Ok I read it nothing that interesting until this 'What?' moment:
'Plastiq—a startup that allows users to pay their bills with credit cards'
Is that a thing in America or someplace you can't use credit cards to pay your bills? I followed the link thinking that it would be something else but it seems that really that's what it does.
My gas[1] and electric[2] companies are both subsidiaries of Avangrid, their websites and payment portals are practically identical, but one of them lets me use a credit card and the other doesn't.
I always wondered if there was a regulatory reason for it, or maybe if it dates back to arbitrarily different company policies when they were independent of each other.
Did not read the article, but a wild guess would be they are just looking for a hook to get users to opt-in to Plaid (presumably before the inevitable media backlash once alarmists figure out what it is.)
This is a surprisingly good article when it comes to describing the needs of a startup, but it makes me wonder why it’s taken for granted that these interview questions have to be asked in subtextual code...
FWIW I run a startup and I think every single one of these questions is terrible.
We can gauge your ambition and your resilience just fine in your first few weeks (in which we've signed you on as a contractor-under-evaluation - if your situation allows). And yes, we've had to learn how to do this the hard way (having kept people on for longer than was good for anyone). But I think it's nuts to think that you can test someone's ambition asking "where do you see yourself in 5 years?".
I'm pretty convinced that that place deep down where people's natural drive comes from is, especially for engineers and other creative roles, very disconnected from conscious career planning. If you want to tap that natural drive, you need to create an environment where it can flourish. I don't think that asking "Why should we hire you?" when you mean "Do you understand what problem we're solving?" contributes to such an environment. You want to be open, clear, direct, and honest, not play guessing games.
We ask the "where do you see yourself in 5 years" question as part of our standard interview. We don't put very much weight on it, but it adds color to the candidates ambitions. Most engineers say they would try to stay as engineers, some say maybe product/project/people management; these are all fine answers.
Then we have the candidates that just blank or give a response like "I just want to make money". Not dealbreakers, but not great answers.
But I've also gotten a surprising amount of really terrible answers, like "at Google" and "ideally I finally paid off all my student debt so I can finally be a comedian". These are the people we don't hire.
> But I've also gotten a surprising amount of really terrible answers, like "at Google" and "ideally I finally paid off all my student debt so I can finally be a comedian". These are the people we don't hire.
I wonder why? Those are the most, if not the only, honest answers you got.
In current state of software industry, unless the candidate happens to believe in your mission, anyone answering they'll want to still work for you is either lying or naïve.
I don't have a reason to believe our candidates are lying when they say they want to continue doing engineering work or move into management. None of that implies they're set on working with us, and that's OK, they're interviewing and know very little about us, so its understandable that they don't know whether it'll work out with us. So by saying they want to continue working as an engineer, we know that as long as they're working with us, we'll have an engineer. With product/project/people managers, we know to maybe keep an open head for a potential internal lateral movement from engineering.
When they say "at Google", that's at least as bad as saying "working for you"; they know nothing about what working at Google is like any more than they know what working with us will be like. Plus, it implies "I don't want to work for you"; would you hire anyone that says that?
Fair enough, wrt. working as an engineer, and not putting too much stock in "working with us" part. Early sign that a person wants to stay on the engineer track, or transition towards management, is useful.
> When they say "at Google", that's at least as bad as saying "working for you"; they know nothing about what working at Google is like any more than they know what working with us will be like.
They do know enough. They know that Google is "hot". Pays well, gives a lot of street cred, opens a lot of doors. Unless you're one of FAANG, they can assume with high confidence that working for Google will be better than working for you. What they may not be confident about is whether they'd pass an interview there.
> Plus, it implies "I don't want to work for you"; would you hire anyone that says that?
IMO the very fact they showed up implies they do want to work for you now. But maybe not in 5 years. For a typical programming job, I don't see a problem. It's unlikely the person will stay with you for 5 years anyway, so why not treat this simply as exchanging professional services for money?
> They do know enough. They know that Google is "hot". Pays well, gives a lot of street cred, opens a lot of doors. Unless you're one of FAANG, they can assume with high confidence that working for Google will be better than working for you. What they may not be confident about is whether they'd pass an interview there.
All valid things to want, but very little of it benefits us. Being a better engineer or manager benefits the company and would convince me to hire. Saying you want to work elsewhere eventually, but not being able to say why doesn't tell me why I should hire.
> IMO the very fact they showed up implies they do want to work for you now. But maybe not in 5 years. For a typical programming job, I don't see a problem. It's unlikely the person will stay with you for 5 years anyway, so why not treat this simply as exchanging professional services for money?
Sure, but they've just raised the bar a lot higher for themselves since now they have to convince us that they can hit the ground running and require minimal onboarding and investment. Very very few candidates can do that, and those that can have no problem getting a job at FAANG. Usually the ones that meet that higher bar are extended an offer, but so far all have declined in favor of a FAANG.
To me stories like this drive home the fact that there is no shortage of programmers. When there is shortage of programmers, no one cares about games like this and hire, because they want non-sociopath to work for them for next year and are drowning in work that needs to be done.
I think that this means there is a lot of good programmers with reasonable personalities around and companies are picky.
Depends on the market. I'm in Los Angeles where there's a ton of people willing to work here, but not very many companies hiring. Bay Area I know is the exact opposite.
All the research into interviewing and cognitive biases says that asking questions for "color" reduces interviewer's ability to predict outcomes while simultaneously increasing their confidence that they've gleaned information from the interview. There are experiments where interviewers are given some diagnostic information (e.g., hard facts about background) and conduct closed-form interviews (yes/no, this/that questions) with the interviewees responding randomly.
They'll still believe they've "learned more" about the candidate than they did from the candidates who they had biographical information alone.
A good interview process reduces an interviewer's exposure to non-diagnostic information and directs their attention so they only evaluate diagnostic information (and do so in a consistent way). This is the basis for the "structured interview".
I would imagine questions about career goals would fall under "diagnostic information" in the context of a job interview. It's not like we're asking for their favorite sandwich.
Why 5 years? Why not 1? Why not 10? 2? 4? How might those numbers change the answers you receive? Are you even testing to see how people's responses change as a function of that number? Are you even coding the responses in a way that would be able to detect how their responses change in response to that number? Is the variance in the responses small enough that you can even come up with a way to code them?
