The Key To Interviewing At Google

Nerval's Lobster writes Wired has an excerpt from a new book of Google-centric workplace advice, written by Laszlo Bock, the search-engine giant's head of "People Operations" (re: Human Resources). In an interesting twist, Bock kicks off the excerpt by describing the brainteaser questions that Google is famous for tossing at job candidates as "useless," before suggesting that some hiring managers at the company might still use them. ("Sorry about that," he offered.) Rather than ask candidates to calculate the number of golf balls that can fit inside a 747 (or why manhole covers are round), Google now runs its candidates through a battery of work-sample tests and structured interviews, which its own research and data-crunching suggest is best at finding the most successful candidates. Google also relies on a tool (known as qDroid), which automates some of the process—the interviewer can simply input which job the candidate is interviewing for, and receive a guide with optimized interview questions. It was only a matter of time before people got sick of questions like, "Why are manhole covers round?"

Comments

First, manhole covers are not always round

[gweihir]

In fact, the ones here are far often square than round, so the answer to that question really is "because otherwise they would not fit the round manhole". Second, It took them pretty long to figure out their interview-questions are bogus. I interviewed there in 2008 on the request of a friend that wanted me for his team. Total failure as I knew far too much about the things they were asking me and the ones asking were not domain experts and hence did not understand the answers. In retrospect, that is fine. I now know several people that left Google, because they did not find the company to their tastes at all anymore.

Re:First, manhole covers are not always round

[dreamchaser]

You're probably better off. Working for the big Silicon vendors is over rated. You can make a six figure income working for a small firm, as well, without the hype and hyperbole.

Re:First, manhole covers are not always round

[hawguy]

So the question can be used to weed out pedants. I guess it is useful after all.

Or to find the engineers that can spot the missing parts of vague software specifications -- just because a user asks for something in the specs doesn't mean that he knows that the case he wrote up doesn't handle all of the options the software will encounter in the real world.

He may ask for software to generate quotes for manhole cover manufacturing, and only ask for a radius because clearly that's all you need to describe a round manhole cover, yet the smart engineer will ask how to handle the other shapes. Few companies want an engineer that blindly adheres to specs even when they don't make sense in the real world... that's more like a job for consultants so they can get paid to do the work and the paid again to do it the right way.

Re:First, manhole covers are not always round

[ShanghaiBill]

Fresh eggs have very small air cells, and ones straight from the chicken coop will have virtually none.

I have some backyard chickens, and I can confirm this. If you boil eggs fresh from the coop, they will crack almost every time, because there is no air pocket to absorb the expansion. Before boiling fresh eggs, it is best to store them at room temperature for at least a week, or longer if you have high humidity. This will allow the air pocket to form, and also loosen the membrane beneath the shell, so they will peel more easily.

If I am ever asked, during an interview, "What is the best way to boil an egg?", I will be ready.

Re:First, manhole covers are not always round

[TheGratefulNet]

let me ask just one more question-

does that apply to front-yard chickens, as well?

Re:First, manhole covers are not always round

[TheRaven64]

That's only one of the reasons (and not the best - the pipes are often circular, so it wouldn't matter if you had a square with a width wider than the diameter of the hole). The others are:

• It's easier (i.e. cheaper) to make circles of metal than squares.
• Manhole covers are often heavy, and if they're circular they're easier to drop back in place because you never have to rotate them in and be very careful of your fingers while getting the correct alignment.
• Manhole covers are often left for a long time and stick (often as a result of corrosion or plan matter growing in the gaps). Being able to rotate them before attempting to lift them makes it a lot easier to remove them.

Drink the kool aide

[Spazmania]

The key to interviewing at Google is to drink the kool aide before you arrive. Download and use the core software they make available. If you're not enthusiastic enough about their tool chain to do that, mere competence won't carry you over the finish line.

Most companies couldn't get away with that but Google is Google. At least for now.

Re:Drink the kool aide

[swillden]

The key to interviewing at Google is to drink the kool aide before you arrive. Download and use the core software they make available. If you're not enthusiastic enough about their tool chain to do that, mere competence won't carry you over the finish line.

