Interviewing people for operations is tough. I’ve written about it before in a long essay saying what makes startup operations different, and some ideas on how to interview for it better. There are many ways to do a bad operations interview, so I thought I'd pen down some quick alternatives to bad operations interview questions.

Interviewing for operations is also really really fun. At least, I think so (I love it!)

You get to re-live really fun problems, which is often a big part of why anyone works in operations ("I love solving problems! And every day being chaos!" = what a star candidate says voluntarily). You get to meet someone new and potentially amazing. And have a pressure-free chat! I mean the pressure is on them. But not on you, for once.

I've written before about how it’s important to not just ask a candidate questions about their experience, but to test them on how they’d apply what they learned to the job you’re hiring for.

"Easy," you think. "I'll just ask them. Like: I see you've worked in manufacturing before. How do you think that will relate to our gizmo service industry?"

You might get a story this way, but you know that with this question, all you're testing for is someone's ability to tell a story. They might be honest. Or, as has happened to me before (and maybe to you, but hopefully not!), you might be hiring a spin-doctor. A charlatan.

So you think you can test them with a problem. Great!

But you have to go way further than a “case” interview (popular with management consulting firms). It’s more about figuring out: could I work with this person? Are they going to elevate me, my team and the company? It’s no easy ask.

Why interviewers ask bad questions

Despite the above being pretty obvious, I still hear really bad interview questions being asked everywhere. I'm sure you do too.

I've been guilty. Perhaps you've been. It's fine.

Bad interview questions are not anyone’s fault, explicitly. This is why it happens:

  • Time pressure: Interviewing is usually 30 minutes sandwiched between other meetings, during a usual stressful day.
  • Numbness: The interview is often the third (or more) candidate being seen for the same role, and you’re usually trying to recruit for more than one role.
  • Logistics problems: There are room issues, teleconferencing problems and the fact that you’ve had zero time to prepare.
  • Emotions: Fear of making a bad decision or of failing to make a good one. Sympathy with a candidate who reminds you of yourself when you were younger. Concern you have a false bias or prejudice.

And probably much more.

It’s no wonder than what usually happens is five minutes after the interview was supposed to start, while you’re still trying to load their profile on Lever or Greenhouse to see what their CV says (let alone other feedback from other interviewers), you think of a great first question:

“So, why don’t we start by you summarising your experience and background for me?”

This is fine, you think. Look, they’re talking! I can read their CV while it happens. You ask a few more vague questions about their CV and they keep talking. Suddenly, you realise you’re run out of time and are late for your next meeting and dash out the door.

I guess she was pretty good. you think vaguely. I wonder what Mike thought? You plan on an informal corridor chat.

And thus the fate of the candidate is sealed. Remember when you were on the other end. You’d hate knowing this is how you’re being dealt with.

(By the way, can you tell I've been that person before? To former candidates from a while ago: I'm sorry! But I'll never, ever do it again.)

So here’s how not to do it.

Question 1 to avoid: “So, tell me about yourself.”

Very tempting. It sounds fine. But it's a mistake.

This introductory question puts you on the back foot, because usually it tells you nothing new, exhausts valuable minutes, and might set you down the wrong course if they say something that catches your ear and sounds interesting (and it becomes a conversation, not an interview). "Oh, you lived in Elbonia? I lived in Elbonia!" It's basically all over now, unless you like to hire people for their ability to chit-chat.

Harsh reality: If you don't already know who your interviewer is and what their background is, then you've been lazy.

Don't blame anyone else, or some system issue. You could have found out by finding their LinkedIn profile.

You can spend literally just ten minutes preparing by reading their CV and the other interviewers’ notes, and then deciding what your approach is going to be (test their intuition? communication? problem solving? See this article on different areas to test in operations interviews).

Do this instead: So instead of that question... you know what, throw the whole idea out the window and ask something totally different.

For example: “I see you used to manage a $15M P&L at that startup. Tell me, what do you make of this P&L?” Then you show them a P&L with something wrong with it, like outsized marketing spend, or customer acquisition cost that’s 3x the LTV, or something. See what more information they ask for, how the conversation goes and whether they’d give you a definitive action plan to fix it.

Show them a crazy P&L like the one below and ask: "Hey is this working? We're kind of thinking maybe this is getting out of control."

What do you make of this P&L? Where would you invest? (This is actually a screenshot of a game, but it'd be fun to see what a candidate makes of it... would take a lot of math!)

Question 2 to avoid: “I’m working on this problem where our utilisation is dropping. How would you solve this?”

This sounds really smart, and can lead to a good discussion.

But the problem with this question is that it’s too leading. You have been working on this problem and you know what the answer is. You basically want them to figure it out in an almost total vacuum, without the huge amounts of context that you have.

I mean think about it. Was it easy for you? Was the answer obvious? And is your answer the definitively best answer, and the one that is the mark of someone you would hire?

Heck no! It took you ages! And it was built on all your experience plus a ton of data you've been familiar with for (probably) years.

Do this instead: Ask this question but in the abstract.

Like “How would you think about this problem? What more information do you need?” and then you have that information on hand. Charts, data, survey results, whatever. You also probably need it to be someone else’s problem, not yours, so you’re detached from the final solution.

