Ask HN: For those trying to hire how hard is it to find people currently?

> A "cover letter" is a terrible prompt to elicit any meaningful information about a candidate, and implies negative things about the company asking for one.

On the contrary, a cover letter is a powerful tool to elicit meaningful information about a candidate’s soft skills, in particular empathy (for the hiring manager) and communication, and refusal to provide one implies negative things about a candidate.

> I think if you want to assess writing skills, you should ask more focused questions, such as "What is the biggest thing you have built in x language", "What is your experience with y"

A good candidate does not need those prompts. Having read the job description, he has an idea of what the hiring manager wants, and he writes about it.

> "write me a thing to tell me how much you want this job"

That’s not a cover letter. I don’t care how much a candidate wants a job. To be honest, neediness is a turn off. I care how effective the candidate will be at the job, both technically and socially.

I feel that perhaps you and others think all a cover letter is, is ‘Company X is such a great place! I want to work at Company X so much! Please pick me!’ If so, then yeah, I 100% agree with you. That’s useless input to the hiring process. But that’s not what a cover letter is IMHO.

As mentioned elsewhere, I assert that a cover letter is the soft skills equivalent to the resumé. The resumé lists relevant skills, education/training and experience. Those are all good things, but they are in the nature of checking boxes. The cover letter is a soft-skills demonstration. It’s also an opportunity for a candidate who is borderline in one area to draw attention to the fact that he more than makes up for it in another.

> > The cover letter is a proxy for soft skills.

> Not anymore. Not with ChatGPT.

I don’t think so. I don’t care if a writer uses spell check, grammar check or ChatGPT: good cover letter is a good cover letter, regardless of what tools the writer used to help.

> In any case, you have completely missed my point, right along with e12e.

What was it, then? That the best candidates consider cover letters nonsense? I disagree: I am asserting that a candidate who considers cover letters nonsense is ipso facto not among the best.

Maybe this is related to a misunderstanding about cover letters? They’re not about ‘I love your company!’ or ‘please please hire me!’; they are about ‘here’s why I believe that my skills, education and prior experience suit me for this opportunity.’ The resumé is what; the cover letter is why. The resumé is facts; the cover letter is judgement.

I used to work in the cleared space -- i.e., people with top-secret security clearance. My first job in this space was at AWS, where we probably had about 1/4 of the qualified, cleared talent pool in our space. It was almost possible to hire people with sufficient technical skill who also had clearance -- but it was equally hard to get people cleared, because we and Microsoft were pushing so many people through the pipeline that clearance decisions were often taking 2+ years. We would have people get hired, get a promotion, and leave all before getting their clearance decisions.

My team started taking a different approach: We started hiring people at the lowest level we could ("associate", below even entry-level engineer), with minimal tech skills, but with active clearance and some adjacent skills. STEM background, or tech writing, or just someone who was a hobbyist. I started interviewing people on troubleshooting -- someone told me a fantastic story about troubleshooting a comms issue while deployed in the middle east. The fix involved copious amounts of superglue on a ground-based satellite. Switch out the words, and you could easily have a story of a grizzled old sysadmin patching in production.

We actually ended up being pretty successful. Sure, a substantial portion of our hires were still misses -- but that was already true. Meanwhile, we hired some really talented people from this route, who I ended up trying to poach years after leaving AWS, and leaving the cleared space entirely, because they were just genuinely great engineers who needed to be given that chance.

tl;dr, I agree, and it's nearly always a good idea to look at expanding the talent base in some way or another if you're struggling. In many cases, it means hiring people who have potential, as long as you have sufficient numbers of senior staff to help get them up to speed.

Out of all the places I've worked, I think my favourite approach was Dropbox's: They ask these algo questions for new grads, because every new grad from a computer science program has been through these classes and it's an easy thing to index on.

But for experienced hires, nobody gets algorithms questions. Instead, you get practical questions -- depending on the role, things like writing an ID allocator, finding duplicate files across a filesystem, demonstrating understanding of concurrency, etc. No "implement this well-understood algorithm that you'll never implement again" questions.

Sought after candidates is not how software hiring works. Maybe start ups flush with cash are an exception, but for everyone else hiring almost always is dominated by perceived compatibility more than anything else, even more than utility need.

That is because there are smart people in the world that can self educate and jump into new technology with minimal friction, but those people are rare. Most employers want employees as early as the position opens and they don’t want to invest in training. Java is super desirable to employers because many developers are taught Java in school, it’s single paradigm, and it’s terse OOP conventions force the vague shape of a uniform architectural style. Most developers only know enough to be hired and not enough to write original software, of any style, so employers have to account for these things as though everyone is a junior developer, least common denominator.

That line of thinking suggests lower expenses, ease of candidate selection, and a wider candidate pool. In reality it tends to result in a less qualified larger labor pool with greater insecurity. This is why everything on the JavaScript side is about giant frameworks and the software is super bloated, because putting text to screen is all that matters but nobody knows how the technology works without a framework.