Hiring advice for startups from Hackruiter (YC S10) | Hacker News Reader

Hiring advice for startups from Hackruiter (YC S10)

hackruiter.com

95 pointsnicholasjbs22 comments

This is all totally sane. Virtually everything on our "careers" page (which is far from perfect) came from feedback from people who passed on us. The single most important factor we've identified so far is responsiveness.

It is so easy to inadvertently go dark on a candidate.

These guys observe that employees are almost never as dedicated to the company as the founders. A similar thing happens during recruiting, except it's the employer who isn't as dedicated as the candidate. For most companies, the simple truth is that your decision on your next hire is not as important to the company as the decision is to the candidate. Your next hire is unlikely to totally transform your company. But it will certainly be transformative for the new hire!

As a result, you should assume time is moving 10x slower for candidates than it is for you.

The only thing I'd add to this set of advice (ie, the only bit of advice I have that I think ranks up there with theirs) is, debrief the people who pass on you. Actually debrief everyone. Find out what people hate about your hiring process, so you can start fixing it. I've found that nobody minds questions like these, so go ahead and ask.

...debrief the people who pass on you.

That's an excellent point. I'll add it. (It's a historical accident it's not there already: We wrote this advice for the companies we work with, and since we ask candidates for feedback on behalf of the companies, we didn't think to tell our companies to do it, too.)

Also, I really like how transparent your jobs page is about how your process works. That's something else we should add: Tell people what to expect, how long it usually takes, etc.

Thanks for the suggestions and kind words!

...debrief the people who pass on you.

Right, I'd be so thrilled to provide feedback because companies have the courtesy to do the same to the candidates they pass on. Back when I was job hunting for 6+ months I would consider myself lucky if I got an almost automated "We're sorry to inform you that..." email, most don't bother even with that.

Fuck them and the horse they rode in on, leave them in the dark.

> Right, I'd be so thrilled to provide feedback because companies have the courtesy to do the same to the candidates they pass on.

They don't, because anything they say is a lawsuit waiting to happen.

> Fuck them and the horse they rode in on, leave them in the dark.

That a personal perspective. Say some big company Google let me hanging for 2 months before sending me an automated "we are sorry" mail, I will be mighty pissed, but I won't take it out on some other company which was responsive.

On a related topic, I've been thinking about the interviewing problem and had an insight that is probably blindingly obvious to everybody else.

In engineering terms, interviewing is benchmarking. Making an effective benchmark is somewhere between very hard and intractable.

A benchmark has to be...

* Difficult enough to clearly differentiate between good and bad yet be simple enough to "run" in 5-30 minutes.

* Repeatable, yet not beatable. Remember the compilers that recognized drystones and compiled them to the final answer? That is what google does to interview "trick" questions.

* Applicable: drystones are totally worthless if your algorithm is dominated by floating point calculations. Having said that, "comes up to speed quickly" is the most applicable metric for a candidate, but is awfully hard to benchmark in a limited interview.

Anecdote: I had lunch today with the engineer that hired me on my first "real engineering" job (I was a freshman in college at the time). He commented that I was "the smartest guy he knew, until he met my younger brother." My brother is several years younger... Bob was working for a different company and my brother applied for a summer job there. Bob's take, and the other folks at the company's take, on my brother's resume' was "meh". Bob told them "never mind his resume, hire him. Just do it! Do. It."

Moral of the anecdote: benchmarks are poor stand-ins for real life. Unfortunately, I don't have a better way. :-/

Good points. Your annecdote about your brother is captures the truism that interviews and resumes are only a substitue for knowledge of someone's real ability. This is why references from employees (smart engineers know others) is much more effective.

"Don't write inflated and ridiculous job posts; they're a turn off to good people."

I have noticed the following things about tasteless tactics: 1) they're often horrifically effective; 2) people with taste will pronounce that they're ineffective without providing any evidence.

You're right, we don't have real data. It's all anecdotal. But we speak to a lot of hackers, and we've yet to meet a good one who's said he likes those posts, and we've met plenty who hate them.

See, that's the thing about tasteless tactics -- the fact that nobody says they like them cannot be taken as evidence that they're ineffective.

Are they effective to the right people though?

