Google's Young Brainiacs Go Globe-Trotting

theodp writes "To train a new generation of leaders, Google sends its young associate product managers on a worldwide mission. Newsweek's Steven Levy tagged along and reports on the APMs' activities, which included passing out candy, notebooks and pencils to poor Raagihalli children, a 'Rubber Ducky' group sing-along at 2 a.m., and competitions to find the weirdest-gadget-under-$100 in Tokyo. The APM program, which seeks brilliant kids and slots them directly into important jobs with no experience necessary, was formed after Google's attempts to hire veterans from firms like Microsoft had awful results. 'Google is so different that it was almost impossible to reprogram them into this culture,' says Google CEO Eric Schmidt of the experienced hires."

Comments

Hiring and capital expenditures

[The Clockwork Troll]

Newsflash:

When you've overspent on hiring and capital expenditures quarter after quarter, it's a no brainer to see that it's cheaper to hire a bunch of young, cheap talent and send them around the world to get them all gung ho and Mouseketeer-y about working 80 hour weeks, than it is to hire senior product management with families and less mental plasticity who turn in mediocre-to-decent performance 9-5 at a $150k base (almost 2x what these APM's are getting).

So what if the APM's fuck up now and then, when your raw productivity is 4-5x that of "adult" talent, you can afford the occasional product airball.

And the reality is they probably even fuck up less.

Re:Hiring and capital expenditures

[Aceticon]

When you've overspent on hiring and capital expenditures quarter after quarter, it's a no brainer to see that it's cheaper to hire a bunch of young, cheap talent and send them around the world to get them all gung ho and Mouseketeer-y about working 80 hour weeks, than it is to hire senior product management with families and less mental plasticity who turn in mediocre-to-decent performance 9-5 at a $150k base (almost 2x what these APM's are getting).

So what if the APM's fuck up now and then, when your raw productivity is 4-5x that of "adult" talent, you can afford the occasional product airball.

As a freelance software developer who often is brought in to clean up the mess which results in having overworked, inexperienced, bright (and cocky) young people designing and developing whole systems, i can tell you that the total costs (including maintenance costs and system improvements costs) of having a system designed and developed by these "cheap young people" far outweighs the savings you get from not including at least one or two experience persons in the team. And this is not even including hard to measure costs such as indirect business costs due to under-performing software (such as the ones you get because the system is 10x slower than it should be at doing time-critical, essential business functions, 'cause the guy that designed it didn't understand database indexes or thought that using remote calls in every layer would be "cool").

Now that i think of it, often enough, even before the project is delivered, the initial development costs when using cheap young people outweigh the cost of having more senior people in the project.

Unfortunately, mediocre managers often fall into the trap of confusing "hours worked" with productivity. Proper measures of productivity - such as: business functions implemented per man hour - actually require having things like requirements specifications and mediocre managers don't use tools like requirements specs ... or any other advanced form of project structuring or planning beyond pretty MS Project graphics.

And the reality is they probably even fuck up less.

Actually, for any piece of software which is in production for more than 6 months, they will keep fucking the support, maintenance and extendability of the software long after they've left the company.

If you're inexperienced:

• You never had to maintain any software so you will have no clue about how design and development decisions affect problem tracking and software extendability
• You will not expect the common sorts of improvement requests you get just after going live, such as the "monthly system usage report" that the Business Unit manager is bound to ask about 1 or 2 months after the system goes live.
• You have never worked anywhere else so you only know one way of doing things. You will not have experience in working in enough environments to know "theres a better way of doing X" or "if we do it this way we'll have to risk Y"
• You will not know on which parts is performance important and on which it is not important. You will spend time optimizing the speed of the monthly report (which takes 1 hour but happens once a month) instead of the data retrieval for the GUI main screen (which takes 5 minutes, and is done average once an hour, per user)
• You will go down design dead-ends and chose under performing technological solutions 'cause you blindly believed the industry hype (and forgot that vendors are in it for their own profit, not yours), only to find out one month into the project that because of that choice the software won't be able to meet agreed performance targets
• You will overextend yourself, going into overdrive and overwork mode early on and, due to being tired, introducing bugs and making wrong design and development decision which result in too much time being wasted in bug-fixing and back-tracking out of wrong design/implement

Re:No experience necessary?