For example, if the interviewee is startup savvy and you ask the 5-year question then they understand the vesting cliffs are at play. Suddenly you're taking into account how savvy the interviewee is about the way vesting works. Was that the intention? If not then you're measuring something without even realizing it Either way: more variance!
Is it clear what the question is trying to get at? Do all candidates understand what it's trying to get at? Could candidate misinterpret the intent of the question given the fact that they're in an interview? Could different interviewers interpret the same response differently?
It's a "bad" question because it controls for very little and therefore inherently produces non-diagnostic information (such is the nature of humans). The level of variance in all the components of this question is exactly the surface area on which cognitive biases become the determining factor.
And the nature of cognitive biases is that you become more confident when they're operational, not less.
You can ask questions about career goals without going at it sideways. Here's some off the top of my head. I haven't thought them through from the above perspective, but as a first pass they control for more, are likely to get higher-signal responses that are easier to evaluate in a consistent, interpretable way.
- At your last job, what career goals did you have when you joined? Did you achieve them? How?
- Imagine you've been working here for 18 months. What would need to happen between now and then for you to not feel disappointed you joined?
- Think back to a manager who helped you define and achieve a professional goal (big or small). What was that goal? How did they help you achieve it? Why did that work?
I usually don't say '5 years' specifically. "In a couple of years" is closer to the phrasing I use. The candidate can pick whatever time period they're comfortable with answering. Most answer with equal fuzziness in their timeline. Every now and then we'll get a candidate who'll have a super rigid timeline, "raise in 6 months, promotion to in 12, second promotion at the 2 year mark, able to buy at house after 3 years, student loans paid off in 47months, after 401k/stocks vest I'll do something else". That's more information than I need and not the answer I'm looking for.
Most candidates get what we're trying to ask right away. Sure, sometimes we get a response like "well my current apartment doesn't have enough power outlets and smells like bread, so I would like to move and get a cat"; then we'll follow up with a surgically specific question.
It's a conversation, not an interrogation. The vagueness usually helps candidates open up and talk. Interviewing is stressful enough without the multi-part hyper specific word problem about something they probably haven't thought about.
We're interviewing people, and people are different with high variance. If we wanted consistency with bias removed, we can just craft a long multiple choice exam and have HR proctor it. Interviewing will always have noise, variance, and bias.
Haha, well, hopefully nobody's opening with "Where do you see yourself in 5 years?" or any of the questions I listed! Doing either of those things is a great way to signal to the candidate that INTERVIEW MODE has been engaged.
The goal isn't to remove noise, variance, and bias, but to control it in a way that produces a better assessment and a better experience for the candidate. This also allows the interview question + overall process to improve over time. Variance is present in every scientific experiment, after all, but the whole process of designing an experiment + evaluating the results is about controlling for known sources of variance and minimizing the ways unknown sources can get in.
For example, one might frame the structured process thus: "We ask an identical series of questions to every candidate so we know everyone's getting a fair shake. It's important to us that we're evaluating candidates based on their actual experience and potential, not on their ability to do well in interviews.
We hate tricky or ambiguous questions because they make the people we interview anxious and produce answers that are hard to evaluate fairly."
You would want to do that before they came into the room, too, to control for the possibility that receiving that information in the interview room itself produced surprising interactions (more variance again).
In general, the candidate should have a clear picture of what will be happening in an interview regardless, to control for the above as best as possible.
(I can say from experience that if the above is done with a friendly demeanor and good bedside manner, the sense of relief in the candidate is palpable.)
Imagine asking a candidate a question about a positive interaction with a previous manager that gets at the thing you're trying to evaluate. The candidate is naturally going to be happy talking about a positive experience! The framing of the question is intended to get them thinking about the moments you want them to be thinking about.
"A couple of years" is not great because now it's a function of how the candidate perceives the phrase "a couple of years." More variance!What if one candidate specifies their impression of "a couple of years" and another doesn't?
Asking the candidate to pick a time period explicitly also not great because now your impression is a function of the (time period, response) they chose and not just the response. More variance, again!
It's better to ask questions structured in a way that gets at what you're really looking for. I don't know exactly what you're looking for, specifically — ambition, reasonable expectations around opportunities for and speed of promotion, alignment with company priorities, something else? — but there will almost certainly be questions that produce better diagnostic information and feel more natural to the candidate.
Inherent in the interview process is the feeling that you, the candidate, want to give the answer that the person is looking for. "Where do you see yourself in (5/a couple/choose-your-time-period) years?" hits right at that anxiety (which is why a blog post like "what are they really asking?" is even necessary).
Are you evaluating their ability to understand the question in the way you intended? If not, structure it in a way that reduces that possibility.
You know, this is why I more and more hate interviewing.
I have enough passive income that I don't need to work. This year I have decided that every cent (after taxes) I make is going to charity. It was a promise I made myself a while ago and want to follow up on it. Tomorrow I am going (after talking with them) spend about 5 grand buying things a local association needs.
If you ask me where I want to be in 5 years I will only say to make more money than I do now. Much more. I don't have goals in my job. I don't want to be a PM or whatever the fuck I should say. I don't need this job. I am trying to do as much good as I can. For a while I was in the top 5% earners in my country (maybe even 3% but data is hard to get by). Now the only reason I work is to give my 'sacrifice' a meaning. When I stopped working because I didn't need to, I started to drink and be a lazy slob, I was drunk pretty much every day, so I use work to keep me in track, but I give every penny away.
I don't care about work. I am a fucking amazing developer. I had offers from top companies I refused as it didn't match my lifestyle (lets say 2 or 3 of the FAANG brotherhood). I just don't care anymore. If you pay me, I do my job, amazingly well. But stop asking this bullshit question and try to infer meaning from the answers. Anyone can read a blog and say 'to grow as an X and be maybe a Team lead/management/suck CEO dick instead of lower management dick'. It is all bullshit and you and me (and the rest of the fucking HN) crowd knows it
FWIW, I think that's a fine answer. Maybe the tone can be less cynical and bitter, but the intent and content is fine.
I have someone on my team that literally said, I'm an amazing engineer. I don't need the money, I just don't want very much managerial responsibility, I want time and location flexibility to raise my kid and fair market pay for my work. You bet your ass we hired her. She's still an amazing engineer and wonderful addition to our team.
