private.

Anonymous said...

It's pretty easy to sniff out what those 10,000 new microsofties do (and it sure ain't development!). Open headtrax, search BillG's org with filters (such as start date, standard title, etc) till you get a reasonably small subset - then export to Excel and use the filter views. My guess is less than 10% will be developers. (Anecdotally, it seems to be easier to hire PMs & test).

Anonymous said...

MSN increased headcount by 44% and now has a hiring freeze. Rumor has it that there will be some layoffs once vista ships.

A homless, jobless tech guy said...

"Excuse me if I do not understand the area of finance. Can somebody please tell me what it means to go private?"

Private? No public stock.

Well, take a look at Gateway stock at less than 1.60/share. Not much value in the total company there, undervalued stock. Should go private until it can get its strength and reputation back up.

MS is a long way from going private. For one. Mini is just blowing smoke up the rumor mill, among other things. The truth is Microsoft would have to get down under $5 bucks for such a consideration. The buy back is a better strategy. Not controlling all stock but the majority of it is the better than private.

I see the buy back as a temporary strategy to bolster the company while it is in a two or three year correction mode on strategy. I.e., Vista was strategized for a different market and economy of function than is the reality today.

Anonymous said...

JonDe and the Engineering Excellence crowd will be COSD x100. The Engineering Excellence group is a joke. It takes months for them to put together a simple presentation. It lives to recycle ideas from niche groups around the company as if they were the next wave of computing. Attend 2 "Engineering Excellence Days" 2 years in a row and you'll see what I mean: same people, same booths, same technology.

I fear for Windows.

Anonymous said...

To answer the question about taking a company private, there are a couple of ways to do it.

The most common way is to find some sort of investment/holding company with sufficient assets or the ability to raise sufficient assets to buy all the stock at a specific price.

Sometimes companies will run this process themselves, so that they aren't controlled by an outside group of owners.

Like mergers, this would typically entail an offer for the stock somewhat above the current market price, to enduce people to sell.

that homless tech guy said...

"Excuse me if I do not understand the area of finance. Can somebody please tell me what it means to go private?"

forgot to mention in my previous response to this that
a) public stock value is based on actual value of the stock driven by market conditions, analyst comments, economy conditions, etc.

b) Microsoft is in a position for the next five years they are not used to: i.e., they are not in control of all the things they are usually in control of. More market factors have entered in decreasing the value of the company and stock, mostly because they core, Windows has peaked and hit a 'life of product' downturn on its product life curve. So what this means is that for five years while putting all of MS's eggs in the Vista basket, the market has changed and has already indicated it wants a different product. MS will come out with a product (Vista) that will have a shorter period of time from which to recoup its huge development cost. This will hit hard in '08 and '09. Some value will be regained, but probably not nearly as much as intended in its initial development strategy.

A private company would give Microsoft enough time to reorganize and retool while being able to offer stock to desired incoming employees at in a more controlled manner. That would be the main driver to going private.

Anonymous said...

Interesting that the speech team still has a Microphone Setup Wizard. Most VOIP software these days seems to be able to dynamically change the microphone gain--conceptually it's not a hard thing to do.

Anonymous said...

"Does it mean a small group of person would buy stocks from all others?"

It means the company would take on debt to buy their shares on the open market and it would be a massively stupid idea. First, even with a premium over current, it would likely be grossly unfair to longterm shareholders who have underwritten the past 5 years of "investments" and massive comp/legal charges via unprecedented losses in marketcap (i.e. current mutual fund holders - many of which are newer owners with lower cost bases - would likely tender sufficient shares at some modest premium that individual LT shareholders would have little choice but to also sell). Indeed, the deal would only be feasible due to the pathetic job management has done in that regard over that period. Second, it would require the company to take on debt for the first time and even at these prices, a significant amount of it. Finally, what would be accomplished? Would a company that is barely accountable to Wall St or shareholders today, suddenly become more focused, efficient, agile, etc. now that they'd be completely unaccountable? Would executive insiders who believe so little in the company that they lead the entire market in insider selling, all of a sudden be willing to put off immediate gratification in return for a stake in the future? Would the current ridiculous compensation disparity between the top 20% and everyone else become less pronounced without external oversight? Puhlease. Current management should be embarassed that their performance over the past 5 years even makes this a potential discussion and imo, general employees should view this as a very stupid option vs some kind of panacea.

Anonymous said...

Some info on going private:
http://mercercapital.com/Publications/ELaw/archive/elaw0202.htm

Essentially, if enough of the shareholders decide to go private, then they simply pay off all shareholders at a given price. I don't believe they need permission from anyone else if they control a majority.

Generally, the price has to be very low, because the company has to pay a good premium above the asking price to avoid problems with the SEC.

Bill Gates's money is very diversified by now. He doesn't keep it all in MS stock.

Also, Bill could make use of the assets of the company to buy back the stock, and he wouldn't have to pay for it all himself. This is done all the time in leveraged takeovers.

Anonymous said...

Mildly offtopic, but there's an interesting article from Mort Meyerson, CEO of Perot Systems.

An excerpt:
The system worked; that is, we got exactly what we wanted. We asked people to put financial performance before everything else, and they did. They drove themselves to do whatever was necessary to create those results -- even if it meant too much personal sacrifice or doing things that weren't really in the best interests of customers. Sometimes they did things that produced positive financial results in the short term but weren't in the company's long term interest. That's a charge you'd usually apply to a CEO -- but I've never heard it said about individuals down to the lowest ranks of a company. Yet my pay-for-performance approach effectively encouraged that behavior from all of our people.

When I came to Perot Systems, what I saw in my six months of listening inside the company convinced me that we were about to make the same mistake. The emphasis on profit-and-loss to the exclusion of other values was creating a culture of destructive contention. We were about 1,500 people, with revenues of roughly $170 million. Our people were committed to growing the company -- but we risked becoming a company where the best people in the industry wouldn't want to work.


Sound familiar? Maybe we need this guy as our CEO.

Anonymous said...

I am a developer and has been with company for 5+ years. I would like to understand why are we so obsessed on hiring developers only? What is wrong with hiring strong PM & Test. So far all the projects & their releases that I've worked with have suffered due to shortage of smart PM & Test folks. Open bug database of any product and tell me how would you even think about shipping that product if the team didn't find & fix those bugs before release. How would you like if your customer's are going to experience that set of defects 'coz your test team was stupid enough to miss them?

I've spoken to bunch of folks who has been 15+ years with company and trust me all of them still seem to be living in the past! For sure those were the golden days when people use to bust their butt 100+ hours a week and MS use to attract best of the best (did I talk about stock going up in that timeframe). Now-a-days MS attracts a ton of average crowd (thanks to our compensation package, stock situation included). Not every dude is a super-star; folks work 9-5 schedule and those who want to work 100+ hrs. they look for work in Kirkland. Given that you tell me how can we dream of shipping world-class products with averge developers and without smart PM & test?

Anonymous said...

For those who have been advocating greater accountability within Windows leadership for the Vista train-wreck, there was simply no way that brianv could stay. If you think he should not be leaving but also want greater accountability, you're contradicting yourself.

Despite being a hero to the Windows org for saving Windows 2000 from itself and then executing really well during the Whistler development cycle (XP and Server 2003), the buck has to stop with Brian when it comes to Vista getting completely out of hand.

While jimall probably deserves nearly as much of the blame for pushing the team to bite off too much at once and constantly randomizing teams over the past five years, brianv is ultimately responsible for the "engineering disaster" side of things.

Also, if you think Brian today is the same Brian of yesteryear, think again - the man has been completely checked-out for the last couple of years, and it shows.

I don't know about this new guy, DeVaan, but there's no question that it's time for some new blood in Windows.

Anonymous said...

Budgets and in some cases hires are frozen until the re-orgs are done and the budgets can be re-examined, at least in Windows.

The meetings I've had recently indicate that many of the folks needed to replace Amitabh and his non build team staff, (most of which are leaving) haven't been chosen yet. My skip-level has no idea who he'll be working for and he's running a team of 60+.

Don't expect many changes until Vista ships.

Anonymous said...

Mini,
Don't give a damn about how many employees you have or what your current stock price is, or how many "process monkeys" there are.
Just want to know when your "live" offerings will work with my purchased server/desktop software?
This will add VALUE to our business and then add VALUE to your/our stock valuation.
Do we have to wait till version 3 to get this interaction between products and services?
Don't think that you have that long to get things turned around enough to compete with your competitors "innovation(s)".
Everyone here depends on our bottom line to get paid. If they don't INITIALLY contribute to that during our probationary hiring period then they work for our competitor(s)or someone else.
Suerly you have people just sitting around "thinkin shit up" (and then actually implementing this daily, weekly, etc) to quote a line from the movie Armageddon.

We are waiting...........

Cutomer

Anonymous said...

FAM speech demo fiasco:

Yes, I agree that MS's speech team has a brain drain problem. Both Xuedong Huang and Kai-Fu Lee are not only their leaders but also world-class speech experts. Kai-Fu and Xuedong are not MS's sales guys such as KevinJo or Mr Ballmer.

Without leaders like Kai-Fu and Xuedong, MS will have a hard time to compete. I think the FAM demo is only the indication of MS's deep problems.

Anonymous said...

"The emphasis on profit-and-loss to the exclusion of other values was creating a culture of destructive contention."

MSFT's massive losses in emerging businesses over the past 5+ years, hardly puts them at risk of being overly focused on profit and loss. In fact, MSFT makes less EPS/share today on 2X revenue than it did in '00.

Anonymous said...

We are a long way from taking microsoft private. Keep in mind that comment was made 5 years ago - probably in the wake of the post dot com and 9/11 market crashes when we were coming off 30%+ growth.

Here's an interesting fact: take a look at the insider transactions for microsoft - where directors and executives have to disclose when they buy or sell MS stock. I challenge you to find one person on that list who is putting their money where their mouth is - and investing their own net worth back into our equity.

Our senior executives have the best perspective on the health of and prospects for our company - and they have unanimously decided it is not a good investment.

Yes, I do understand all of the benefits of diversification (caveat: I am a non-developer MS employee with a - don't hate - MBA). But I will quote Warren Buffet - "diversification is only required when investors do not understand what they are doing".

I will feel really good about the company when I see Raikes, KJ, Robbie et al flip the switch from being aggressive sellers of our company into opportunistic buyers. For me that is the signal that they actually believe what they are telling us.

So, if you are a senior executive and you are reading this - can you get up in front of us and say that you believe so much in the company that you are buying the stock with your own money? If you can, then I believe what you are telling me.

I am not in Redmond. But I would LOVE to see at every town hall, every big meeting that the first question be "are you buying microsoft stock with your own money, and if not, why not?" That would be excellent!

Anonymous said...

I love Brian. I think he's a great guy, and he's been a pretty good leader over the years. But with Vista, he's really been little more than a cheerleader. He has NOT "called a spade a spade". He parrots the same "We're almost done with this milestone!" BS that almost every other manager sends out. For several Vista milestones, he was
completely in denial up until the last possible moment. To me, that's not an example of good leadership.

Brian has both led and personally contributed to the (recent) success of Vista. His mark can be seen everywhere.

Please provide some examples.

Anonymous said...

The meetings I've had recently indicate that many of the folks needed to replace Amitabh and his non build team staff, (most of which are leaving) haven't been chosen yet. My skip-level has no idea who he'll be working for and he's running a team of 60+.

-
There is no replacement for Amitabh and the people that are leaving. Build is most likely to report to the WTT team or some one in that hierarchy.

Dont underestimate EE lions. They will tame the windows group.

Anonymous said...

Open bug database of any product and tell me how would you even think about shipping that product if the team didn't find & fix those bugs before release.

Most PM's don't care about fixing bugs. After they get people working on the V1 release of something, they're already off working on V2. That's why it usually takes us 3 times to get something right. We don't have a chance to get customer feedback incorporated until the 3rd attempt.

And devs would LOVE to not have test. Then they could just go fix a bunch of bugs without ever filing any in the bug database. Then management looks at the number of bugs filed against their code and thinks they're superstars.

Wouldn't it be nice if orgs were FORCED to hire 1 tester for every 2-3 developers that got hired? Way too often new dev headcount gets opened, but nobody bothers to sync with the test group to make sure they have matching headcount opened to handle the influx of new code.

Anonymous said...

Anonymous said:
It's pretty easy to sniff out what those 10,000 new microsofties do (and it sure ain't development!).
....
My guess is less than 10% will be developers. (Anecdotally, it seems to be easier to hire PMs & test).


Looks like this guy hasn't been responsible for hiring. Dev pipeline is usually quite full all the way. Dev hiring is quite easy compared to PM. Harder still is QA hiring. QAs keep moving to dev because devs are the prima donnas and walk away with most of the credit.
(Formerly dev, now PM and never been in QA myself but know the hiring pain in that area well)

Anonymous said...

>> PM & Test

I have nothing against hiring good testers. In fact, if our team boosted test headcount by, say 25% with _good_ hires, I'd be happy. But PMs - that's another story entirely. There are way too many of them as it is. So to justify their existence they create process and "report status" to each other. In fact, I'm convinced one could hire solely PMs and they'd all be "super busy" without the team doing anything of value. Answer this (particularly if you're in one of "web" teams) - when was the last time you've seen a spec which you could say was not a waste of everyone's time?

Anonymous said...

Anonymous said:
For those who have been advocating greater accountability within Windows leadership for the Vista train-wreck, there was simply no way that brianv could stay. If you think he should not be leaving but also want greater accountability, you're contradicting yourself.

Or maybe you are in SQL and think that those responsible for the "four year grand success of WinFS" (like Peter Spiro and Quentin Clark) need to be rewarded. C'mon, did you miss the frank discussion about accountability is for L65- ?

Anonymous said...

I don't know about this new guy, DeVaan, but there's no question that it's time for some new blood in Windows.

-----------

Well i can tell you about DeVaan .. but before i do here is a snippit... he used to work in Microsoft TV and we all know "how" successful that was. Thankfully Moshe L. took over and stopped the bleeding and invested somewhat wisely given the unfair advantage Windows has on various technology patents and business funding over the previous MSTV CE based crud and distrust in the CE TV space

Anonymous said...

While jimall probably deserves nearly as much of the blame for pushing the team to bite off too much at once and constantly randomizing teams over the past five years, brianv is ultimately responsible for the "engineering disaster" side of things.


As another blogger stated, brianv came to notoriety when folks put in 80 hour weeks and worked in teams where everybody essentially knew everybody else. His style was perfectly suited for that. His wacky sense of imagination and natural esprit d'corp were perfect for breaking the tension and removing the stiffness from the product development process. In Windows the scope grew huge and required constant factoring out of features and factoring in of competitive challenges (essentially building on shifting sands.) In the end, it was hard to rein it all in. This will provide some welcome new challenges for DeVaan and Sinofsky, and a welcome change of scenery for brianv.

Anonymous said...

I'm glad Google should "watch out" and Microsoft now "gets" the net. I'm an ex-MSFT now working on search-related businesses. AdCenter 1.0 is same 'ol, same 'ol Microsoft v1 - it doesn't come close to working like it should. And we hear this from a lot of other companies. I guess Microsoft will rally and get v3 right, but it's painful right now

Anonymous said...

With regards to there being too many PMs reporting status to each other and creating process: I'm a PM and admit this is a fact. My project currently has 3 PMs and 0 (that's right 0) devs assigned to it. How can a software project not have any devs assigned to it?

To the person who said that specs were generally a "waste of time" I'd say this is sometimes true and sometimes not. It's useful to have a spec that is a basic outline of what's going on. After dev kicks in the spec goes out the window, so I'd say anything more than the raw basics is pretty much a waste of time. Of course I'm a bit biased since I started as a dev and moved to PM to lower the # of hours worked per week. I have no complaints it's easy being a PM :)

But if I'm CEO, PMs are the first to go.

Simon G said...

Mini - don't forget Dynamics. When MS got into ERP and CRM, they took on a business which is far more people intensive than desktops or servers require.
FYI - in a small operation like here in South Africa, the Dynamics group just increased by 40-50%.

Anonymous said...

On the matter of devs vs test... how many people remember when the core NT team (call it NT OS/2) didn't have any testers that did component-level testing? The accountability (enforced by DaveC) was tremendous: devs didn't check in shit. It almost always built and triple-booted. If a dev checked in crap, there was a line outside his door offering helpful advice on how to be professional.

Devs need to be accountable too. Oh, and I'm a dev too.

Anonymous said...

"C'mon, did you miss the frank discussion about accountability is for L65-"

Don't underestimate the problem here - if you compare ratings across levels, you will find a healthy distribution of exceeded / achieved / underperformed in levels <65 (20%/60%/20%).

But if you look at the distribution above level 65, it's more like (40%/60%/0%).

The issue is that it's difficult to track failure to management, as opposed to blaming the ICs. Management needs to be reflective enough and upstanding enough to blame themselves officially for impact on their review - and in reality, the human and economically sensible response is to not do that.

Anonymous said...

I am in field sales and I can tell you that they keep throwing 'overlay' sales folks at us (by ‘us’ I mean the front line -> talk-to-the-customers-every-day folks). For two years we have been telling upper management that we DON'T NEED more overlays to execute. Instead they hire all of these overlay people and now they are telling us our cost of sales is too high. So what are they doing? Cutting our MARKETING MONEY – far less executive events, far less schmoozing (means our customers are in the sole hands of IBM, Oracle, SAP, etc). STOP HIRING PEOPLE in SALES! PLEASE!!!!! Instead ASK THE FRONT LINE PEOPLE what we need to win and grow the business. The MBAs are constipating the sales process, raising our costs and HARMING our ability to sell and win. We don't need more people, but we do need motivated and well funded (marketing) sales experts who can hand Google, IBM, Oracle, and other competitors their A**es when we go toe-to-toe in a sales opportunity.

