Tech companies can't find good employees, and it's their own fault | Hacker News Reader

Tech companies can't find good employees, and it's their own fault

inc.com

51 pointsCorrado39 comments

I suspect at least half the problem is that their recruiting and filtering practices are utterly broken. Like, you go for a job that requirements wise you seem like a perfect fit for, then get rejected for seemingly no reason? And then the job you meet maybe 1 requirement for offers an interview? While a bunch of others just ghost you altogether?

I'm not sure how CVs are being filtered out, but it feels like it's completely broken to the point of absurdity.

It doesn't help that the interview process is often ridiculous, and feels more strenuous than the actual job. Like, I've had take home tests which were basically 'build a minimum viable product', of the kind that would take a good week or two of work at the actual company. Or times where a very junior job was throwing Hacker Rank type puzzles at people for a job making WordPress sites.

Meanwhile, if it's not too hard its too easy, or completely divorced from the work required, or run on a platform that's so paranoid that any sort of window change flags the user as a cheater or the interview process takes weeks on end...

Is it any surprise companies are struggling to find employees so badly? How many great engineers are going to put themselves through this sort of crap?

It's utterly insane, I've been on this for 20+ years, in the early 00s the interviews seemed much more related to the job, you'd have a chat, explain what you've done before, get some curious questions about how that went; dive a bit deeper when the interviewer got interested, maybe showcase some portfolio if that was possible. It felt much more similar to how architects (real ones, not software) and other creative technical professionals interview.

Then Google started the trend of grilling people on trivia, puzzles, and a whole industry followed suit "it's the Google way! They have the best people!", the cargo culting was ridiculous, when some not well funded startup wanted to grill you on puzzles of how many window panes exist in Manhattan and bullshit like that.

Nowadays it's an exhausting game, I don't have the patience anymore to study just for interviews, it's not like I can practice that on my day-to-day, the knowledge rots after a few years on a job where almost nothing of those questions matter, solving algorithmic puzzles from Leetcode don't motivate me at all and it's absolutely useless for any other part of my life except doing these bullshit interviews.

It's broken beyond repair, it's just a game (mostly numbers): study some interviewing book for 3-6 months, apply for N jobs, go through the lottery and find a job when you don't fumble some puzzle, get interviewers that aren't assholes, and for some reason they didn't fill the position before you went through the 4-7 stages of interviewing.

An absurd waste of resources for everyone, for what? So companies can increase a small sliver of a percentage of finding some wizard kid that will perform 2-5x better than the average worker? Fuck off...

There's the pain of having an unfilled position with work that isn't getting done. There's also the pain of making the wrong hire and having to let someone go.

Most of the people I've spoken with on this issue find the latter to be much more painful. Therefore, they end up over-correcting with an overly arduous hiring process.

I think what they aren't giving enough weight to is the fact that the hiring process might actually scare away the best candidate.

As another example of how whack the job market is, recently I wrapped a year-long job hunt. When I tally my offers, I got offers from mid-level to Staff, and the highest offer was 2.5x higher than the lowest one.

The fact that different companies can have such a radically different opinion of the same person tells you how much of a crapshoot hiring in tech is.

The wide range of offers you received may not necessarily reflect a "crapshoot" in tech hiring. Instead, it likely demonstrates the diverse needs, prioritizations, resources, and role definitions across different companies. Especially if the openings varied in company size/location/industry/division/business area/etc.

For example, one hiring team maybe prioritized specific technical skills for a certain role, which affected your outcome in that particular process.

RTO hasn’t helped.

4+ rounds of interviews, panel interviews to make sure coworkers like your personality and you look a certain way. Take home tests, no feedback on why they didn’t pick you, and stupid questions like what’s your hobby, what book did you read last, do you watch sports.

RTO for giants is a must because they die without the office. Offices are the last meaningful dense-city economical activity, if you cut them there is no more economical reason for dense cities. With dying dense cities the service industry will collapse because if you live in a single family home it's unlikely you buy from Just Eat or get an Uber car, similarly if you WFH you unlikely buy trendy business dress, shiny mobile bugs/smart-devices and so on, you also do not see physical ads, witch are shop windows, by commuting by feet downtown.

Essentially if we WFH for all eligible jobs we re-create a SMEs economy killing the giants, a thing very positive for the middle class and the society at a whole but certainly not something the big of finance capitalism want.

> Offices are the last meaningful dense-city economical activity

Yup. there's no other money that gets spent by people living in a city. No cafes or stores of any kind or theaters or nightclubs or anything. we should all just pack it up and go off into the wilderness. there's no reason why people want to live conveniently near other people.

some people aren't city people, and some people are. trying to forecast, thinking that everyone's just like you isn't a winning strategy.

