When I conduct an interview…

This post is a follow up to a thread (“How do you conduct technical interviews?”) that carried on for quite a while on the Oracle-L mailing list (you should consider joining if you aren’t already on the list). Here is my contribution to the discussion that started with the eternal question “How do you find the person with the right attitude, not just technical skills?”

As has been said often, there are no silver bullets and your “gut” feeling has to play at least some part. Here are some of my thoughts (in no particular order) on how I typically conduct interviews (I interview consultant candidates, but IT candidates wouldn’t be much different):

  • My mindset in the interview is about determining whether this person has a high capacity to learn new things quickly and apply/adapt them. This is almost always more important than the knowledge they have in their head now, because at least some significant part of the detail they know will be obsolete or changed in the coming months/years.
  • I don’t care if they can recite syntax or quote intricate details (as was mentioned–that’s what reference books and docs are for). If they can explain HOW things work (things like the process of a commit, how data is read into the buffer cache, how read consistency is achieved), that’s more important than whether the command to switch a logfile is alter system or alter database (I always have to stop and think about that one). If possible, in a face-to-face interview, I like to have them draw pictures to illustrate the process. If they can’t communicate something to me in a way that makes me understand it, they may have some communication issues–that’s at least an orange flag for me when hiring consultants where communication skills are absolutely critical.
  • While past experiences are important and things like “tell me about the worst project experience ever…”, I recognize that people change as fast or faster than technology does sometimes. So, I’m more interested in very recent history. My questions include “What did you do at your current job in the last week?”, “What did you learn in the last week?”, and/or “What did you teach someone else last week?”.
  • Every interview candidate has to answer “I don’t know” or “I’m not sure” at least once (for a technical question). I follow up with “Ok, how would you find the answer?”. See my point #1 above. No one knows all this stuff and if they think they do, then they probably don’t have the right attitude for my team.
  • Again, not an absolute, but those that talk about the server they have at home or the VM images they maintain to play with new stuff often have a predisposition to learning new things without being told to. This is a Good Thing(tm).
  • Recognize that employing the people with the “right” attitude and presenting them with routine, boring work will eventually lead to them being unhappy and leaving. There has to be a strong manager in place to ensure that the path of learning continues or these individuals will often become restless and look for a better place to work. Short version: employing the “good” ones is often hard as they require challenges to remain engaged. At least I do. And I am hard to employ…ask any of my current/past managers! :)
  • On an interpersonal level, I like to hear a story about an encounter with someone that had a “difficult” personality and how the situation was handled. We all have them and while not a technical issue, we all know that the non-technical issues often hurt work environments worse than technical ones.

Of course, as I’ve published here previously, I always include my “signature” (technical) question: Can uncommitted data be written to Oracle data files? (Of course, I always like the follow up why (not)?)

And, starting from now on, I’ll be asking candidates that interview with me if they’ve read this post or not. :) I don’t necessarily like anyone to think I’m famous or anything, but I always Google the name of any person I interview with to see what they do/publish online. It’s always nice to know what the other person is thinking about before/after they talk to me.

  • Venkat

    Great post!. Point 2 is the best one.

  • http://www.dannorris.com/ Dan Norris

    Thanks for stopping by Venkat. Hope your baby is doing well!

  • Padraig

    Cool stuff Dan. I'll have to remember these next time I am being interviewed. That second last point is so true. There is nothing worse than being bored!

  • http://www.dannorris.com/ Dan Norris

    You'll never actually go back to work again, will you? I figured you'd
    be one of those perpetual researchers…probably work for the government
    and have your fingerprints removed or some crazy thing. Thanks for
    stopping by!

  • joel garry

    I've seen gut interviews just be totally wrong. Most managers overrate their ability to judge others ability to fit in, so _cause_ the revolving door. Some of the best workers for a technical position are the quiet ones – I know I'm terrible at interviewing, these manufactured problems are usually just plain bad. As someone on oracle-l mentioned, you just can't know until they sit in the seat a while. A common misconception is that someone needs to be able to articulate something to prove they know it. There are many ways of learning, and many ways of doing. The problems arise when the expectations don't match that. Don't we all know competent DBA's who have been in their positions for years, but are really, really quiet? When they finally leave, you get this revolving door problem?