Anonymous said...

Mini, I encourage you to look at this blog post by peter norvig if you haven't yet.

http://googleresearch.blogspot.com/2006/03/hiring-lake-wobegon-strategy.html

Anonymous said...

Just to add more dead wood on the dev vs. test vs. pm fire, I am a tester and we need accountability too.

Did the dev or PM ever read my test plan?

Did my manager ever review my test code?

Did anyone pretend to even care?

Hell, my manager told me to stop filing bugs!

I guess that is why I stopped caring about my job and was 3.0-ed out the door.

Testers can make busy-work just like everyone else. We just had accountability on metrics that didn't matter but that were easy to count.

Anonymous said...

http://www.cnn.com/2006/TECH/ptech/08/03/microsoft.hackers.ap/index.html

Whoever is responsible for this initiative and bold statement, THANK YOU!!!

Shows that someone somewhere there apparently has the cajones and belief in company/job to take advantage of those who would seek to destroy.

Customer

Anonymous said...

Well, I was hired to track you down...

Anonymous said...

Wait, I thought I was the only on the super-secret team hunting you down? Could you clear this up for me? Ask the previous commenter to post his/her alias. Thanks!

Anonymous said...

While MSFT is loosing some good people to Google, so is the Open Source crowd:
Andrew Morton gets googled

Andrew Morton is considered by many to be Linus Torvalds' right hand man.

Anonymous said...

I keep getting this haunted feeling on the back of my neck, you know like when the hair stands up because of something dangerous is out of place?

So it leads me to ask a couple of questions. First, did Microsoft decide to hire 10,000 people before or after awareness of Google's challenge? Answer will answer a lot of questions and provide a map to what may be.

Second, does Microsoft (maybe to complex a question for the total Microsoft but, certainly by division,) ever ask its customers what they want? This question was driven by other questions, like has anyone ever really done a focus analysis to determine what aspects of live services, web based computing, etc. are really wanted and needed? Or does Microsoft just build stuff, you know, like when a customer asks for a product and buys a hundred of em and you then sell it to everybody? My own observation is that products are born at Microsoft because someone else is doing it, has done it and well, by golly, Microsoft better shake a stick at it.

Does Google actually have an OS or usable product for the likes of most Internet users? I am aware that there is a lot of activity, buildings, hiring, ideas, but where is the revenue stream returned from all of that? Or has the whole IT market gone nuts and the new strategy for the 21st Century is some kind of lose money like forever man because investors are morons? The same will be asked of Live. Really, what DO your customers want? Do any of you even know?

Anonymous said...

As far as dev vs. test vs. PM, let's put it like this:
If I RIF'd 50% of them tomorrow and didn't replace the headcount, how long would it take for the company to adjust to still produce at the same level we are at now?

PM: 0-2 months
Test: 1-2 months
Dev: Never

EE Deadwood said...

Engineering Excellence group is a joke. ... It lives to recycle ideas from niche groups around the company as if they were the next wave of computing.

Well, in a way, yeah, but (naturally) I'd spin it differently. We try to identify best and emerging tools and practices from inside and outside the company, pilot and train on those tools and practices, and evaluate their effectiveness and applicability at Microsoft. We want to drive to a leaner, more agile way of doing things, but more importantly, arm teams with proven tools and practices that are relevant, real, and not burdensome. All while we "rest and vest". :) (Hm, wait, vest -- vest what exactly?) :)

Anonymous said...

Brian, thank you for your service. It's a testament to your track record that you've remained in your position this long, given the challenges with Vista.

Thank you for your service, this shake up is what we need to be effective again. Don't worry, you've got alot of stock, we'll make you much richer.

Anonymous said...

>PM: 0-2 months
>Test: 1-2 months
>Dev: Never

Aren't you devs the ones putting out all of this late, low quality code and embarrassing bugs that is causing the loss of customer confidence in the first place?

Just asking.

Anonymous said...

First, thanks for the information on taking a company private.

Second, to the person who posted a link to a graph showing Google's hiring policy is the best, Peter Norvig, Director, Google Research did not do simulation properly. It is mathematically faulty. Following is the reason. In keep hiring above the mean, sure the mean increases. This is a simple mathematical fact and need no explaination or simulation, unless of course you have no common sense. But the down side of hiring above the mean policy is that it does not allow you to hire a large number of employees.

Since Google is hiring a large number of employees, a simple mathematical theorem (proof by contradiction) is that Google does not follow the policy of hiring above the mean.

Following is a proof and fault in Norvig's simulation.

A random perspective employee is rated between 0 and 100. As the standards keep increasing you would only be able to hire one in a large number of employees.

This problem is not reflected into Norvig's simulation because he did not set the sumlation properly. First, he assumed that a random perspective employee has a rating of 0 and 100. Second, he compared the three graphs after hiring N people in all three situations. This is not the same input to all the three policies. If he treats this as the same input, then let me give him an even better policy - hire above the max. If any interviewer points to a current employee better than the persepective employee then automatically decide "no-hire". If Norvig runs the simulation he will find that this fourth policy will beat Google's current policy. (Google please route a million dollars in my bank account for proposing a much better hiring policy than you have!)

The same input to the policies will be that you pick from a pool of N available people. Say N persons with a random rating between 0 and 100 comes one by one. Then you will see that although the graph with Google's policy will go higher than the other two but the graph will be shorter (i.e., it ends on more left than others). The graph will go even higher with the policy I proposed but it will be even shorter.

So what is the true Google's policy? Well every company gets to realize a thresh-hold, which is typically not explicitly states, but gets there as the company's need and market supply changes. So the policy really is that Google has a thresh-hold rating say x. If a perspective employee is perceived to be above x he/she is hired else not.

This is Google's policy and this is Microsoft policy too. Our demands are different but supply is the same so I could believe that x could be higher for Google. Though in practice I know that x is lower for Google. I have seen a lot of people rejected from us and accepted by Google. Some of my friends with less IQ but more preparation (like finding solutions to Google puzzle before hand) got into Google whereas some other friends with high IQ but low preparation did not make it.

I like to make a point about no-hiring manager policy too. Norvig claim that we hire first and then decide what to do with the new hire. This is plain stupid (I mean mathematically, no insult intended). This must not be true under the hood. This impression is probably being created to look cool as with many other Google things which are shaky under the hood.

Well here is a problem with this policy. If the current Google policy allows then Google would hire accoring to the supply side of the labor market. It will get more people in certain area then it needs. Let us say, for the purpose of explanation, it will get more chef's then it needs. It will hire more HR people then it needs. It will hire more directors of Google research then it needs. And since Intel recently fired 1000 managers, hire according to the supply side means Google gets more managers than it needs.

An intelligent policy in any transaction is to buy according to demand and not according to supply. Would you buy because something is a good deal, say an item is on sale, or would you buy because you needed that thing.

If one hires according to the supply side and ask them to chose a role among the available roles then clearly a talent will be wasted. For an example a person with no skills of making a correct set of simulation would end up running one.

This simulation is of course not scientifically set up but the person who did (or whose reportee did) is the director of Google Research!

Anonymous said...

"As far as dev vs. test vs. PM, let's put it like this:
...
"

GASP!! Devs would actually be accountable and responsible for quality?!!

This might work if no one below 63 was allowed to check in code.

Anonymous said...

"While MSFT is loosing some good people to Google, so is the Open Source crowd:
Andrew Morton gets googled

Andrew Morton is considered by many to be Linus Torvalds' right hand man."

You obviously did not read the article. Morton is working at Google to continue unabated his previous role as 2.6 kernal maintainer. He is still Linus's right hand man: "I shall continue my maintainership role with the Linux kernel--there should be few if any visible changes in this function," Morton stated.

Anonymous said...

To the anon commenter above me:

Idiots like you are the reason why our products suck. You think you can manage by RIF'ing testers? Who is going to make sure the crap you write doesn't make it past PRS?

Anonymous said...

This question was driven by other questions, like has anyone ever really done a focus analysis to determine what aspects of live services, web based computing, etc. are really wanted and needed?

At this point, do we have a choice? The state has clearly moved from the desktop to the 'Net. Let's go with the flow. We know we have to service-ize our offerings, make a lot of them down-loadable for pay, another good lot of them advetiser supported, and some perched in between. There is no question about this. Yeah, should we extremely smart and analytical about how we do this? You bet. But at least we have a direction. (Actually, we - most of us - have known that this was the right direction to head in, but we kept tractor-beaming back to the PC. (Thank you, Ray!)) Do we need to hire a *bunch* more people? I don't know. But fresh-thinking is imperative at this point!

Anonymous said...

> While MSFT is loosing some good people to Google, so is the Open
> Source crowd:
> Andrew Morton gets googled

Google will be paying Andrew to continue the same kernel work he's been doing for years. How is that losing him from Open Source?

Since this is a MS-related forum, you might want to ponder why Andrew and Linus 100x more productive than even the top MS kernel engineers though.

Mostly it's prolly due to less dev process overhead, but it's stars like that you need to recruit now, and in the future. Good luck with Balmer at the helm.

Anonymous said...

As far as dev vs. test vs. PM, let's put it like this:
If I RIF'd 50% of them tomorrow and didn't replace the headcount...


I'm a Dev here. How about instead we RIF all the PMs and keep the testers? And then, instead of having gobs of testers reporting through their own chain, we put a few devs together with a few testers, give them all the same manager, and tell the manager,

"Here, you have some people who are good at writing code, and some people who are good and analyzing and finding defects in code. Use them wisely, make a quality product. You will be reviewed on it."

Anonymous said...

Just for a bit of perspective.. Apple has project managers. Their job is to monitor the progress against the ship date and the feature/bug list. They are not in charge of the developers, and typically they have no direct reports. They basically keep track of the schedule so that the engineering managers, directors, and VPs know the status in some detail.

The Safari team has one project manager. The Mail team shares their project manager with two other projects. Apple's got processes too, and yet they are very careful to avoid letting the process get in the way of the real work: writing the code.

Anonymous said...

About the above comments on Google's hiring policy:

First, Peter Norvig's simulation demonstrates the simple statistical results of two hiring approaches. It's simple mathmatics and nothing wrong. So the person above claim it is faulty will never see the one million dollars in his account.

Second, there is no doubt that Microsoft's hiring bar is currently much lower than Google. The average quality of Microsoft new hires in last two years are clearly lower than Googles.

Third, the no-hiring manager policy is creative and great! I see so many weak people get hired into Microsoft simply because the hiring manager want fill his openings. The result is bloated organization with low quality employee and efficience.

Ihar Filipau said...

> Devs need to be accountable too. Oh, and I'm a dev too.

I would love to be accountable. If somebody ever looked at what I as developer did. I'm not a msft employee - but that's industry-wide disease.

Keeping devels accountable is only half of story. How much of the devels work was thrown out of window? I feel the pain of the people who started Vista development - work they did now is completely disregarded, forgotten.

There is nothing worse for developer for his work to be disregarded. How would you expect people to work 80+/week after that? They are already exhausted.

I somewhat understand companies like Apple, which keep their development very secretive. I've been in that situation once too: as developer it allowed me to concentrate on work (instead of long meeting with higher-ups about how important the work I'm doing). In accidental meeting with management by coffee machine we've been solving about 90% of all problems.

MSFT's problem in part that you need something to keep talking about in press to maintain company's image. That's why developers find themselves under double/triple pressure: deliver on schedule and keep the high profile image.

Test guys are not helping in that image. You Americans like to boast numbers like "millions lines of code". But the last thing anybody want to brag about is "thousands unattended issues in bug tracking system".

Having little experience as PM, I can only advise to hire guys as "developers" but really giving them QA jobs. I did that once. I have received junior programmer and given him task to develop alternative parser implementation. Officially he was still doing "software development", in reality he was doing test suit. He wanted to be "software developer", management wanted to speed up later project with "more developers". IOW, everyone was happy ;-)

Charles said...


This question was driven by other questions, like has anyone ever really done a focus analysis to determine what aspects of live services, web based computing, etc. are really wanted and needed?


At this point, do we have a choice?

Absolutely you have a choice. Several choices. It is the wrong choice that is to be avoided.

The state has clearly moved from the desktop to the 'Net.

Is that so? And the data that confirms this is.... where?

And if true to some extent, then why? Making the "right choice" for the wrong reasons will just as quickly lead to a dead-end as the wrong choice. If the underlying market forces/motivations are not clearly and correctly understood, all but the most nimble of offerings will fail for lack of customers, revenue, or execution.

And what metric is being used to conclude The state has clearly moved from the desktop to the 'Net.? Dollar profit or just revenue, or only quantity of users without regard to profit. And what is the opportunity cost for the scale of revenues that are significant to Microsoft? Just because SaaS is in fact a new growth market doesn't mean it's growth supercedes or otherwise impairs the growth of licensed software. And while Microsoft's growth prospects for licensed software may be impaired on the PC (largely due to Vista-related customer-issues), that also means that licensed software revenue could grow more readily were those issues fixed (recall, NT used to be ported to the DEC-Alpha iirc) and an entire new revenue stream could be opened if Microsoft products were available on other platforms (as are Oracle's, IBM's). The cost of a Microsoft SaaS offering should be weighed against the cost of restoring portability (and adding virtualization) to Microsofts existing product line.


Let's go with the flow. We know we have to service-ize our offerings, make a lot of them down-loadable for pay, another good lot of them advetiser supported, and some perched in between. There is no question about this. Yeah, should we extremely smart and analytical about how we do this? You bet. But at least we have a direction.

Actually, no you don't have a direction, at least not yet, and you likely don't have some key prerequisites.

What software does Microsoft have that would make a viable service (read profitable, competitive business offering)?

For example, the Office suite is not viable as a service offering. There are not sufficient 'users' who will pay to do their spreadsheets and documents over the web. While Vista raises the TCO of doing that on the desktop, don't presume that 'customers' will see their choices as restricted to Vista or Microsoft Live.

Downloading and/or playing games, music, entertainment etc, are obvious pay-per-use services. But that is a consumer market.

In the business market, Microsoft Dynamics (CRM, ERP) is an obvious choice for example, until the pros/cons of an externally hosted application are considered. For example, from Microsoft's perspective the ability to scale-up an unshared customer instance (customers demand privacy) while sharing the application(s) code across multiple platforms and multiple customers (multitenancy) requires virtualization - a technology well beyond Microsoft's grasp at present due to Vista architectural complexity. Also, from the Customer's perspective the ability to integrate CRM/ERP into a BI/DW with customization that gives the Customer competitive advantage is key. There are some apps to which the customer can easily adapt its business practices (such as sales expense reporting) and other apps which must accommodate the customer's business (supply chain management), and a few apps that deal with proprietary intellectual property (e.g. R&D, customer list, sales pipeline) where data must be current and immediately available to authorized users and secure from theft.

There are more considerations, but the question "has anyone ever really done a focus analysis to determine what aspects of live services, web based computing, etc. are really wanted and needed" goes to the heart of what can Microsoft offer cost-effectively, who would want to buy that from Microsoft, what are they willing to pay, what are the barriers to entry of competitors for that offering and customer profile, and what are the prospects for customer retention and usage growth. For example, some SaaS models 'bootstrap' the customer into an operational service of integrated and customized offerings, which at some point that makes business sense to the customer, they can take the entire operation back in-house and pay license fees. What starts as an SaaS customer migrates/morphs to an licensed in-house software customer.

Then too is the issue of does Microsoft really want to be in the business of providing the service(s) or might Microsoft be better advised to enable its partners to provide the service(s) using Microsoft product offerings and/or assistance?

(Actually, we - most of us - have known that this was the right direction to head in, but we kept tractor-beaming back to the PC. (Thank you, Ray!)) Do we need to hire a *bunch* more people? I don't know. But fresh-thinking is imperative at this point!


Fresh thinking is indeed needed. Correct fresh thinking even moreso.

Anonymous said...

Andrew Morton finally realized that the wife and kids need to be fed. OS fervor does not pay the rent

Oh he is not the first. Linus did same with Transmeta

Anonymous said...

Hey Mini,

Please keep the comments on-topic and don't let it degenerate into PM vs Dev vs Test again.

Thank You

Anonymous said...

the current team's collective IQ is probably 30% of its competitors.

Yes, Microsoft can hire 10,000 more people to match with Google, but Google only hires the cream of the crop. With Kai-Fu and other top guys leaving MS for Google, how could MS match with Google?

Anonymous said...

"The state has clearly moved from the desktop to the 'Net. Let's go with the flow. We know we have to service-ize our offerings, make a lot of them down-loadable for pay, . . "

Wow, talk about making my point for me! "go with the flow"? For those of you who are confused, software has always been a service, unlike your WGA crowd who seem to take that to mean service 'to the company bottom line' not to the customer.

The fundamental point that I keep seeing over and over again on this and other MS blogs starting with Mini is that Microsoft is not a customer driven company. It is really more like a kingdom of Lemmings and every thirty years a king dies and all the Lemmings head for the cliff. Customers are NOT Lemmings.

Just so you know, as a customer, I, my employees and my customers will never:
1. buy subscription based OS or small apps.
2. port my data to a server farm of any kind unless it is in my intranet.
3. pay as I go for services to anyone, Google or Microsoft.
4. trust Microsoft to be a customer driven company. (search QFD, Voice of the Customer).

I will ask you all again, has anyone at Microsoft asked the customers what they want? The question and answers you will find are very complex, detailed and not what you think.

