Does communication matter in technical interviews? Here's the data
blog.interviewing.io
blog.interviewing.io
The article concludes
> The bottom line: it is far less damaging to score 1 rating point lower on Communicate, and it is much more beneficial to score 1 rating point higher on Code and/or on Solve. Talk is indeed cheap, and indeed, coding and problem solving is what you need to show in a coding interview.
which is the opposite of what the two current HN responses predict.
Or were you saying that at some point you stopped caring about coding ability and valued strong communication even in the absence of the ability to problem solve and write code?
I think it's close to this, but not quite as it's a sliding scale and not a step function.
From a business perspective, there tends to be diminishing returns as one's coding skills improve. And that's the point where returns on communication skills tends to become more important. So more often than not the higher the level the more important communication skills become over coding skills. This is not saying a high level engineer can be all talk and no ability, but rather a great communicator/good coder might become more valuable than a good communicator/great coder.
Apologies for all the hedging in the above. I'm trying to take into account that there are always exceptions to the above. Some positions need that absolute wizard. It's the exception to the rule, but it exists.
I was being a bit tongue-in-cheek there, but I will point out that folks who are good coders and great communicators often find roles that are more rewarding in every way outside the IC SWE universe, and so are probably poorly represented in this sample…
For context, I'm in fairly mundane web app development, and technical issues (getting stuff to work at all, or perform well enough) have hardly ever been the main problem on my projects. The hardest part is almost always figuring out what to build.
The "Does communication matter in technical interviews?" is not even deserved. It's a terrible article, most likely an ad, with an artifical conclusions about a low quality biased data source composed of a one-sided, subjective star post interview star rating.
As for my experience, I've been hired to perform technical interview on behalf of companies, I hired people for my companies, and I've been interviewed many time.
And the answer is: it depends.
It depends of the person, of the job, of the context, and what you consider communication.
I have an ex that is incredibly incompetent and not very bright, but she is amazingly sexy and dresses well, and gets hired very easily. I have another one that is the best at her job, she is an angel for customers, but she is shy, doesn't explain things well, and is not as sexy. She tanks interviews regularly. Is that communication?
When I was young, I traded my contact lens for glasses and noted that, as a geek, people took me more seriously in interviews with glasses. Is that communication?
Last month, I interviewed a guy who said "I like to work at home, this allows me to be at ease when I want to code 12 hours straight". I knew he was boasting - people capable of focusing for 12 hours are very rare - and rejected him because I want to avoid BS. But he came from a trading firm, where this may be what you are supposed to say. Was it good communication?
"Communication" is not unidimensional. It's what you say, and don't say, when you say it, how and to whom. It's also what you convey by you body, your look, your voice. Being late is communicating something. Saying "I don't know" is communicating something. And the context in which you are saying it is as important as what you are saying. A good job here could be a social mistake there.
A few star rating is communicating to me that one knows nothing about the topic.
Then there's the crowd which will kneejerk towards "rather someone who communicates than a jerk who won't communicate". Really, most people qualify for the former regardless of their technical capabilities. It's a comparison which barely happens in real life. Most "toxic" people are very easy to filter. Most people don't lack communicative skills. Job adverts and interviews very clearly filter out technically deficient people (why else do you think many of them are looking for i+2 people for i skill level jobs?), and don't optimize for quick, communicative learners at all.
That's not to mention communication skills are much more difficult to quantify than tech skills. One's "he's poor at communication" is another's "he's great at communication". See endless discussions about meetings and meetings being held as a "good at communication" standard.
Being able to connect with and clearly communicate to your interviewer will absolutely increase your chances.
I've found interviewing via video call is absolutely terrible compared to in person.
So damn hard to describe architecture without a whiteboard. And humans just seem to communicate better when you're in the same space.
Most fundamentally, of course the data shows that “would hire” correlates to “appears to have the skill this interview is specifically intended to assess”. These are not independent variables, nor are they objective measures. Don’t waste your time analysing them as if they are.
From a candidates point of view, the question is “what can I spend time on to improve my chances of getting the job I want”.
I’ve never met anyone who’s preparation is analogous to the scenario presented, “I decided to convert a communications point into a coding point.” Developing complimentary skills is rarely a zero sum game.
The data appears to show that unless you are top ranked on two different technical criteria, poor communication skills will prevent you from getting the job. Data-wise, we’re starting to see the impact of having a variable that isn’t almost perfectly correlated with the outcome. That’s less likely to mean “talk is cheap” than “there’s an easy opportunity to improve your performance; you’d be a fool to dismiss it”.
I would also suggest that earning a high score for problem solving is going to benefit significantly from good communications skills. Asking good questions and explaining your process, whether or not you could code the solution, goes a long way.
So sure, if you are extremely skilled on the technical side, and the company hires based solely on the technical interview, and that company is offering you the best of all the potential roles you could take, better communication skills aren’t necessarily valuable to you.
For everyone else: career tip, of course you should be applying for jobs where your skills are relevant to the role, but also remember that general communication skills are valuable.
The problem is the reverse is probably as common, but, honestly, equally valuable: if you have a person who simply shows amazing tech skills, and talks when prompted to, I mean, will you reject this person in favour of a less stronger candidate just because that weaker candidate was more outgoing?
I mean, in technical interviews you have to be technical and solve your stuff, that's it.
FAANG system design interviews actually somehow require some clarifying questions to clear out ambiguous requirements or probe for missing data, but, I have seen and witnessed people immediately tabling X, Y and Z assumptions and go "assuming these....." and immediately craft a perfect solution - pass with flying colors.