That's why we ask these type of open ended soft questions, people are different and have different wants and needs; we want to have that conversation before hiring to raise the probability that it'll work out.
Sorry if I sounded cynical and bitter. I was, and probably still am. This time of the year really does make me like that.
I just hate this whole huppflah about hiring when I have seen companies miss amazing candidates because of these questions, and hire shittier ones because they 'studied' the right books/blogs. (I work with various startups here and help them with interviewing and building teams)
I will be honest, I am pretty bitter and is probably one of the reasons I am doing what I am doing now, but I do feel my reply to you was very 'directed' and I was venting more in general about the process and 'focusing' on you in particular. I am sorry for that.
You're right, and I too have run a startup and came to the same conclusion about contract-to-hire being the most reliable way to ensure employee quality. Unfortunately it's just not an option for most candidates (especially for more senior people in high CoL places like SV).
The fundamental challenge of interviewing is that no matter what you focus on whether it's technical or social skills, an interview is not long enough to give a proper indication of how someone will perform day in and day out. As an interviewer the best thing you have are proxies such as work experience, and yes, squishy behavioral questions of dubious value—you still have to ask something. As an interviewee there is no way around learning the subtexts and getting good at answering them. In many ways it's a silly social dance, but when you get good at it you actually can learn a lot about companies as well, and improve your chances of getting hired at a place you'll really like.
You can get pretty far with precursor in-person (or video call in a pinch) personality profiling as a screening, some informal probing for motivational biases, a half-day or so of pairing w/ a couple different teammates on the problems they're actually working on, and including the candidate in a directed meeting.
This assumes as a hiring manager you've done your work to fully understand the need you're hiring for, not just the skill set. That you know how this need fits into the composite model of the team and the company. And, that you've prepared and trained the team to be able to assess and give you the specific kinds of feedback you need to derive from tbeir working alongside the candidate for such a limited amount of time.
There's no better similation or signal for reality than reality itself. You can't ramp a team or company at "hyper-growth" speed this way, but I'm of the opinion doing that sort of thing is almost never a good idea anyway.
The research says work-sample tests and structured interviews are the way to go and they're well within reach for a startup. In fact, they're easier for a startup in many ways: half the difficulty in implementing a structured process is the internal selling + training + reinforcement necessary to roll it out across multiple teams/functions/etc.
Behavioral questions don't have to be squishy. There's a lot of literature (like, over a century) about conducting and evaluating behavioral interviews.
Here are some example questions one might ask a programmer:
- How do you know when your code is ready to ship?
- What methods or systems do you use to keep track of requirements?
- Of all your past managers, who got the most out of your potential? What made their approach effective?
- Imagine your professional life a year or two from now. What do you want that to look like and how do you hope this job would contribute to that?
- If you realized you couldn't meet an expected release deadline, what would you do? What have you done in the past?
- Tell me about a coding mistake you made in a past job that caused a serious problem. What happened and what did you learn from it?
- If we talked to your past managers, what would they say your strengths are? What would they say you need to work on?
BTW, if this stuff is interesting to anyone, I'm the head of instruction at Strive and we run workshops for engineering managers (including one on evidence-based interviewing). Email me jesse@strive.co if you want to learn more!
> We can gauge your ambition and your resilience just fine in your first few weeks (in which we've signed you on as a contractor-under-evaluation - if your situation allows).
We're a small company. So far we've been lucky that everyone we've wanted to hire was flexible for some reason (just out of college, been freelancing, huge amount of holiday days saved up, etc). We've yet to face the harder situation where something like this isn't possible.
But even then I promise we're not going to ask "where do you see yourself in 5 years".
To be clear I have no problem with companies hiring from the pool of people that aren't currently working. On the contrary, it strikes me as pro-social. I'm just not sure it's the best move from your perspective as an employer.
if someone offered me 'contractor-under-evaluation' employment, I'd likely laugh in their face.
in the US (where I work) or in the NL (where you work), my employment is contingent on a visa, which is just way too much effort for a "maybe," position. Once you've ascertained I can program and don't appear to be an asshole, most other things can be worked out.
There is an enormous amount of consideration and nuance being projected by this article onto these questions and the context they're often asked in.
I've been in this industry a long time, I've been in startups and in big enterprises, and I've been a hiring manager in both cases and an executive in the former.
Very, very rarely would I say that my hiring manager or executive peers (especially of the typically inexperienced startup variety) would be asking questions like these with the awareness of intent and interpretation being claimed here. Mostly these questions are being asked because, just like the fairly pointless "whiteboarding" questions, they're just the questions everybody thinks they're supposed to ask. That is of course until an alternative model is presented and nurtured.
If you find a great hiring manager who is actually putting in a lot of forethought, intent, and consideration into their hiring and interviewing process, then do whatever you can to work with them. But, one of the ways you'll be able to detect such a person is they won't likely be asking cookie-cutter questions or giving you cookie-cutter quizes.
As the interview was further and further analyzed by both The Company and the proletariat, an arms race developed. The Company asked esoteric questions and the proles studied. The proles learned to FizzBuzz and The Company developed FooBaz. Whiteboards gave way to take home interviews which gave way to probationary offers. Personal questions were supplanted by a form of subtextual sparring in which small facial movements, pauses, and hand movements spoke volumes.
By 2034 the majority of proles were on illegal performance-evaluation enhancing drugs, necessitated by the fact that the average prole was also on The Company's indefinite performance improvement probation. They occupied over 20% of their working hours proving their worth to The Company, which had hired Prefects to wander the endless halls. The Prefects started administering the Voight-Kampf employee aptitude tests on proles, but soon began testing each other on their own effectiveness in administering the test.
Before The Fall, the end state of the system was that The Company existed only to vet its own employees, all of whom were in charge of secretly monitoring other groups of employees. They had their nominal tasks such as programming or planning, but employee evaluation was based solely on the ability to evaluate others in their nominal tasks.
Jane Adeline Jones was 27 and has already been evaluated as an adult by both State and Federal governments. Not a probationary adult. Full adult. She was even allowed to drive during the day.
As a Level 4 Artificial Learning engineer with Gamazo people always assumed that she was a Perfect. She wasn’t. She was a hard working proletariat.