It does not matter what the 'flow' is regarding that question. What matters is what your customers want. You are confusing self reification of ideas (i.e., following the group which reinforces itself with itself regardless of reality) with a seemingly non-existent strategy for MS for the next era. If Vista does poorly (and it will) it will be because of its cost, long delayed one product fits all characteristic. Had anyone done a focus test you would have found that your customers wanted a) cheaper more plentiful upgrades, probably on an annual basis, probably under $39, $49 for the OS itself b) a more modular approach that would allow system builders to easily configure the OS for different purposes and c) open sourcing of file formats.

The primary reason Microsoft did not do the focus testing was because the company has always been driven by a business model which sets fincancial targets and then conspires on ways to achieve those targets. The customer rarely enters into that formual except as the patsy.

Anonymous said...

Hey Mini,

Please keep the comments on-topic and don't let it degenerate into PM vs Dev vs Test again.


It's such a persistent theme in postings that Mini has to address it sometime.

Aside from that, Mini keeps letting these postings through, so it's clear that this rivalry is felt within him to some extent.

Anonymous said...

The Andrew Morton thing is a small piece of a big problem for Microsoft. Here's why.

Google uses a lot of Linux. It is in their interest that it continue, and that people continue to work on it. So they hire Morton.

IBM hires some kernel developers.

Linus has a job (I forget with who). Somebody is paying him to work on the kernel.

But each of these outfits that care a lot about Linux only hires one, or a few, people. But since they get the results of everybody who contributes to the kernel, they each get many people working for the price of hiring very few.

In Google's case, hiring Morton is insanely cheaper than buying a copy of Windows for each of their machines. In IBM's case, hiring some kernel developers is far cheaper than the development team that works on AIX.

The economics of this situation are so radically against Microsoft that I don't see how MS is going to have any chance to deal with it. The obvious answer is "better products and more features", but Linux is good enough for most things (gaming being the obvious exception), and Vista took 5 (or was it 6?) years to produce, and I have yet to hear a straight answer on what the compelling new features are. So Microsoft isn't doing too well at the "better products and more features" approach to competing with Linux, and it can't compete on a purely economic basis.

All that leaves in Microsoft's favor is inertia. This is not a long-term solution.

And the long-term problem is only going to get worse. Both operating systems and office packages are becoming commodity technologies rather than leading-edge wonders that require awesome technical ability. What is Microsoft going to do in response?

MSS

Anonymous said...

Second, there is no doubt that Microsoft's hiring bar is currently much lower than Google...

Poppycock! I know several of the people Google has recently hired away from Microsoft. Some of them are really good. Others, not so good. They've hired some devs whose output I would describe as more quantity than quality.

I also know several of the people we've hired at MSFT recently. Some of them are really, really good. Others are just okay. My group hasn't hired a complete bozo dev in the last couple of years, and we've had three super-star industry hires.

I've been impressed with our college hires too, though I don't know any Goog college hires, so I can't compare.

But from what I know of industry hires, Microsoft and Google seem to be about equal.

How we manage that with our 3rd-class salaries, I don't know, but we do.

Anonymous said...

"Just want to know when your "live" offerings will work with my purchased server/desktop software?"

Why ask mini, like he'd know the answer. We don't know what he does at Microsoft, if anything. LOL

Those of you that think "going private" is even a remote possibility (and it's sad so many of you are actually entertaining that notion), come back to reality, quick. Better yet, quit the company; you're too delusional to be of much value to Microsoft.

Oh, I noticed that GOOG went into freefall in July, despite reporting double earnings over last year. And Apple has a stock option scandle on its hands; they've had to withdraw all earnings reports back to 2002. See, nobody's perfect.

Anonymous said...

>> This might work if no one below 63 was allowed to check in code

I (a 61) routinely have to rewrite code checked in by our L63 (or even 64 - he's been around for a while) "senior" developer. The guy can't code his way out of a wet paper bag. It's cut and paste all over the place, no error checking anywhere and exception handling seems to be an alien concept to him.

But he's always in our director's office "discussing things" (and taking credit for work he didn't do, if doors are closed). Seniority doesn't mean shit these days, particularly if a person has been around for 10+ years.

Anonymous said...

>>>>First, Peter Norvig's simulation demonstrates the simple statistical results of two hiring approaches. It's simple mathmatics and nothing wrong. So the person above claim it is faulty will never see the one million dollars in his account.<<<<<

You will get better statistical result with my fourth policy. Hire above your current max!

What's wrong. Any scientific experiment which compares A, B and C give equal statistical input to all three. Yes I agree your average quality goes up. But then my fourth proposal would make average quality go even more up! The fault - the length, left to right, of the curve decreases. Overall the area under the curve may decrease too.

>>>>>>>Second, there is no doubt that Microsoft's hiring bar is currently much lower than Google. The average quality of Microsoft new hires in last two years are clearly lower than Googles.<<<<<<<<

I accepted this fact. If you have not noticed. This is simply because Microsoft's demand is much larger than Google now. Google development needs are still less than quarter of Microsoft.

>>>>>>Third, the no-hiring manager policy is creative and great! I see so many weak people get hired into Microsoft simply because the hiring manager want fill his openings. The result is bloated organization with low quality employee and efficience.<<<<<<

No. This way you can't hold the manager accountable. He will simply say that he got screwed up people and he could have done better. Further a manager may need a top class candidate in skill A then but he may instead get a top class candidate in skill B.

And do not forget that without the hiring manager their interview process is broken. Many times they leave the gems and hire the bozos.

Sure their hiring policy is different but it is inferior. They are hiring better than a 70 thousand people company. But they could have done even better if they were following a time tested approach in hiring.

It is unfair to compare Microsoft hiring with Google to justify Google hiring. Google being smaller naturally gets better candidate. A start up hires even better than Google! Or for that matter, the current hiring of Google is lower than what Google was hiring two years ago. Another proof that this simulation is broken! Because this simulation says average quality of hiring keeps increasing which if you see is false in Google's case.

Anonymous said...

>Andrew Morton finally realized that the wife and kids need to be fed. OS fervor does not pay the rent

>Oh he is not the first. Linus did same with Transmeta


Hey, no ragging on Torvolds or Morton. They are damned fine engineers by any measure and the kernel that they created is an equally fine achievement, as demonstrated by its ability to displace the other Unixes and Windows Server. I guarantee they're better than 99% of the developers here at MS.

They also do not have "OS fervor". They use the GPL but are not fanatical about it, as Stallman is.

I'm ashamed that you work at the same company I do and I sure as heck hope your head in the sand attitude isn't reflected across the Windows division or we can kiss MS's future goodbye.

Anonymous said...

To the person who said that we should just RIF PMs: I'm a PM and you're correct.

Why don't we just start with the PMs who cannot write code? It's pretty easy to identify these bozos:
ask everyone some standard easy coding question. If the PM comes close to getting it right (maybe it's o(n^2) instead of o(n)) they stay, otherwise they are fired, no questions asked.

This is a crude tool, but it is objective and we have to start somewhere.

Anonymous said...

WAIT WAIT stop the presses! I realize off topic but did you guys see that Ken DiPietro was hired by Lenovo?? By a former colleague! I wonder if Martin Taylor will be joining that club?! Talk about old boys network. How many subordinates do you need to boink before you DO become unemployable - particularly in a function as sensitive as HR!?

Anonymous said...

"10,000 More Microsofties - What Do They Do?"

Maybe they are needed to wrestle and juggle the automation panacea that will automatically and painlessly generate bug-free sofware for all eternity.

Collision Domain said...

Just so you know, as a customer, I, my employees and my customers will never:

1. buy subscription based OS or small apps.
2. port my data to a server farm of any kind unless it is in my intranet.
3. pay as I go for services to anyone, Google or Microsoft.


Explain Salesforce.com then.

No matter how much some of us like the old model, the future is hosted applications and services. The long-term economics and complexity management of buidling your own data center to host applications and services is stacked against firms that cannot afford multi-million dollar operating costs.

4. trust Microsoft to be a customer driven company. (search QFD, Voice of the Customer).

People said the same thing about IBM in the 80's.

Anonymous said...

Why don't we just start with the PMs who cannot write code? It's pretty easy to identify these bozos

I'm a PM (lead, actually), and I admit that I can't code Windows apps worth crap.

And I don't think I should have to -- I represent the customer and business requirements, not the implementation design. Sure, knowing a little coding will help in triage, but decisions should be made about what's right for the customer or business; coding is a result, and the dev's job. In fact, I think knowing too much about how to code is a liability to being a successful PM.

I'm sorry, but I have to say it's not a good time to be a SW dev in the U.S. I've worked extensively with teams in IDC and they do just as well, at half the cost. Dev is becoming like customer support -- skill set that is easy to train, essentially the "blue collar" of 21st century.

Anonymous said...

"10,000 More Microsofties - What Do They Do?"
-
A good number work in HR and finance serving the numerous new VPs, GMs and Directors that have been hired.

Anonymous said...

I will ask you all again, has anyone at Microsoft asked the customers what they want? The question and answers you will find are very complex, detailed and not what you think.

I think Microsoft does a reasonable job of listening to customers. If you poll customers and build a composite of what they say they want you will wind up with a useless contraption. The same customers will laugh at you and make a beeline to your competitor. Apple did not take a poll before building itunes or ipod. You obviously don't work at Microsoft or you would know that we take into account issues of detail and complexity daily.

Anonymous said...

> The customer rarely enters into that formual except as the patsy.

Um, that's how most big businesses work. It's not just Microsoft.

Think Ford and GM care about what you need? Nah, it's more profitable to churn out SUVs based on trucks. And convince you that you need them. Done well, haven't they?

And in our world, take Apple for instance. Awesome marketing. Variable products. Worsening bundling and tie-in. (Nothing personal, I'm typing this on a '96 vintage Powerbook ;-))

Anonymous said...

>In Google's case, hiring Morton is insanely cheaper than buying a copy of Windows for each of their machines. ... The economics of this situation are so radically against Microsoft that I don't see how MS is going to have any chance to deal with it.

Plausible, but even if true, is unlikely to to be one of the key factors in MS' demise. Our structure provides many countervailing effects like the close co-ordination made possible by centralized teams and the economy of scale that permits across-the-board deployment of the infrastructure (e.g. tools, internal training, etc.) necessary for good software development.

I suspect we are more likely to be brought down by basically a lack of desire to win, if you will. Many of the posts above clearly paint a picture of cultural ossification instead of seeking innovation, inter-disciplinary and inter-divisional infighting instead of unity in the face of competition, cronyism instead of meritocracy and a host of other ills. This internal rot matters far more than the other factors, and will hand victory to our competitiors, just as our past competitiors mis-steps handed victory to us.

Anonymous said...

I\'m tired of all this praise for Kai-fu and XD by people who obviously have never worked with either of them. I\'d like to clear up a few things about Kai-fu because whatever genius he may have had in research he was an absolute disaster as a VP.

Get this clear: the reason he left is because his entire division - the Natural Interactive Services Division - was about to be disbanded for its abysmal performance. It was already being picked apart when he left (you guys who used to work one of the 4-5 different NUI groups know what I\'m talking about. Hope you got rehired.) Kai-fu was on his way out, he simply chose to leave on his terms. His poor decisions and lack of planning make him personally responsible for the loss of billions of dollars and untold man-hours. I won\'t go into the numerous bozo projects he launched and sunk, but I will say that even his biggest success, the Speech Server, was a bottomless moneypit and has no hope of recovering development costs. Cool product, no buyers. The only thing he was good at was using his research background to fool people like BillG and Sinofsky into believing in his pipe dreams. But research is not product development. Far from it. BillG figured KFL out and destroyed NISD, and as for Sinofsky: notice that speech reco is no longer in office.

Kai-fu was a terrible leader. What he lacked charisma, charm, and personality he made up for in arrogance and narrow-mindedness. Now I realize I can\'t offer proof for this, but just talk to anyone who used to work under him. And add spite to his list of qualities because the whole Ballmer \"chair throwing\" rumor was started by Kai-Fu during his testimony.

XD is a whole \'nother rant.

My point is this: if these are the kind of people Google is \"stealing\" from us, then we have absolutely nothing to worry about. They are filling their executive ranks with some of the worst we have to offer. And probably the lower ranks too.

Think of all the people you personally know who have \"switched sides\" and ask yourself this question: Are we better off without them?

Anonymous said...

On the topic of hiring, I have interviewed with both Google and Microsoft (multiple times). I have to say that Microsoft hiring process comes across as much more organized and professional.

I also know a large number of people who work at Google. An overwhelming number of them are super stars - the kind I would hire in a nano second for my own company. But I also know a few whose value is quesionable. It would be interesting to see how Google handles these people as it grows. Whether these people change Google's culture of whether the Google culture is strong enough to weed them out.

Not having hiring managers in the loop seems like a negative thing to me. Most companies train managers to hire for the company rather than the group (I know this is true for Microsoft). Either Google does not trust its managers or does not train them well if they have to resort to such measures.

Finally, Norvig's simulation seems like more of a publicity stunt designed to get the "wow"s. This looks great on charts but in practice is not very feasible to implement. Most companies (including Microsoft) use a more practical strategy to keep the average high -- managing out the bottom 10%.

Anonymous said...

check the jawadk commitments. judge for yourself if this person is worth paying millions for what he is doing.

Anonymous said...

"Those of you that think "going private" is even a remote possibility (and it's sad so many of you are actually entertaining that notion), come back to reality, quick. Better yet, quit the company; you're too delusional to be of much value to Microsoft."

Good to hear, because the last thing that MSFT stock needs is for shareholders who've long questioned management's focus on enhancing shareholder value (with good reason based on performance), to conclude that in fact management is actually working against them in order to take the company private on the cheap.

Anonymous said...

"I will ask you all again, has anyone at Microsoft asked the customers what they want? The question and answers you will find are very complex, detailed and not what you think."

Yes, we asked customers what they wanted. And we came up with a next version product that would meet their needs. Then the VP came along and saw how long the dev cycle would be (~2+ years to implement all features to meet customer needs; that included a two-tiered product system)and told PMs and Devs to come back with a product that could be released sooner AND that both tiers would have to be released at the same time. Okay, what to do to meet that requirement? You strip away features--including some that the customer required and has asked for.

So we will be left with a diluted product when it ships. And no, it won't meet all customer needs. Not that we don't know this! However, the VP wants something "out there" to not let a vaccuum exist (read: other companies will fill that vacuum while we are still churning out product).

I don't like either option. I guess we could throw devs at the product to get it to ship faster (and with the required features), but we're struggling to meet milestones as it is. Taking the time to hire and train more devs would definitely hinder progress.

Truth be told: we know our product won't be exactly what customers want. Tell that to the VP and the Senior VP, okay?

Anonymous said...

Dev is becoming like customer support -- skill set that is easy to train, essentially the "blue collar" of 21st century.

Most ridiculous comment. You probably have no idea on what it takes to be a great Dev.

PMs with your attitude are scum that MSFT does not need.

On a related note, Google has 1 PM for every 10 Devs.

Anonymous said...

As far as dev vs. test vs. PM, let's put it like this:
If I RIF'd 50% of them tomorrow and didn't replace the headcount, how long would it take for the company to adjust to still produce at the same level we are at now?

PM: 0-2 months
Test: 1-2 months
Dev: Never


Produce what? Useless specs? Test automation for the customers?

Anonymous said...

To the PM Lead who can't code:
Congrats on "representing the customer and business requirements"

The PMs on my team claim to do the former and don't even try to do the latter.

Of course we need someone in the PM role, but they need to have a clue about what writing code is like (that is our business, we have graphic designers for the nonsense you are talking about)

But I ask you, do we need a 1:2 PM to Dev ratio?
Not a chance.

Anonymous said...

I will ask you all again, has anyone at Microsoft asked the customers what they want? The question and answers you will find are very complex, detailed and not what you think.

Well, we asked customers and we spec'd out a product and timeline that would meet customers' needs for a next-generation product. Well, to meet all the needs (very complex), it would take almost 2 years, so we spec'd a two-tier product (to be able to release a standard product earlier). So what happened? The VP and Senior VP indicated that this was not a good solution and reduced the timeframe for the product! We just had to get something out there in the market, you know, to fill the vacuum. Because a competitor might jump in during that time and grab our customers.

The result is a watered down product that does not meet all our customer needs. Did we ask? Yes. Did we spec the product? Yes. Will we deliver what the customer asked? No.

I guess we could throw devs at the product to deliver a more feature-rich product sooner, but we're struggling to meet existing milestones, so how do we take time to hire more people? It's really a Catch 22.

Don't say we didn't ask, because we did.

Anonymous said...

"Explain Salesforce.com then."

Do you mean the part about Gateway Computer on their website front page talking about what a great job they are doing for Gateway? I guess thats why Gateway stock is down to a whopping $1.58 per share, from a whopping $2.50 a share six months ago. Gimme a break.

Without clear definition of your products, quality and accurate response to customer needs, a sales force won't have anything to sell, like the precarious situation that Microsoft seems to be embracing. Don't forget the gee-whiz factor too. As in "gee-whiz, sez customer, this is great, how did you know what I wanted?" If your customers are NOT saying that, your competition will easily see that they just need to lower their rifles a little bit to knock out their adversary. That's why they call listening to customers a FOCUS study.

The interesting thing about the `rational' manager is that he tends to wallow in areas of self justification as bullets whiz perilously close overhead.

Just remember this and you will do alright: Give the customers what they need, give the customers what they want; and the collective wants and needs of your customers are absolutes.

Anonymous said...

"You obviously don't work at Microsoft or you would know that we take into account issues of detail and complexity daily."

Yes, obviously.

