Are you Smart Enough to Work at Google?
- William Poundstone
"You are shrunk to the height of a nickel and thrown into a blender. You mass is reduced so that your density is the same as usual. The blades start moving in 60 seconds. What do you do?" - the blender question.
Primarily about the interview process at Google and other popular tech companies Are you Smart Enough to Work at Google is a must-read for anybody interested in interviewing for most of today's big companies. Each chapter ends with a series of interview questions in the form of puzzles. Answers are given a thorough development in the appendix. Entertaining and thought provoking.
Though the book centers on Google's culture, it does have insight into other important companies both in tech (Apple, Microsoft, Oracle, etc.) and elswhere (J.P. Morgan Chase, Nordstrom). The non-tech company info is a bit limited.
The value of these types of interviews is questioned, but a better discussion of job interviewing in general with supporting research is in Daniel Pink's Drive. (my short notes on Drive here.)
Some simple advice is provided (136), and I think it works well across all job interviews:
- Use trial and error. Think out loud. These questions don't usually have one correct answer. The interviewer wants to see your thought process as you evolve an answer. These questions are in part looking for creativity - they want to see you bump into the wrong answers or blind alleys, and continue to recover.
- Don't hesitate to ask for clarification.
- Listen! The interviewer may provide a hint.
- Use the whiteboard if you think it is permissible and will help; it illustrates your thinking.
A list of question categories is provided with discussion.
- Classic logic puzzles. Step-by-step deduction and a logical process are key
- Insight questions. Too hard to answer through deduction... needs the Aha!
- Lateral thinking puzzles. Depend on verbal ambiguity
- Tests of divergent thinking. Compare two processes.
- Fermi questions. Asks for a quick estimate of an unknown quantity. (see cheat sheet below)
- Algorithm questions. How would you perform a task - efficiency counts.
Salvaging a doomed interview:
- Restate the question in your own words. It will clarify for you what you need to think about, and ensure the interviewer knows you understand the question.
- Disambiguate. Ask for clarification of details ("Am I fully functioning despite being the size of nickel?")
- Describe why the obvious answer fails.
- Analogize. ("When shrunk to the size of an ant I have the relative strength of an ant!")
- Brainstorm. ("I can hang on a blade. I can duck below the blades")
- Critique. Tell why some of your brainstorming won't work - ("The blade will spin me off. I can't stay below the blades indefinitely")
Interview cheat sheet: primarily for Fermi questions which ask you to quickly estimate an unknown quantity.
You are probably expected to know the following as general knowledge, input in to the question as needed. (insert your company's info in place of Google)
- Population of the world - 7 billion
- Gross domestic product of the world: $60 trillion
- Population of the US - 300 million
- GDP of the US - $14 trillion
- Federal minimum wage - $7 (actually $7.25)
- Population of the San Francisco metro area including Silicon Valley (insert your company's city data here)- 8 million
- Google's stock market value: $100 billion
- Google's annual revenue: $25 billion
- Google's annual profit: $10 billion
- Google's stock price per share: $600
- Number of spheres that fit in a big volume, with random packing: 1.2 times the value you would assume with a cubic lattice. For the "How many tennis balls could you fit in this room?" type of questions.
The Blender Question: Answer developed in the WSJ article by Poundstone here.