(I interview software engineers at Google)

This really isn't true. I mean, certainly some level of interest and enthusiasm is important, but the interview process doesn't really focus on that. SWE questions are pretty much all technical, about algorithms, data structures and coding. Not to test your knowledge of those topics (Google isn't really concerned with what you know, but with how smart you are) but to see how well you can solve problems on your feet. There is a significant component of the interviewer's report that covers "Googliness" which probably partially covers enthusiasm, but is much more about whether your personality is a good fit for the culture -- are you a nice person, friendly, interested in technology and solving problems, etc.

Feynman interview joke & manhole covers

[Anonymous Coward]

  Probably this was brought to the mind of many people reading the article, so I might as well post it.
http://www.sellsbrothers.com/posts/details/12395

I know!

[zieroh]

I couldn't resist answering this:

Why are manhole covers round?

Because if they were square, they could be turned sideways, rotated 45 degrees, and dropped through the hole. As it turns out, this holds true for any shape with an even number of sides, until the length of each side drops below a threshold that's related to the lip of the hole.

The only way to win at Google?

[bobbied]

Is not to play.. .

When they start to reduce the process of interviewing down to a standardized series of questions and tests, they remove the human from the process too. Who wants to work for a company that isn't about HUMAN interaction first, that isn't willing to treat their employees less like interchangeable cogs and more like unique individuals.

This is the end of innovation and uniqueness for Google, or at least a sign that it's falling out of favor. This is the MBA mindset of trying to remove the variables in the process, standardize on some ill fitting solution in an attempt to be efficient. This means that they won't get innovation because failure is becoming something to avoid, taking risks leads to mistakes that cost money and time. When this becomes the prevailing attitude at a company, that company then becomes risk adverse and innovation slows down.

The problem here is Google is nothing but a search engine and software development house if it doesn't continue to innovate. It will die like Yahoo, AOL and all the others if it doesn't stop this.

Re:The only way to win at Google?

[swillden]

When they start to reduce the process of interviewing down to a standardized series of questions and tests, they remove the human from the process too.

(I do interviews at Google)

Google doesn't use standardized questions or tests. The app mentioned just provides some decent questions. At least for software engineer interviews, though, the interviewer would be foolish to use a question read from an app on the spot. The Google SWE interview questions are complex technical problems, designed to give the interviewer a chance to watch the candidate solve problems on the spot, and write code. To do that effectively, the interviewer has to know the question well, and to have explored most of the potential answer space, and to have some idea about how different kinds of candidates will respond to it.

Googlers call the process of exploring the answer space "calibrating" the question, and it's a pretty important and serious process. Generally it starts with grabbing a few other Google SWEs and doing mock interviews to see how they handle the question, and ultimately interviewers like to use the same set of questions with many candidates because seeing how several candidates handle it really nails the calibration down. I have a couple of questions that I have so well-calibrated that I can make 90% of a hire/no-hire decision in the first five minutes. Basically, good candidates blow through the first stages in a couple of minutes, while poor candidates struggle for a half hour. I don't make the hire/no-hire decision in the first five minutes, though, because there are exceptions. Some people just take a while to settle down / warm up, which is cool.

I suppose you could use an uncalibrated question from an app during an interview and then calibrate it after the fact. I've done that (without the app), asking a question that I haven't already calibrated, then after the interview getting some of my teammates to solve the same problem. It's not nearly as good as going into the interview with well-calibrated questions, though, because you don't understand the solution space well enough to effectively direct the candidate.

Actually, I just looked up qDroid and it's specifically for non-technical interviews. I had it run up some questions for a sample position, and they actually look pretty good. All open-ended, exploratory stuff, with lots of suggested followups.

This is the MBA mindset of trying to remove the variables in the process

FWIW, I'm sure Google employs some MBAs, but I've never met any of them. Google is an engineer-driven company, top to bottom. All eng managers are required to be competent engineers themselves, and for most engineers their entire management chain, up to and including the CEO, is all technical. There are negatives to this SWE-heavy structure, but it's far better than any other company I've worked for (and I've been around the block).