    The real problem with DBA hiring is that the position itself requires contradictory skils – you say “determining whether this person has a high capacity to learn new things quickly and apply/adapt them,” and yet, so much of what we do requires a mindset of carving things in stone. DBA's have to be able to switch hats quickly – how do you interview for that? A couple of the oracle-l posts really pissed me off because they said things like “the first thing the person does is surf the net.” Well, what are they really doing? Checking their stocks and running another business? Or checking oracle docs, oracle-l archives, etc., looking to not reinvent the wheel?

    Which leads into another common hiring goof, which is that the hiring person is looking for someone like themselves, when what they really need is as broad a set of skills and world-views as they can get – they should be looking for people as different as possible! Even more dumbass-green-MBA, is trying to fit people into pigeonholes.

    Well, by now you've googled me. Am I good or bad? Huh, I only just noticed Burleson called me a guru! http://www.dba-oracle.com/t_funny_job_interview

  • http://www.dannorris.com/ Dan Norris

    Agreed, gut feelings are just a contributing factor, not the only one. Your comment about requiring articulating knowledge to prove they know it is interesting to me. If they can't articulate their knowledge, how can you evaluate them? And, in my case, I'm interviewing consulting candidates and in my work, if you can't explain what you know in many different ways (technical, non-technical, business analyst audiences), then you generally fail. So, in my case, if the candidate can't explain something to me clearly, they'll fail in most of the consulting scenarios that they'll experience with my team. Maybe in a small IT shop where there's just one DBA, you can afford to have a Milton-type person that doesn't say much, but gets the job done well/right. Doesn't work in my world.

    As for switching hats quickly, I do that in the interview. I ask a detailed question about performance tuning, then jump to a personality-related question, then ask about archive logging, then about RAC installation, then how to create a status report, etc. It isn't perfect, but it's the best I can do.

    All in all, I agree that people just need to be in the job for a little while before you'll know. It's absurd to think that you will guarantee a perfect hire based on a few 1-hour interviews with different people.

    Being called a guru by Burleson isn't something I'd openly volunteer. Guess you know which side of that ball I'm on :). If I could figure out how, I'd remove the link to dba-oracle.com as I try never to visit or link there. I use the plugin from http://ora-safe-search.sourceforge.net/ daily.

    Thanks for sharing your thoughts!

  • joel garry

    Sorry, I forgot about people not liking links there. I would have put a rolling-eyeballs smiley on it if I had thought of it and knew how. I'm not all the way on your side of that ball, but mostly, and not afraid to say it publicly.

    Then again, I was making observations about OCP not testing the right thing when it was originally being discussed in the last century. For a while after I finally got it, I wasn't sure if I should advertise the fact or not. I finally decided those who don't know much about it deserved to be impressed. :-)

    But it all eventually boils down to how well interviews, resumes, school transcripts, handwriting analysis or what-all predicts a fit for a job, and personally, for me, I've concluded that it doesn't work any better than just coming in and starting work, and in general in the HR industry suffers from the same sort of silver bullet handwaving as the wrong side of the ball. Plus, they spam.

    On the other hand… http://www.google.com/search?hl=en&q=san+franci

  • http://www.ece.umd.edu/~posulliv/ Padraig

    Well you know, I'm actually thinking of just finishing up with a Master's at the moment. Apparently, programming is not research and I seem to enjoy programming a lot!

    I'm thinking of applying for a job with Oracle out in CA in product development. I always thought it would be cool to actually develop some of that stuff like RAC and ASM. We'll see anyway, I'm still thinking about it.

  • http://www.ece.umd.edu/~posulliv/ Padraig

    Well you know, I'm actually thinking of just finishing up with a Master's at the moment. Apparently, programming is not research and I seem to enjoy programming a lot!

    I'm thinking of applying for a job with Oracle out in CA in product development. I always thought it would be cool to actually develop some of that stuff like RAC and ASM. We'll see anyway, I'm still thinking about it.