[frognumber]

One of my top lessons I've learned doing successful startups:

1. Recruiting is the #1 job of any startup CEO, and the #1 determiner of corporate success.

2. Up market, down market, side market, it doesn't matter: You will get better employees if you treat candidates with respect and you will be more competitive.

3. It's a lot of work for the 95% of clowns out there you interview, and there's a push towards automated process, but it will hurt your business.

4. There's a lot more to recruiting than just treating candidates with respect. It involves how you present yourself as an employer (participating in conferences / meetups / etc.), how you find candidates, checking references, reviewing github repos, etc. It's a crazy amount of work.

5. This is hard, but if you can do this, you will have a huge edge.

The flip side is that as an employee, doing a good job interviewing / recruiting, especially at a big company, is one of the lowest value-add tasks you can bring on, from a purely selfish / incentive structures perspective. This friction, I think, is a major reason why recruiting is handled so badly. There is absolutely no upside to doing a good job, and it takes a lot of time to do so.

[fnordpiglet]

Almost every place I’ve worked (early at startups later at megacorps) there were never enough recruiting resources. So eventually I just got my own recruiter account on LinkedIn and started using my knowledge of the industry and state of hiring to make my own search queries and reached out directly. I had an incredible conversion rate - the senior person actually hiring saying “hey you look like you could fit in here” was very powerful. It was a hell of a lot of work, but I could fill seats faster than anyone around me with high quality people and my projects landed successfully. At a certain point I became too senior for that as my direct teams became smaller and more senior, and I’ve never been able to convince a single other manager to do it. Not a single person. I’ve since switched back to IC at a super senior level so am even further removed from direct recruiting, but I can see even more broadly. All I see now is engineering managers whining about recruiting, and I’ve still yet to see another manager take direct control of their recruiting. Why? The only thing I can guess is they don’t actually care about being successful at what they’re doing, they just care about successfully doing what they were hired to do, which doesn’t include recruiting.

[gustavus]

I have a story from a client I worked with. They had 3 or 4 positions they needed filled for their team ranging from Jr to Sr level. But because they worked at a big company they couldn't do the recruiting directly, everything had to be through HR. While the problem was their assigned HR person was unresponsive and slow on the uptake and then took a huge vacation. So they just started looking at the applications that had been submitted through the portal and asking the division's secretary to reach out and start setting up appointments. Time to interview and make a decision on people went from literal months to 2-3 weeks.

Unfortunately this isn't a happy ending to the story. HR threw a hissy fit that they were being sidestepped (because they were completely incompetent ninnies.) And management had to come down on the manager who was doing it and tell him to follow the process.

[mikestew]

My observation has been that if the company has fewer than 100 employees, yet has a director-level or VP-level "...of HR", you will watch candidates one after another slip through your fingers. When you don't have enough employees worthy of a "VP of HR", then that VP will find busy work to justify their existence.

My poster child is a company of about 30 that had a "VP of HR". I can't count how many candidates took another offer while she sat on her ass "checking references".

[plugin-baby]

100% with you, but it’s not always easy to get control of your own recruiting. I’ve also seen HR get upset when their involvement is not what they expected.

[ineptech]

Agreed, which is insane! We'll spend untold zillions fine-tuning and A/B testing the process by which we get users, but when it comes to getting coworkers it's "Well HR says the applicants have to make an account to apply so I guess that's just the way it is".

[mschuster91]

It will be that way inevitably as an organization grows, usually it's an investor requirement to prevent outright nepotism hires or discrimination. No one cares too much for the first few dozen employees but anything larger usually will come with "ffs go and implement at least some basic hiring standards to reduce the legal risks".

[frognumber]

For what it's worth, in the US, the legal risks are negligible.

So long as you don't discriminate against a protected class, you're usually good to go. Nepotism is bad business, but it isn't illegal.

Investors do care about bad business decisions, so it makes sense for them to require good hiring practice. However, if a sole proprietor -- or a business with all its shareholders -- decides to hire inept cousin Vinny to keep him from idling about the house, that's usually legal. In a family-run business (where there is a paper-thin wall between business and personal finances), it may even make financial sense.

This differs from almost any other jurisdiction in the world.

[fnordpiglet]

Note I used HR for everything other than sourcing. The bottleneck is usually identifying people and initial phone screens. The rest of the process is legitimately HRs job.

[ivan_gammel]

This. I gave up working with external recruiting agencies even if budget allows it for exactly the same reason. They just can’t beat me and my team in productivity and opportunity costs are high. Investing 40 hours of personal time to close a vacancy 1 month earlier means being able to delegate 1 month worth of work earlier.