When I was in college a very long time ago, around 1974, before Microsoft or Apple even existed, before the personal computer era had nary a sniffle, I worked graveyard in a factory where we built APCs for the military. My job as a CNC machinist was to run aluminum and titanium parts for the vehicles being used in Vietnam. When you run a CNC part (and the parallel to today is any digital exercise of a processor), the most important thing to do is set your zero point very accurately, as exact as possible from a single point. If your are off, the likelihood of a failed build of the part is high as well as the possibility of ruining the tool bits and the machining head and table.

The analogy does translate to management objectives and corporate culture. If either are tweaked out of sync with customer needs (your zero point) the likelihood of failure is high. Let me describe what that looks like: a highly detailed and extremely expensive development process, a complex product that essentially has no use or value. Again, you are, rationalizing the definition of products. A very dangerous process and methodology.

Anonymous said...

"New research on formal composition methods for building software systems will better equip companies to deal with escalating complexity, as physical engineers and architects have learned to do with their creations, [Craig Mundie] said."

Blarg. When will the last person realize there is no magic process bullet? Or that if there is a magic bullet it consists of getting people who really know people who know software development to run the show?

Anonymous said...

"Um, that's how most big businesses work. It's not just Microsoft.
Think Ford and GM care about what you need? Nah, it's more profitable to churn out SUVs based on trucks."

Sir, (forgive me if you are a babe) think clearly about the jaded perspective from which you write. I was around for two different disastrous periods of both of those companies. The first, climaxed in 1989 when the Japanese taught us about things like JIT manufacturing, statistical quality control (Taguchi methodology) and TQM. All offshoots of an American in Japan Dr. W. Edwards Deming whose ideas were ignored by his own people until the 90's. You may remember Ford was the first to embrace new quality and engineering techniques and it saved the American auto industry when they did. . .for about fifteen years. Old managerial traditions die hard, and American auto manufacturers are right back where they started, simply because of what you already stated in your post. Look at the might GM vulnerable to foreign takeover and pretty much a company on the blocks.

Microsoft is no different. Ignore your customers and I promise you this. . .they will ignore you.

Anonymous said...

>>>>>>Finally, Norvig's simulation seems like more of a publicity stunt designed to get the "wow"s.<<<<<<<

Exactly! Google has been doing a tons of things just for wow factor. Doing things just for wow factor only is a typical sign of short stamina.

In the end quality of product and making money on it matters. Everything else is a waste. Google search is as good today as it was two years ago. Everything else is a wow factor. Their stock did not grow because of any of the things they are doing for two years. It grew because the market in which they are the leader grew naturally.

Anonymous said...

Just to change the subject a little bit, it looks like the latest "death to MSN, long live Windows Live" relaunch has sputtered spectacularly:

http://www.eweek.com/article2/0,1895,1998792,00.asp

"Microsoft's Aug. 1 launch of its Windows Live Spaces blogging/social-networking platform has not been a smooth one.

"Customers of Microsoft's successor to MSN Spaces are reporting a variety of performance, publishing, rendering and browser-compatibility issues since Microsoft rolled out the final version of Windows Live Spaces on Aug. 1."

Personally, I haven't been able to access my spaces site since this new launch. And judging by this article, I'd wager I'm not the only one. Way to impress, people.

Anonymous said...

>> Google search is as good today as it was two years ago

No, it's much, much better. Folks in MSN Search, who know quantitative statistics will tell you that. Google relevancy is ahead and they're making the gap wider, slowly but surely.

Anonymous said...

"Why ask mini, like he'd know the answer. We don't know what he does at Microsoft, if anything. LOL"

So then I'll ask YOU,
When will our server/desktop purchases payoff with your current "Live" beta service offerings?
Are you waiting for billions and billions of us to chime in?
Will it take till v3 to get this operational?
If so your wasting your time and investment. Or more to the point, you'll be behind AGAIN!
I can see how it COULD work for us.
Apparently you cannot since you did not provide an answer or any information, as well, not even any speculation of a time frame. You know, the reason why we buy your products/services/stock...
This is exactly the difference between life there and life out in the real world.
Don't care if Mini can shoot bolts of lightning out his/her arse, At least he/she has balls enough to bring Microsoft to the customer/shareholder in a totally new light.
Thank you for your time Mini.

Customer

Anonymous said...

"Wow factor" is why most of the top talent out of college is heading to Google instead of Microsoft.

Think about Gmail: Almost no one uses it, failure right?
Wrong, a good chunk of the people who DO use it are CS majors at top schools. How many of those people use hotmail? Try about 0.

Anonymous said...

>I'm a PM (lead, actually), and I admit that I can't code Windows apps worth crap.

>And I don't think I should have to -- I represent the customer and business requirements, not the implementation design. Sure, knowing a little coding will help in triage, but decisions should be made about what's right for the customer or business; coding is a result, and the dev's job. In fact, I think knowing too much about how to code is a liability to being a successful PM.


Your response is such a depressing commentary on the state of MS today that I must respond:

You and your non-technical brethren are failing miserably at figuring out "what's right for the customer", as demonstrated by our many, many flops, because you haven't the faintest clue what they're like. You don’t understand the views of developers, systems administrators, IT professionals, or even the views of ordinary users of advanced features in Office, because you have no understanding of it. You might have some claim on understanding the needs of the general computer illiterate population, but remember that they are rapidly being replaced by a generation who grew up wired and they are quite distinctly not of the opinion that MS’s offerings are “cool”. Ouch.

Then consider too that non-technical PMs are no more able to perceive technology opportunities any more than a blind person can direct a ballet. A recent Joel on Software blog entry said it far better than I could:
Watching non-programmers trying to run software companies is like watching someone who doesn't know how to surf trying to surf. "It's ok! I have great advisors standing on the shore telling me what to do!" they say, and then fall off the board, again and again. The standard cry of the MBA who believes that management is a generic function. Is Ballmer going to be another John Sculley, who nearly drove Apple into extinction because the board of directors thought that selling Pepsi was good preparation for running a computer company? The cult of the MBA likes to believe that you can run organizations that do things that you don't understand.

Page and Brin didn’t need PMs to “represent the customer and business requirements“ when starting Google nor do they feel the need to keep that many PMs around now. Note that they are unequivocally kicking our butts in search.

Still proud of your ignorance? What will you do when our "business needs" dictate that no longer need dime-a-dozen technology-illiterate PMs?

Anonymous said...

"I guess thats why Gateway stock is down to a whopping $1.58 per share,"

Whoops. My error. Gateway down to 1.40 on Friday (down 11.95% in one day).

Anonymous said...

"Google relevancy is ahead and they're making the gap wider, slowly but surely."

Not true. Google relevancy remains stuck for most of the last two years. Improvements are only recent. This is only after a lot of pressure from wallstreet when the Google trio need to come out and make the clarifying statement that 70% of Google is still working on search. Such clarifications are needed only when there are evidence of the opposite.

""Wow factor" is why most of the top talent out of college is heading to Google instead of Microsoft."

Not really. Our division hardly lost a candidate to Google. If the talented attracted to wow factor is again start working only on creating a wow factor then it is pretty much like a pyramid. That's why I said, short stamina and not no-stamina. Pyramid schemes are great initially. But for the leaf nodes it sucks.

Google climbing at $400 is great for starting employees. But it getting stuck is depressing for new employees. At some point, Google has to come out and say to itself - no more wow factor but real innovative products only could take this $400 to $600.

They have to say that product needs to be high quality and in desirable state to be out of beta.

Charles said...

Watching non-programmers trying to run software companies is like watching someone who doesn't know how to surf trying to surf.

Let me correct Joel's statement to illustrate a problem: Watching non-programmers trying to *micro-manage* software companies is like watching someone who doesn't know how to surf trying to surf.

The job of executive level management is *not* to plan, design, implement or debug offerings. That is what the development, marketing, and QA staff are hired to do.

The job of executive level management is to see to it that competant development, marketing, and QA staff are in fact hired; that they are resourced commensurately; that supply chains and sales channels are (or will be) in place to manufacture and distribute the product or service; that a competititve strategy is formulated and followed to build on the companies strengths and correct weaknesses; that everyone gets paid; books are kept; regulations complied with; that competant managers and technicians are promoted within their respective roles to increasing levels of responsibility; and that all these differing functions are organized such that they execute together towards profitable, sustainable growth.

Running a software company is *not* programming and programming is *not* runnng a software company.

Microsoft, however, is indeed a textbook case (likly to be taught in B-schools someday) of how inexperience got this badly confused. Inexperience in operating system development; inexperience in corporate management. But inexperience mitigated by serendipitous cash flows (derived from the market-share IBM ceded over), the lack of which cashflow forces most companies to learn from and correct their technical and managerial mistakes - or go under (as do 9 out of 10 young companies).

I [a PM] represent the customer and business requirements, not the implementation design. Sure, knowing a little coding will help in triage, but decisions should be made about what's right for the customer or business; coding is a result, and the dev's job. In fact, I think knowing too much about how to code is a liability to being a successful PM.

All projects and people need to be managed (even the CEO is managed by the board, shareholders and customers) and complex projects require steady communication of requirements, results and status. Monitoring and communicating is generally the job of program/project managers who report to a product manager whose responsibility is to budget, staff and deliver the product. The role of "PM" at Microsoft seems to vary from powerless status-reporter to power-usurping decision maker, and while the customer and business requirements do need representation on the project, that would normally be the purview of a "marketing" representative or liaison. A PM ought reasonably to report to the product manager whenever the stated customer and business requirements are not being met (assuming the "marketing" representative has raised a flag), but the PM should not be deciding what correction is made. That corrective decision should be between the product manager and his technical staff, with marketing's informed concurrence.

If, in truth, PMs represent the customer and business requirements, then that is a marketing failure. Marketing's job is (in part) to ensure products cost-effectively meet the productivity needs of the customer and the competitive business needs of the company.

Executives shouldn't program or micro-manage. Developers shouldn't macro-manage. PM's shouldn't "triage" or arbitrate.

Anonymous said...

I see so many weak people get hired into Microsoft simply because the hiring manager want fill his openings. The result is bloated organization with low quality employee and efficience.


We had that happen in our group. We had a crack team that consistently accomplished the impossible. Money was thrown at our group and we had a big project coming up, so our manager (who had not built the original crack team; he'd come in after the fact to manage it, and was familiar with managing more average-performance teams doing what we do) hired a bunch of average (and, sadly, below -- and yes, I voted no hire on them) performers because he was under pressure to fill the reqs before end of FY06.

The cache' of our team went down. VP trust in our team went down because we couldn't train the new team members quickly enough to keep up the level of quality our group was known for.

I'm leaving a group I thought I'd be in for the next 10 years, because I can't stand the mediocrity and the associated 75 checks and balances now built into our project to account for the mediocrity. Funnythingbut, we didn't used to need all that crap when our team was only superstars we could trust to do the work right, and I daresay we did 80% of the work that a team 3 times our original size does today.

Please, somebody, preserve whatever superstar teams the company has left. Adding non-superstar-potential people to them only drags the team down, and overtime the quality and quantity of output go down even more. I'll also add that breaking up a superstar team with the idea that these people will energize the groups they're split out to, is bogus reasoning. Add one superstar to a group of 7 average performers with no motivation and not much potential to do better. What do you really think the likely result of that is? It's hard to win a 7:1 game. Plus, again, the group of 7 average folks has those processes that superstars don't want to deal with, so now instead of 7 unmotivated people, you've got 8.

Anonymous said...

I see billg sold 7 million shares of MSFT end of July

http://moneycentral.msn.com/investor/invsub/insider/trans.asp?view=All&Symbol=msft

...and the other thing I notice there is that no MS executive has BOUGHT any MSFT stock in a long, long time.

Anonymous said...

"Think about Gmail: Almost no one uses it, failure right?
Wrong, a good chunk of the people who DO use it are CS majors at top schools. How many of those people use hotmail? Try about 0."

Gmail is not a failure, it's just somewhat of a missed opportunity. What made Gmail really stand out from the competition when its beta started was storage space (1GB, vs. 10MB for free Hotmail at the time). The problem is that the beta and invite system have lasted so long that the competition has had time to match Gmail's main advantage. Now there is little incentive for the avrage Hotmail or Yahoo Mail subscriber to switch.

Now imagine if Google waited a year or two, and released Gmail and Google Talk simultaneously, without the invitation system ...

I don't get the Darling treatment that Google receives. Look at their latest offering (code hosting); had Microsoft released that, they would have been ridiculed shitless by the press.

Anonymous said...

"Yes, Microsoft can hire 10,000 more people to match with Google, but Google only hires the cream of the crop. With Kai-Fu and other top guys leaving MS for Google, how could MS match with Google?"

And yet, all Google seems to be able to do lately is *buy* web versions of Office programs (without even 1/10th of the functionality). *yawn*

Anonymous said...

"No, it's much, much better. Folks in MSN Search, who know quantitative statistics will tell you that. Google relevancy is ahead and they're making the gap wider, slowly but surely."

Let me know when they end the policy of shoving so many search indexes into their search results. You do a search, you get a bunch of pages that are nothing more than search indexes, which lead to other search indexes, and on and on, before you finally get to a page with real content.

Anonymous said...

"Let me correct Joel's statement . . .
PM's shouldn't "triage" or arbitrate"

Remove 99% of the words and what do you have the same thing with all the words. Yada, yada, yada. No wonder Microsoft seems lost in the ozone. I'm not Joel, I'm a customer.

Anonymous said...

And yet, all Google seems to be able to do lately is *buy* web versions of Office programs (without even 1/10th of the functionality). *yawn*

Before too many people try those web versions it would be nice to launch Word and Excel online. A single sheet of either could be free. If you want to create workbooks etc, the added functionality could be paid for. Microsoft clearly owns this space - how will they deploy? That's the big question.

Anonymous said...

Let me know when they end the policy of shoving so many search indexes into their search results. You do a search, you get a bunch of pages that are nothing more than search indexes, which lead to other search indexes, and on and on, before you finally get to a page with real content.

What the hell are you talking about? I use google about 100 times a day and have never seen "search indexes" appear in results?!

What I find really annoying about MSN is the ads -- most are completely irrelevant or look relevant but simply lead to pages with more ads. In contrast I often find Google ads useful, especially when I am looking to purchase something specific.

Anonymous said...

"Add one superstar to a group of 7 average performers with no motivation and not much potential to do better. What do you really think the likely result of that is? It's hard to win a 7:1 game. Plus, again, the group of 7 average folks has those processes that superstars don't want to deal with, so now instead of 7 unmotivated people, you've got 8."

Actually, it's worse than that. In most cases, the other 7 (or more specifically the 2-3 ol' boys who set the pace for the 7) resent the 8th because they perceive that he/she makes them look bad with management and/or just makes them feel less adequate by comparison. So then #8 has to decide between doing or contributing less in order to placate them, or leave MSFT. In the end, no superstar will work at less than 100% for any extended period - so MSFT losses another high performer. Meanwhile, any members of the 7 who might have been capable of reaching a higher level of performance (likely at least 1-2), get the message - keep your head down and get along with the ol' boys. Been there.

Anonymous said...

This past Friday over in Xbox-land, Shane Kim was made Corporate VP as recognition for his stewardship of Microsoft's first-party games. I particularly loved the reference to his "acquisitions/divestitures" to make MGS a "stronger" first-party contender.

Give me a break. The only acquisition that's occurred since Shane Kim took over from Ed Fries was Lionhead, and that may be the only software company in the world whose promise to delivery ratio is exhorbitantly higher than Microsoft's. He certainly wasn't part of the front-line effort to pick up Bungie, Ensemble, or Rare (who, say what you will about them, at least can produce more than one game at a time and have yet to produce an embarrassment on the level of Fable).

Divestitures? Yeah, as part of the "let's cede the lucrative sports genre to EA" drive, Shane decided to lop off Indie Games, which was probably our best internal studio as far as being able to produce a number of quality titles at once (with the money they take in, I think a lot of studios could match Bungie's record of producing one good/great game every three years). And that's in spite of the fact that MGS killed the cash cow that was PC Links in order to get Links on the Xbox.

Perhaps "divestitures" also relates to the gutting of MGS itself as roughly a third of the FTEs were laid off (and all the "STE" level contractors were replaced with economy-priced "STAs" who generally don't even file bugs).

Aside from that (actually even including that), I really don't see how Shane Kim's performance merits this promotion. The guy can't play games due to his motion sickness, and he certainly hasn't bothered to learn anything about the industry (I've related here in the past some of the gaffes he's made in interviews and I could easily relate another dozen or more anecdotes). The whole MBA mindset of not needing to understand the actual industry you're managing could not find a better poster boy than Shane Kim. Spend a few minutes with him some time and see if you don't walk away with the impression that he thinks he's a few orders of magnitude higher on the evolutionary scale than the customers (and employees) he supposedly represents. That is if you can actually catch him in the office and not on the golf course.

Aside from Peter Moore, does anyone actually believe the MGS first-party lineup is in better shape under Shane Kim? Remember when Xbox 1 launched and we had five good-to-great games on day one, followed by six more within the next three months (although I'll admit a couple of those later games were dodgy). Contrast that to the 360: Three games at launch, none of which were the groundbreaker that Halo was, and nothing in the nine months since.

(Yeah, I know Gears of War is coming, a year after the 360 was launched, along with such "gems" as Viva Pinata, but you don't make money in this business publishing three titles a year, not when you're losing money on every console sold. Sony and Nintendo understand this, and I was sure that MGS in the pre-Kim days did too. Now, I'm not so sure).

A couple years from now, when Microsoft announces they're getting out of the games business for whatever spin reason they come up with, you can look back on the moment Shane Kim became a VP after a career of being completely worthless as the last nail in the coffin.

5 billion or so down, how many more to go?

Who da'Punk said...

