Are your hiring filters working?
How often would an otherwise great candidate simply not continue the application process due to your filtering criteria?A quick reminder: Today at 17:00 CET (Europe) / 8:00 AM PT (USA), I’ll be doing a free, live video Q&A session answering your DevOps career questions. Space is limited! Register up now!
There are many opinions about the best way to hire people. Most are unsubstantiated for the simple reason that it’s exceedlingy difficult to do objective studies on something as complex and subjective as hiring.
A common trope in these discussion is “X is a good practice, because it filters out candidates.”
I want to drop a bomb on that trope today, and hopefully get you to look at such claims with a more critical eye.
True story: A few years ago I worked for a company that added as a required field on their job application “Why are you excited to work for [company X]?” I submitted feedback saying “I don’t think this is an appropriate question, as most applicants don’t know this company from the next at the point they fill out an application. It would be better to remove this question, or make it optional.”
The CEO replied: “It’s a filter.”
Yes. Yes, it is a filter. That’s obvious.
The question we ought to be asking is: Is it a good filter? Is this filter actually filtering out candidates we don’t want to hire?
There are two parts to this question. First, how well or poorly does this filter eliminate bad candidates? Second, how well or poorly does it inadvertently filter eliminate good candidates? In statistics, these two charactaristics are referred to as the sensitivity and specificity respectively. If you’re curious about these details, here’s a short video for you.
But without understanding the statistics behind filters, you can do a lot intuitively to ask two simple questions:
- How likely is a bad candidate to pass this filter?
- How likely is a good candidate to not pass this filter?
And of course this is where subjectivity comes into play. Perhaps it is useful to filter out candidates who have never heard of your company prior to applying for a job. If you’re Apple or Google, maybe that’s acceptable. If you’re a small startup (as was the case for the company in my story), maybe that’s more questionable.
Another similar story I’ve heard repeatedly, both with past clients, as well as in various conversations, goes something like this:
“Take-home coding challenges are a great filter, because only 20% of candidates pass the screening stage.”
Once again: Are we filtering out candidates we don’t want to hire? How often would an otherwise great candidate not pass this technical screening for some reason that’s not actually related to their job performance? Additionally, how often would an otherwise great candidate simply not continue the application process when asked to do a take-home coding challenge?
This won’t solve the hiring problem, but asking yourself these questions will help you be more aware of the unintentional bias your hiring process may have.