[ylee]

>Almost every place I’ve worked (early at startups later at megacorps) there were never enough recruiting resources. So eventually I just got my own recruiter account on LinkedIn and started using my knowledge of the industry and state of hiring to make my own search queries and reached out directly.

I've never worked for a tech company. But I have worked for two bulge-bracket investment banks.

In the first case I'm pretty sure the person looking to fill the role personally contacted my college's recruiting service (because she had received her MBA from there). In the second case I saw a Craigslist listing either by the person himself or his assistant, talked to someone I knew at the firm who worked near him, then reached out to the person directly. In both cases the person hiring did my first (and, in the second case, the only) interview, and I never talked to a true HR person, let alone recruiter, before receiving the offer. Does this sort of hiring never happen in tech?

[JohnFen]

Yes, it happens in tech. It's not even all that rare, really (at least outside of the FAANG crowd). I've reached a place in my career where I'm well-established and known enough that I don't really need to hustle to find good dev jobs anymore, but for most of my career, I got most of my jobs by being my own recruiter in that way.

[gumby]

> and I’ve never been able to convince a single other manager to do it.

I agree with everything you wrote, but if your company is big enough to have in house recruiters they probably don't want you doing it!

Not because they don't want to be able to hire great candidates efficiently. But because you presumably don't maintain the statistics they need to keep. For example, say (I don't know you and am just making this example up), you might without realizing it only be contacting candidates who have names that you recognize as male, while they want to make sure they are pulling from a wider pool to improve their chances of getting the best candidate They may have other reasons for collecting such statistics (SEC filings? Who knows?).

[fnordpiglet]

There are only a few bigger companies than the ones I’ve been at.

I only do the sourcing, and the HR side I leave to recruiting. The bottleneck in the process is the sourcing, and I have no desire to do HRs job.

On statistics, they only care about the statistics for their own sourcers and processes. There is no regulatory reason for collecting such stats.

Another thing to remember is HR is the weakest organization in any company, and recruiting the weakest in HR. Any manager worth their salt pushes recruiting around, not the other way around.

[gumby]

Ah, you're doing the sourcing part, which indeed is the thing i-house recruiters are worst at (despite being compensated just for that!). Sorry I misunderstood.

With reference bonuses, I think most companies do appreciate having their employees do sourcing, which makes your observation that you can't get others to do it all the stranger.

[fnordpiglet]

Almost all companies prevent hiring managers from receiving bonuses for their own referrals. I was hiring my own team, rather than using internal or external recruiters.

[datadrivenangel]

If you want something done right you often have to do it yourself.

Especially when the communication required to delegate it well is more work than just doing it.

[benhurmarcel]

My experience in a large corp is that whenever we can hire, we never have any issue finding very good candidates. The difficulty is getting approval from HR and management to hire externally.

[oramit]

In the last year of my last job I took on recruiting and mentoring tasks as one of my goals for the year. Management encouraged it, I got great feedback all year, and in my annual retrospective I highlighted it as a big win. When the annual review came around though it got no mention and I was basically punished for taking on those tasks because my billable time went down a small amount, even though we all knew (and agreed) that would happen. That, and other reasons, are why I found a new job the next month.

My experience is the same as yours OP. Hiring just isn't treated with any respect and your career will probably suffer if you take it on.

Now that i've finished complaining... I think there's a good reason for this behavior. In the US, where most of our posters are from, you can fire people for any or no reason. It's true that you can put in more effort at the beginning of the hiring process to find better candidates but you won't really know if they are a good fit until they work there for a while. You can have great candidates on paper who don't work out in person, and terrible candidates on paper who are great on the job. This randomness to the hiring process means that people don't treat it as a real discipline. And if you do hire a dud, you just fire them. Is it really any wonder that most recruiting processes are so callous?

[mschuster91]

> This randomness to the hiring process means that people don't treat it as a real discipline. And if you do hire a dud, you just fire them. Is it really any wonder that most recruiting processes are so callous?

It's just the same in Europe with way higher standards on recruiting (i.e. anti-discrimination laws are actually enforced) and employee protection laws. Recruiting, accounting and IT are usually seen as a "cost center" by the remainder of the employees instead of being respected as vital contributors to the business, so it's inevitable that people eventually "check out".

[tracerbulletx]

You'd think if professional sports could figure out how important this is we could too, but for some reason there's a tendency to give it lip service then put absolutely no effort into doing a good job. Scouting and screenings are done by people that companies treat like dirt, we incentivize them only for body count, interviews are done by low experience employees and employees who don't even like interviewing. We don't know how to identify and cultivate talent in this industry yet, it's clearly dysfunctional and deprioritized.