“That’s actually a funny way to put it” she thought to herself as the words came out of her mouth. “Hard working proletariat”. Not perfect. Just hard working. “Interesting concept” she thought and stuck a fork in a salad bowl.
The salad was a 13 out of 10. The restaurant could do better. The sauce was a solid 15, the bread here was always over 100. But the lattice was soggy. Could it evaluate to less the 100%? She, obviously, wasn’t certified to evaluate that but Danny was. “Danny, what you think of the lattice?” she asked. Danny looked up from his own bowl of salad and gave her a thoughtful stare. She put her head down and kept munching.
It was a nice day. The sky was blue. The sun was warm. Warm but not too warm. “17 at least”. Jane cleared her mind and laid back, as she waited for Danny to answer. “Life was good with high probability of getting better” she thought. She smiled. Closed her eyes.
When she opened them again, Danny has yet to answer. Instead there was a man standing there right in front of her.
Not bad, but you need more visibility for your promotion. Maybe you could put it up on Confluence and highlight which Company Value each paragraph supports?
"The theme is familiar to anyone who's tried to join a country club or high-school clique. It's not supposed to make sense. You are expected to conform to the rules of The Culture before you are allowed to demonstrate your actual worth. What wearing a suit really indicates is —I am not making this up— non-conformity, one of the gravest of sins. For extra excitement, the rules are unwritten and ever-changing, and you will never be told how you screwed up."
I've noticed this phenomenon seemingly distilled down to its purest form when fundraising especially.
Once you accept it for what it is, the process doesn't become anymore sane or reasonable, but it does at least become less disorienting and demoralizing.
Thank you, very nice article, actually this sentence made my day:
>The pioneers who are connecting the global human family and removing barriers of time and space won't take you seriously unless you brunch at the same restaurants they do.
On an interview, I was once asked, "Why shouldn't we hire you?". This got me totally confused, I was speechless for like 10 seconds. I don't remember the answer, it wasn't really anything remarkable. Still, I got the job.
I always thought "why should we hire you" is the new "tell me about yourself". Just an open, completely dumb question to estimate the ability of candidates to generally speak.
There are a lot of ways to translate these questions. As an example: "Tell Me About A Time You Failed". It can mean "Can you stomach failing until we figure things out?" or it can mean "Are you willing to be vulnerable by exposing your failures?" or "What kind of problems you perceive as a failure?". I can probably write another half a dozen of translations of such questions.
I don't think there should be a right/wrong translation. And based on that, there are rarely terrible answers.
These questions are open-ended on purpose, so people will talk and say things which are essential to them.
As an example: “I didn't get along with a coworker.” isn't a terrible answer. It means that interpersonal relations for this person are important. “I struggled to balance work and life.” isn't a terrible answer, but rather show that he/she prefers an environment where the work doesn't occupy all the time/energy.
The only answer which I wound consider bad is a dismissive answer (trying to ignore/deflect the question).
I'd say there's very few terrible answers, since terrible answers are the ones that dodge and deflect and most people don't do that. However it's still very easy to for a good answer to reflect poorly on your candidacy.
"I was terminated due to repeated negligence causing millions of dollars of damage" is a wonderful answer, it just torpedoes your chances of getting an offer.
Perhaps I have a vastly different idea of what these questions mean (though reading other replies here, I don't think I do), but the "translations" to these questions are just horrible.
"1. Tell Me About A Time You Failed
Translation: Can you stomach failing until we figure things out?"
No. No it doesn't. It's far more likely that they're asking you if you can identify that you've got something wrong in the past and have learned something from it. Rambling on about things that you didn't get wrong, but someone else did and were out of your control will only serve to make the interviewer wonder why you're avoiding the question.
"2. Where Do You See Yourself In Five Years?
Translation: Are you ambitious enough for a startup?"
Again, no. This is a simple question to find out what sort of direction you want your career to go in. It's a common question to help evaluate whether the company and you are going to be a good match. If you want to be a CTO and they want someone who's OK staying more hands-on, this is the sort of question to help make sure you're not going to be unhappy later on.
"3. Why Should We Hire You?
Translation: Do you really understand the problem we're solving?"
I'm really unsure how the author got to this conclusion. This is the same old question that's been asked for donkeys years, and the underlying reason for it (if any) is that they want you to sell yourself to them. Any halfway competent interviewer is going to have some much better questions for finding out if you understand the problem they're solving than this would give them.
Now, of course, different hiring managers are going to disagree with what I've just written there, and that's just fine! We're all going to have different motives for interview questions, whichever side of the table you're on. Point is, you can't "translate" common interview questions like the article attempts to do as you have no idea what the interviewer is actually looking for. Be honest, be friendly, and be yourself.
Tell Me About A Time You Failed Theirs: Can you stomach failing until we figure things out? Mine: When the shit hits the fan, do I have to worry that you will engage in defensive CYA?
Where Do You See Yourself In Five Years? Theirs: Are you ambitious enough for a startup? Mine: What are your long-term values? (For the record, I think that this is a terrible question. But I answer the translation honestly.)
Why Should We Hire You? Theirs: Do you really understand the problem we're solving? Mine: What value do you think you bring?
* the question
* Why we ask this question
* What a good answer looks like, and often what a bad answer looks like; and occasionally follow up questions depending on their answer.
This is nice because then we can all internally debate about columns 2 and 3, and any new question you want to ask has to come with a 2 and 3 yourself. We actually learn a lot about our own values with this method.
Disclaimer: I work at Strive where we run management workshops, including ones around evidence-based interviewing. If you're interested in learning more, reach out to me at jesse@strive.co!
An evidence-based approach (aka the scientific method) + a little knowledge about cognitive biases shows why questions like "Where Do You See Yourself In Five Years?" are not great.
All the research into interviewing and cognitive biases says that asking questions for "color" reduces interviewer's ability to predict outcomes while simultaneously increasing their confidence that they've gleaned information from the interview. There are experiments where interviewers are given some diagnostic information (e.g., hard facts about background) and conduct closed-form interviews (yes/no, this/that questions) with the interviewees responding randomly.
They'll still believe they've "learned more" about the candidate than they did from the candidates who they had biographical information alone.
A good interview process reduces an interviewer's exposure to non-diagnostic information and directs their attention so they only evaluate diagnostic information (and does so in a consistent way). This is the basis for the "structured interview".
