Hacker News Reader

Comment by asdfologist | Hacker News Reader

I disagree.

All companies have only a finite amount of time/money to spend on hiring, so they're often forced to use crude heuristics to filter out resumes, especially in companies that get a huge number of applicants every year, like Twitter. Maybe it's just me, but based on my experience, filtering candidates (as a first pass) based on school or previous role does yield many false positives and negatives, but it still works pretty damn well compared to other equally fast/cheap filtering methods.

View on HN

I can get why managers think the important thing is to save managerial time. But as somebody who has just gotten done with a round of hiring developers, I believe this is a false economy.

If I can hire just one good developer, I'll have added thousands of hours of effort. If I can find one who settles in, it could be tens of thousands of hours. If I choose especially well, I'll get somebody who acts as a multiplier on those around them. If I choose poorly, I could get somebody who's net negative, making messes and driving away good people.

Given what's at stake, I can spend what seems like ridiculously large amounts of time on hiring and still come out ahead. Maybe if we were drowning in good people, I could still take the risk of tossing good people in the trash. But demand for developers is incredibly high.

Actually, I take it back. HN, y'all should definitely choose people only with top-3 CS degrees. Because then while you're all fighting over a small number of Stanford job hoppers with crazy demands, I can keep hiring the less obvious candidates, ones who are happy to stay for years.

Realistically, I think this policy probably makes sense at a company the size and visibility of Twitter. They probably get enough applicants from the schools listed to fill their needs. I worked at a company of comparable size, and they actually stopped giving referral bonuses because it stopped being useful to pay people bonuses when they had 100+ people apply for every available position.

I suppose. But it's based upon a couple of theories that I disagree with: a) the people you need are the ones who apply unprompted, and b) it's ok if you throw out your best potential hires.

My experience is that large companies mostly end up being seas of mediocrity, and I think lazy hiring is a big cause of that. I think one of the drivers of it is companies seeing programmers as fungible resources, so hiring becomes less about getting the best possible people carefully fitted to teams and more about getting adequate people with low managerial effort.

A couple of things:

1. Limiting your hiring to a tiny set of schools makes about as much sense as only interviewing candidates with names that start with A-F or candidates that come from California. You're just going to end up with a smaller list, not a better list.

2. I thought there was this dire "shortage of engineers", at least that's taken as a given on HN. Why the need for an arbitrary filter?

So you're implying that there's no correlation between academic pedigree and work performance. Maybe that's true, but not what my personal experience would suggest, at least in the realm of software engineering. I'd be happy to change my stance if there's data to the contrary.

Google has studied this.

http://mobile.nytimes.com/2014/02/23/opinion/sunday/friedman...

Didn't Google drop their degree requirement entirely?

how would you even measure that without the causal effects of "we only hired people from these schools" not being an interfering factor?

my own personal experience (from > 20 years in professional workforce) is that there's generally very little correlation between work performance and academic pedigree. For entry level positions, I've seen a bit, but after, say, 5-8 years, people are generally either really good at what they're doing, and are getting better, or they're struggling. I've not seen any correlation between the strugglers/non-strugglers and where they went to school, or indeed, to some extent, what they studied.

Not true. Google probably receives far more applications every year and still takes the time to screen every single resume and give them a shot. I graduated from a relatively unknown university in a relatively unknown part of the world, but they looked at what I did and not what stamp I had.

> other equally fast/cheap filtering methods

Like what?

Reply on news.ycombinator.com