Just of our curiosity: does anyone think that the above Shane Kim (oy, I'm going to be repeating that name all night now) comment crosses the line?

Would it have been better suited on the Cutting Room Floor or is it fine to challenge the VPs and what their contributions have been?

Is it fine to challenge them and let defenders rally about them, explaining why there's more to the stories and filling in the details around positive accomplishments?

Just looking for feedback from anyone that feels especially passionate about the comment.

Mini.

Anonymous said...

To the person who said that we should just RIF PMs: I'm a PM and you're correct.

Why don't we just start with the PMs who cannot write code? It's pretty easy to identify these bozos:


Well, tell me if I am right: you are either an idiot or a newbie PM (or faking to be a PM). Look at the PM CSPs and tell me where it says that PM should know how to code. PMs fulfill a very important role at Microsoft and if you want to understand what it is, go thro the PM CSPs or spend a day with a PM.

I work in DMD and I am happy to say that there arent bozos like you in my team. In fact, knowing how to code is considered somewhat of a disadvantage. And oh, lest you think that I dont know how to code, I was a developer at Microsoft for more than 5 years before I became a PM.

Anonymous said...

The guy b*tching about Shane Kim's promotion is probably the same guy who was b*tching about Robbie Bach.He's probably just mad he got RIF'd and decided to get 'pay back' on this blog.

Anonymous said...

The comment about Shane Kim had some reasoning behind it. It was not a "he just sucks" kind of comment.

About crossing the line, it's up to you.

Anonymous said...

>Just of our curiosity: does anyone think that the above Shane Kim comment crosses the line?

I'm sure that there are, but I am not one of them.

Even in its current weakened state, MS still has the resources and talent to build any software conceivable. That we fail to do so is a matter of direction and leadership.

If one of our leaders is inadequate, let the reasons be stated loudly and clearly so that each of us may judge the truth and take appropriate action.

Anonymous said...

So then #8 has to decide between doing or contributing less in order to placate them, or leave MSFT. In the end, no superstar will work at less than 100% for any extended period - so MSFT losses another high performer.

This #8 is not going to do the former. My resume is already in to GOOG. After two gold stars in two years, and a small opportunity to change the world that I savored as much as any challenge so far in my career, I'm asked to be a gear in a very large machine. I love this company (TM SteveB), but I won't accept a slow-track position to stay in it. Two colleagues from the original superstar team are pursuing internal transfers due to similar thinking. That leaves a couple good engineers and a bunch of deadwood in the area in which we used to work. The remaining good engineers haven't yet done their year in the group, so they're stuck with the status quo for a while.

So again I ask anyone from management who's listening: When you have a team that's gelled, is accomplishing great things, and its members are pulling in more great hires that MSFT otherwise would never have had a chance at, PLEASE DON'T DESTROY IT BY BRINGING IT DOWN TO THE STATUS QUO. Teams like this are like the Microsoft of old; it's these teams, not the current fad of detailed process, that provide the soul that will get you real wins. Process may get you pretty pie charts and "scorecards" (another internal fad) that make you look good to your VP, but it's engineering soul that makes you look good to your VP, the CEO, shareholders and the world. Oh, and should YOU ever be looking to leave Microsoft, accomplishing a paradigm shift doesn't look too badly on one's resume, either!

Anonymous said...

> Blarg. When will the last person realize there is no magic process bullet? Or that if there is a magic bullet it consists of getting people who really know people who know software development to run the show?

You won't. The idea is just so seductive: go with THIS methodolgy and you don't need superstar developers/testers/whatever: think of the money you'll save! And my God, were the VPs sold some snake oil for Vista (e.g. WTT). Your final sentence was right on. At the end of the day, it's having folks who understand how to actually design, and actually implement, and actually test, that makes the difference. Shame our decision-makers don't or won't just GET this...

Anonymous said...

To be a contrarian...

To set context, I'm sympathetic to the worries on this blog, have followed it closely, and commented a few times. But I also drink deeply of the company kool-aid.

But I have had a turn of heart recently, to possibly irrational exuberance, though I'm not normally irrational.

In particular, the company is taking personnel issues seriously, really building buildings for the long term, hiring field guys to make the next waves happen, reorganizing senior mgmt to correct problems (even if done in a very nice Microsoft way), attacking on all technical fronts (big and small), sincerely addressing the EU's documentation concerns, building a coherent marketing message (people-ready), fixing the obvious nasty Vista bugs at a remarkable rate, getting cool with XBOX phones TV and devices, prepping a truly impressive Office release, making very solid profit and gross income increases (what happened to the argument about the 'law of big numbers'?), building some truly cool software (check out MGX keynotes, etc.), envisioning wonderful things, buying back a real amount of outstanding shares, and on and on it goes.

I'm sure we could quibble about the excellence of the execution on any or all of these points, but the reality is that we're executing. IMHO, we're not just doing a superficial bit of marketing on these points - there are trucks moving dirt, 10k of real people hired, beloved execs moved to obscure positions, real software running on my machines, etc.

If we make half a dozen of these changes well, we change the game. But I could see a dozen of them being done well. And probably there is a ton of changes going on that I as an obscure IC don't see.

Do we have bad management in some places? Too many layers? Too much bureaucracy? Contradictory projects? Missed opportunities? Yes! Big time. Point them out. But the reality is that every organization, of all types and at all times, have had those. But let's acknowledge when things are going well more than they're going badly. And then let's add to the goodness. Let’s not just earn our good wages but also change the world while we’re at it. (And sure, enrich the shareholders too).

The Microsoft dream is still alive, and I'm living it. How about you?

Anonymous said...

An earlier commenter said:

Gmail is not a failure, it's just somewhat of a missed opportunity. What made Gmail really stand out from the competition when its beta started was storage space (1GB, vs. 10MB for free Hotmail at the time).

Nope, the 1 gig storage was just something easy for the press to pick up on. What makes gmail stand out is that it is a browser email interface that does not suck. Let me know when Microsoft's stock webmail packages for exchange are similarly pleasant to use.

Anonymous said...

Just of our curiosity: does anyone think that the above Shane Kim (oy, I'm going to be repeating that name all night now) comment crosses the line?

Mini, I'm the one who made the comment and while I definitely do hold that honest opinion of Shane Kim (nothing personal, he never stepped on my dog or anything), I actually didn't originally intend for that comment to be the simple (pointless?) rant it turned out to be (and if I were in your position, I'd have been tempted to cut it as well).

My original intent, after giving some background for my opinion, was to ask if any of the contributors here, without breaking NDA, could provide a valid reason for Kim's continued ascendancy. Unfortunately I went back and added a new section toward the beginning and forgot to add that part.

If you do decide to gut both that comment and this followup, I'll understand and consider it a dead issue at least as far as this blog is concerned.

On the other hand, if you feel like indulging me, I really, really would like to hear some kind of justification for this move (again, not asking anyone to break NDA. I've just had enough interaction with him and heard enough anecdotal corroboration that I can't see how this is anything but the old boy network in action).

This is a critical time for MGS. If the Xbox effort is ever going to become profitable, it needs leadership of a sort that Shane Kim seems neither willing nor able to provide.

Anonymous said...

Just of our curiosity: does anyone think that the above Shane Kim (oy, I'm going to be repeating that name all night now) comment crosses the line?

Mini, I have long given up on you. The inconsistencies, the flip-flops and the fanning of the discipline wars (Dev vs Everyone) has taken a toll on the percieved relevance of this blog.

Of course you know this Shane Kim post crosses the line by several kilometers. This is apparently written by somebody who has been shafted rightly or wrongly and is lashing out viciously. Writing about unsubstantiated medical issues? That one takes the cake.

At the end of the day, Mini, it is your blog. But if you care to know, I wonder (like several others) if this blog is helping or hurting now (after the comeback of the towels:)).

Anonymous said...

Did my manager ever review my test code?
Forget your manager...have your peers reviewed your test code? Does *anyone* review test code? I've been with MS for almost 5 yrs now. In that time, zero test code review meetings have been scheduled in my group.

Hell, my manager told me to stop filing bugs!

They don't care about quality. They care about bug count.

If I RIF'd 50% of them tomorrow and didn't replace the headcount, how long would it take for the company to adjust to still produce at the same level we are at now?

PM: 0-2 months
Test: 1-2 months
Dev: Never


Let me guess...you're a dev? You're fooling yourself if you think that any team could ship a quality product without a competent PM and test team. Also, PM and dev are very easy to hire. Test is much more difficult (mainly because it's hard to convince people to take a crappy job.

check the jawadk commitments. judge for yourself if this person is worth paying millions for what he is doing.

Who should we compare them to? Brian? KJ? I assume you think the commitments suck, so how about some details? Instead of posting confidential info here, you can just reference them by number.

and the other thing I notice there is that no MS executive has BOUGHT any MSFT stock in a long, long time.

Why would they? They already have shitloads. And they aren't as dumb as they seem. They know the stock is going nowhere fast.

Anonymous said...

mini - yes, the attack on Shane Kim should've been saved for the CRF. I'm pretty sure
it's been posted in a previous comment stream anyway. That, or we have another VP
over in Xbox that has motion sickness :)

I thought you mentioned previously that the personal attacks like this wouldn't be posted. If not, let me know, and I'll open up on the old Scobelizer :P

Anonymous said...

"Nope, the 1 gig storage was just something easy for the press to pick up on. What makes gmail stand out is that it is a browser email interface that does not suck. Let me know when Microsoft's stock webmail packages for exchange are similarly pleasant to use."

I agree that Gmail is MUCH snappier than the Live Mails and Yahoo Mails. The point remains that the main incentive for the Gmail early adopters *was* storage space. Now that Yahoo and Microsoft both offer 2GB, nobody begs for Gmail invites anymore (and in my experience, many people receive invites and just ignore them). Topping that (e.g. offering 4 or 10 GB instead of 2) won't be of much help either, because the average internaut isn't going to use 100% of 2GB anytime soon. Google had a real opportunity there. Do you really think that that many early adopters would have bothered to switch if Gmail's initial storage space was merely on par with Hotmail or Yahoo Mail?

Anonymous said...

My resume is already in to GOOG. After two gold stars in two years, and a small opportunity to change the world that I savored as much as any challenge so far in my career, I'm asked to be a gear in a very large machine. I love this company (TM SteveB), but I won't accept a slow-track position to stay in it.

You had better hope I don't interview you because if it comes across you loved being a toady to a law breaking monopolist then you will definitely get a no hire from me.

There mini, dare you to let this one through. Sure, it's harsh but it's the truth and if you can't deal with the truth, what can you deal with? But even if you do blackhole this honest comment, at least you read it. Think about it.

Anonymous said...

"Look at the PM CSPs and tell me where it says that PM should know how to code. PMs fulfill a very important role at Microsoft and if you want to understand what it is, go thro the PM CSPs or spend a day with a PM."

Congrats on the circular reasoning here. The whole point is that the current system is broken.

Again, explain how a 1:2 PM:Dev ratio makes sense. Think about it, what percent of the MS workforce actually writes code that goes into our products?

I would love to get an actual percentage on this.

Anonymous said...

Maybe you don't have to be able to code, but what ARE the qualifications for being a PM? I'm often amazed by how out-of-touch, uninformed, and disengaged my group's PMs are. When our competition releases a new version of their product, half the devs in my team download it, try it out, and spend the next day talking about it. Most of our PMs don't follow tech news enough to even realize there was a release, and I wouldn't be surprised if many of them have never used any version of the competing product. Is this usually the case with PM teams??

Anonymous said...

I'm a PM (lead, actually), and I admit that I can't code Windows apps worth crap.

PMs can provide value without being technical or knowing a lot about the undeerlying code. For example, I don't think Jensen Harris needs to know C++ to do what he does for Office. On the other hand, PMS who brag about how ignorant they are of software development practices turn my stomach. For every Jensen Harris I've come across at MSFT, there are two idiots like the above poster who think that being able to use the application they are supposed to be designing is the only technical qualification they need to be in charge of its design.

PS: Yes, I'm a technical PM so I might be biased.

-- Dare

Anonymous said...

The Microsoft dream is still alive, and I'm living it. How about you?

Yes, and no.

There are many exciting things going on, but one problem keeps the dream out of reach:

Reward opportunities are still limited, especially for the people we need most - Level 63 through 65 Devs. These are the people who make the product what it is. All the folks at those levels make decent, but not outstanding, salaries, and level compression means most of them aren't getting promoted. The very people most important to the product have the least incentive.

We're going to become a farm team for other companies - we train devs until they are highly productive, then put them in dead-end jobs, encouraging them to go to other companies. Please stop.

Ray Myers said...

Hi, Mini!

Sorry you didn't see fit to post my "Welcome to Microsoft" comment; I am sure you have your reasons.

I became insanely interested in Microsoft's inside workings after submitting my idea to "save the world" to Adam Butterworth in your Business Manager department. He bumped it up to "Legal" and that's where it sits, now.

Based upon Vista's chances of success, I am very surprised that you guys aren't at least "stealing" it.

Check it out here:

http://rmyers.spaces.live.com/PersonalSpace.aspx

Anonymous said...

"Mini, I'm the one who made the comment and while I definitely do hold that honest opinion of Shane Kim (nothing personal, he never stepped on my dog or anything), I actually didn't originally intend for that comment to be the simple (pointless?) rant it turned out to be (and if I were in your position, I'd have been tempted to cut it as well)."

Kudos for admitting that. You made several important points in your original post and that, combined with the critical nature of the Xbox investment overall, imo warranted Mini's allowing the post. But by treating yourself to a personal character attack as well, you detracted from your message and risked it not being heard at all. The message to others should be clear: stick to business implications and leave the personalities out of it.

Anonymous said...

"You had better hope I don't interview you because if it comes across you loved being a toady to a law breaking monopolist then you will definitely get a no hire from me."

Thou should love one's employer. That could only be considered positive in my book.

Do not worry about this huy. If I find he/she messes up your interview with us, I will make sure he/she is disciplined. We ofcourse needs people from a monopolist who have learnt their lessons. We ourselves heading to be become one.

Two gold stars from Microsoft - impressive! Most our interview candidates from Microsoft have no stars.

Anonymous said...

Of course you know this Shane Kim post crosses the line by several kilometers. This is apparently written by somebody who has been shafted rightly or wrongly and is lashing out viciously. Writing about unsubstantiated medical issues? That one takes the cake.

I wrote the original comment. You can continue to assume I'm some disgruntled "shaftee" if that's the knee-jerk reaction that works for you. I'm sure there were plenty of people on the Titanic who initially thought the whole "iceberg" brouhaha was just a vicious rumor some junior officer made up to slander the captain.

As for the "unsubstantiated" part (and the main reason I'm replying), give me a break. Kim's motion sickness is public knowledge and has been mentioned in several interviews and profiles. Again, you assume I must be either making up some terrible lie or divulging a closely-held personal medical secret because actually giving me the benefit of the doubt would go against the "irresponsible malcontent" role you've assigned me.

Whatever. As I indicated in my followup, I'd really like to hear something positive (and substantiated) about Kim's performance at MGS to address the many, many doubts I have about his ability and commitment. So far the closest anyone's come is attempting to shoot the messenger.

(I appreciate the respondents who understood the intent of my original post - including mini - but I'll probably step back and let the flames die down unless someone makes another "allegation" along the lines of the above or actually steps in with some substantive pro or con argument on the subject).

Keeperplanet said...

"I really, really would like to hear some kind of justification for this move (again, not asking anyone to break NDA."

No NDA here, I'm a customer. I know ditz about Mr. Kim, but if you have been paying attention to Mini blogs for the last few weeks, Xbox has been the subject of the following weirdness: its losing money, big time. Regardless of rationalized justifications, there is no justification for a business losing money, ever.

Microsoft could make a fortune on its XBox games porting to the pc, would cover all your losses almost immediately. Microsoft is propping up a horrific business model with Xbox and will do the same with Zune.

Now, if Microsoft goes into hardware development like Apple, builds bundled computers with its OS, that is a huge huge departure from a 25 year business model. If that is where you are going, well I have no more to add and will watch the slaughter, but if Microsoft has come to its senses and realized that:
1. Xbox competes with its core business of selling software and
2. Zune competes with its failed subscription strategy for music, and 3. that both those self-competitive strategies have limited growth as well, then the next step would be a business model that actually makes products that make money--that is done by giving your customers (individual and OEM) what they are asking for.

In the game business that would be a strategy of porting all games to PCs and in the music/movie download business that would be a non subscription model that opens up the formats so that when one downloads a piece of media, one 'owns' it in the same way one 'owns' a CD or DVD purchased at Walmart.

Like I said, I don't pretend to know what Microsoft management is up to but that had better be your strategy or XBox division is finished.

Anonymous said...

"Google had a real opportunity there. Do you really think that that many early adopters would have bothered to switch if Gmail's initial storage space was merely on par with Hotmail or Yahoo Mail?"

I agree that the commercial success of gmail is questionable. However, as a tactic that cost them little, resulted in MSFT and YHOO scrabbling and spending significant money to match, and gave them media atta boys for being "innovative" (however ignorantly that term is used notwithstanding), it was a very smart move indeed. Then again, is that a tribute to their brilliance or the dog-like reponses of MSFT and YHOO (aka GOOG-envy)? I'd say a healthy dose of the latter. Bottom line, MSFT and YHOO should focus on doing a better job of solving the customer's need using their respective strengths and leave GOOG to disprove Ellison's recent contention that "they're just a one-trick pony, but it's a hell of a trick".

Anonymous said...

>You had better hope I don't interview you because if it comes across you loved being a toady to a law breaking monopolist then you will definitely get a no hire from me.

Okay, Mini, now *this* definitely crosses the line.

Anonymous said...

