Data, Precision, and Results
Team building means selection. The more precision in the selection process, the better you can guarantee a result. I have spent the greater part of the last decade building teams that create solutions, and prior to that, I contributed as a team member. Both as a contributor and a manager, I have seen how many organizations perform this process, and have had the fortune to learn from many mistakes, a lot of them my own. One thing I can say with certainty is this: the more data you have at your disposal, the more time you can spend evaluating what really matters, without the noise of perception and natural bias. Stated another way: I do not want to waste any time gathering information in an interview that I could gather differently, and in which my asking the question and capturing a response may influence the actual data behind it.
Managers often place a lot of emphasis on “cultural fit,” which is something that I admit is still confusing to me even though I have performed these kinds of interviews. In fact, they can be pretty enjoyable as the tone is more about getting to know a person than a probe of their knowledge or capability. But do they really indicate whether someone will perform? How accurate are our perceptions, and are they always captured in a way that can be measured?
The irony is that these are often the last interview in the hiring process, and in this context, it is often the last impression that counts. This automatically gives it more weight, like a final go/no-go. Instead of placing 80% weight on the data that matters and 20% on our perception of “fit,” we often do exactly the opposite!
To quote the Harvard Business Review:
If you’re a hiring manager, you’re probably happiest getting a sense of a candidate through unstructured interviews, which allow you to randomly explore details you think are interesting and relevant. (What does the applicant think of her past employer? Does she like Chicago? What does she do in her downtime?) After all, isn’t your job to get to know the candidate? But while unstructured interviews consistently receive the highest ratings for perceived effectiveness from hiring managers, dozens of studies have found them to be among the worst predictors of actual on-the-job performance — far less reliable than general mental ability tests, aptitude tests, or personality tests.
Now I do believe that for someone to thrive, there has to be a professional environment that promotes positive growth. I also think the demonstrated ability to adapt and communicate across cultures is a better indicator of success than whether a candidate is a match for what may be one’s own unique culture (or interpretation of it).
Now let’s take the scenario of a technical interview, where a structure is essential. A typical pattern in organizations where decision-makers have a technical background is to have an interview where the questions vary from free-form “tell me about what you do/did with X company or Y technology” to very the specific and granular, which can at times feel like a game of trivial pursuit. Best case, these interviews have a standard structure and questions so that all candidates are guaranteed the same interview experience and are evaluated equally. In practice, however, it is more likely that a uniform set of questions with expected successful responses is not used. A highly skilled engineer is not necessarily a highly skilled interviewer, although the assumption is that he or she will excel in both domains.
Taking the structure a step further, by an ad-hoc interview and creating standard questions, opportunities for automation abound. All of those data points that a human is collecting can be done more efficiently by a machine. That leaves more time to ask the really good questions – the ones that show how people actually solve problems. The best technical interviews I have taken or given have this in common – the opportunity to whiteboard a problem, discuss approaches and reveal a thought process. This is what humans should spend time analyzing.
To sum up:
- A cultural fit is a nice-to-have but ultimately does not have much data behind it to support it as a success factor (I invite someone to prove me wrong).
- It may be both biased and incorrectly weighted.
- A structured process driven by data is not only more efficient but lessens bias and is a better indicator of whether a candidate is the right selection.
What we do at BairesDev
What differentiates BairesDev from its competitors is precisely the methodology behind the hiring process. This consists of exams, both technology-specific as well as general aptitude, that not only ensure that evaluation is non-biased but also allows us to select from a pool of over 145,000 candidates in a given year.
Technical interviews are structured around a specific set of practical examples specific to projects. These include written questions for solving complex problems and often last much longer than “normal” interviews (up to four hours).
Finally, algorithms are employed as a final stage to best match a candidate’s skill and experience to projects that will best suit her.
If you would like to know more, please do not hesitate to reach out.