Of course there are many ancillary activities that could exists outside dense cities as well, because hey, if you like a restaurant you like it both in a big city, a small village and in isolate countryside. What's change outside the city is that you typically know the restaurants around by name, so you call them directly not ordering food via some apps. Such restaurants tend to be familiar activities not large enterprises owning many identical places in the world. You also need dentists both in the city and outside, the sole difference is that instead of having 100 dentists in a small area you have one per small village (for instance) but the number of people needed them does not change, so their business does not change much. Long story short ALL business existing in a city can exists outside except those of big actors who need the city scale to exists.

> there's no reason why people want to live conveniently near other people.

I've left the big city for the Alps and well... I have MUCH more social life now than in the city, simply because being spread means being much less choosy, so instead of isolated ourselves in a bubble of few friends ignoring the rest of the crowd, we almost all meet almost anyone else. Instead of going together somewhere to consume a service like going to a shopping mall you go together in nature or in some friends home for a home party. A much more social activity alone.

> some people aren't city people, and some people are. trying to forecast, thinking that everyone's just like you isn't a winning strategy.

That's not the point, the point is IF the modern city is sustainable and my answer is no. We can't evolve it, there is no green new deal future for dense cities, while there is a foreseeable future of crescent poverty, issues, unrest and so on. Personally I dream a starship to travel the universe, but it doesn't matter because I can't have it. The point is what we have and what we can or can't have. Giants need the city, but we can't have it in the modern time, so giants will disastrously fall in a decade or so, probably with a world war in the middle, because current economical model can't keep up. Current society can't keep up.

Do you mean that there is a secret quarterly meeting where Starbucks are lobbying RTO policy to Apple in order for "us big sharks to thrive"?

Believe it or not, once you start hanging around in management circles, you'll learn there are modes of communication that despite not involving direct exchanging of words between two individuals, nevertheless, conveys information. In fact, people are selected for the management class many times for their lack of acknowledgement of said mechanisms, as it gives everyone else the cover they need to play ignorant when, having toyed with the levers of power causing something bad to happen, they can play the ignorance card.

So no. They are not asserting that everyone has a secret meeting at midnight. They are pointing out the incentives and dependencies that lead to a pathological state; one that creates a hell of a pot of institutionsl inertia against a WFH paradigm shift.

I pointed out the lack of description of those incentives.

I think not, there are some official meetings, like those hosted by the WEF, casually renamed https://www.worldgovernmentsummit.org to be a bit more modest but there is no special need to meet someone like you to understand your common interests. If you are a blacksmith you know your "cohort" interests without the need to meet some competitor/colleagues. You might occasionally meet them anyway, for lobbying, for a common cause (like the farmers against John Deere for the right to repair their tractors, or to be more precise the tractors they have bought but they do not really own anyway), some meetings might be public, some others behind closed doors, but the common interests of a category are well known by all belonging to that category.

The cohort who seems unable to understand it's own common interests are the 99% of the people who apparently, as usual, fails to see where are their enemy and their potential allies, regularly fighting against their own interests...

I suggest if you have a bit of patience an old book from 1841 by Clinton Roosevelt, The Science Of Government, Founded On Natural Law, it's strange at first bug get clear quickly and it's very fast to read: https://dn790002.ca.archive.org/0/items/sciencegovernme00roo... try and you'll see in a succinct and crystal clear the TODAY world, economy, with anything we have. Another good reading would be Eduard Bernays Propaganda and to complete the game a bit of modern network theory like some Albert-László Barabási.

Sorry, something was probably lost in interpretation. You wrote:

> RTO for giants is a must because they die without the office.

Are you referring to tech giants here? That's what the thread was about I believe (interviewing for tech roles, at least). Just to make sure - you are not including Starbucks here.

> Offices are the last meaningful /.../ [stuff that I probably agree with] /.../

> Essentially if we WFH for all eligible jobs we re-create a SMEs economy killing the giants

This is what sounded like conspiracy to me. Would RTO due to "saving Starbucks" really be an argument line that Apple pursues?

> Are you referring to tech giants here?

In particular, because they are the most impacted, but it's still valid for all big financial actors starting from real estate.

> This is what sounded like conspiracy to me.