>Of course you know this Shane Kim post crosses the line by several kilometers.

Hogwash. He laid out his reasoning for his disapproval of Mr. Kim clearly and in great detail. So far I see a complaints but nothing resembling a reasoned rebuttal.

Anonymous said...

Well there you have it. I'm one of those 4K new people at Redmond, and I must say that the previous guard hasn't exactly rolled out the red carpet.

How about this - I have over 20 years of *real field experience* while people like you sat at Microsoft and told us how lucky we were to have you around.

You complain about workload, then when they attract some of us in to help you, we get the "I'm better than you because I've been here longer" attitude.

I'll tell you what the 10K of us that are starting are doing: we're after your job. And maybe we won't be arrogant jerks to the ones that come in after us. Why don't you go get a job somewhere that you like?

Anonymous said...

>>In Google's case, hiring Morton is insanely cheaper than buying a copy of Windows for each of their machines. ... The economics of this situation are so radically against Microsoft that I don't see how MS is going to have any chance to deal with it.

>Plausible, but even if true, is unlikely to to be one of the key factors in MS' demise. Our structure provides many countervailing effects like the close co-ordination made possible by centralized teams... Many of the posts above clearly paint a picture of cultural ossification instead of seeking innovation, inter-disciplinary and inter-divisional infighting instead of unity in the face of competition, cronyism instead of meritocracy and a host of other ills.


And do you see no connection between the "close co-ordination/centralized teams" and the ossification? To me, it looks like the close co-ordination and centralization directly lead to the ossification.

Of course, I don't work there, so what do I know...

MSS

Anonymous said...

To Mini-
"10,000 More Microsofties - What Do They Do?"

As 1 of the new 10,000 hires I can only answer as to what I intend(ed) to do: bring passion to my work and to do the best damn job I can possibly do. But here's the problem, when I look on this blog and see a bunch of cynical and arrogant employess questioning the validity of the new hires I can only surmise that I might be committing career suicide by joining the company (especially since I will be joining as a "dreaded" SDE/T). I am seriously considering rejecting the offer, although technically I have already "singed on the dotted line". And that would be sad, because there appears to be a lack of passion and belief in the company which is something we 10,000 hires would have injected into the company. Call it naiveté, but perhaps that is what Microsoft needs right now.

Also, do you think I want to join a work environment where I'm viewed as someone who was herded in to fulfill a head count? No way. Nice job of making the new guys feel unwanted before they even step foot into their roles. And while this might be viewed by you as a positive (one less "incompetent" to drag the company down), what you have done is driven away a 10 year veteran who is very passionate about solving problems and shipping quality software and believes in Microsoft and that it can continue to be great. I passed my Microsoft interviews just like you but I now know that I won't be viewed in the same light as a "real" Microsoftie. Funny, I experienced the exact same thing as a cadet at the Air Force Academy: every class above you thinks that the standards for admission were tougher when they applied. In reality, it's just a bunch of inflated egos that are unable to cope with the possibility that there are other people in this world that are just as smart as they are. Sad little person you are indeed.

Anonymous said...

RE: Just of our curiosity: does anyone think that the above Shane Kim (oy, I'm going to be repeating that name all night now) comment crosses the line?

I think the number one reason to not post opinions about VPs is that, by doing as such, you effectively ensure that they will never get fired, even if they are ineffective. They then fall into a "protected" status with which it would be a PR problem for Microsoft to fire them. Microsoft will only let go the VPs that either (a) do something incredibly stupid (a la Martin Taylor) or (b) can slip away unnoticed.

As an alternative, it might be more effective to post facts about how well a division or group is doing. For example, Division X spent $2,000,000 on something useless.

In sum: Easy targets are the first to go. Easy targets are nameless.

Anonymous said...

Look at the PM CSPs and tell me where it says that PM should know how to code

-
I am a PM in IE. All leaders in IE are PM. PM provides vision and idea. Dev writes code. PM has difficult job.

Anonymous said...

I've been trying to get to my spaces for 3 days and can't get thru. Can't view it and can't edit it. What the fuck are we doing? Do we have no testers? no SE guys? devs on vacation?

Why do we switch from MSN to Live? Is MSN having such a bad repo that we need to change to Live?

Do you know about staging? If things screwed up - roll the fuck back.

Or maybe all those guys from Spaces got promoted and the newly hired engineers are dealing with their mess? Is this the way to compete with other guys like GOOG?

I really want to hear from someone from Live or MSN about this bad transition.

c said...

Answer this (particularly if you're in one of "web" teams) - when was the last time you've seen a spec which you could say was not a waste of everyone's time?

I work in a web team (in the Office org) and our specs are usually very good.

Anonymous said...

The developers that make comments like fire all PMs, Testers etc are such a jealous bunch of infants that it is not funny. They somehow think everybody else is having a wild time except them. Or is it the 'Manager' in PM that they are envious about?

Everybody else should be fired because they don't code. Really?

Isn't it really a sad commentary on the quality of these Dev beings that these people don't understand the role of Technical PMs, Process PMs, Community PMs, Combo PMs (more than one flavor), STEs, SDETs, Product Planners, Consultants, Project Managers, UE, UA, Technical Writers, Design, TAMs, Consultants, Sales, Technologists, Executives, Support, etc, to software engineering?

Is this a failure of the colleges, the cubicles in MS, NEO or just plain lack of social skills? Should these people be this naive or ignorant?

Well the only answer to give to such uninformed developers is that none of these roles mentioned above cause bugs. Only Developers cause bugs.

Maybe that's what they should be proud of.

Anonymous said...

You had better hope I don't interview you because if it comes across you loved being a toady to a law breaking monopolist then you will definitely get a no hire from me.

So much for Google hiring the best and brightest, lol.

Anonymous said...

>> Look at the PM CSPs... blah blah blah

No shit, dude. I was freakin' shocked to find out what our PMs responsibilities are. They don't do a half of what CSPs say they should do. Now that CSPs are being rolled out everywhere in the company, I hope we get some 3.0 tough love on PMs who don't conform to the formal description of responsibilities at their level. In fact I'll be hugely disappointed if we don't.

David Notario said...

"Nope, the 1 gig storage was just something easy for the press to pick up on. What makes gmail stand out is that it is a browser email interface that does not suck. Let me know when Microsoft's stock webmail packages for exchange are similarly pleasant to use."

Have you tried Exchange 12's OWA? Feel free to drop an email to me or to any of the internal OWA aliases with your suggestions on how to improve the product.

Anonymous said...

I am a PM in IE. All leaders in IE are PM. PM provides vision and idea. Dev writes code. PM has difficult job.

STOP THE PRESSES!!

We can end the dev vs. test debate now. Turns out it's PM that's fucking us all.

Ok, continue pressing now...

Anonymous said...

On a completely different topic, our SVP just pushed down a CURVE on Commitment Ratings for our review model, contrary to everything being touted about myMicrosoft. We're being 'instructed' to keep our 'exceeded' ratings under 50% irrespective of how people actually performed with the result that many people who bust a gut this year and deserved an 'exceeded' are now going to get an 'achieved'. What a sham! I just lost my remaining bits of optimism, loyalty, and faith in the system...

Anonymous said...

when I look on this blog and see a bunch of cynical and arrogant employess questioning the validity of the new hires I can only surmise that I might be committing career suicide by joining the company (especially since I will be joining as a "dreaded" SDE/T).

And how many comments here are "plants" from an OSS crowd that has no better way to spend their time? You were hired for your expertise, trust me. Microsoft has its challenges but its still damn tought to get hired here. Congratulations.

Anonymous said...

I'll tell you what the 10K of us that are starting are doing: we're after your job. And maybe we won't be arrogant jerks to the ones that come in after us. Why don't you go get a job somewhere that you like?

You bet, that's the spirit! You're broke, smart and hungry. Knock those arrogant jerks back on their heels. By the time they figure out why you were hired it will be too late. Damn straight.

Anonymous said...

It seems reading comprehension isn't in the CSPs for PMs either. I don't think anyone suggested that all PMs be fired, just that we have too many of them. It's what known in the business world as "overhead." You need some of it, but not nearly as much as Microsoft has.

Anonymous said...

Only Developers cause bugs.


You can cause plenty of bugs with stupid specs. Did not learn that in college?

Anonymous said...

Mini, it's that annoying open source developer again. You asked:

"Interesting that he brought up a worry over developers and how he sees developers gravitating towards an Eclipse / Open Source environment. What do we do to get developers excited and engaged?"

Firstly, cut the crap in the Platform SDK license, about just when and where one is allowed to use it. If I need highly specific Windows SDK information to write a GPLed piece of Windows software, you don't want to tell me to eff off, do you? Or do you? Under the current PSDK license, I'm not allowed to use that PSDK to write GPLed Windows software using it.

Secondly, consider the Visual [programming language] Express series. (The Express series caused quite a sensation.) I've found out that each [PL] Express requires and incorporates its own IDE. I think, what a waste. How many megabytes does each Express IDE take up? Do I care for the official reasons? All I see is bloat. And that makes me think that the rest of the Microsoft offerings are going to have the same sort of bloat. Bloat kills cattle, if you really want to know. You would know more about bloat and companies ...

Thirdly, how much of the Visual [PL] Express is truly private Microsoft-only Intellectual Property? And how much of it is budgeted to bring in Cash-On-Delivery payments, according to the old way of doing business? It seems to me that it's intended to be a loss-leader.

Well, take it all the way, then. Microsoft has got a couple of useful Shared Source licenses that are recognized - by the Free Software Foundation of all people - as being Free and Open Source Software licenses. Release the Visual [PL] Express source code under the Microsoft Community License, including it in the downloadable cdrom images. Microsoft would need to make a few things clear though - any "fork" would not be able to use the "Visual [PL] Express" branding; Microsoft would appreciate bug fixes; bug fixes would make bug fixer a place in the credits; etc.

Well, you asked, and I have answered. If you can't beat 'em, join 'em. You're welcome to dump this whereever you please.

Anonymous said...

PMs...blessing or curse?

The real answer is neither. I came to MSFT as a PM after more than 10 years in the industry, as a dev.

I have worked with old time NT devs and newer ones. The role of PM has shifted over the years, and as process taxes are being added within the different groups, PMs have taken on the brunt of that work. In my previous teams, the technical PMs took this on in order to free up the scarce resources of their dev and test counterparts.

One cycnic observed that although the complexity required to ship code has grown tremendously, SWI, specs, scenarios, personnas, etc...let alone the LH basics, we seem to imagine that all this extra process can be accomodated without extra process bearing headcount.

To those of you working in orgs with PMs that you believe have no value, I think you are part of the problem. If you get PMs onboard, have low demands, and do not know how to grow the, then you get what you pay for. I have managed multi-disciplinary teams across the company and generally found a 1:3:3 ratio effective as a guideline (PM:Dev:Test). Some times we hit 1:4, some times 1:2. It depends upon the deliverables, the amount of coordination, coding, etc...

The important thing is that each job discipline has a place at some stage of the product lifecycle, and if you fail to realize that, then you have a deficiency in your understanding. If you comprehend it but are arguing that your particular PMs suck, grow them!

Anonymous said...

Whatever. As I indicated in my followup, I'd really like to hear something positive (and substantiated) about Kim's performance at MGS to address the many, many doubts I have about his ability and commitment. So far the closest anyone's come is attempting to shoot the messenger.


Not sure if it was a personal or not. (I think it happened about 100 comments ago :) I will say ... any correlation you can make between some negative in one manager's style that is broadly held as a 'best practice' by multiple managers at Microsoft is worth pointing out. I'm sure LisaB has someone on her staff watching this blog, looking for ways to eradicate that sort of inbred and destructive behavior (product of a monopoly) which is simply not going to serve Microsoft well going down the road. Its one area where personal attacks are less useful than generalizations.

Anonymous said...

I am a PM in IE. All leaders in IE are PM. PM provides vision and idea. Dev writes code. PM has difficult job.

Someone eariler posted that PMs are representing customer side in software development cycle. You wrote how PMs provide 'vision and idea'. Why then we here all the time comments like this 'Is Microsoft purely incompetent and tone-deaf to customers — or simply counting on IE's non-compliance remaining a de-facto standard' (Taken from here)? And what's the 'vision and idea' behind IE7? I don't see much innovation there, to me it looks like a copy of the Firefox.


Is this a failure of the colleges, the cubicles in MS, NEO or just plain lack of social skills? Should these people be this naive or ignorant?

Well the only answer to give to such uninformed developers is that none of these roles mentioned above cause bugs. Only Developers cause bugs.


Hmmm if i remember well my classes from Software Engineering the most expensive bugs to fix are spec bugs (yes, there could be _spec_ bugs) found late in the cycle (becasue cost of fixing the bug grows exponentially with time between 'creating' the bug and finding it). Just take a look at Vista. Mr. Ballmer promised that this won't happen again, but with folks like you on board i really doubt it ...

Anonymous said...

I am a PM in IE. All leaders in IE are PM. PM provides vision and idea. Dev writes code. PM has difficult job.

Horseshit. I am a Dev in IE. Many of our PMs are useless, but none write as badly as you do.

And if one of my PMs had a vision for the product, I'd probably fall out of my chair in shock. Sorry Tony.

Isn't it really a sad commentary on the quality of these Dev beings that these people don't understand the role of Technical PMs, Process PMs, Community PMs, Combo PMs..

In order:

"Technical PM" is, by and large, a joke. But the small handful of true Technical PMs are the only PMs I'd keep around. They are worth their weight and then some.

We don't need process monkeys. Good god we don't need process monkeys. A "Process PM" is what a failed "Technical PM" becomes. Just look at the processes we have in place today and remind yourself, "Process PMs" created them. And they suck.

"Combo PMs"? Huh? Are these the drivers of integrated innovation? If so, I'd rather have the process PMs.

Test? I love Test. I wish we had more testers, paid them better, and gave them more respect. I wish Test was more integrated into product development, and had better career paths. I wish we promoted more people from Test into GM positions. If we fired the PMs, we could do this.

Sales, TAMs, UA, and the rest of your laundry list. Nice straw man. Who was calling for the sales guys to get axed? TAMs do a critical job. etc.

Sorry Bub, it's not about arrogant devs thinking the world revolves around us. We don't want to get rid of Test or Sales or the cafeteria staff. We just want to get rid of process PMs. Combo PMs, and the other make-work PMs, because y'all take more than you give. We have Dev Managers and Dev Leads. We don't need non-technical flyweights pretending to manage the Dev staff. The company would be a better place if we recylced half the PM headcount into Test.

Anonymous said...

>>My resume is already in to GOOG. After two gold stars in two years, and a small opportunity to change the world that I savored as much as any challenge so far in my career, I'm asked to be a gear in a very large machine. I love this company (TM SteveB), but I won't accept a slow-track position to stay in it.

>You had better hope I don't interview you because if it comes across you loved being a toady to a law breaking monopolist then you will definitely get a no hire from me.

There mini, dare you to let this one through. Sure, it's harsh but it's the truth and if you can't deal with the truth, what can you deal with? But even if you do blackhole this honest comment, at least you read it. Think about it.

The poster talks about "a small opportunity to change the world".

Did it occur that this might simply mean pride in the ability to exert technical and cerebral abilities? I'm sure it did, but you couldn't resist a gratuitous snipe, it seems. Don't forget to refer to "M$" and "Windoze" in future posts for added cred ;-)

Not sure what you folks are smoking over there, but you can keep it to yourselves.

CorpMonkey said...

I am a junior PM in a team with roughly a 4:3:1 Dev:Test:PM ratio. We work on an enterprise application. For folks wondering what a PM does; I offer some anecdotal evidence - my observations of our best level 59 (Jane Doe), and our worst level 63 (John *annoyed grunt*). Your mileage may vary.

Jane was a fresh college hire from a public university with a moderate level of coding knowledge. She brought with her a strong desire to accomplish, and an aggressive personality. Jane is responsible for driving two of our 5 core feature teams, consisting of a total of 8 mid-level developers and 6 low-level testers. With the help of a technically strong dev lead, she was able to pick up scraps of early design and drive to a define all of the functionality needed to satisfy all key customer requirements. Jane worked 1-1 with executives from our largest two customers, and provided them with guidance on how to best use our product for their needs. She even implemented a significant portion of the functionality for one of them.

On any given day, Jane can be found doing any combination of the following:
- Designing and documenting functionality for upcoming milestones
- Driving the investigation and resolution of day-to-day issues
- Driving the triage of incoming bugs
- Shielding her teams from unnecessary distractions; representing her teams at managerial and cross team meetings
- Mentoring the new testers and walking them through the product functionality and the related business scenarios
- Delivering keynote demos of the product at our largest trade shows and conferences
- Helping Customers, Sales, and Marketing teams debug issues with the beta software over the phone
- Working with dev to investigate and prototype new features
- Configuring servers for various executive demos
- Meeting with peer product teams to define SLAs for interop features

Jane’s typical day at the office starts at 8am. She often leaves after midnight. Some of these hours are spent managing the chaos she sometimes creates with her own inexperience; most are heads down quality work. It is always a pleasure to work with Jane.

Jane reports into John. John is an industry veteran with 5 years internal, 10 years external experience. In his “15 years of industry experience, he has developed an intimate knowledge of the problem space and customer needs.” John is a lead responsible for managing Jane and a former co-worker (Bert). He is also directly responsible for 1 of our core feature areas – a team of 3 mid-level developers and 4 mid-level testers. John drove the creation of the feature area from scratch, and to date, no functionality has been delivered. Instead, a set of APIs to “enable others” to develop the features was delivered. While this work may eventually prove useful, he so effectively shielded his team from external interaction that all of our other core feature areas ended up having to code up their own versions of what the core feature area was supposed to offer. Every one of the original members of his feature team has since left.