"But Jesse," you say, "'Where will you be in 5 years?' seems diagnostic to me!"
Questions aren't diagnostic; responses are! So you have to think about the nature of the responses and how those responses will be evaluated.
Here are some alternate questions. Consider the responses you might get to these vs. the 5-year one. Are those responses more diagnostic? (Note: I just thought of these off the top of my head and haven't fully evaluated them).
- At your last job, what career goals did you have when you joined? Did you achieve them? How?
- Imagine you've been working here for 18 months. What would need to happen between now and then for you to not feel disappointed you joined?
- Think back to a manager who helped you define and achieve a professional goal (big or small). What was that goal? How did they help you achieve it? Why did that work?
As a first pass, these questions control for more and are likely to get higher-signal responses that are easier to evaluate in a consistent, interpretable way.
How can you tell the original 5-year question isn't great, though? Think like a scientist.
Why 5 years? Why not 1? Why not 10? 2? 4? How might those numbers change the answers you receive? Are you even testing to see how people's responses change as a function of that number? Are you even coding the responses in a way that would be able to detect how their responses change in response to that number? Is the variance in the responses small enough that you can even come up with a way to code them?
For example, if the interviewee is startup savvy and you ask the 5-year question then they understand the vesting cliffs are at play. Suddenly you're taking into account how savvy the interviewee is about the way vesting works. Was that the intention? If not: more variance!
Is it clear what the question is trying to get at? Do all candidates understand what it's trying to get at? Could candidate misinterpret the intent of the question given the fact that they're in an interview? Could different interviewers interpret the same response differently?
I notice the goals when asking the rote questions posted by angel.co and others in the thread, while very sensible, differ quite a bit.
And many interviewers are less capable and just ask these questions because those are the questions you are supposed to ask. They make up a reason for them after the fact. Like so much of hiring, it seems a fair example of a cargo cult practice to which a few smarter interviewers have invented a tolerable use.
For example, in a workplace that shall remain nameless, the question "Tell Me About A Time You Failed" really was to naively see how badly the candidate messed up in the past. And for "Where Do You See Yourself In Five Years" there was one correct answer: "manager" even though there was no expected need.
"Make Your Questions As Good As Your Answers" sounds like exactly the opposite of Spolsky:
> Some interviewees try to judge if the candidate asks “intelligent” questions. Personally, I don’t care what questions they ask; by this point I’ve already made my decision.
I maintain my belief that nobody in software has a clue how to do technical interviews.
I'm finishing my degree, haven't dealt with hiring yet but, god. Couldn't everything be easier if one could just say he wants the job for the money and wants to do a really good work for every cent?
Modern society is getting annoyingly over-complicated.
One of the reasons software engineers in general are underpaid relative to their value to a company is because collectively not only do we not understand the social aspect of employment but we deride it as beneath us.
Engineers need to ask a startup why the position is open. If you find out there were 3+ engineers in that position in the last 2-3 years then run away.
I can't stand this type of "gotcha" questions like "tell me about a time you failed". They're almost designed to make the interviewed misstep and, in so, waste everyone's time.
"Why should we hire you?" Is a weird one, specially if it was the company who contacted the interviewed. They should already know why they want to hire and instead ask specific questions to corroborate.
In general, this article assumes that the company holds all the cards and the applicants should be grateful (and waste plenty of time doing the "homework") just because they were given an interview. This is very often not the case.
80 comments
[ 6.7 ms ] story [ 253 ms ] thread'Plastiq—a startup that allows users to pay their bills with credit cards'
Is that a thing in America or someplace you can't use credit cards to pay your bills? I followed the link thinking that it would be something else but it seems that really that's what it does.
I always wondered if there was a regulatory reason for it, or maybe if it dates back to arbitrarily different company policies when they were independent of each other.
[1] https://www.soconngas.com
[2] https://www.uinet.com
We can gauge your ambition and your resilience just fine in your first few weeks (in which we've signed you on as a contractor-under-evaluation - if your situation allows). And yes, we've had to learn how to do this the hard way (having kept people on for longer than was good for anyone). But I think it's nuts to think that you can test someone's ambition asking "where do you see yourself in 5 years?".
I'm pretty convinced that that place deep down where people's natural drive comes from is, especially for engineers and other creative roles, very disconnected from conscious career planning. If you want to tap that natural drive, you need to create an environment where it can flourish. I don't think that asking "Why should we hire you?" when you mean "Do you understand what problem we're solving?" contributes to such an environment. You want to be open, clear, direct, and honest, not play guessing games.
Then we have the candidates that just blank or give a response like "I just want to make money". Not dealbreakers, but not great answers.
But I've also gotten a surprising amount of really terrible answers, like "at Google" and "ideally I finally paid off all my student debt so I can finally be a comedian". These are the people we don't hire.
I wonder why? Those are the most, if not the only, honest answers you got.
In current state of software industry, unless the candidate happens to believe in your mission, anyone answering they'll want to still work for you is either lying or naïve.
When they say "at Google", that's at least as bad as saying "working for you"; they know nothing about what working at Google is like any more than they know what working with us will be like. Plus, it implies "I don't want to work for you"; would you hire anyone that says that?
> When they say "at Google", that's at least as bad as saying "working for you"; they know nothing about what working at Google is like any more than they know what working with us will be like.
They do know enough. They know that Google is "hot". Pays well, gives a lot of street cred, opens a lot of doors. Unless you're one of FAANG, they can assume with high confidence that working for Google will be better than working for you. What they may not be confident about is whether they'd pass an interview there.
> Plus, it implies "I don't want to work for you"; would you hire anyone that says that?
IMO the very fact they showed up implies they do want to work for you now. But maybe not in 5 years. For a typical programming job, I don't see a problem. It's unlikely the person will stay with you for 5 years anyway, so why not treat this simply as exchanging professional services for money?
All valid things to want, but very little of it benefits us. Being a better engineer or manager benefits the company and would convince me to hire. Saying you want to work elsewhere eventually, but not being able to say why doesn't tell me why I should hire.
> IMO the very fact they showed up implies they do want to work for you now. But maybe not in 5 years. For a typical programming job, I don't see a problem. It's unlikely the person will stay with you for 5 years anyway, so why not treat this simply as exchanging professional services for money?
