A humble alternative to technical interviews

By zwf

I've got a proposal for an alternative to the dreaded technical interview loop that could make life easier, better and cheaper for people and their companies. The idea came from a conversation with coworkers I overheard a few years back: INT. OPEN OFFICE - DAY Two software engineers, GEORGE, a senior engineer and highly-skilled technical leader who people love to work with, and JOHN, a mid-level engineer, banter in their San Francisco office, surrounded by drink fridges and ping-pong tables. The sound of tapping on mechanical keyboards fills the background. GEORGE: "Who's the interview candidate today?" JOHN: "They're Ringo's referral." GEORGE: "Hmm, I referred a few people a while back. But all my referrals were rejected—they didn't make it past the technical interviews." JOHN (frustrated): "Man, I feel like if George refers a candidate, we should make them an offer pretty much automatically." FADE OUT.

I can relate to John's feelings of resignation. When someone I know is great at their job recommends another person, I trust them. It's pretty straightforward to gauge if someone is competent once you've worked with them for a little while. Contrast that with the many interview debriefs I've been in where the difference between hire and no-hire turned on subtle impressions or  trivial technical details.

So here's the proposal: companies should designate a set of technical employees as trusted referrers (for example: the set of senior engineers or higher with a tenure of at least two years). And when a trusted referrer submits someone for a software engineering role, the technical interview loop is waived. That's it! That's the whole plan! There are a couple details to iron out (you shouldn't give out referral bonuses; that's a messed up incentive. And you should still do a couple non-technical interviews as well, both as a sanity-check and to sell the role. After all, any half-decent candidate is sizing up you as much as you're sizing up them.) Is this a good plan? More specifically, is this a plan which will allow your company to hire good and great software engineers more quickly and cheaply, while keeping the rate of bad hires unchanged? What we've done here is create an alternative path to a software engineering role at your company. It won't be the most common path: most people you hire probably haven't worked directly with your trusted referrers. But for a small set of ostensibly low-risk candidates, it creates a low-friction route. Her's why that's a good thing:

1. The market for skilled candidates is extremely competitive, and companies that move quickly in the hiring process have an advantage. How frequently does your company lose out on a potential hire simply by moving too slowly? 

2. It creates a no-lose situation for highly-skilled candidates investigating a job with your company. Technical interview loops, on the other hand, come at a massive cost to the candidate. Just a few of the risks a candidate takes on:
  • The risk that they will waste a day on a fairly stressful activity which provides no benefits.
  • The risk that they will be humiliated at some point during the interviews.
  • The risk that they will get rejected after the interviews.
  • The risk that their contacts at the company will think less of them because they were rejected.
3.  It's easier on your employees. Interviewing is time-consuming and many people find it unpleasant. 4. It's novel. I've never heard of a company that does something like this, outside of tiny startups that basically have no choice but to hire people known by the founders. Novel ain't always a good thing, but in this case there's an obvious benefit to candidates—the main question is the cost. The obvious question is whether this plan can be implemented without hiring bad candidates. I'm guessing this would work pretty well for some companies; it would crash and burn for others. Fundamentally it is an issue of trust. Are your trusted referrers competent enough at assessing someone's skills that they can pick out great people from their previous experiences? Are they credible enough to only refer people that are up to par? There are other risks, too. Maybe the pool of people known by your trusted referrers is not large enough to make this plan worth it—the cost of setting it up could be more than the benefit it provides. It could also sew some bitterness within the engineering team: it might be seen as unfair that some people can get hired through the nepotistic route, while others have to do the tech loop. Would companies ever adopt this? I doubt it, unless it started from the ground up. It's an extraordinarily risky proposition—it's been pounded into our heads through thousands of blog posts that FizzBuzz and its compatriots are the only bastion that keeps your team from hiring an ocean of dunces. And hey, they could be right.

But I suspect that in the right organization, this would work pretty well, and actually be a pretty effective innovation in the technical interview space. A little less friction can go a long way.