[nradov]

In professional sports there are objective metrics and performances are largely in public. In knowledge work there are no reliable metrics that apply on an individual level and you have to rely on candidates themselves (plus perhaps some unreliable references) to understand work history. It's a fundamentally harder problem.

[tracerbulletx]

I agree it is much harder, but no less important. Engineering orgs at most companies are no where close to even being good at identifying success or talent in their existing employees.

[Solvency]

And yet those sports teams still spend millions of dollars and expertise on it (recruiting). So what's techs excuse? It's harder so we don't bother?

[LargeTomato]

We can watch players play previous games. We can see their progression as a player over time.

For tech, it's hard to know if someone is good in just a couple hours of interviewing.

[azemetre]

Why could you use someone's previous work or portfolio?

[nradov]

The best developers don't have any portfolio that they can share. It's all locked up in former employers' code repositories.

[nradov]

It's not an excuse. Everyone in tech acknowledges that effective recruiting is important. But beyond doing basic stuff right like not ghosting candidates, that knowledge isn't actionable. No one has found a reliable, repeatable, scalable solution. If you can figure that out then you'll be a billionaire.

[x0x0]

> We don't know how to identify and cultivate talent in this industry yet, it's clearly dysfunctional and deprioritized.

Do you even watch professional sports? Professional sports is not great at identifying or cultivating or recruiting, and the incentives there are far simpler, and the performance metrics generally easier.

Take the NBA. A handful of teams are famous for cultivating talent, but mostly because the modal nba team is terrible at it.

Even the best regularly completely mess up. The team of folks that put together the Warriors -- a franchise that dominated the nba for a decade -- completely blew a #2 draft pick, who is close to being out of the league. They gave Jordan Poole a huge contract and then were forced to trade him because he decided to stop playing defense and start taking terrible shots. He's busy being a tank commander in Washington.

Hell, Michael Jordan -- my take for the greatest of all time, and, at worst, the 3rd best basketball player ever -- famously didn't go #1, and that's with one of the best college coaches of all time (Bobby Knight) telling anyone who would listen that he was an extraordinary basketball player. Hakeem went first (ok, that's not a disaster) and a complete bust went second (complete disaster).

Lots of GMs struggle with really basic roster construction issues (Russel Westbrook on the Lakers). etc.

iirc, only 4 of 30 coaches (Pop, Spo, Steve Kerr, Malone) have held their jobs for more than 4 years.

edit: It's very common for top-25 all time players to not be drafted first, or often, even all that high. Steph Curry, with a decent shot at top 10 all time: #7. Jokic: 41st (!!! -- essentially every team passed on him). Giannis: 15th. Luka: 3rd, after winning euroleague mvp at 18 (yes, I'd confident he'll end top 25). etc.

[philwelch]

In professional sports you are selecting one of the few hundred or thousand best people in the world. Its not even remotely the same impact.

[tracerbulletx]

Every successful athletic team makes player scouting and development a core function, not only the elite levels. In the US that means minor league baseball teams, university teams from volleyball and fencing, to football and basketball. Pro cycling teams that pay almost nothing and aren't competitive world wide.. It's not just the elite teams that make this top priority. They ALL know that the most driven talented players they can get at their level is what will make them win.

[philwelch]

Most of your examples are talent pipelines for the top leagues. Regardless I think you made my point for me at the end:

> They ALL know that the most driven talented players they can get at their level is what will make them win.

This is not how normal employment works. Sports are competitive by nature so the difference between a 60th percentile player and the best player is the difference between winning and losing. But for most businesses, the difference between the best frontend developer and the 60th percentile frontend developer is close to zero.

[tracerbulletx]

The difference is being Stripe where part of the reason they won was that they are considered extremely excellent at developer experience and technical execution. Or Netflix where they beat all of the legacy companies to a great platform doing something no one had done before, and retained the advantage to the point where they seem like they're going to make it through the die off of streaming platforms.

[philwelch]

Neither of those examples is particularly convincing. Stripe succeeded because it tackled a famously difficult and annoying set of business problems; the technology is important and they're reputed to be top rate, but that's not why they've succeeded as a business. Netflix has no technical moat either; there are half a dozen streaming services that, on a technical level, are completely interchangeable with Netflix. The only difference between them is their respective content catalogues, and while Netflix probably has some advantage in being able to drive content decisions with customer data, that only gets you so far.

[realusername]

> There is absolutely no upside to doing a good job, and it takes a lot of time to do so.

I've also learned this the hard way. I've conducted about a 100 interviews in 2 years and didn't get compensated any for it despite being one of the most critical part of the company.