Sure, but they've just raised the bar a lot higher for themselves since now they have to convince us that they can hit the ground running and require minimal onboarding and investment. Very very few candidates can do that, and those that can have no problem getting a job at FAANG. Usually the ones that meet that higher bar are extended an offer, but so far all have declined in favor of a FAANG.
I think that this means there is a lot of good programmers with reasonable personalities around and companies are picky.
Because those are the answers you clearly want to hear.
"If they are willing to lie to us, they are willing to lie for us"
Apparently, is the paradigm business runs on these days.
They'll still believe they've "learned more" about the candidate than they did from the candidates who they had biographical information alone.
See the Dilution Effect (https://psychology.iresearchnet.com/social-psychology/social...) and https://en.wikipedia.org/wiki/Sensemaking.
Here's a fun paper: http://journal.sjdm.org/12/121130a/jdm121130a.html
A good interview process reduces an interviewer's exposure to non-diagnostic information and directs their attention so they only evaluate diagnostic information (and do so in a consistent way). This is the basis for the "structured interview".
For example, if the interviewee is startup savvy and you ask the 5-year question then they understand the vesting cliffs are at play. Suddenly you're taking into account how savvy the interviewee is about the way vesting works. Was that the intention? If not then you're measuring something without even realizing it Either way: more variance!
Is it clear what the question is trying to get at? Do all candidates understand what it's trying to get at? Could candidate misinterpret the intent of the question given the fact that they're in an interview? Could different interviewers interpret the same response differently?
It's a "bad" question because it controls for very little and therefore inherently produces non-diagnostic information (such is the nature of humans). The level of variance in all the components of this question is exactly the surface area on which cognitive biases become the determining factor.
And the nature of cognitive biases is that you become more confident when they're operational, not less.
You can ask questions about career goals without going at it sideways. Here's some off the top of my head. I haven't thought them through from the above perspective, but as a first pass they control for more, are likely to get higher-signal responses that are easier to evaluate in a consistent, interpretable way.
- At your last job, what career goals did you have when you joined? Did you achieve them? How?
- Imagine you've been working here for 18 months. What would need to happen between now and then for you to not feel disappointed you joined?
- Think back to a manager who helped you define and achieve a professional goal (big or small). What was that goal? How did they help you achieve it? Why did that work?
Most candidates get what we're trying to ask right away. Sure, sometimes we get a response like "well my current apartment doesn't have enough power outlets and smells like bread, so I would like to move and get a cat"; then we'll follow up with a surgically specific question.
It's a conversation, not an interrogation. The vagueness usually helps candidates open up and talk. Interviewing is stressful enough without the multi-part hyper specific word problem about something they probably haven't thought about.
We're interviewing people, and people are different with high variance. If we wanted consistency with bias removed, we can just craft a long multiple choice exam and have HR proctor it. Interviewing will always have noise, variance, and bias.
The goal isn't to remove noise, variance, and bias, but to control it in a way that produces a better assessment and a better experience for the candidate. This also allows the interview question + overall process to improve over time. Variance is present in every scientific experiment, after all, but the whole process of designing an experiment + evaluating the results is about controlling for known sources of variance and minimizing the ways unknown sources can get in.
For example, one might frame the structured process thus: "We ask an identical series of questions to every candidate so we know everyone's getting a fair shake. It's important to us that we're evaluating candidates based on their actual experience and potential, not on their ability to do well in interviews. We hate tricky or ambiguous questions because they make the people we interview anxious and produce answers that are hard to evaluate fairly."
You would want to do that before they came into the room, too, to control for the possibility that receiving that information in the interview room itself produced surprising interactions (more variance again).
In general, the candidate should have a clear picture of what will be happening in an interview regardless, to control for the above as best as possible.
(I can say from experience that if the above is done with a friendly demeanor and good bedside manner, the sense of relief in the candidate is palpable.)
Imagine asking a candidate a question about a positive interaction with a previous manager that gets at the thing you're trying to evaluate. The candidate is naturally going to be happy talking about a positive experience! The framing of the question is intended to get them thinking about the moments you want them to be thinking about.
"A couple of years" is not great because now it's a function of how the candidate perceives the phrase "a couple of years." More variance!What if one candidate specifies their impression of "a couple of years" and another doesn't?
Asking the candidate to pick a time period explicitly also not great because now your impression is a function of the (time period, response) they chose and not just the response. More variance, again!
It's better to ask questions structured in a way that gets at what you're really looking for. I don't know exactly what you're looking for, specifically — ambition, reasonable expectations around opportunities for and speed of promotion, alignment with company priorities, something else? — but there will almost certainly be questions that produce better diagnostic information and feel more natural to the candidate.
Inherent in the interview process is the feeling that you, the candidate, want to give the answer that the person is looking for. "Where do you see yourself in (5/a couple/choose-your-time-period) years?" hits right at that anxiety (which is why a blog post like "what are they really asking?" is even necessary).
Are you evaluating their ability to understand the question in the way you intended? If not, structure it in a way that reduces that possibility.
I have enough passive income that I don't need to work. This year I have decided that every cent (after taxes) I make is going to charity. It was a promise I made myself a while ago and want to follow up on it. Tomorrow I am going (after talking with them) spend about 5 grand buying things a local association needs.
If you ask me where I want to be in 5 years I will only say to make more money than I do now. Much more. I don't have goals in my job. I don't want to be a PM or whatever the fuck I should say. I don't need this job. I am trying to do as much good as I can. For a while I was in the top 5% earners in my country (maybe even 3% but data is hard to get by). Now the only reason I work is to give my 'sacrifice' a meaning. When I stopped working because I didn't need to, I started to drink and be a lazy slob, I was drunk pretty much every day, so I use work to keep me in track, but I give every penny away.
I don't care about work. I am a fucking amazing developer. I had offers from top companies I refused as it didn't match my lifestyle (lets say 2 or 3 of the FAANG brotherhood). I just don't care anymore. If you pay me, I do my job, amazingly well. But stop asking this bullshit question and try to infer meaning from the answers. Anyone can read a blog and say 'to grow as an X and be maybe a Team lead/management/suck CEO dick instead of lower management dick'. It is all bullshit and you and me (and the rest of the fucking HN) crowd knows it
I have someone on my team that literally said, I'm an amazing engineer. I don't need the money, I just don't want very much managerial responsibility, I want time and location flexibility to raise my kid and fair market pay for my work. You bet your ass we hired her. She's still an amazing engineer and wonderful addition to our team.
