[Google] has created an automated way to search for talent among the more than 100,000 job applications it receives each month. It is starting to ask job applicants to fill out an elaborate online survey that explores their attitudes, behavior, personality and biographical details going back to high school....What Googlish methods were used to come up with the questions and the formula for weighting them? They surveyed the employees they already had with various questions like "Is your work space messy or neat?" and "Are you an extrovert or an introvert?" and connected the answers to the 25 different measurements they had about each survey-taker's job performance. This gave them 2 million data points to analyze.
The answers are fed into a series of formulas created by Google’s mathematicians that calculate a score — from zero to 100 — meant to predict how well a person will fit into its chaotic and competitive culture....
What did they figure out? Nothing, really, as far as I can tell. High grades in school don't guarantee success at work. Who is surprised? You can tell if a person was once both smart, goal-oriented, and hard-working. But how they're going to behave once they land the job is a different matter altogether.
The "vice president for people operations" is quoted saying: "Interviews are a terrible predictor of performance." That's not surprising either, is it? And it's a good thing too. How awful it would be if an interviewer could see into your soul.
I suppose I like the idea of a long survey that is individualized to the conditions and requirements of a particular workplace. It lets different people rise to the top and crack through the layer normally occupied by the applicants with the highest grades. But it seems as though people who want jobs will figure out ways to ace the survey, won't they? And some people are just good at surveys. Why would that make them good workers?
But maybe you think that Google has such brilliant ways of coordinating vast numbers of data points, and whatever they do will somehow -- amazingly, magically -- work.
Posts
11 comments:
Before an applicant is allowed to post the answer to the question, put them through the Word Verification Routine. That will separate the chaff from the grain.
It's eHarmony for jobs.
I doubt Google has discovered the way to discover the folks who are real corporation-kilers. Not deadbeats or shirkers, not even unfreewheeling types.
Sociopaths.
Google's money is blood in the water to these types, social chameleons who will be able to play by whatever rules Google devises, and better than they can. And they rise to the top, slaying all who oppose them, even the current leaders. Then they kill the business, and move on. Think Enron or WorldCom.
We'll see if they've figured it out. They should invest less in finding folks able to orbit the giant hairball than in finding the Voigt-Kampf test that weeds out a corporate Ted Bundy.
I Want to work for google : )
Ann wrote: "But maybe you think that Google has such brilliant ways of coordinating vast numbers of data points..."
This they most assuredly do; it's how everything from search to checking spelling and beyond works at Google.
Ann: "...and whatever they do will somehow -- amazingly, magically -- work."
Given enough accurate, uncorrelated data (and that's obviously the hard part of crafting the survey), it can almost assuredly work better than traditional interviews alone, and Google doesn't do traditional interviews anyway—the most intense on-the-spot programming exercises I've done have been for Google interviews.
Anyway, it's important to understand that Google's entire business consists of optimizing sets of thousands of equations of thousands of variables. The HR process is just one more domain in which that might prove useful.
How awful it would be if an interviewer could see into your soul.
Imagine what interviews would be like:
Welcome to ESP Industries, Mr. Jones. I'm Claire Voyant. Oh, I see it says here that you are Nigel B. Jones, but I perceive that your name is really Neville P. Jones.... Now, for the first question, Why do you want to work for ESP Industries?"
Well, Miss Voyant, I want to go to work for a leader in the---
"Yes, I know. In the field. They all say that. And cut the crap, I know very well that you are unemployed and desperate....
Next question: Why did you leave your old position?
I left my old job because it just didn't seem to fit someone with my set of skills. So I quit in order to look for a position where I could really contr----
"No, Mr. Jones, you didn't leave your last job because you didn't feel like you fit there. Your last supervisor fired you for getting drunk at the Christmas party and puking on the xerox machine. And I see clearly that you begged him and begged him and hugged his leg until he had to call security and have you dragged out the door....
Now, Mr. Jones would you care to elaborate on why you want to work for our company rather than our competitors in the industry?"
Uh, I have done extensive research on ESP industries and I believe it presents exactl---
"Mr. Jones! You googled our company this morning before the interview, and that's the first time you researched it....
Mr. Jones, what are your expectations regarding salary?
Well, I'd like to make $40,000 per year to start--
Mr. Jones! You'd be willing to work here for $30,000! Maybe even less than that. We'll be sure to let them know that in personnel...
Mr. Jones, what do you feel is the biggest asset that this company.... what? MY LEGS?!? I'm glad you think they look 'hot,' Mr. Jones, but I'm a married woman and I think you look kind of revolting, myself."
Job interviews would never be the same again.
We'll let's just hope they don't have beer parties on Fridays, beanbag chairs, and wearing baseball hats backwards....like the dotcoms of the 90s.
Youth culture.
Total lack of sophistication.
Whatever happened to savoir faire, and debonair Bon Vivants?
Peace, Maxine
I don't know what it's like in law, but from my own observations and discussions with friends and acquaintances, the big successful companies around here like Google and Genentech place huge emphasis on personality in hiring. Competence is quickly determined, and then it's all personality fit. Getting called back for third and fourth interviews over a period of months is becoming the norm, even for non-mgmnt positions. (Although Manager or Director seems to part of everyone's job title now even if noone reports to that position.)
Given the number of math majors working at Google, this shouldn't be too surprising. A few of them are bound to have taken some psychometric classes.
Who needs psychometry? They could provide those same tests to people already in the company and use it as training data for some unsupervised learning algorithm like kernel PCA.
Working as a Director of HR for a Biotechnology company I am curious if Google is going to be conducting these tests for all employees.
Specifically, admin staff, facilities staff, finance etc. Many of these positions require a much different skill set than an engineer or mathematician. I don't think these tests would predict an excellent secretary.
My current employer, which has over 10,000 employees does not require any type of GPA for our hires. I believe a GPA is completely BS when determining who to hire.
Past business/work experience successes/accomplishments is our top indicator of future success. Also the company where they have worked is incredibly important-we can get a better understanding if it is a similar culture.
We hire tons of sales candidates and we want to see how they rank in the country, their sales numbers/% quoto they met and how well they achieve their bonus targets.
I understand that Google has a reputation as an innovator but this sort of thing really isn't anything new. In certain situations employers are concerned with person-organization fit. By attempting to match applicant's personality traits to organizational culture, Google is just trying to maximize the likelihood of hiring someone that will not only be good but will be more satisfied and therefore less likely to leave. Many large companies do this sort of thing at some level.
Post a Comment