[douglasjsellers]

All that this article says is that past performance, in terms of GPA, is the best performance of future performance - rather than a 30 minute interview predicting future performance. This sees like a basic truism to me and the main lesson that tech hiring processes can take away from this article.

In my experience (having hired > 100 engineers) one of the basic problems that tech hiring, as a whole, has is that it misunderstands the point of a technical interview. Organizations and hiring managers see the interview process as a way of improving the brand of the engineering organization - "We have super high standards and to prove this our interview process is really hard - therefore if you think you meet these standards you should apply". This leads to the current interviewing trends of super academic/puzzle/esoteric technology based interviews. Applicants leave those interviews saying that it was super hard reinforcing the brand messaging (classic marketing).

Rather, in my experience, the best results come from viewing the hiring/interviewing process for what it is - an attempt to predict future performance (and specifically performance at your organization) using a variety of techniques which interviewing is one. In this context, of attempting to predict future performance, interviews are not a great tool - better to look at specific past performance.

Past performance is always the best predictor of future performance and the point of a technical interview, in my mind, is to critically inspect that past performance to understand how closely it relates to the future performance that your organization needs.