On any given day, John can be found doing any combination of the following:
- Delegating key investigations to Jane
- “OOF until noon on personal business”
- Punting interop. feature requests back onto the other core teams
- Approving SLAs drafted and negotiated by Jane
- Holding ‘brainstorming meetings’ for already implemented functionality
- Mentoring Bert by shadowing him in meetings and interjecting occasionally to show “how a lead PM should respond”
- Collecting and reporting on Bert’s status

John’s day usually begins at 8:30. His day always ends at 4:30. If you need to reach him after that, he is fine with giving you Jane’s office number and personal cell number. John has been publicly recognized by executive level management for his great efforts in driving Jane to secure and deliver on critical SLAs.

Talking with John, he is quite happy with his job and the work he has done to date. He is openly looking forward to another promotion again this year, and is happy to offer advice to his directs about how he achieved a 4.0 last review season.

Recent discussions with Jane show she is looking at opportunities outside Microsoft. She has been with the company for just over two years, and is quite unhappy with her manager and the lack of recognition to date. She has also looked at other teams, but is afraid that her average 3.5 LRA (“Welcome to MS” 3.0 + “good job” 4.0) will limit her options.

I suspect that many of the posts against PMs on this blog are because of folks like John. For those of you who have worked with a Jane, send them some lovin’ – a good PM is a terrible thing to waste!

CorpMonkey said...

Only Developers cause bugs.

I gotta call BS on that one. Bad planning and leadership cause 10x more bugs than any developer could willfully cause.

Anonymous said...

Isn't it really a sad commentary on the quality of these Dev beings that these people don't understand the role of Technical PMs, Process PMs, ........, Technologists, Executives, Support, etc, to software engineering?

No, it's sad you think this is the only way to organize a software company. You're guzzling the Microsoft kool aid. Most companies have far lower test:dev and PM:dev ratios. It really makes sense. If you were running a small company and didn't have money to burn, who would you hire? Self-sufficient devs who can design their own features and test their own code, or PMs, who can... well, what IS a PM supposed to be able to do that a good dev can't do himself, anyway?

Anonymous said...

"Only Developers cause bugs."

As much as I would like to agree, I have to admit that as a tester I have played "dev yo-yo" when the spec was vague or missing.

Yo-yo rules:
1. Find a vague area in the spec.
2. File a bug causing a change in that area.
3. File another bug that forces the dev to reverse the previous change.
4. Repeat until the dev gets pissed off and writes a spec or until all of the vague details are defined by bugs.

Anonymous said...

So here I was trying to get a comment on the Cutting Room Floor and end up quoted on the main site in my own post instead.

Sorry Mini, I got a little hot under the collar there. The guy trying to impersonate an IE PM sent me over.

Attempting to be productive this time, the problems with PMs start when someone running process or doing cross-group collaboration slips into thinking they are the actual team leaders, even though they aren't doing any of the actual engineering work and are nothing more than glorified secretaries. Suddenly there's lots more process and more cross-group collaboration and less engineering. The solution to that problem is two-fold: first, insist that PMs have real engineering work-product (specs, designs, usability reports, competitive analysis, scenario story boards), and second, have officially designated team leaders who do cross-discipline management at a much smaller scale than GM so PMs with no pressing work commitments are not tempted to abrogate the role for themselves. Those team leaders should be drawn from Dev, Test and PM, with no silly prejudices that Devs or Testers "don't make good managers."

The problem with Dev is when devs fall to either side of the sweet spot of engagement. Devs that are unfocused (overloaded with work requests, time-sliced onto too many sub-projects, working on grossly understaffed projects) write terrible code, and do it inefficiently. Devs that are overfocused (in their silo, not aware of how the technology they work on fits into any customer scenario) make nice shiny bits of code that are useless and don't connect well with anything or anyone else. The solution to this problem is better feature discipline and an expectation that Devs understand the scenarios their work contributs to. Conveniently enough, the solutions to the Dev and PM problems are complementary. PMs who know they have large personal (as opposed to team) deliverables for every feature will be less inclined to the feature bloat that defocuses devs.

The problem with Test is when monolithic test orgs disconnect from the feature teams actually delivering product. Testers (correctly at the moment) believe they have limited career opportunites in actual product development, and so create career paths outside the actual product development. Huge boondoggles (WTT) and insufficient actual product testing are the result. The solution is to create a smallish service organization for labs and test automation, and to raise the status of Testers in the feature teams by giving them bigger strategic roles and promotion opportunities in product development.

Anonymous said...

I don't understand why we have people whose only job is to test software. Granted it has been a while since I was an OS dev but we managed to release very reliable software without test specialists or much in the way of dedicated PMs. I think we had one for the whole operating system. When I read Adam Barr's book Proudly Serving My Corperate Masters and found out how MSFT developed software it scared me silly. I hope things have changed since he wrote that book but unless they have changed a lot I think that explains why Vista is so late.

Anonymous said...

"You had better hope I don't interview you because if it comes across you loved being a toady to a law breaking monopolist then you will definitely get a no hire from me."

Thou should love one's employer. That could only be considered positive in my book.

Do not worry about this huy. If I find he/she messes up your interview with us, I will make sure he/she is disciplined. We ofcourse needs people from a monopolist who have learnt their lessons. We ourselves heading to be become one.

Two gold stars from Microsoft - impressive! Most our interview candidates from Microsoft have no stars.



Thanks for the props, guy (or gal).

It can be said that, in the area in which I work, we are playing catch-up to GOOG (and YHOO and others). And yes, we've learned at least MOST of our lessons on things like tying, pricing agreements, API disclosure and documentation and such.

No matter how well-intentioned MSFT is, there still a person/segment of a chain of command that periodically comes up with some great idea that croses the line. Most recently, I believe the person who caught public flack for something was a relatively new FTE who wasn't well-versed in how NOT to conduct business as a convicted monopolist; they hadn't realized that conduct appropriate anywhere else in the business world was not appropriate here. They probably had a chair thrown at them in private... the last thing this company wants is for that type of behavior to reappear. Microsoft, at least the parts of it that I see, is trying to be a better corporate citizen.

And thanks for letting me know that ex-MSFT are welcome at GOOG. From that last guy's post, my impression was, "sounds like someone who makes decisions on religion rather than technical merit and business case." I was beginning to worry that it could be a step backward career-wise to consider GOOG if that was indicative of my potential coworkers. Glad to hear rationality prevails.

A. Nonymous IV said...

This 10K increase in headcount may not be as bad as you think; we've had several acquisitions over the last year, each of which brings its own set of happy campers to the mix. Any idea (Bueller?) how many are acquisition people vs. true new hires?

btw, I work at WinSE as an SDET, and love working with our PMs. Good folks most of them. Sometimes I think WinSE is the best org in all of MS, despite the fact that we just fix vs. writing new product, which would be more fun.

Anonymous said...

10,000 New Microsofties, more than half are in the Services Business in the form of consultants and application developers. Unfortunately, that is because we would rather hire in new expertise than invest in the people that have made this company great.

Regardless, I came across the article that really captures it for me.

http://seattlepi.nwsource.com/fun/mgg.asp?date=20060802

Anonymous said...

I'm curious... I'm an Escalation Engineer and as most of the EEs we are/were developers... but we develop small tools on our free time and spend most of the time debugging... sometimes debugging customer's applications without symbols or source code access. (scenario where we have most fun! :) ) I know we are a very small number when compared the ammount of EEs against SDE/SDET/PMs... My question is... how do PM's, SDE's and SDE/T's see us? (I bet most of the Microsofties don't know what is an EE :) )I'm just curious... :)

Anonymous said...

OK, Windows Live, what gives. After a year of 43% growth, suddenly it is shrinking by attrition.

Since when is shrinking by attrition a good way to manage teams? As soon as someone important leaves any smaller project, that whole project becomes a risk. This causes a cascade effect of other people wanting to leave. And, no, the rest of you can't go anywhere because, well, there is a headcount freeze.

Now the unrestrained growth of last year is looking really quite stupid.

Anonymous said...

Since when is shrinking by attrition a good way to manage teams? As soon as someone important leaves any smaller project, that whole project becomes a risk. This causes a cascade effect of other people wanting to leave. And, no, the rest of you can't go anywhere because, well, there is a headcount freeze.

Now the unrestrained growth of last year is looking really quite stupid.

>

You hired to meet your quota. You hired HR, finance and other junk. Bloat your budget with junk and now cut your devs.

Ex MSFT SDET (VS) said...

CorpMonkey,

You just described all the lead PM's I've ever worked with.

The "best" of them turned a 500 page API into nearly a programming language that was never implemented.

The hard copy of the spec was so big you could stop the heavy doors in building 40 with it.

Needless to say, only about 5% of that spec got implmented, and the 2 years of 12 developer's time wasted on it got thrown into the trash when the PUM found a sexier project to lead, and the covering fire dissapeared.

Anonymous said...

the problems with PMs ...bleh, bleh bleh, wah wah wah,.... with no silly prejudices that Devs or Testers "don't make good managers."


I knew it. Pure jealousy of the PMs is driving all this infantile drivel (an envy of the "Manager" in PM, as someone posted earlier).

PMs are natural born leaders (even the fresh outta college ones), who lead by influence and not by hierarchical management. That is why you don't report to PMs (PM is a very flat organization) and why Microsoft invented the triumverate core business unit of PM/Dev/Test. The absence of a reporting structure betweeen the three disciplines is to foster more sincere input in creating great products and to avoid the fear of criticism in consideration of possible review retributions.

Still everyone knows who the unofficial leader of the business is. It is the PM. In some cases, it can be the Dev Lead too if they possess skills beyond coding. When GMs or PUMs or other execs need to know anything about the product, they go to the PM. That is the leader, no usurping of the role - it is just that way. The skill set required for PMs are so diverse that sometimes the job posting clearly states that the team is looking for a superstar, someone who can walk on water. That is one of the reasons why it is more difficult to hire PMs. The interviews are tougher (not just a "reverse an array" kind of deal)

No doubt most of the Devs posting here are the low level 59 - 61s. To make matters worse the Dev org is structured hierarchically where the Dev Lead or Manager's word is law and low level Devs cannot challenge any decision or orders. They are buried in whatever narrow silos their Leads created for them. PMs on the other hand have free reign (with the requisite responsibility) to take their creativity to any level. I would be jealous too if I were a grunt Dev.

My advice to these low level Dev grunts? Keep working on your soft skills (interpersonal, communication - written and oral, presentation, personal hygiene etc) and your PM might give you a chance at other tasks that are not coding specific. Devs with these skills get to go to customer sites, conferences etc as a reward.

But if you don't recognise your natural leader (PM) because of whatever reason, chances are that you will remain a grunt for a long time.

The above is spoken with Love.

Anonymous said...

The 9:50PM comment above (DATE THE BLEEPING POSTS) about PM's being lords of the universe, is a troll. If authentic, the company is toast.

Anonymous said...

My question is... how do PM's, SDE's and SDE/T's see us? (I bet most of the Microsofties don't know what is an EE :) )I'm just curious... :)

As a Redmond SDE who sometimes does SE work (IE team), I see you as awesome, kick-butt contributors. If every issue on my plate was as well-researched as the ones I get from EEs, my job would be a piece of cake. Keep up the good work. Our customer sat is rising, and you deserve a lot of credit for it.

Anonymous said...

Sorry Mini, I got a little hot under the collar there. The guy trying to impersonate an IE PM sent me over.

Attempting to be productive this time, the problems with PMs start when someone running process or doing cross-group collaboration slips into thinking they are the actual team leaders

---------------------------
It is ok to follow leaders in cross-group collab or in process without being resentful. All devs will grow faster if they are open and respectful of their leaders.

Anonymous said...

Test lead here. At review time we do lifeboat drills only within each discipline.

In a lifeboat drill involving the whole product unit- Dev, PM and Test, how many PMs would you jettison before touching a Dev or Tester?

Anonymous said...

The skill set required for PMs are so diverse that sometimes the job posting clearly states that the team is looking for a superstar, someone who can walk on water.

PMs might be superstars, but problem is we don't need *that many* superstars reporting status to PUMs and being "leaders of the business". We need good dev and test grunts to ship quality software.

1:2 PM to Dev ratio is a ridiculous amount of bloat, even if all the PMs are able to walk on water.

Anonymous said...

Re: CorpMonkey's Jane/John analogy

This is so true!!! I would like to add that this does not apply to just PM positions, but elsewhere at the company.

And, following the "rules", we all know what will be the fate of Jane if she dares to ask for permission to interview. If she asks before her evaluation for FY'06 is final, she will get a less than stellar review so that John can keep her around a little longer. If she (wisely) waits for permission to interview until after her evaluation, she had better get the job or her career at Microsoft is over, unless John is miraculously run over by a bus. In the meantime, her only real option is to schmooze up to John and pretend like everything is just swell.

Now, if Microsoft would just get rid of the permission to interview on the other hand...

Anonymous said...

"PMs are natural born leaders"

LOL! If that was the case, we probably wouldn't be having this discussion.

In the last few years I have seen some great PMs, who really thought through a vision of what should be done during a release, met with customers, drove getting actual pain points from customers incorporated into feature specifications that were executed on and delivered.

I have also seen PMs who you see generally sitting quiet at meetings "Not sure, I'll find that out...", and report status to each other about all the nothing which is happening every day while the component quality and usability languishes.

I have also seen PMs who scream obscenities in a most unprofessional manner when they feel threatened by their own incompetence rather than discuss the situation like a rational human being.

Every single person I describe is still employed at Microsoft to this day.

Yes, there are indeed some natural born leaders, but I sure wouldn't generalize that statement to all PMs at the company.

They are people with personalities, just like everyone else and their value to the company should be assessed according to their actual and potential contribution, not which cabal they are a part of, or if their component's dev lead saved their butt by taking up the slack and making the component successful.

The negative feedback towards PMs you read about here is NOT directed out of "jealousy" towards your "natural born leaders", it is directed at the RIF-raff who don't pull their own weight and the product suffers as a result.

Those times where I have seen components with tight pm/dev/test collaboration, the difference in execution was impressive to say the least.

Microsoft should continue to explore ways to adequately resource component development using smaller more tightly linked pm/dev/test teams who are all accountable to each other and rewarded as a group for deliverables of the highest quality on time.

One can dream right? :-)

CorpMonkey said...

You just described all the lead PM's I've ever worked with.

That is quite unfortunate. Aside from "John", I have also had the pleasure of working with at two rock star PM Leads in my short time at the company. It is incredible to see these folks in action because they have an incredible ability to energize employees, customers, partners, ISVs and still manage to deliver quality content for the core engineering teams.


... But if you don't recognise your natural leader (PM) because of whatever reason, chances are that you will remain a grunt for a long time.

The above is spoken with Love.


Whoever posted that comment needs to get over themselves. The PM title doesn't grant mystical leadership qualities. If it did, I'd fear for the social well being of some of my less-than-stellar peers if they were to lose that title. Nor do soft skills make a leader, despite the high value we place on said skills.

Leadership is demonstrated by earning the respect of your peers through your actions. Leaders are those folks who risk failure to stand a chance at success, and inspire a team to move forward despite barriers. My favorite leaders are those folks who encourage others to try and exceed their own expectations of success, and then go out of their way to help them achieve.

I've seen leaders in all disciplines at Microsoft regardless of how many times 'lead' or 'manager' appears in their title. Unfortunately, I have also seen those very same leaders actively blocked from making progress out of fear or jealousy - especially by folks who felt a successful 5 hour interview loop was irrefutable proof of leadership ability. The review system of recent years had actually encouraged that behavior with the result being the mediocrity and bureaucracy celebrated today.

Prakash said...

I see a lot of posts on the role of PMs and their value in the whole scheme of things. I was a dev at Microsoft for 5+ years and then moved into PM. I have shared my experiences at http://prakashsayeth.wordpress.com/2006/07/14/program-management-at-microsoft-demystified-part-1/ and hope that it might be of some value. I pretty much felt the same way about PMs as many of the folks on this forum do...until I actually moved into the position of one.

A better description of the PM role can be found at http://blogs.msdn.com/techtalk/archive/2005/12/16/504872.aspx.

Hope this helps,
Prakash

Anonymous said...

To the guy who posted the "PMs are natural leaders" essay:

Do we need 1 "leader" for every 2 devs?

(note I'm not even counting Dev Leads, PUMs, etc.)

Drei said...

To the commenter at 9:50 pm, 6/8/06: tsk, tsk, tsk. Dude, if you have a failproof system for detecting "natural born leaders", make the world a better place and publish it.

Back on track (BTW, I'm slightly disappointed that the one time this sensitive nerve of dev/test/pm has been officially touched by Mini, it's in a post with comments disabled, and quickly shadowed by the departure of a high level, highly wailing doofus ^H^H^H pardon me, I meant "superstar")..

To spare you reading my own sob novel, I declare that test devs are unequivocally 2nd class citizens. I'm afraid it will remain that way for as long as Microsoft produces software. (That is, produces software the "traditional" way.) When the lines between disciplines will blur, so will the stigma. But such blurring carries the cost of reduced "measurability" and apportioning blame - meaning assigning responsibilities. Which is clearly not the direction we're apparently headed.

