How would you move Mount Fuji?

Technical interview fascinate me. I have been in countless interviews, on both sides of the table, actually. I have been interviewed with companies such as Google, Yahoo, Microsoft, Electronic Arts, IBM, Sapient, Goldman Sachs, Motorola, Bell, Sybase, Nortel, Epson, etc, etc. On the other side of the table, I have interviewed at least over a hundred of technical career seekers as a technical recruiter, hiring manager, or simply helping out my colleagues to "screen out" or "get a feel" of the candidates. I enjoy hiring. Hiring right is a very satisfying experience. Of course, bad hire sucks, badly.

I like using brainteasing question to evaluate candidate, especially when I need to hire based on raw talent. Brainteasing question is a great way to understand the candidate's analytic skills and lateral thinking process. Among all types of brainteasers, I did not see much point of asking popular questions such as:

• how many piano tuners in the US?
• how many gas station in the state of California?
• how many tennis balls can fill in this room?

… but that was before.

Recently, I have encountered a situation that asking these questions does have a point -- to access whether the candidate can guesstimate with some grounds, give a ballpark figure with some reasoning and assumptions.

Why is this guesstimate skill is important? The World is not just black and white. We don't usually have the luxury to make decisions based on a complete picture. It is not surprising that clients have allocated a huge budget for us to build something and ask us:

How long would it take?

Of course, the inner-self (the one with just black and white) would argue that there is no way I can tell how long to build certain thing until I know absolutely every bit of detail in what we are building? In reality, I have never been in a project that we know all the details before we estimate the LOE (level of effort).

Good estimate is essential in the world of software engineering. But estimate is not a science; it's an art. It comes with experience. And experience is not free. Poor estimate is the price to pay. It's important for us to reflect our estimates and the actual LOEs regularly.

When asking these guesstimate interview questions, the interviewer can gauge how industrial-savvy the candidate is. Is the candidate smart enough to ask follow-up questions to increase the ballpark accuracy? Could the candidate compose a convincing and logical reasoning to support his/her estimate? Could the candidate able to leverage known/accurate information to solidify his/her arguments? Could s/he confidently present the estimate? Could the candidate identify where the areas of the greatest risk for error? Could the candidate identify the most crucial determining factor that affect the accuracy?

One thing for sure: if the candidate refuse to answer the question simply because s/he thinks it is a stupid question. We may pretty well end the interview right then and there!!

Some professionally qualified recruiter may not agree with me. Some even argue that it is illegal to ask such type of questions in an interview. I do not hesitate to go into court to justify my motive in asking the question and I am insisting that this type of question is highly relevant to conduct our daily business -- software engineering.

Argue all you can, but this is my blog. I am always right.

Back to the title of this entry: how would you move Mount Fuji? I don't know what kind of answer you are expecting. I read that book long long time ago. I don't remember the answer. In fact, I never try to remember any brainteaser answers. I derived them myself!

To me, I can move Mount Fuji unlimited number of ways. I can step back by one feet and I have "moved" Mount Fuji one feet further from me. I don't think it's a trick question, it's a question of relativity, think out of the box, lateral thinking, change of perspective, etc.

Wow, I completed this post in one train ride, impressive!