A conspiracy typically is made by some who want to overthrow and substitute a power not by the those who rules against their subjects, that's instead called political agenda. As opposite to a conspiracy it's mostly public, the "sharing economy" is described everywhere, the industrie 4.0 is even a book (more than one), and so on. The agenda can be described simply as "you'll own nothing" where the "you" part means those who pronounce such sentence count to own anything and "rent" anything. In the IT world the cloud+mobile "the sole integrated platform" (actually not at all, but that's what advertised) have already "users" who own essentially nothing and very few giant owners, big enough to steer the IT world simply with their developments, in the physical world it's slowly happen "hey, try our new shiny autonomous taxis", "you do not need a car, just use our app and a car will arrive to bring you whatever you want", obviously the toward-rent trends in cities for both commercial and personal real estate that obviously does not fit much in a spread society, the ready made food downtown for the commuters instead of eating at home what personally prepared, to the point many new city "goshiwon alike" condos have essentially no kitchen but just a small fridge for bottled drinks and something to quickly heat already made food, no dishwasher, common washing machines with some kind of card-based auth to use in the condo basement and so on even "smart locks" for apartments and main entries. It's definitively NOT a conspiracy since it happen under plain Sun light publicly advertised as a good and needed evolution.

But it's not that easy to made people owning nothing, accepting things like abolish the concept of inheritance (a relatively recent PR trend, with some famous testimonials) before, stating that anyone should start with his own arms in an open arena, now stating that successions taxes should exists and be higher and higher, selling "property rights" time limited at 99-years and so on. Remote work is the rock in a running mechanism, because if it spread it obviously push most remote workers outside the expensive, degrading and chaotic city, and actually in the western world potential remote workers are a very BIG slice of the population, also a wealthy and acculturated slice. So the obvious political need is to stop it or the entire big actors common interest informal agenda will derail.

Speaking as someone who is currently hiring, you have to go through a vast number of resumes from completely unqualified people via sites like LinkedIn. Too easy for people to just push the button to apply on the off chance. Working with a good recruiter is better since they do some of that filtering for you.

Once you have some candidates, most of them are exaggerating various aspects of their experience. The only way to figure this out is to talk to them, so interviews are necessary.

First is just a quick screening call to see if they have any idea what they're talking about. If they pass that, you need more of a deep dive with an SME. This takes time from your talented SMEs (they have a job to do!) so you only want to pass them actually possible candidates.

After that (depending on role) the company may want interviews with senior management, and HR usually want to weigh in on one too.

After all that, it still does not produce reliable results, although I suspect they would be even worse without them.

Open to other ideas on how to do this effectively.

What if we just threw the whole "process" away and hired a bunch of applicants at once? Give them 90 days to prove themselves (smart, friendly, knows how to write documentation, etc.), keep the ones like and let the other ones go.

I'm kinda only half joking.

Wouldn't work in practice for all kinds of reasons, but it did make me smile :)

we do that for very jr roles over 10 wks, it's called an internship.

What you described is perfectly logical and should be standard practice.

Some key pieces left out are time frames and human factor involved.

If you actually want to hire for any role, like you actually want to put a body in that role and you’re not just doing busy work, then each candidate should only take 2 Weeks to go through the process you described.

The people you have internally interviewing candidates should also want to fill the role and not be gatekeeping against good candidates.

Good point about keeping the time down for each candidate to go through it.

Gatekeeping won't be a problem for us, but also a good point in general.

Do you find screening calls helpful? If possible, I’d always prefer meeting people in person instead - changes everything.

I do think they're helpful, yes. The main idea is that they are fast and efficient to get the candidate list down. Meeting people gives much better signal, but takes more time and committment from both parties.

They can tell you who shouldn't be in the set with reasonable probability, but not who should be employed Sort of like a bloom filter ;)

While I agree the hiring pipeline has become much more difficult, employers have also become much more risk averse because a bad hire can absolutely slow down deliveries and have a net negative result org-wide.

It's a low trust situation on both the employer and employee fronts with bad actors tainting each other's opinions.

I'm not sure there's a way to fix this aside from relying heavily on referrals and hiring known quantities within your network.

That said, I absolutely agree with his opinion about not cargo culting BigTech style interviews - they are in a position where they can use it to choose candidates. Other companies may need to tailor interviews to meet their own needs (and most do ime).

And finally, we are now in an actually global hiring market. This isn't like 25 years ago when tools like high bandwidth internet, Zoom, Slack, Confluence/Notion, GitHub, and other productivity tools for asynchronous teams didn't exist, and more critically - plenty of hires from CEE, Israel, and India who worked in the US have started moving back to the Old Country, meaning there are now employees who can bridge the management gap across the globe.

The downside is now the labor market has become globalized and Ron in SF is competing with Radosław in Warsaw, Ramon in Costa Rica, Ra'an in Tel Aviv, and Raghu in Hyderabad.