For example, take this problem, for which I made up the below chart using some (frankly, amazing) automatic dummy data generation.

I'd show this to a candidate and just ask: "My boss showed this to me this morning. Is this bad? What's going on here?"

Like memes on Instagram showing snippets of a much bigger situation, charts like these are engaging and get the candidate thinking. You can have a huge conversation just based on the chart below.

A P&L to show a candidate with a simple question: "Is this bad?"

Question 3 to avoid: “How fast are you at responding to emails?” (or any other question about "how" they are)

Yikes! I was asked this question by a highly-strung CEO who explained that he gets anxious as minutes go past between when he sends an email and gets an answer, no matter what the hour.

I think my face said it all. He sounded like a nightmare to work for, and I instantly was terrified I’d get the job (luckily, and unsurprisingly, I didn’t).

I know what he was getting at. It was: "Will we be able to work together?" Everyone wants to know this.

Do this instead: This might sound crazy, but you can simulate this environment. It’s amazing how you can weed out the wrong people from the right ones.

I'd only do this in specific situations, where you want to work very closely with one person. It would probably be like a COO hiring a Head of Ops, or something, but you're not quite sure if their experience would make them a good fit for the crush of daily operatioins.

Here's how you simulate the above situation.

Give the candidate some information, like an investor presentation and some made-up data to support it. The kind of thing that’s general and easily understandable.

Then at some point tell them you’ll email them a question and want an answer as soon as possible. Do this a few times, asking questions about the data, and check the answer. Is it correct? Succinct? prompt? If so, you’ve got yourself a winner.

The people you eliminate in this way are people who’d have hated working for you. You’re doing them a favour by not considering them. Not a favour they’ll love, but one that they might appreciate in the long run.

Question 4 to avoid: “How many tennis balls would fit into a Boeing 747?”

Avoid this, or any other brainteaser.

Brainteasers rose to popularity in the 1980s, I think. They were a way of testing particularly how programmers would solve real world problems. They test math, intuition, an ability to estimate and even a few real world things like knowing roughly how big a tennis ball is.

Some of that is relevant to realiity. But regardless, mostly you'll be screening for people who are good at brainteasers. And probably someone who has practised brainteaser problems in the even they’d get this very interview question.

Unless you're interviewing for a position in a brainteaser-writing blog or something, this isn't going to give you very useful information.

Do this instead: Our dashboard stats have been taking a dive! Can you tell me why our scores are dropping??

Then show them an example of a dashboard that your management team looks at every single day.

This is not just realistic… it’s completely real. I have started so many jobs where on day one I was asked to look at a bunch of dashboards that had metrics I had never heard of, that painted an incomplete picture of a business and that I didn’t even need to understand the business. They looked cool though.

In fact, you can talk about some problem that’s unrelated to the dashboards. For example “Our NPS scores are dropping! Here’s what our dashboard looks like” and then show them dashboards on rides, utilisation metrics, post customer-service survey satisfaction or whatever.

How often has this happened to you? You get some question and think "Damn it! None of these carefully polished dashboards by the analytics department are relevant to this real-world operational question I have to resolve!" But guess what, you're terrible at making dashboards and have to work with what you're given.

A good candidate will immediately ignore everything irrelevant and ask for the good stuff that you — “oh yeah!” — have in your back pocket. A bad one will flounder or make up something bogus. (Very awkward to watch.)

Show them this dashboard and ask an only tangentially related question, like about NPS or satisfaction.

By the way: The answer to “how many tennis balls fit in a 747” is ~10,500, plus or minus 10%. (I’m kidding, I have no idea and I don’t care.)

Question 5 to avoid: “Do you have any questions for me?”

I mean, this can work sometimes. But it usually just shows you’ve run out of things to say, or want an opportunity for the candidate to show off.

What inevitably happens is a candidate asks me something about the future of the personal mobility industry or something else very broad. Every time this happens I get disappointed.

Firstly, how the hell should I know? We’re just trying to survive week-to-week here! I’m happy to explain what I think, but really, almost nobody’s thoughts (except maybe investors who bet on it, and then, only the investors who in hindsight bet correctly) are worth a damn.

And secondly, I get sad because it's a lost opportunity for a real conversation.

Think about it like a date (if you need  to, think back a while!). An awkward date. The most awkward thing was when your date asked you something that was clearly unrelated to anything you know. "What are your thoughts on Chechnya?" Could be a great question, if you're a policy expert, or Chechnyan or something. But otherwise... crickets.

Do this instead: Help your candidate by prompting a better question. Give them an idea by feeding them some background or context. Have a conversation you actually want to have!

For example:

  • "I'm the head of recruiting but I'm also really involved in HR. Do you have any questions about company policy on diversity or culture?"
  • "I'm head of marketing so my daily job is working on customer acquistion. Do you have any questions about campaigns or ads you've seen, or any ideas you want to share?"

Or anything else.

The key is to give the candidate an idea, so they can ask you something that's not generic. They usually don't know much about you. They might have looked you up, but usually people's backgrounds on LinkedIn are out of date and sparse, so the more you can tell them, the better.

I hope some of the above was useful. If it was, drop me a line.