Conducting interviews is also very tiring and time consuming, I'm estimating that two interviews in a day and your day is gone. I also evaluate it 2x more tiring than coding personally.

It wasn't a complete waste of time though, I got a lot of experience from that which will be very valuable in future management positions.

[1920musicman]

I have the same experience. Interviews are very time consuming (prep 30 min, interview 1h, fill out the feedback form 30m-1h), and having several interviews each week means I spend ~1 day weekly on something that's not going to benefit me directly in any way (excluding the benefit of potentially working with good engineers I helped hiring).

So unless the incentives change, I don't see this process improving in big tech.

[realusername]

Yes exactly, the interview itself is a bit less than half of the work surprisingly and then you do need a real break after all of that very intense concentration anyways. 1 interview = roughly half a day gone, that's what I've experienced.

And then it's indeed never valued inside the company, worse than that, it might be counted against you since you will achieve less in your team where all the evaluation will take place...

I really don't understand why companies don't value engineers capable of conducting interviews because it's really not an easy task, you need much better than average interpersonal skills and much better than average tech knowledge as well.

[1920musicman]

Agreed. I mostly conduct system design interviews which already have a smaller pool of interviewers at my company. This contribution has been included exactly zero times in the countless review cycles I went through.

[varispeed]

> and didn't get compensated any for it

So why would you do it? Such a big red flag. It means that employers will expect doing work for free (in some countries this is illegal) and potential employees should know about it.

[realusername]

I'm doing it for my future career, not for them that's for sure!

Despite not helping in this current company, conducting a lot of interviews taught me a lot for sure.

[varispeed]

Fair point.

[sokoloff]

I think the last paragraph is blunted in companies where teams recruit/interview for their own team (as opposed to recruiting/interviewing for the company in general).

In the former, if I do a good job, at least I get better colleagues on my team, making my daily life better and giving my team more chance to be seen as successful.

[varispeed]

You forgot about fair remuneration. Most employers think that they are doing potential employee a favour and the employees should kiss their feet just for reading the offer.

I don't respond to job offers that don't include _minimum_ pay (not "up to") and I don't respond to low-ball offers.

Fair remuneration should also include true equity in the business as otherwise it is exploitation, as at software business the profits are typically not linked to rewards for people creating those profits - it's all get creamed by shareholders.

So if your company generates millions or billions and you only offer a salary? No thanks.

[tayo42]

> is one of the lowest value-add tasks you can bring on, from a purely selfish / incentive structures perspective.

At one place I worked, interviewing was one of the check marks you could participate in for promotions.

[1920musicman]

But that's exactly what it is then: a check mark. I can interview dozens of candidates, and then add a total count to my self review. There is no incentive to do a good job (how would that even be evaluated?..).

[filoleg]

> how would that even be evaluated?

Depends on the company. At my current workplace, the candidate is given an option to provide feedback at the end (either through a form or an email), and all interviewers are also required to submit written notes on how everything went down.

Given that a candidate at the onsite will get interviewed by 4-5 people, with each of them providing a very detailed set of notes, it would be fairly trivial to smell out a misbehaving interviewer, if one cared to do so. What actually happens at the end of the day with those notes and candidate feedback, that’s the part i am not sure about. Once they get submitted to the hiring committee (or HR), it is out of my hands.

But just saying, they do have ways of evaluating it, just on a less precise scale and more on a “bad/good enough/amazing” scale. With only the “bad” outcome raising any eyebrows/having any meaningful effect, and with 99% of them getting the “good enough”/“amazing” ratings. And how often the signal for that “bad” rating gets caught is also not something I know much about.

P.S. Your assessments and notes are all preserved in the centralized hub, so you (and some others) can always access them later as well. And, sometimes, you indeed have people checking them out for assessment or such. Especially during your first couple interviews, you have a person supervising you and taking notes in parallel as well, and then you discuss them and they give you improvement suggestions and such.

[1920musicman]

I am sure what you are describing is being tracked at my company. The concern with misaligned incentives I have is that there is a vast gap between conducting a passable interview in terms of engaging with the candidate and actually investing yourself in the process. So it's not really about being a bad interviewer (as in rude, openly biased, etc) but about having the energy to do the best job you can - which all candidates deserve IMO.

[ivan_gammel]

>It's a lot of work for the 95% of clowns out there you interview, and there's a push towards automated process, but it will hurt your business.

If 20 candidates enter your funnel before you hire one in a screening-technical-offer flow, that’s maybe 5-6 hours per week spent on recruiting in 1 month. You can even send a few personalized rejections to worthy but unfit applicants, sending 80% of applications to trash. The effort is noticeable, but it is not a lot of work with good automation, planning and interview design.

