Hiring for technical or data roles is a tricky process. And the biggest barrier, whether for candidates or those doing the hiring, is the interview.
Often, companies fail to invest enough thought and resources into the hiring process, either losing promising candidates in the process or ending up with sub-par hires. Inertia also works against transforming the interview process, with many companies continuing to interview the way they do simply because that’s how they’ve always done it.
This article looks at some of the common interviewing styles, and why some are more important for technical roles. While interviews typically involve many different styles, it’s also important to consider dumping some common elements and retooling others for more optimal results.
The traditional style
The default option, the traditional style involves the interviewee walking the interviewer through his or her resume. While it might feel tempting to use this technique to break the ice, the traditional style simply yields too much filler and not enough useful information. It’s much more useful for interviewers to get a sense of the candidate’s past experience in pre-interview research and tailor questions to dig deeper into items that catch their eye. Indeed, a 1998 meta-analysis of 85 years of studies on assessment techniques found that traditional, unstructured interviews can only explain 14% of an employee’s performance1.
The abstract style
Theoretically, asking what kind of tree interviewees would be or how many golf balls could fit into a Boeing 747 is supposed to show how they think on their feet and how they process problems. In reality, as Google’s Laszlo Bock states unequivocally in the article mentioned above, brain teasers serve only to make the interviewer feel smarter.
The technical style
At its core, one major purpose of the interview is to ensure that a candidate has the subject knowledge and competencies needed to fulfil the role being offered. In that sense, the technical style is a no-brainer. Yet, there are several ways in which companies fail to appropriately deploy the technical style of interviewing.
Chief among these is the whiteboard test. Yes, it is useful to watch a candidate work his or her way through a problem and to have them talk you through the process, so you can judge their technical skills as well as their ability to communicate the process. However, there’s no reason why this must take place on a whiteboard in an age when everything is digital, unless it’s because that’s how it’s always been done:
@sarahmei tech interviews depend a lot on hazing mentality: we went through this, so we have to make them go through it toi— Amy Hendrix (@sabreuse) March 20, 2015
Even if you dump the whiteboard, tasks the candidate might undertake on the job would rarely involve working under the pressure of someone watching over their shoulder. Instead, it makes sense to ensure more conducive conditions. While take-home tests are common, it’s also important to include a data day or a coding day, where the potential hire gets the experience of a potential day on the job and gets to show off skills under more realistic conditions. This also means that such assessments should stick as close as possible to actual issues and concerns that the company or team has faced or is facing.
The behavioral/situational style
It’s vital that interviewers get a sense of how the potential hire would react to realistic conditions. That’s why behavioral and situational questions add value to the interview. While the former focus on examples from the candidate’s past experience where he or she tackled a particular problem, the latter pose hypotheticals that mirror the kind of tasks or situations he or she would face on the job.
Such questions are most useful when they’re not asked with a single answer or a rigid checkbox in mind, but have room for digging deeper and understanding the candidate’s methods or approaches comprehensively. But such questions are also tough to develop. They require time and effort for conceptualization and testing, and for interviewers to get familiar with them. And they have to be frequently updated so that candidates can’t simply memorize a desired response.
Some final thoughts
When it comes to fields like programming and data sciences, there’s no one ideal candidate mould, especially when it comes to formal qualifications. That means there’s a greater onus on interviewers to get a clear and comprehensive picture of candidates’ knowledge, skills, competencies and fit with the work culture. This requires a rigorous and structured, but not formulaic approach to interviewing. At the same time, it’s not the smartest, most creative questions that ensure successful hiring, but the most effective and consistent ones.