That's why we ask these type of open ended soft questions, people are different and have different wants and needs; we want to have that conversation before hiring to raise the probability that it'll work out.
I just hate this whole huppflah about hiring when I have seen companies miss amazing candidates because of these questions, and hire shittier ones because they 'studied' the right books/blogs. (I work with various startups here and help them with interviewing and building teams)
I will be honest, I am pretty bitter and is probably one of the reasons I am doing what I am doing now, but I do feel my reply to you was very 'directed' and I was venting more in general about the process and 'focusing' on you in particular. I am sorry for that.
The fundamental challenge of interviewing is that no matter what you focus on whether it's technical or social skills, an interview is not long enough to give a proper indication of how someone will perform day in and day out. As an interviewer the best thing you have are proxies such as work experience, and yes, squishy behavioral questions of dubious value—you still have to ask something. As an interviewee there is no way around learning the subtexts and getting good at answering them. In many ways it's a silly social dance, but when you get good at it you actually can learn a lot about companies as well, and improve your chances of getting hired at a place you'll really like.
This assumes as a hiring manager you've done your work to fully understand the need you're hiring for, not just the skill set. That you know how this need fits into the composite model of the team and the company. And, that you've prepared and trained the team to be able to assess and give you the specific kinds of feedback you need to derive from tbeir working alongside the candidate for such a limited amount of time.
There's no better similation or signal for reality than reality itself. You can't ramp a team or company at "hyper-growth" speed this way, but I'm of the opinion doing that sort of thing is almost never a good idea anyway.
Behavioral questions don't have to be squishy. There's a lot of literature (like, over a century) about conducting and evaluating behavioral interviews.
Here are some example questions one might ask a programmer:
- How do you know when your code is ready to ship?
- What methods or systems do you use to keep track of requirements?
- Of all your past managers, who got the most out of your potential? What made their approach effective?
- Imagine your professional life a year or two from now. What do you want that to look like and how do you hope this job would contribute to that?
- If you realized you couldn't meet an expected release deadline, what would you do? What have you done in the past?
- Tell me about a coding mistake you made in a past job that caused a serious problem. What happened and what did you learn from it?
- If we talked to your past managers, what would they say your strengths are? What would they say you need to work on?
BTW, if this stuff is interesting to anyone, I'm the head of instruction at Strive and we run workshops for engineering managers (including one on evidence-based interviewing). Email me jesse@strive.co if you want to learn more!
I once replied to that "as the CEO of this company". The CEO wasn't amused.
So only not currently employed people?
But even then I promise we're not going to ask "where do you see yourself in 5 years".
in the US (where I work) or in the NL (where you work), my employment is contingent on a visa, which is just way too much effort for a "maybe," position. Once you've ascertained I can program and don't appear to be an asshole, most other things can be worked out.
I've been in this industry a long time, I've been in startups and in big enterprises, and I've been a hiring manager in both cases and an executive in the former.
Very, very rarely would I say that my hiring manager or executive peers (especially of the typically inexperienced startup variety) would be asking questions like these with the awareness of intent and interpretation being claimed here. Mostly these questions are being asked because, just like the fairly pointless "whiteboarding" questions, they're just the questions everybody thinks they're supposed to ask. That is of course until an alternative model is presented and nurtured.
If you find a great hiring manager who is actually putting in a lot of forethought, intent, and consideration into their hiring and interviewing process, then do whatever you can to work with them. But, one of the ways you'll be able to detect such a person is they won't likely be asking cookie-cutter questions or giving you cookie-cutter quizes.
By 2034 the majority of proles were on illegal performance-evaluation enhancing drugs, necessitated by the fact that the average prole was also on The Company's indefinite performance improvement probation. They occupied over 20% of their working hours proving their worth to The Company, which had hired Prefects to wander the endless halls. The Prefects started administering the Voight-Kampf employee aptitude tests on proles, but soon began testing each other on their own effectiveness in administering the test.
Before The Fall, the end state of the system was that The Company existed only to vet its own employees, all of whom were in charge of secretly monitoring other groups of employees. They had their nominal tasks such as programming or planning, but employee evaluation was based solely on the ability to evaluate others in their nominal tasks.
As a Level 4 Artificial Learning engineer with Gamazo people always assumed that she was a Perfect. She wasn’t. She was a hard working proletariat.
“That’s actually a funny way to put it” she thought to herself as the words came out of her mouth. “Hard working proletariat”. Not perfect. Just hard working. “Interesting concept” she thought and stuck a fork in a salad bowl.
The salad was a 13 out of 10. The restaurant could do better. The sauce was a solid 15, the bread here was always over 100. But the lattice was soggy. Could it evaluate to less the 100%? She, obviously, wasn’t certified to evaluate that but Danny was. “Danny, what you think of the lattice?” she asked. Danny looked up from his own bowl of salad and gave her a thoughtful stare. She put her head down and kept munching.
It was a nice day. The sky was blue. The sun was warm. Warm but not too warm. “17 at least”. Jane cleared her mind and laid back, as she waited for Danny to answer. “Life was good with high probability of getting better” she thought. She smiled. Closed her eyes.
When she opened them again, Danny has yet to answer. Instead there was a man standing there right in front of her.
"The theme is familiar to anyone who's tried to join a country club or high-school clique. It's not supposed to make sense. You are expected to conform to the rules of The Culture before you are allowed to demonstrate your actual worth. What wearing a suit really indicates is —I am not making this up— non-conformity, one of the gravest of sins. For extra excitement, the rules are unwritten and ever-changing, and you will never be told how you screwed up."
http://carlos.bueno.org/2014/06/mirrortocracy.html
Once you accept it for what it is, the process doesn't become anymore sane or reasonable, but it does at least become less disorienting and demoralizing.
>The pioneers who are connecting the global human family and removing barriers of time and space won't take you seriously unless you brunch at the same restaurants they do.
I don't think there should be a right/wrong translation. And based on that, there are rarely terrible answers.
These questions are open-ended on purpose, so people will talk and say things which are essential to them.
As an example: “I didn't get along with a coworker.” isn't a terrible answer. It means that interpersonal relations for this person are important. “I struggled to balance work and life.” isn't a terrible answer, but rather show that he/she prefers an environment where the work doesn't occupy all the time/energy.