I appreciate your selection of names appropriate to each location haha

Geeze. Imagine that. Could it have anything to do with their brutally medieval interviewing cycles?

I’ve hired at a tech company, and it’s definitely weird. The problem may be non-tech employees involvement in the process.

For example, we needed a part time dev. A few days later we had a skilled candidate with a decade of experience in our exact tech stack. No brainer hire, right? Nope. HR and finance person thought he was too expensive ($20k/yr too expensive) and scared him away with clumsy negotiating. Months later and several failed offers later, that job wasn’t filled. How much money did that non-hire cost the company? Far more than the little bit extra compensation they should have paid.

Another piece of the puzzle is where and how job listings are posted. LinkedIn, Indeed, and the others are good at generating a massive pile of resumes. The question non-tech employers need to ask is if great candidates are even on those sites looking for work. They aren’t part of the culture, so they wouldn’t even know to post on HN “Who’s Hiring” threads, or sponsoring a local meetup, or one of the more tech-focused job boards. Great technical employees aren’t commodities, and you don’t get them the same place you can hire forklift drivers, dental assistants, and office managers.

It is probably worth noting that great technical employees aren't necessarily obsessed with tech in their day to day life, and don't necessarily follow the community online like a hawk. So while it's definitely worth posting job listings on Hacker News or sponsoring meetups, it's also worth noting that you'll still need to post on more traditional job boards to catch applicants that aren't all that interested in the social side of things.

I understand that companies need some sort of SAT/ACT-like method to rank candidates. I completely understand that part since there aren't really licenses in software engineering like other engineering fields. But I feel like a better way would be to present candidates with questions and problems actually related to the job. For example, if the job requires an understanding of Pandas, perhaps giving candidates actual tasks in Pandas would be better than asking them to traverse a binary tree.

The licenses associated with other engineering fields are optional and largely unrelated to the skill and expertise of the engineer.

There are many fields of engineering where practicing without a license is against the law, with licensing exams directly related to education and skill.

Technology has more or less co-opted the term but it used to be that “engineer” was fairly synonymous with licensed professions.

Where is this? My experience is in the US and my background is physical engineering. A license does not confer competency. Competency is not testable except at the most rudimentary levels. I wouldn't trust an engineer with anything important because they were licensed. Licensing is about liability, not competency.

It would be useful if we could devise a filter for engineers in software but it hasn't proven effective in any other engineering discipline in my experience. Software is even more prone to cargo-culting fashions than more traditional engineering disciplines and I would hate to see them enforced as "best practice" with the force of law.

They don't want to hire someone who knows Pandas specifically. They want someone who can identify this technology, learn it and then possibly implement it. And more importantly - someone who can do the same when the situation calls for some other technology. This is the crux imo.

I truly believe if you were trying to invent the WORST possible way to hire people it would look very close to what they've implemented.

Is it their fault? I regularly see college students cheating on their CS homework/projects using LLMs. They probably use them on interviews as well. In response I think companies will have to make the process of getting a job more difficult and drawn out.

Agreed. The hiring process is a total nightmare these days.

Broken job boards, A.I-CV-scraping, soulless video interviews, ridiculous noob “tests” for experienced folks, endless interview rounds + bad communication.

It’s really no surprise most tech companies can’t find anyone.

Even if you’re talented and have a lot of experience, they manage to break you during the process.

Few notes:

- from the first part of the article I feel a sort of hidden critic of automation and it's outcomes, well, while people should ALWAYS remember the Ironies of Automation by Lisanne Bainbridge [1] it's obvious that as much as we automate as less humans we need to do something and similarly that we loss diversity/leveling anything toward some "standard solutions". This is to be corrected with openness and experimentation, meaning investing more in research and development, with long terms goals, without managerial practices, not to be avoided trying to stop the history course;

- from the conclusion, well, I read only a simple point: to get works done companies need to know how to work, this it not much a single policy fault, it's the byproduct of modern universities and schools with their target to create replaceable useful idiots instead of technicians because they are more manageable. It's WAY TOO LATE to correct the course changing policies, we need SOME GENERATIONS growing in new-old schools whose target is forming Citizens not Ford model stereotypical workers and we should remember a thing very clear: intelligence alone is like an engine, useless without fuel/energy. It's easy to loose and very hard to make. So the actual situation can't be corrected quickly nor easily nor without an IMMENSE effort with a gazillion of false start and issue before reaching the goal.

[1] https://ckrybus.com/static/papers/Bainbridge_1983_Automatica...

Sounds like Darwin will solve the problem!

Reply on news.ycombinator.com