The drivel starts here: I'm Anonymous, and I've been an SDET for my 7+ years in MS. I've had ups and downs and, I think, ample opportunities to analyze my track record, my future path and compare that with those of my peers, co-workers and friends. The main metric I use on myself is the respect of those I work with, and, as seen in various comments above, it does seem some devs are certainly respecting their testers, as well as wishing them well (better career plans etc). However, the simplest of tests of this respect is having the tester disagreeing with the dev in a relatively public forum - be it in a meeting with the PUM, or on the same thread with a 3rd party entity. Set this up and feel the sense of unity building a priori in support of the dev. The universal acception is that testers are really, really well informed users, users that cheated by reading the source code - and little else. It might be/is true in the majority of cases, which further proves the point of _all_ testers being pariah. The selection level (for the same group) is different between these two disciplines. If you said "but it's elementary/perfectly normal" to that, then you're sharing this view and thus, promoting it.

The day this stigma will disappear will be the day when the hiring decision is taken for the entire component unit, with interchangeable roles assigned on aptitudes. In my utopic vision, everyone (dev + test) working on a component should be able to exhibit similar coding/problem solving skills. Are your design chops/experience a bit more prominent? You get to "architect" in addition to coding. Are you constantly looking for flaws? Then, in addition to coding, you will try to destroy it. Pair programming is exactly that - a dialogue of proposals countered by constructive criticism. (I wanted to add here something disparaging about PM roles, feel free to insert your own nugget.) The team is only filled when all the roles have been assigned. Certainly, there are abundant reasons for which this method is flawed, but it looks to be the only one that prevents the onset of the caste system.

As it is, ascending to higher levels in test (simply compare the ladders for 59 -> 60 for dev and test) is slow(er), more of a function of chance and more biased towards non-technical criteria. (For instance, my (test) group has a policy of 2 years/level regardless. Make that "at least 2 years", lest you understand that everyone is promoted when their level clock beats 24 months.) Past L62, the ICs path ends abruptly, especially for those not born leaders naturally. Please keep in mind that, in plenty of cases, the technical abilities, the execution capability and comprehension of a test dev is no less than that of his/her friendly peer dev. Au contraire, it's quite possible that the tester has a feel for problem areas, possibly making him/her a bit more careful of a developer. Where I've studied computer science, there were no separate seminars for future devs and testers, nor were there different assignments given/classes taught/grades awarded.

Lastly, in case of certain components, the test developer needs to understand the space in more detail than a dev that possibly translates calls from an API to an underlying driver, because of the need to develop tests assessing standard worthiness. I'm willing to bet there are plenty of test teams around Windows that emulated drivers/devices fully for their WHQL/WDK certification tests. (In support of this idea, I also refer you to the latest talk announced by Engineering Excellence, which features quote from our own Adam Barr - MS will enter an era of the tester, whereby testing requires more knowledge/equal skill as developing a high quality piece of software.)(Other than that, the parallen between software engineering and environmentalism seems more mildly amusing than useful.)

Given all that, an IC tester simply reconciles with the status quo and remains an under-leveled worthy target of dev/PM lip service, moves to dev or embraces his/her managerial side. The first option fosters cynicism and the following two result in a loss for the component.

The testers in Windows have been hit particularly hard these past two years. The introduction of half baked (if well meaning) processes, tools and protocols caused a great deal of effort to be spent only to integrate these new concepts, while the product itself remained largely untested. I've spent more hours than I'd like to admit in public making crusty old tests work in the confines of new frameworks, for an overall gain of nothing. When the warts of the unnamed hurdles were tamed, bugs sprang like daisies, and were either punted vigorously or overloaded the devs against tight deadlines. Ripple effects are slowly subsiding. Various PMs, driving status meetings and connecting inter-disciplinary coworkers gained a lot of visibility in the process (load thy browser with the list of Vista heroes), as did (rightfully) the devs who fixed the damn thing. The grunts... Which is too bad, 'cause the effort and the hours were on par.

I'm nearing the end of my initial cynicism supply, so I'm looking forward to transitioning to dev after Vista. I'm positively certain the other side features better saturated colours.

Anonymous said...

>> Still everyone knows who the unofficial leader of the business is. It is the PM.

This is so ridiculous, I don't know how to reply even. Mark my words, "natural leader", when it will come to pink slips (and it will), you, "leaders", will receive the first round.

Anonymous said...

Mark my words, "natural leader", when it will come to pink slips (and it will), you, "leaders", will receive the first round.

Thereby showing that they are leaders by leading the way out the door...

(Sorry, couldn't resist.)

I actually think that the "natural leader" post was just a well-done troll. If not...

Leaders lead by getting people to want to go where the leader thinks that they need to go. "I'm the leader, because I'm a natural leader, whether you realize it or not" means that you're not a leader. All the pride and arrogance does is make you act like a jerk - and your post shows it. It certainly doesn't make people want to go where you want them to go. If anything, it makes them want to sabotage you, just as a way of sticking an upraised middle finger into your assumed superiority.

Harsh words, but I'm genuinely tring to help you see what your attitude really does. If you actually have any genuine leadership ability, the arrogance strangles it.

That's assuming that the post wasn't a troll. If it was, my compliments on a really well-done one.

MSS

Anonymous said...

I spent 5 years in devdiv as a SDE and SDE lead, 3 years as a PM, and now a couple years as a Dev, so I think I can offer some perspective on the different disciplines.

But I'm not going to.

Sure, I can tell you which job I found harder, but I think that's besides the point.

The problem with the way the majority of the MS teams run is that they are not teams, they are jammed together groups of devs, pms, and testers. Each discipline has their own charter, their own goals, and is rated towards those goals rather then the success of the overall team.

There's a ton of "not my job" going on. My job is to write the code, not to make sure it works. My job is to test it, not to work directly with customers. My job is to write the specs, even if they can't be implemented.

Anybody arguing for the primacy of one discipline over the other is part of the problem.

Let me clue you in. If your product is late, too slow, doesn't meet customer needs, or gets bad review, it's *your* fault.

Get off your high horse, figure out what the customer really needs, then work as a team to build it and ship it.

Anonymous said...

>PMs are natural born leaders (even the fresh outta college ones), who lead by influence and not by hierarchical management.

I would find that more credible if I hadn't just spent the last few years watching PMs spec out ill thought out products slammed together under the excuse of time-to-market pressure that none of our cusotmers seems in any hurry to buy.

Most non-technical PMs are as replaceable as paperclips and serve a similar purpose: to keep the paperwork from getting out of hand for the people who are doing real work. The good ones serve as liasons between the team and the rest of the organization and dedicate themselves to making things run smoothly. The bad ones operate under the delusion that the project is their own personal fiefdom for them to display their own self-aggrandizing achievements, like dubious innovations or completion on unrealistically short schedules.

>No doubt most of the Devs posting here are the low level 59 - 61s...

If it comforts you to believe this, you may continue to do so. Personally, though, there is a quote that seems rather appropriate here.

Now, if you'll excuse me, I have some appointments to kidney punch each of the "natural leaders" who blithely assured me two years ago that neither Longhorn support nor 64 bit support would ever be an issue for our product.

Anonymous said...

These are the words from our leader Steven Sinofsky. It shows PMs dont have limit to their potential where as dev has limit. How many UI controls can one write?


Up front I would say that the PM role at Microsoft is both unique and one that has unlimited potential -- it is PMs that can bring together a team and rally them around a new idea and bring it to market (like OneNote, InfoPath). It is PMs that can tackle a business problem and work with marketing and sales to really solve a big customer issue with unique and innovative software (like SharePoint Portal Server). It is PMs that can take a technology like XML or graphics and turn it into an end-user feature benefitting 400 million people (Office Open XML format or the new graphics functionality in Office12). I could go on and paint a very emotional picture, but let's spend some time digging into the more analytical aspects of the job.

Where developers were focused on code, architecture, performance, and engineering, the PM would focus on the big picture of "what are we trying to do" and on the details of the user experience, the feature set, the way the product will get used. In fact the job has matured significantly and it is almost impossible to document a complete list of the responsibilities of program management

Anonymous said...

Of course, Sinofsky's words tend to favor PMs. He had this to say about himself:

"I think that was a good move for me because while I like to think I was competent at development, I was probably never going to hit it out of the park with my ability to write code, but I think what I wasn’t able to do in writing code I made up for with the skills required of a program manager :-)"

Anonymous said...

I\'m tired of all this praise for Kai-fu and XD by people who obviously have never worked with either of them.

Wow, why did you sue Google if Kai-Fu Lee is so bad? Do you have any better known speech expert than XD?

Charles said...

Mr. Sinosfky writes an elucidating piece on Microsoft's unique implementation of program management. Insofar as his elaboration goes on the generalities of why program management is good and the problems it solves, it is informative.

However, program management is hardly unique, not even in the software industry. PM was born of the matrix organizations in the aerospace industry in the '60s and '70s, and was adapted long ago by Microsoft's various competitors. They just haven't elevated it to the level of prominence that Microsoft has. But what makes Microsoft's implementation of PM unique is placing the "specification" responsibility on the PM, rather than on engineering. There is another anomally noted by another poster I'll touch on in a moment.

Mr. Sinofsky is certainly correct in that specification of features, functionality, interfaces, etc. must be planned out in detail and then implemented and tested, and someone needs to coordinate this across organizations. In the successful product development efforts I have witnessed that role is performed by an engineering development manager. The dev manager is tasked with producing a spec that satisfies marketing's function and feature set, and is technically feasible to implement within the engineering budget & schedule allocated. If not done by the dev manager himself, a PM-like person is usually staff to the dev manager to perform the coordination, communication, status tracking and facilitate status meetings and follow up on action items and problem resolution.

But unlike Microsoft, this PM responsibilty is part of (hired by and reports into) the engineering group, not marketing or some generic PM org/silo. The engineering manager is tasked and measured on keeping marketing, customer support, sales support, finance, etc all satisfied that their various inputs/requirements are being met. Succssful software product development is essentially (if not purely) a programming effort and those programming skills (commensurate with developing quality products) are in the engineering development staff. When tradeoffs are to be made or problems to be avoided, it is essentially programming experience and judgement - the more senior and seasoned the better, but programming experience, not management experience.

Which brings me to the anomaly noted by another poster:

The problem with the way the majority of the MS teams run is that they are not teams, they are jammed together groups of devs, pms, and testers. Each discipline has their own charter, their own goals, and is rated towards those goals rather then the success of the overall team.

There's a ton of "not my job" going on. My job is to write the code, not to make sure it works. My job is to test it, not to work directly with customers. My job is to write the specs, even if they can't be implemented.

Anybody arguing for the primacy of one discipline over the other is part of the problem.

Let me clue you in. If your product is late, too slow, doesn't meet customer needs, or gets bad review, it's *your* fault.


I wholeheartedly agree.

Again, in the successful development efforts I've seen, the difference from Microsoft is that the engineering manager owns the responsiblity and authority to deliver. It is his/her fault when that doesn't happen. It is his/her fault if marketing's feature set is shaved at the last minute. It is his/her fault if the product is buggy. It is his/her fault if the product is poorly documented, slow, bloated, unfriendly, late, costly, etc. yada yada yada...

The engineering development manager was given the resources to plan, develop, and ship the product and the authority to make the judgement calls along the way. They are also expected to keep their marketing, finance, support peers in the loop as the development effort progresses and the product content/schedule naturally changes (nothing is perfectly planned) but there is no excuse for surprises.

Program management aides the engineering development manager and normally, for large projects, one or more PM people report to the engineering development manager in this regard.

Again, in the successful development efforts I've seen, engineering is responsible for quality. The QA people may report into a different org, but their job is to find development's mistakes and development is also tasked with the job of making that easy as well as fixing bugs reported. *Engineering* means designing products that can be tested easily (as well as maintained and enhanced later). Testability is as critical a feature as any and needs to be designed-in from the start. If designed properly, the testing can be almost entirely automated, with test suites to be written by QA folk based on product specs as well as their experience.

The point of all the above is that project management as Mr. Sinofsky describes is indeed critical, but ownership of the resources, responsibility and authority to deliver the product even moreso, and the skills Mr. Sinofsky cites in a good PM are indeed necessary, but not sufficient. As product development is largely a technical problem, the PM issues at bottom
are largely technical in nature, their solutions consequently technical and the experience necessary to select correctly from a number of arcanely complex competing "solutions" is largely technical.

The distinction I'm drawing is that, yes the PM job nees to be done as Mr. Sinofsky describes, but it is a job that needs to be done from engineering, with engineering commensurately resourced and tasked to satisfy the aggregate requirements from other organizations.

If I had to "pinpoint" Microsoft's mistakes in this regard, I'd pick two:

1) Responsibility and authority to deliver a product is dispersed across the organization. The PM has responsibility to pull everyone together, but no authority. That works when exceptional people rise above their differences, but it doesn't work in general for average people, from which average people companies of 70,000 are composed. Yes, it's supposed to be cooperative teamwork, but teams still have captains or quarterbacks who call the plays, sideline (and armchair) coaching excepted.

2) Were one to casually browse Search Jobs - Microsoft Careers listings for program, project, product and devlopment managers, one can't help but note an appalling overlap of responsibility to "drive" design and direction with an even more appalling lack of responsibility to set budgets and staffing levels normally commensurate with the responsibility to "drive". Even more noteworthy is the absolute absence of any metrics the successful candidate ought to have in terms of skill and experience, for example 'managed annual budgets in excess of $1M and teams of 10 people or more and products of over 500 KLOC'.

Is it any wonder the people 'driving' Microsoft offerings are always stepping each others toes and yet don't seem to have more than a few years 'driving' and seemingly with little prior experience at budgeting, staffing, and leading?

Anonymous said...

On a completely different topic, our SVP just pushed down a CURVE on Commitment Ratings for our review model, contrary to everything being touted about myMicrosoft.

Please raise your hand if you're truly shocked by this.

Anyone?

Didn't think so.

I think the SVP had two choices: 1) Give lots of people "exceeded", which makes them feel better about the work they do, but a small raise
2) Give more people an
"acceptable" that will go together well with the incredibly tiny raise the budget allows.

Anonymous said...

These are the words from our leader Steven Sinofsky. It shows PMs dont have limit to their potential where as dev has limit. How many UI controls can one write?

Then he too will fail, if he believes in that and rewards people based on it.

Microsoft sells software. There are many, many roles to play in building that software, and very important contributions that come from the people we call PMs and Testers. But the Devs write the code, and the experienced devs write most of it. And the most important parts of it. If those devs come to believe they have no future at the company, they will move on to other employment, and the PMs will be left with incompetent or inexperienced engineers who simply cannot build the products the PMs envision.

At that point, the decent PMs will leave out of frustration and the company will indeed be run by "natural born leaders" such as our earlier troll, fighting like jackals with each other over the rotting carcass.

Know what you want to sell, know how you make it, reward the people who make it possible. Is that really such a hard concept?

Anonymous said...

You know, there are very few absolutes in the business world, but one of them is this: You get the behavior you reward. If you reward teamwork, creativity, and innovation, you get more. If you reward politics, empire-building, and emphasizing profits over customer service, you get more of that instead.

What does YOUR manager reward?

- Not an MS employee, actually a UNIX bigot who'll admit that Windows 2003 server doesn't suck

Anonymous said...

emphasizing profits over customer service, you get more of that instead.

--

We emphasize revenue over profits.

Anonymous said...

Dev vs. Test vs. PM vs. Managers

I wish that mini would open up comments on the topic. It is not so much that Devs hate Testers or that either is stupid or irrelevant.

The angst is a symptom of other problems.

1. Management levels go too high before Dev, Test, and PM reconnect to a common goal.

2. Scheduling screw-ups inflate inter-disciplinary angst. If Dev slips, dev goes on to the next project or maybe even has some slack-time while test is pressured to squeeze their schedule to make up the lost time.

3. There are genuine problems. Some PMs try to pass off horribly incomplete specs. Some Devs try to push the check-in window to the last second. Some Testers spend more time writing automation than looking at the product. Some Managers try to pretend that problems don't exist and hope that they will go away. Some people try to abuse the lack of oversight and see how long they can go without doing any work at all.

4. There is a technical descrimination within Microsoft. The more technical a person appears to be, the more he/she is acknowledged. People who appear to be less technical (MSN, Tech Writers, some Testers) are often marginalized or ignored.

5. Management is bombarded with requests for un-ending reports and presentations while they are forced to fight so many fires caused by other groups that quite a few things are always falling on the floor.

Anonymous said...

"Some Testers spend more time writing automation than looking at the product."

Well DUH!!! Where have you been? We're not supposed to test the product anymore: automation does that. So testers automate and test/fix the automation. Then there is automation to check the result of the automation and automation to record the result of the check of the automation. You gotta automate that too.

Thankfully there are many tools to do all that that, and as soon as they are debugged and work as advertised everything will float like a butterfly and sting like a bee. Ooooh yeaaaah baby!!!

It's a beautiful system. Sure Vista will be a bit late but it won't have any bugs so it's all worth it.

Anonymous said...

"4. There is a technical descrimination within Microsoft. The more technical a person appears to be, the more he/she is acknowledged. People who appear to be less technical (MSN, Tech Writers, some Testers) are often marginalized or ignored."

Just out of curiosity, the subsidiaries from Latin America are the right opposite! If you are technical you have lower levels, status and acknowledge, if you are part of management, marketing, MSN, etc... you can have a great career. For LATAM the more technical a person appears to be, the LESS he/she is acknowledged. Mini this is a good subject for a new article...

Anonymous said...

But if you don't recognise your natural leader (PM) because of whatever reason, chances are that you will remain a grunt for a long time.

Dude , Respect should be earned but not given through a 5 hour Intervew loop. In my last 6 years at MS, I have worked with a lot of PMs who don't even know how to use their feature and these people represent the feature at GM\PUM level. There were only 2 PMs that I was truly impressed , both of them worked as Devs for at elast 6 years before they become PMs. We are Technology company and PMs should be technical.

«Oldest ‹Older   1 – 200 of 209   Newer› Newest»