How I Interview

Danbo and Danbo
Credit: Flickr / Takashi Hososhima

I have interviewed a lot more people than I have myself done interviews. This is mostly because I am terrified of interviewing. But I also used to be extremely nervous to be the interviewer. It’s a lot of pressure to try and evaluate someone in 30-60 minutes, and it matters a lot to me to be as good an interviewer as I can. I’m not claiming that I am a good interviewer, just that it is something that I really care about doing as well as I can, and work really hard to be better at.

This is the process that I follow.

Before

I personally do not give more than a cursory glance at resumes. There are two reasons for this. Firstly, and most importantly, I want to go in without preconceptions. Secondly, I found that when I studied resumes I would frequently start to worry that I wasn’t qualified to interview that person, which would put me in a bad headspace for the interview itself. Obviously for some kinds of interviews this isn’t possible, but for technical interviews, or behavioural based interviews (such as I did for a not-for-profit organisation I’m involved with) I haven’t felt the lack of information.

When someone is job hunting, an interview is probably one of (if not the) most important thing they are doing that day. I feel like it is only fair to reflect that in your schedule as an interviewer – jamming someone inside a 30 minute window is not cool. In as much as I have control over scheduling, I leave an hour before and at least an hour (preferably 2) after open.

I don’t think I am naturally a warm person, but as an interviewer I try to do my best impression of one. I found The Charisma Myth helpful for articulating what this means, but essentially it’s about being in a state of goodwill towards the world in general but particularly the person you are interviewing. This is a mental state, but also a physical one – being fed, hydrated, and having slept, for example. For interviews, I think this is about entering the interview in a state where you expect the interviewee to be a smart, capable person who can do the job that you are interviewing them for.

The Question

These comments apply more to technical interviews.

When choosing an interview question, I think it’s important that everyone can achieve something with it. This doesn’t mean that it’s trivially easy, it means that there are gradations of an answer. It shouldn’t be all or nothing, genius or fail. Someone who is incredible needs to be able to showcase that, but someone who doesn’t really have the knowledge or the experience to give even an okay answer needs to be able to achieve something.

I firmly believe that questions should be domain free where possible. This is because the interviewee is already nervous, and springing an unfamiliar domain on them is liable to induce a state of panic. The classic response to this is that it’s just an example, but if it doesn’t need to be tied to a specific thing, why tie it? For example – any question that gives a video game as an example. No. Anyone who doesn’t play games is liable to be intimidated by that, or need extra time to understand it.

Esoteric knowledge and gotchas, no. If a question requires someone to know some deep detail of the JVM or GCC, it’s a bad question.

At the Beginning

Be warm. Ask them how they are. Mean it.

Be considerate. Ask if it’s still a good time. Mean it.

Explain what is going to happen. How long it will take, what you’re going to do, if there will be time for their questions at the end, anything else they should know. For example, when I do phone interviews I always explain that typing is me taking notes (not doing email or coding!)

During

Conversation over Interrogation

Some interviewers seem to take the approach that they are a standardised test in human form. I favour conversation over interrogation. There are two parts to this – the first is to encourage them to talk, not just to answer. I think part of this is taking a conversational tone in your questions, and encouraging the interviewee to elaborate. The second thing is that it’s weirdly unnatural to spend an hour with someone and never reveal anything about yourself. So try and be a bit human. If they mention a place, and you went there for some reason, say so. If they pick a language you aren’t familiar with, express interest and say (if it’s true) you’ve been meaning to try it. As an interviewer, you shouldn’t be dominating the conversation. But if they leave feeling that they know no more about you than they did when they entered, you probably didn’t seem that friendly, or make your interviewee feel very comfortable.

Hinting is an Art

The thing I still find hardest is when to hint, and when to wait patiently. Conversation is 2-way though, and if someone isn’t clear on something it could be their understanding… or it could be your communication of the problem. One thing to start with is to assume that your communication is at fault, and clarify that they understand what you are asking. The second thing is to be very general, and get more specific. E.g. Ask first “are you sure you’ve covered all cases?” then “what if the number is negative?”