[TheCaptain4815]

This is so true it hurts. If you get burnt once with hiring, it’s something you’ll never forget but could destroy the startup/small business in the process

[drewcoo]

> treat candidates with respect

> 95% of clowns out there you interview

What's your company affiliation?

[LargeTomato]

I worked at a rocket startup. 1/3 ex SpaceX. 1/3 ex Google/Facebook.

I interviewed a guy and asked him, essentially, "find the biggest number in a 2D array". This guy spent half an hour struggling because he "wasn't sure how to look through the grid in a circle pattern".

You'd be incredibly surprised who gets interviews.

[filoleg]

> "wasn't sure how to look through the grid in a circle pattern"

Now I am curious to know whether I am too dense to get it or if the candidate was just that off the mark.

Is circle pattern sorta like iterating through a 2d array like a spiral (i.e., outer layer of the 2d-shape first, then one deeper, etc.)? And if yes, why would that ever be useful for just searching for a specific value in a 2d array?

I get how it could be useful for some more niche/specific problems where the layering of the 2d array would actually matter, but is it just entirely off the chain to recommend it here? Because I cannot for the life of me figure out why you would want to do that instead of just iterating, especially considering how significantly less trivial it is to code-up that “circular” iteration (as opposed to just a regular linear iteration).

Sidenote: Is there even a more efficient way to solve that problem, other than just sequentially iterating through the 2d array and simply tracking the value/position of the largest number until you finish iterating over the entire 2d array (assuming it is non-sorted)? It seems way too simple, so I feel like either I am missing something about the problem statement or there is a better solution than the one I proposed.

[LargeTomato]

This was part 1 of a 2 or 3 part question. First we literally iterate through each element, keeping track of the max. I use this question for interns, too, so it's intentionally super easy. It gets harder in part 2.

For clarity, the pseudo code solution to the question is

    for row in a[0]:
        for e in row:
            max_so_far = max(e, max_so_far)
    return max_so_far

No tricks. Just an initial weeder question for interns before we move onto the real question.

[filoleg]

Ah, makes perfect sense, thank you for clarifying. Your pseudo code solution confirms that I understood the problem statement correctly.

Out of pure curiosity, what were the follow up (part 2 and 3) questions? Not looking for a solution, but if you could post the problem statement, it would be very appreciated. If you feel uncomfortable sharing it publicly out of some concern, that’s entirely fair, no worries.

[chrismcb]

I think the point of the OP was A the person wanted to solve this in a spiral pattern, which isn't necessary. And B didn't know how to do a spiral pattern, which is fairly simple. As to your other question, unless it is directed you need to look at each element. The real question is how to iterate over a 2d array. There are several methods and really that is all this question is looking for.

[icedchai]

Most day-to-day coding is simple and boring. Your interview questions should be, too. I've interviewed 100's of candidates over the years. Many of them had trouble writing a couple of for loops. This was for a somewhat similar problem. I would stress "we don't need an optimal solution, we just need a working solution."

[1920musicman]

Honestly, at this point I stopped making strong conclusions about candidates based on a single interview. I won't recommend to hire but also won't judge their abilities overall. Interviews can be very stressful, candidates overthink and often get fixated on a random solution they think the interviewer expects, etc.

I had several odd experiences myself in the past, as a candidate. The funniest one was when I interviewed at a prestigious company I thought was hiring only top talent. I spent an hour trying to come up with the most efficient Sudoku solver, got completely stuck on some arbitrary algorithm that I came up with on the spot. It wasn't a "circle pattern" but close to that. Wanted to impress the interviewer and also did not sleep the night before overthinking the process.

[FredPret]

You can and should treat someone with respect, even if your expectation is that there's a 95% chance they will turn out to be a clown.

[foldr]

IMO this way of looking at it is indicative of a lack of respect for candidates, even if communication is superficially respectful. If someone performs badly in an interview then by all means don't hire them, but it's both unkind and irrational to jump to the conclusion that they're a 'clown'. If someone is making such a harsh judgment about 95% of their applicants, on the basis of an interview process which we know to be highly imperfect, then I would not want to work for them.

[michaelt]

He works for Clownr, the Uber of children's party entertainment.

[1920musicman]

I had a good laugh reading your comment and then found this: https://www.clownr.com/

[arcanemachiner]

> Are You a Clown?

> Say hello to you new best friend.

[geraldhh]

your last paragraph resonated very well with me as i can't remember the last time i spoke to a recruiter that seemed to have a horse in the race.