[Broken scope]

Yes, and I can see why the hospital has older folks who can handle the project.

Re:No experience necessary?

[spxero]

While your boss may not know how to take out an AGP card, I'm sure he knows a heck of a lot about policies and procedures... specifically when it comes to user IT management. IT is more than just a field of working with computers- it's about working with users to help them and show them how technology can impact their jobs.

And while some of those people may not be in exactly the correct position, some of them are there (as you mentioned) because they can handle a project. They can't plug/unplug AGP cards, but they can make the system work well.

Here's my theory on Google's hiring...

[xxxJonBoyxxx]

Here's my theory on Google's hiring plan up until today.

1) Hire anyone who seems to have any technical talent, lives only for work and/or could be useful to any competitor.

2) If an employee is not part of the core search project, give them some random B.S. to do. Also provide benefits out the ying-yang so competing offers look silly. Just make sure the B.S. provides our minions with no useful experience, exposure to real-world requirements or any tools outside the Google universe. This way, if they do decide to leave us, they will be unable to set up viable companies on their own or provide any value to our competition.

3) If anyone from the core search project (our only source of profits) tries to leave, kill them.

...the APMs' activities, which included passing out candy, notebooks and pencils to poor Raagihalli children, a 'Rubber Ducky' group sing-along at 2 a.m., and competitions to find the weirdest-gadget-under-$100 in Tokyo.

Yeah...I still like my theory.

Re:Inbreeding

[gwern]

Yeah... it actually reminds me very strongly of Enron - because of their cult of talent, they had a similar program where the best and brightest were encouraged to transfer from disparate area to disparate area, regardless of how little competence they actually had in the new area. This Google program isn't identical to Enron, AFAIK, but I find myself wondering what other similarities there might be between the two companies.

MS and Google Culture... separated at birth?

[Mongoose Disciple]

I think this is sort of interesting (ironic?) because I'd say the corporate cultures of Google and Microsoft (at a developer kind of level -- not necessarily CEO etc.) have or had a lot in common.

I interviewed for a job at the Microsoft campus back in the 90's, before the dot com era made pampered developers more of a common phenomena. This is also before any of the MS monopoly suits -- the company just wasn't seen as an evil empire by most people in the kind of way it can be now. The whole first round of interviews was composed of logic problems and puzzles to test your ingenuity/creativity. They had a hell of a campus and all kinds of unusual perks I wouldn't see again until the dot com boom. It was pretty clear that their strategy was to try to pull bright people straight out of college, give them 'fun' and pampered environments, and basically work the hell out of them. Not that anyone would demand an 80 hour week from you, exactly, but more: you've taken this new job in a city where the only people you know also work at Microsoft, you see your job as something kind of cutting edge / geek-cool, you're provided with this office and cushy work environment and any meals you care to eat at the office (and their cafeteria was pretty much the best I've seen anywhere before or since, not that they wouldn't also order out as appropriate)... you're with this team of people all fired up about how great Windows 98 is going to be, and they're all working late, and maybe you'll just stay long enough to get that free dinner...

Anyway, damn near everything I remember from that visit and everything I hear about the interview process and corporate culture at Google today is very, very similar.

Does Microsoft still try to do this? I have no idea. Of course, time does strange things to a company's culture despite its best intent. I know a guy who took a job there out of school and lived that kind of culture; today he's still there, married (his wife also works there), is a manager, and has kids. Even though a guy like that may have worked under a very similar culture to modern-day Google for years, he's not going to be the same guy and he's not going to see that kind of glorification of young genius the same way. Most likely he's seen projects where it helped a lot but also projects where it went horribly awry, and his inclination as a manager is probably not going to be to allow everything he had.