Reading a recent piece in the FT “Wall St turns to machines to find better-behaved bankers” I felt conflicted. The article focuses on Koru, a platform that’s using the same technology as dating sites to match candidates with employers. Their algorithm produces a benchmark, or ‘fingerprint’ for each of its corporate clients by identifying the character and preference traits of their longest standing best performers. The idea is that they map the unique DNA of the ideal candidate for each individual employer and its beauty is that it identifies traits in a candidate pool that can be way broader than the ones classically linked to specific elite business school grades. It’s the digital matching answer to the ‘grit over grades’ thesis expounded by Angela Lee Duckworth and it purports to accurately predict performance of candidates before hire. Koru claim that customers have seen between 30 and 60% increases in high performing hires as a result of using the test.
So Koru come to me and tell me what is ExecutiveSurf’s optimal candidate profile and it then shows me real candidates who look like that profile. That’s amazing isn’t it? It’s like the driverless car: technology that’s able to learn from the mistakes of its proxy colleagues around the world will one day make no more mistakes. Job matching technology that ‘learns’ from its universal mistakes will one day produce nothing but perfect hires.
So why am I skeptical? Is my nagging doubt that this is another example some clearly amazing technology still managing to dumb down what is a spectacularly complex process – the elixir that means that things work out great with the new hire? I don’t think so. I think it’s more that even if that were possible, in creating the ‘perfect’ profile, companies will sleepwalk into a future where all their employees are clones (sex, race, academics, social background etc may be hugely and heart-warmingly diverse but clones they will be) and therefore they will remove the very thing they seek to instil: diversity.