The only answer which I wound consider bad is a dismissive answer (trying to ignore/deflect the question).
"I was terminated due to repeated negligence causing millions of dollars of damage" is a wonderful answer, it just torpedoes your chances of getting an offer.
Because you need a programmer and I am one?
Stupid questions get you stupid answers.
"1. Tell Me About A Time You Failed
Translation: Can you stomach failing until we figure things out?"
No. No it doesn't. It's far more likely that they're asking you if you can identify that you've got something wrong in the past and have learned something from it. Rambling on about things that you didn't get wrong, but someone else did and were out of your control will only serve to make the interviewer wonder why you're avoiding the question.
"2. Where Do You See Yourself In Five Years?
Translation: Are you ambitious enough for a startup?"
Again, no. This is a simple question to find out what sort of direction you want your career to go in. It's a common question to help evaluate whether the company and you are going to be a good match. If you want to be a CTO and they want someone who's OK staying more hands-on, this is the sort of question to help make sure you're not going to be unhappy later on.
"3. Why Should We Hire You?
Translation: Do you really understand the problem we're solving?"
I'm really unsure how the author got to this conclusion. This is the same old question that's been asked for donkeys years, and the underlying reason for it (if any) is that they want you to sell yourself to them. Any halfway competent interviewer is going to have some much better questions for finding out if you understand the problem they're solving than this would give them.
Now, of course, different hiring managers are going to disagree with what I've just written there, and that's just fine! We're all going to have different motives for interview questions, whichever side of the table you're on. Point is, you can't "translate" common interview questions like the article attempts to do as you have no idea what the interviewer is actually looking for. Be honest, be friendly, and be yourself.
edit: formatting and duplication
Here are mine.
Tell Me About A Time You Failed Theirs: Can you stomach failing until we figure things out? Mine: When the shit hits the fan, do I have to worry that you will engage in defensive CYA?
Where Do You See Yourself In Five Years? Theirs: Are you ambitious enough for a startup? Mine: What are your long-term values? (For the record, I think that this is a terrible question. But I answer the translation honestly.)
Why Should We Hire You? Theirs: Do you really understand the problem we're solving? Mine: What value do you think you bring?
* the question * Why we ask this question * What a good answer looks like, and often what a bad answer looks like; and occasionally follow up questions depending on their answer.
This is nice because then we can all internally debate about columns 2 and 3, and any new question you want to ask has to come with a 2 and 3 yourself. We actually learn a lot about our own values with this method.
An evidence-based approach (aka the scientific method) + a little knowledge about cognitive biases shows why questions like "Where Do You See Yourself In Five Years?" are not great.
All the research into interviewing and cognitive biases says that asking questions for "color" reduces interviewer's ability to predict outcomes while simultaneously increasing their confidence that they've gleaned information from the interview. There are experiments where interviewers are given some diagnostic information (e.g., hard facts about background) and conduct closed-form interviews (yes/no, this/that questions) with the interviewees responding randomly. They'll still believe they've "learned more" about the candidate than they did from the candidates who they had biographical information alone.
See the Dilution Effect (https://psychology.iresearchnet.com/social-psychology/social...) and https://en.wikipedia.org/wiki/Sensemaking.
Here's a fun paper: http://journal.sjdm.org/12/121130a/jdm121130a.html
A good interview process reduces an interviewer's exposure to non-diagnostic information and directs their attention so they only evaluate diagnostic information (and does so in a consistent way). This is the basis for the "structured interview".
"But Jesse," you say, "'Where will you be in 5 years?' seems diagnostic to me!"
Questions aren't diagnostic; responses are! So you have to think about the nature of the responses and how those responses will be evaluated.
Here are some alternate questions. Consider the responses you might get to these vs. the 5-year one. Are those responses more diagnostic? (Note: I just thought of these off the top of my head and haven't fully evaluated them).
- At your last job, what career goals did you have when you joined? Did you achieve them? How?
- Imagine you've been working here for 18 months. What would need to happen between now and then for you to not feel disappointed you joined?
- Think back to a manager who helped you define and achieve a professional goal (big or small). What was that goal? How did they help you achieve it? Why did that work?
As a first pass, these questions control for more and are likely to get higher-signal responses that are easier to evaluate in a consistent, interpretable way.
How can you tell the original 5-year question isn't great, though? Think like a scientist.
Why 5 years? Why not 1? Why not 10? 2? 4? How might those numbers change the answers you receive? Are you even testing to see how people's responses change as a function of that number? Are you even coding the responses in a way that would be able to detect how their responses change in response to that number? Is the variance in the responses small enough that you can even come up with a way to code them?
For example, if the interviewee is startup savvy and you ask the 5-year question then they understand the vesting cliffs are at play. Suddenly you're taking into account how savvy the interviewee is about the way vesting works. Was that the intention? If not: more variance!
Is it clear what the question is trying to get at? Do all candidates understand what it's trying to get at? Could candidate misinterpret the intent of the question given the fact that they're in an interview? Could different interviewers interpret the same response differently?
It's a &...
And many interviewers are less capable and just ask these questions because those are the questions you are supposed to ask. They make up a reason for them after the fact. Like so much of hiring, it seems a fair example of a cargo cult practice to which a few smarter interviewers have invented a tolerable use.
For example, in a workplace that shall remain nameless, the question "Tell Me About A Time You Failed" really was to naively see how badly the candidate messed up in the past. And for "Where Do You See Yourself In Five Years" there was one correct answer: "manager" even though there was no expected need.
> Some interviewees try to judge if the candidate asks “intelligent” questions. Personally, I don’t care what questions they ask; by this point I’ve already made my decision.
I maintain my belief that nobody in software has a clue how to do technical interviews.
It's 404. Can somebody provide a working URL? Or is finding one is itself the exercise and you are supposed to hack around the error someway?
They need you more than you need them. (And if not, then you made bad life choices and need to recalibrate yourself)
"Why should we hire you?" Is a weird one, specially if it was the company who contacted the interviewed. They should already know why they want to hire and instead ask specific questions to corroborate.
In general, this article assumes that the company holds all the cards and the applicants should be grateful (and waste plenty of time doing the "homework") just because they were given an interview. This is very often not the case.