December 11, 2004
Google Labs Aptitude Test: The Cats are Firmly in Charge
One of the most vexatious tasks and probably the most important is the construction of the team. Google is on a big hiring binge at the moment, and as the only 'cool' company in town, and the only one flush with IPO dollars, they are having it all their own way in scouring for top software developers. I went to a recruiting session loosely camouflaged as a 2 hour talk on search engine technology recently, and got turned away at the door! Dozens were milling around wondering whether to stay for the 'reception' and chance their luck.
In another human resources effort, Google are running online quizes and brain teasers, trying to attract those that think. LIke they do, presumably. A flyer slipped into the ACM's Communications and other prestigious mags is called the GLAT - Google Labs Aptitude Test, as a sort of play on the series of standardised tests that the Americans run for their graduate schools. You can find the whole thing written out on various blogs like Relax and Amit's.
It's inspired human resources thinking, and well worth reverse engineering (a fun pastime). The GLAT includes these gems: "You are in a maze of twisty little passages..." and also "What's broken with Unix? How would you fix it?" Those with a sense of history will tag the archtype as the West Coast / guru / long hair long beard crowd. My first lesson in HR as a teenager with way too much confidence (arrogance?) was when I bounced into these people, decided they were even more arrogant than me, which was sobering, and also singularly unproductive for their salaries, which was confusing. So what do you do with people that are smarter than yourself? That latter challenge has bugged me all my working life.
There are lots of multiple choice questions which can only be described as leading and presumably designed to tell you how to act when you get to the next stage. See above link for why this is. And of course the "you already work for us for free" questions like the griller on "What will be the next great improvement in search technology?"
The GLAT is a Great Laugh for All Triers, but a pile of bad news for shareholders. Lots and lots of "problem solving questions" means that Google is intent on stuffing the company to the gunwales with a) problem solvers and b) problem solvers like ourselves. So we can take the monoculture thing as a given, but also, we can observe that it's a feline monoculture. The cats are will go forth and multiply. Even "What's the coolest hack you've ever written?" whilst an opportunity to really express, has a powerful conflicting force deep in the guts of those of us who have spent years herding the cats.
(The explanation for the "cats" metaphor is that someone once said that managing programmers is like herding cats.) So how many actual non-feline questions are there? One:
What is the optimal size of a project team, above which additional members do not contribute productivity equivalent to the percentage increase in the staff size?
[The answer is below.]
Now that's a good question! In fact, it's a great HR question. One could have a lot of great followup discussion on that. Issues like just why would google be better off employing even one more cat when they already have more than gomorrah spring gracefully to mind.
Shareholders watch out! This is one stock where the ceiling is firmly in view. I'll leave you with one more question:
"Which of the following expresses Google's over-arching philosophy?
A) "I'm feeling lucky"
B) "Don't be evil"
C) "Oh, I already fixed that"
D) "You should never be more than 50 feet from food"
E) All of the above
The answer is of course all of the above as right now, google can afford to have as many over-arching philosophies as it desires. Live well, live young!
The answer to the above question: 1. Anything larger incurs the costs of external coordination.
Posted by iang at December 11, 2004 07:37 AM
mate, the answer to this is simple..
> So what do you
> do with people that are smarter than yourself?
you hire them !
typically, "smart" folks don't have much of a clue about money, business or life so you just hire them. There are no or few examples of incredibly rich, empire builders who got that way by being "smart". As Henry Ford commented "PHD? Sure, I hire them all the time. What, do we need more?"
That Robert Kiyosaki guy has a good, excellent, explanation of all this (that changed my life in its clarity of explanation!) You may just think of those books as rah-rah get rich books, but one section (I think in the second book?) explains nicely the four types of people. One is the "perfectionist-professional expert" which I feel is what you are alluding to when you say "smart". They basically feel, often correctly, that they are the only person who can do something right and best.
In terms of Kiyosaki's topic, i.e. starting your own business and so on or getting in to business, he points out that the particular demon these blokes face in business is that they try to do everything themselves, can't delegate, etc etc.. It's an excellent exposition of the idea!
LoL! Yes, that's what google is doing...
But there is a thing about having too many. In general, a smart person has lots of smart ideas, but too many to implement. So they are wasted. The thing is to build a team of do-ers around a few architects, and that's where the challenge lies: in software engineering there is a pervasive belief that everyone is architect material, so how do take people who want to be smart and get them to do?
This is why companies with lots more money like banks employ a lot of contractors. Contractors are those who are smart enough to be smart, but instead take the money to do. So they have found a way to allow others to be smart, without themselves 'losing face' to use the Japanese term.
I will look out for Kiyosaki; there is always more to learn about building startups. I suggest you look around for Belbin; this is the most influentual model that I have come across for team building.
Belbin's model has 8 roles in it, and suggests that for success each team has to have representatives of each of the role; the experimental data has it that the ideal team size is 4, where each member nicely adopts two roles. Most people have a major role and a minor role, and the truly good people have an ability to suppress their natural roles and fill in the missing roles.
Unfortunately, the inventor seems to sell the model as a product so it isn't easily found on the net.
(If this comment appears!) .. I think Kiyosaki is totally irrelevant to the discussion at hand, sorry, I just mentioned him in passing.
There is a whole INDUSTRY of "Kiyosaki criticizers". You can find any number of web pages, such as the "John Reed" guy mentioned below. Most of them are authors who sell Invest In Real Estate, or similar, books, but only sell tiny numbers of books, who are horribly angry that Kiyosaki sells 100s of times more books. So (amazingly) they have nothing better to do than spend days asembling minutely detailed "I Hate Kiyosaki" pages on the internet. Perhaps this will boost the sales of their own How To Invest books.
The specific thing I mentioned in Kiyosaki was his description in book #2 which exactly meshes with the sense Iang meant about "smart" people. Anyways.
Iags original blog entry was awesome and Iang's blog is awesome. And I just learned about RSS, which will fail within a year or two like all these other pseudo-web standards attemts. Boring web and email will continue it's crappy plodding dominating course! :) JP
> and also singularly unproductive for their salaries, which was
> confusing. So what do you do with people that are smarter than
> yourself? That latter challenge has bugged me all my working life.
So, they are kept on the payroll just in case they are needed for something nobody else can figure out?