Whew, holy smokes, talk about some busy commenters over the holidays... 175+ comments on that last real post regarding The Passion of the Googlers. If you find yourself with some spare moments, it's worth bringing it up and reading through the discussion. While some folks saw the 4am working late as the focus, MarkL followed up saying no, no, it's about the passion. How is the passion at Microsoft?
From my perspective: you treat brilliant people like cogs, they transition to cog-like passion. And passion certainly gets drained into the grayness of pulling a paycheck once they get clued into the grotesque difference in compensation and accountability between individual team members and upper management.
I've spent the time flying home re-reading about 350 comments (out of the 4600+ comments so far... please, Blogger, can we have a comment web feed soon?). Some of the recent interesting ones follow...
This comment certainly read my mind... from the middle:
Why can't a quick decision be made:
1 - Realize the 65% percentile pay target relative to the industry worked well when our stock options were splitting every 2 years, now it simply doesn't work. A change is needed.
2 - To afford the change, e.g. costs, something else has to happen - some cost cutting somewhere else.
3 - What's everyone complaining about? The fat cats, middle- and upper-management that are not incented to being risk-takers. They want to minimize the decisions they have to make to ensure they can stay at MSFT as long as possible. This the the best target for reducing costs.
4 - Biggest, quickest change: Offer incentive package for the "fat cats" to take early retirement. If they don't, make it clear their role, success measures, and compensation will be changing, i.e. they will need to start justifying their salaries. Not only make it clear, make it so obvious they will not succeed in the "business as usual, rest and vest" fashion that 80% of those offered the option of leaving w/the package take it. Then simply bump the 65% percentile to 75% or higher.
One comment from someone who used to work with MarkL ends with:
OK, now put MarkL aside. Look at Longhorn. Definitely not his fault. A lot of interesting bottom-up ideas that could make a great release. New file system, new desktop composition, sounds great. But that wasn't enough for the execs engorged on Powerpoints. DotNet Everywhere was the command - it's ready because it *has* to be ready. It's a Big Bet. You cannot question it... until 3 years are gone by and the Big Bet isn't paying off, and the top-down architecture is a piece of shit.
I do not think MarkL is 100% correct. But (apparently unlike most other commenters) I can see a true picture being drawn. At Google you can engineer your way if it excites you. Your peers do the reviewing. At Microsoft you have entrenched management controlling almost all visible initiatives. [That sad little live.com was probably just a small team doing something cool in a short time; now they have to carry the mantle for Ray Ozzie and a company wide intiative. God bless them.]
- New products need small teams getting work done.
- Coordinating with fiefdoms can be the death of small teams lacking in political talent (NOT engineering talent).
- Engineering talent is disgruntled and increasingly jumping.
- Key recruits are no longer at the IC level.
- The focus is on process because the other mechanisms of maintaining quality have failed.
- Growth has slowed and progression through the ranks is more political as openings are scarcer in the older businesses.
Another that worked with MarkL during the Hailstorm days posted their perspective on Hailstorm, including the following bit:
Hailstorm was a mess because it was trying to fix too many problems at once for people that didn't want them to be fixed. It was orginally trying to fix the online storage problem. It was trying to fix the proprietary interfaces and PR problem. It was trying to be the first web service. It was trying to get all the MS and industrywide apps to use the same data schemas. It was trying to make IIS based servers scale to millions of users.
Innovation is almost impossbile to sponsor at MS because whenever you want to solve a customer problem, you're stepping on the toes of three other products that are already offering something to that customer or selling something into that market.
I spent the last period of my employment with MBS, and that just summed it up and made it very real for me (so in many ways, thanks for illuminating that it was time to find greener pastures). Man, talk of a place that needs a very serious shake-up. I have never seen so many people and teams trying to f*** it up for each other. SAP and Oracle aren't the biggest challenges this unit faces - they are their own worst enemy, and until they get it sorted internally, MBS will never fulfill the potential, which is there. I am amazed that noone has done anything about it. Hopefully someone will, now that Doug Burgum has stepped down.
And finally, the MarkL poster loaded up both barrels when responding to the Robert Scoble criticism (and be sure you flip your sarcasm bit for that first paragraph):
Any finally Robert, Sorry about the mess I left behind, thanks for cleaning up after me. It was my idea to flip off the united states of america by giving them a version of windows that wouldn't boot, by telling the attorney general to go to hell, and by manufacturing phony demos. And yes, it was my idea to put our customers on a 3 year subscription cycle knowing that there was no way in the world that we would ever deliver them updates within their subscription window. Oh yeah, it was also my idea to crush all competition in the browser market and then once we one the war stop working on ours. Oh yeah, Vista is my fault too. Bill and Steve wanted to ship it 18 months after XP. I convinced them that an OS every five years is more than enough. Sorry, forgot about stack ranking... Me and a few other guys thought it would be a fun to give huge options and bonus awards to 1/3, give a cost of living adjustment to another 1/3, and then screw the bottom 1/3 with no raise or bonus. But agin Robert thanks for stepping up and offering to clean up this mess. I am sure that your non stop blogging (hey, how come your blog isn't on msn spaces?) is exactly the right way to fix this stuff. Shoot, if I knew you while at Microsoft, I would have voted you in as a partner and awarded you a 200,000 share restricted stock award. After all, as a partner, your contributions are certainly worth a cool $5mil right?
And just one small correction in your post. I was not a manager at Microsoft and I am not a manager at Google. No one has to kill themselves to get my respect. They just have to be smart and creative, willing to work the entire software stack from metal thur ui, and finally, do what they say their gonna do and do it in the timeframe they say their gonna do it in.
Another year. The pipeline is flowing. The MSFT stock went down 2% during 2005. Todd Bishop has his own stock comparison post today. I can only hope and pray (for what, the fourth year now?) that it's upward bound from here, though our leadership doesn't gauge success from our stock price.