We know technical interviewing is a problem but rather than asking interviewers to do better, a lot of suggested solutions push that problem off onto people we interview rather than those who are doing the interviewing.
This comes up a lot because the hiring process is the second most popular place to improve “diversity” after teaching children to code; the hiring process is the end of the pipeline.
Bad hiring processes disproportionately affect people who don’t pattern match. Better hiring processes benefit everyone, although people who usually pattern match may disagree. I’ve seen a number of white dudes react angrily when it’s explicitly stated that diversity is valued – they take it a sign that they are unwelcome. The way the use of words like “hackers” and “rockstar”, while calling attention to pool table and the beer keg and amid endless reference to “he” and “guys” have made people who aren’t white dudes like them feel so unwelcome for so long. Except to them “unwelcome” seems to mean “not centred”.
I have made these points general where I can, but some of them highlight the effects of seeking “diversity” without first understanding inclusion or the broader problems arising in a homogenous, often insular environment.
Internships Rather than Full-Time Jobs
Whether hiring new grads or for “returnships”, I understand why companies would like to have a three-month job interview rather than a single-day interview. But consider this: when you onboard a new full-time engineer into a team, they have a manager who (hopefully) has some experience and some training. Interns, on the other hand, tend to be assigned “mentors” who might have gone to some one-hour training sessions where nothing substantial had been covered. If companies are serious about internships like this, they need to set interns up to succeed. A big part of that would be to stop using the word “mentor”, start using the word “manager”, and to train, incentivise, and hold accountable accordingly.
I understand the appeal of take-home assignments from both sides: candidates get time to think and code without being watched, and companies can ask them to solve a bigger problem (and don’t have to allocate engineering time to watching someone solve it). The earlier take-home assignments are in the process, the less I like them. (Some well known companies have people complete an assignment before they’re given the chance to speak to an engineer at all).
My reservation with take-home assignments is that you end up asking for the most work from the people who you’re least likely to hire. An assignment that may take an average person two hours may take a super star one hour. Someone who really doesn’t have the experience for the role might find themselves spending all weekend on it. (In comparison, at least technical interviews are time-boxed.)
I worry that companies avoid training and incentivising their engineers to interview candidates well, and that they’re instead reducing engineers having to talk to potential colleagues. If your selection process is riddled with bias (and it is), the solution is to reduce your bias, not shove it onto other people with strategies like “have every $diversePerson who applies do the take-home assignment”.
The other benefit of technical interviews is that they force some to decide whether a candidate is worth spending the most valuable resource your engineering team has – time. This is powerful from the perspective of the candidate—a take home assignment can feel impersonal. But also because it forces you to consider if it’s worth your spending time to continue the process. In my opinion, there’s a lot to be said for a fast, respectful no. Giving someone a chance that isn’t really a chance is a waste of everyone’s time.
I want to be very clear about what this means: if you’re confident that someone doesn’t have the skills or experience that you need, it’s better to cut things off sooner rather than later. But where does that confidence come from? The more objective you can be, the better. Is it a vague feeling? That’s likely your bias talking. Or is there a concrete skill that you’ve identified you need, that you have given them the opportunity to demonstrate—or asked specifically about—and the answer was no?
Whose Job Is It, Anyway?
A little past the hiring process, but the other thing that I worry about is seeing junior engineers (particularly women) being given the burden of fixing the process that they just went through. Rather than hiring junior women and expecting them to do the thankless emotional labour of fixing your hiring process, first consider why there isn’t a woman who’s senior (and interested) enough that doing so might be part of her actual job.
The answer might tell you a lot about your culture and what people are rewarded for.
People throw around “diversity is a competitive advantage”, but it’s not clear that they know what that means. Here’s the reasoning I embrace: Inclusivity is a competitive advantage because it yields a diversity of ideas and insights at all levels within an organisation.
A better hiring process is necessary but not sufficient in building an inclusive workplace. The hiring process is actually the least of it. I can accept that interviewing for jobs is inherently awful, and that as an interviewer or hiring manager I may only be able to make it less awful. But what I no longer accept is that being an engineer whilst female is inherently awful.
So with that caveat, here are some ways that you can create a competitive advantage in hiring:
- Check and reduce bias (for all interviewers).
- Be transparent about your hiring process (and for startups, that’s especially true if you’re still defining what that process is).
- Treat interviewees time with the respect it deserves. (Paying people to interview is one strategy, and it’s a good one.)
- Be clear about what’s important to you—what kind of person do you need?
- If you’re hiring to fill fewer, more specialist openings, be very clear about what kind of skills and experience you’re looking for. Importantly, ensure they’re realistic—does that kind of person even exist? And will they work within your constraints of geography and salary?
- If you’re hiring “smart generalists” this is much fuzzier prospect, so define what you care about. Be mindful of what kind of criteria you use to assess what a “smart generalist” looks like, and what kind of problems you present to evaluate them.
Structural and Individual
I used to work at a place where I interviewed mostly women. There was no incentive for me to be a good interviewer, but I worked on that anyway. A bad move for my career at that company. But a good move as a human being. Not a rational decision, then. A moral one.
Hiring processes are structural and individual. We design a process structurally—influenced by our bias to create a process we’d do well in—but interactions are individual, typically one-on-one. There’s no structure that eliminates the need for one-on-one interactions because we’re hiring humans—not building robots—and because the hiring process is a two-way evaluation where candidates learn about the company while the company learns about them. The people you hire aren’t going to work under the thumb of an automated computer evaluation system—they’re going to work with other human beings.
If you design a hiring process that ignores the individual side of the equation, you’re just shifting the part of your process that’s broken. And if you consider the structure without the individual, then your theory will fail at the first interaction with someone who operates rationally rather than according to a moral code.