That's it. They end up hiring people who think they're ninjas and rock stars. If all you need to do is fill a seat (traditional HR or recruiters) then this ad is probably pay-dirt - after all, there are a lot of people who want to be special.

In looking for a place I'd like to work I look for ads that say they understand what I'd be doing.

Also, there are two ways of asking for everything under the sun. The first, which rockstar/ninja companies ask for, is "7+ years Java, 3+ using ANT, 5+ in etc". You can never satisfy the requirements of this kind so they only get scammers or the truly naive who qualify but only because they once did all that stuff 10+ years ago.

The second way, which hacker companies use, is to say "We use X, and do hard-thing-Y in it. If this is your speed, join us. Bonus points for things we don't know about." These are not only clear and precise to the people they want but easily lend themselves to entrance interviews.

I finished up a job search a month and a half ago, and there were several places I didn't bother to apply to because I found their job posts annoying.

Pretty much anything than mentioned being a "ninja" or a "rockstar" was ignored immediately, and there were a few others I discarded out of hand, too for similar reasons.

Other than that, I'd just like to echo the point about responsiveness. There was one company I'd have loved to have worked whose job description I fit very well, but they took almost three weeks to respond to my application, and by then I'd already had an offer I really liked and I'd already written them off completely. There were a couple of others that said to bug them if I hadn't heard back in a while, and I mostly just stopped paying attention to those too.

They're horrifically effective at what? I kind of doubt the answer is "matching the job with a great candidate quickly." I would imagine it's something more like "getting a lot of responses," which IMO is the opposite of what you should really want out of a hiring call.

This ^^

Also, horrifically effective sounds actually like a bad thing. As in, they're incredibly effective at getting an obscene amount of poor job candidates that it's horrifying.

> We can't recruit for you, so please don't ask. We hope you find the advice below useful nevertheless.

Does this mean they're closed to new clients? I've never seen a recruiting agency do that.

Edit: they are. http://news.ycombinator.com/item?id=2979874

Interesting that they're full up already.

I guess that's either a good sign for their business or a bad sign for their business model :-)

Honestly, it's a little of both.

I'd kinda suspected. I've spent a lot of stray moments thinking about the whole "recruiting is obviously broken" thing (and congrats for taking it on). And I have yet to come up with a genuinely good solution.

I put together something called "wantmyjob" that used the OKCupid solution with making and answering questions about your workplace. But that really won't work for startups since they don't have a bunch of existing employees to recommend them. I'd also probably have to commit the cardinal sin of trying to charge the job seeker -- not so good.

Similarly, there are a lot of places that will certify that you (as an individual employee) aren't a bozo. Those don't scale terribly well, but worse, they also charge the job seeker.

You can make money but not do the world much good by charging (probably bad) employees a small fee to apply or charging bad companies a lot for dubious access. But while that's probably where the money is, it does the least good for the world of almost anything you could do in recruiting. Maybe a "schlemiel match-up service" for mediocre-to-bad companies with exactly-equally-bad employees? Don't know. I'll bet you guys wouldn't be passionate about that one ;-)

And yet, how do you make enough money and still scale well on matching up the best companies (who don't have to pay as much to hire) and the best candidates (who should never, never be charged anything at all)?

It's a hard problem. Glad you're thinking on it.

I put together something called "wantmyjob" that used the OKCupid solution with making and answering questions about your workplace.

Funnily enough, the idea we applied to YC with was "OkCupid for jobs."

It's a seductive-sounding model, isn't it? I couldn't find a way to make the business model work. It sounds like you've departed from that model pretty completely (not a bad idea, I say now that I've tried it out).

Yes, it is seductive-sounding. And yes, we switched to a very different idea shortly after YC started.

I still think "OkCupid for jobs" has merit, but it has some serious problems, too, not least of which are that it only gets really valuable after lots of companies and people are using it (and we didn't have a good way to bootstrap, at least from the job-seeker end).

Another issue is that matching isn't the really hard part of hiring -- the hard part is finding good people.

One of the lessons of the mortgage crisis is that a rating service whose paying clients benefit from inaccurate ratings will become worthless. I wouldn't trust a bozo filter that takes money from bozos rather than me.

Reply on news.ycombinator.com