Bungled interviews go back a long way, and a mythological example comes to us from The Ring of the Niebelungen. Wotan, the God who had Valhalla built, and Mime, a dwarf, have a mutual interview that goes awry. At the start, they agree to bet their heads on their abilities to answer each other’s questions. Things do not go well.
Mime goes first. Perhaps he was overly influenced by the advice to attorneys to only ask questions to which they already know the answers, and being a stupid guy he asks three questions that almost anyone can answer, especially Wotan. When it is Wotan’s turn to ask the questions he chastises Mime with: “Was zu wissen dir frommt, solltest du fragen: Kunde verbürgte mein Kopf.” which translates roughly to “You should have asked me what you needed to know because my head was on the line,” instead of the Google Translator’s humorous version, “What avails you to know, you should ask my client authenticated head.”
In search of the perfect question
This is very much the situation with the one hour interview, well documented by Joel Spolsky, Paul Graham, and many others to be a difficult and synthetic encounter. In my experience, management interviews have generally been conducted to a high standard, but when I have worn my programmer hat, I have personally been asked all kinds of things the interviewer could not possibly have needed to know.
My personal favorite was being asked to write code that implements the itoa() function [integer to string representation thereof]. Almost no one has ever had to write itoa() for any business motivated purpose. If you are going to ask about itoa(), and you only have a few minutes, you are better off asking “What are some of the pitfalls associated with its unrestricted and unchecked use?” If we are programming in C, then we should be sure that the programmer who writes code that calls itoa() knows the answer to that question.
So, I propose Flanagin’s Axiom of Interviewing.
The perfect question is one for which there are a variety of degrees of correctness, and a variety of ways in which the answer may be framed. In a perfect question, both the correctness and the delivery should inform the interviewer.
If we accept the axiom, interview answers can be plotted on two axes. The correctness axis is correlated to the question in some obvious way, but I have had to think about the other axis. To give appropriate credit on two levels, Scott suggested that I read this article and I suggest you do, too.
The types of answers
There are five categories in my classification of the way the answer is supplied. Perhaps we should just call them “types of answers.”
- Bullshit: I don’t really need to describe it, do I?
- SWAG: There are candidates whose friends or headhunters have told them to never say “I don’t know,” apparently overlooking the fact that there are many interviewers who respect a display of honesty. And so they guess. Guessing is more a strategy for test taking than interviewing, because it is known what happens if you guess wrong on a test, and unknown what the penalties are in an interview. For example, you may appear to be misinformed.
- Cautious: Some candidates want more than anything to avoid being wrong, and so they offer an answer that is diffuse, and non-refutable. Doing so successfully is a significant talent not to be overlooked, and caution is not the same as being partially correct.
- Pragmatic: This is the kind of answer I tend to give, and it does not impress the interviewer in environments where the inmates run the asylum, whether or not they do it well. Pragmatic answers tend to negate the value of the existing social order, and the person in the room with the stake in the existing social order is the interviewer. In management interviews this type of answer is better appreciated because managers have more stake in the P&L than they have in proving they are correct.
- Psychological: Simply put, the candidate tries to figure out what the interviewer thinks is the correct answer, and provide it. Of course, for such an answer to be successful in the interview situation it must be substantially correct, as well. This is the answer that does interview well in many cases because the interviewer is satisfied with his own cleverness.
In search of Genius
There are moments of instant insight, but you cannot run a business by depending on it. In an interview, the thing not revealed is Genius, which is the thing we all think we are looking for. In management roles, I have been most impressed over the years by my employee who takes on a problem, and then calls me later in the day, perhaps on the way home from work getting a Slurpee at a Stop-and-Rob, with an answer that is filled with the insight gained by reflection, or unfocussed awareness if you happen to be Zen practitioner.
I believe the search for Genius is best handled by Joel Spolsky’s question in Guerilla Interviewing where Spolsky suggests asking the candidate to describe a previous, completed project and how the solution was arrived at. A candidate who has had a personal epiphany will usually be able to tell the story in such a way that I, as the interviewer and listener, also have that epiphany and share it.
The entire premise of interviewing is grounded in a belief that the past really is the best predictor of the future, unlike universities who believe that the SAT scores of incoming freshmen (people who have never taken a class) predict how well the university prepares its graduates for the world. If I can be made to share the candidate’s epiphany, there is a good chance the experience can be repeated.