The rest will be gauged in behavioral interviews and in the final decision meeting to put forward an offer or not.
But in technical interviews? Just solve the thing.
If they are the opposite, they will not ask questions (but instead make assumptions) and refuse to admit when they're wrong. Yes I would absolutely reject such a person regardless of technical skills, I would even fire them.
These people can have a very negative influence on the culture and turn it into a competition of ego instead of actually understanding and solving problems, which is also stopping everyone else on the team as well from actually solving problems, because this person will act as a huge blocker. Technical ability doesn't help you at all if you are insisting on solving the wrong problem.
> But in technical interviews? Just solve the thing.
Yes sure, but the issue is that there is no non-technical part.
> You must live in an all-white people world
is kinda ill-mannered considering half of white folks don’t speak English. You might be conflating the Anglosphere with all white folks.
No but being quiet unless prompted to speak, will make a team work environment dysfunctional.
You're overestimating the proactivity of most people, and how important that proactivity is beyond making an impression one is proactive at a job interview (read: faking it). And the willingness of others to endure what a large crowd of truly proactive people would do when constrained by a huge management overhead: complain a lot until they are fired or finally given the power to solve their own problems.
Yeah, for example the military and factory work. But not team work environment, it does not function this way.
Most people aren't proactive. End of. To think otherwise is to dismiss what modern schooling does to individuals and to forget the downside of proactivity: increased friction when people can't agree on what they want. If we agree communication is important, surely you're not going to dismiss the obvious that most people suck at trying to come to an agreement when everyone is trying to play proactively.
There's a reason management methodologies are so tight on synced up moments of conversation and prodding people to talk even when they themselves like there's nothing to be said. It's all to push people into talking. You can do that with reactive people. It's not optimal, but it works and is a lot more feasible than trying to create a culture with mostly proactive, total strangers who somehow agree enough not to devolve into death by committee.
The assumption is that you have a proactive team work environment, for example scrum, which is how the vast majority of all programmer jobs are, and that the people involved can all speak fluently in a common language.
The ESL argument is pulled in not because of language, but because some cultures prefer reserved people over proactive people. Not sure how you missed this.
"Why team work itself sucks" is a huge leap in logic from "reserved, reactive people can function in a team". Taking your own example, a military with bad team work would leave its country ripe for takeover.
There's nothing about software development in particular which makes reactivity dysfunctional. What is true is proactivity is preferred when applied in a certain way. Turns out many proactive people are proactive for reasons which make them butt heads whenever they disagree. Your "proactive but also collaborative" people are unicorns, because the very things which make them proactive either cause friction, or aren't rewarded appropriately.
>for example scrum
If anything, SCRUM shows exactly how we really think about proactivity. Scheduled rituals which prod people to respond, trying to get some semblance of that "proactive but collaborative" gold, only for most people to still feel suffocated by management when it turns out that proactivity includes futile efforts in telling management to change what is preventing the team from working more efficiently. You can call this dysfunctional SCRUM, misapplied or whatever you wish, but the far majority of these places still turn profits and create products despite of this.
I probably can’t be convinced that individuals who keep to themselves will create a dysfunctional work environment.
If I’m reading your statement as strict as you wrote it, I’ll mention I have never worked with someone who felt they could not speak unless spoken to.
From my experience:
- being willing & able to "bring along" the interviewers, thinking out loud
- being easy to follow (clear phrasing, logically chained etc.)
- asking for help/input when needed
Without defining what exactly 'communication' means it's impossible to measure. Even then, it's highly subjective - that doesn't make it less important though.
Indeed, at best, this is not about how well a candidate does at interview depending on how good at communicating they are.
This is about how companies rate canditates on communication depending on if pass their interview or not.
You can't post a comment on HN without people obsessivelly asking proof about your opinion, numbers, studies, data...
And this reach the frontpage. How ironic.
I think mock interviews are great, but do we have any solid data on how well the "advance to next round" flag for a mock interview correlates to advancing to the next round in a real interview?
If I'm conducting a real interview, I'm going to be thinking about what it would be like to work with this candidate on a daily basis (or at least have them somewhere in my organization). I would think I would be more concerned about communication skills in a real interview, even if it's unconscious. With a mock interview, I'm in no danger of working with this person. I can be focused completely on their technical solution without thinking at all about what it would be like to work with this person day to day.
From what I understand, they do a lot of calibration to make sure the "advance to next round" flag correlates with their clients' version of that flag.
All that said, the interviewers they work with are all also interviewers are FANGs and similar -- the likelihood of working with someone you interview at a company with 10000s of employees is...about as low as for a mock interview.
You've definitely identified a possible confound, but I think the i.io team is aware of it and does a lot of mitigation (for many reasons, not just for this one report).
But we know the brain is terrible at assessing its own decision process.
The hiring party could very well have not like the person, or be attracked sexually to the person, and felt that it was all about communication.
Post decision star rating is junk science.
Looks like as usual, "more research is needed".
Unless it's clearly defined I don't see how anyone can measure it in a meaningful way.
This seems to be treating communication as some kind of qualitative abstract.
"On a scale of 1 to 4 stars, how much did your spouse love you last Wednesday?" That tells me more about the esteem of the respondent than it does about the state of the person we're trying to evaluate.
I think I would have done better if I had taken the advice to communicate less during the code review and explain after I solve.
Who thinks up such a question? It’s like asking, “Is the sun hot” and then writing a blog post investigating the obvious.