Don’t Twist the Knife

When I co-organised an interview training event for women students, one of the things I warned interviewers about was that if their interviewee cried I would hunt them down. As I did this, the men in the room mostly looked at me like I was being completely unreasonable. A day after the event, one of them pinged me to tell me that it had been really useful as a warning – because he had known that it could go that badly, he was able to prevent that from actually happening. So when his interviewee stood up to walk out of the room, he was able to redeem the situation. I was really pleased that he had handled that situation, and really happy that he felt he could tell me about it.

It hasn’t happened often, but there have been a couple of cases where someone did so badly on the interview that I switched from trying to evaluate them (I had all the information I needed) to trying to make them feel OK about it. Giving them softball questions that they could answer, or coaching them through the question step by step. I think this is especially important when you are later in a day of interviews, and those interviews have been going badly. Before you give up on someone, it’s important to be sure. Plenty of people are nervous at first and warm up as it goes on. I can think of 2 – extreme – examples in well over 50 interviews.

After

Thank them. Mean it.

Make sure they know what is happening next. If they have another interview, see if they need a break or want to get water.

Feedback

I like to write up my feedback immediately, but definitely that same day. I think it’s important to have a fresh memory of what happened. It’s also the kind of task that will weigh on me and be on my mind if left undone.

Typically feedback gets written up under headings, although I usually include my complete notes somewhere. Often I start with an idea of what should go in these sections, but I don’t completely trust those initial ideas, and instead go through my notes again. I often note times in my notes, so I have times to refer to instead of feelings about how long things took.

Once I’ve written my conclusions I go through and comb them for bias. Years of reading original research and writing about and being a woman in tech myself have given me an attuned antenna to subtly gendered or racial comments in feedback. For example, the use of shy. Suggesting someone is more junior than they actually are. Penalising someone for not being confident enough… or for being too confident. How to de-bias your feedback could be a book, but one thing that you can try is going through your feedback with someone (a colleague) who won’t be involved in making that decision, from that perspective. If I’m not confident that something is unbiased, I tone it down or just delete it. It is possible that this makes my feedback for underrepresented minorities slightly more positive than is deserved, but given the evidence about the level of bias we have, I 1) doubt this is actually true and 2) expect that other people will compensate because most people do not go to these lengths.

Terribly, in tech this idea of “culture fit” is often mis-defined as liking that person. I once saw feedback which said that someone was a good culture fit because they “liked board games and sci fi”. Sometimes disliking is relevant – for example I have disliked men I have interviewed because they made sexist or patronising comments to me. But in a short time period, this is usually down to the level of connection you felt with that person, which is fully 50% on you (more, the person with the most power in a relationship has the most influence over it – this applies to managers and directs, and interviewer to interviewee). And secondly can be really influenced by things like language barriers. So my final pass of my feedback is to make how much I liked the person less relevant, and make sure that I haven’t let what I wanted to see influence what I did see. If I didn’t feel like we connected, I don’t want that to show up in my feedback. The level of connection someone creates in an interview doesn’t really have a lot of bearing on how they will work with people over time. They might have come across as quiet or a little reserved, but you don’t know what they are like when they warm up – so don’t assume you do.

Closing Thoughts

Every so often I read these posts about how interviewing is broken, and actually what we should be doing is something else entirely. And I don’t really believe in, or advocate for the standard industry process of whiteboard interviews or whatever.

But I do wonder how big a problem is the process we follow at a macro level, compared to the attitude we take as individuals.

6 thoughts on “How I Interview

  1. Greatly enjoyed this article. I’ve been trying to wrap my head around interviews and how to properly conduct them. This makes a lot of sense.

    [WORDPRESS HASHCASH] The poster sent us ‘0 which is not a hashcash value.

Leave a Reply