Should big data analytics decide whether you get promoted?

Human Resource | Sectors   |   
Published February 8, 2014   |   
Tomas Chamorro-Premuzic

Although some organisations are more political than others, a fully apolitical organisation seems unthinkable, at least for anybody who’s ever had a job. However, in the age of big data, where supercomputers and machine-learning algorithms track, predict and influence a wide repertoire of human behaviours, this could well change. Indeed, a data-driven approach to selecting, developing, and engaging employees should, in theory, leave very little room for office politics. Furthermore, if the intuitive art of managing people can be turned into a robust science of people analytics – to use Google’s favourite term – that would surely mean the end of subjectivity, irrationality, and bias in human management – or would it?

Not quite.

First, it is not easy to put numbers of people. Although less skilled jobs, like call-centre operator, may be repetitive enough to allow managers – and computers – to identify fairly robust indicators of performance (e.g., number of calls per day, customer feedback, time away from desk, and visits to the bathroom), professional and skilled jobs are generally too complex to quantify through data. Even in highly analytic work domains, such as sports, stats don’t tell the full story. That’s why we may dispute Cristiano Ronaldo’s Ballon d’or despite all his goals and assists: e.g., Ribery won more trophies, Messi is more skilful, and if we go by individual analytics, there are probably some players with better stats than Ronaldo anyway. Besides, key organisational goals are accomplished by teams, rather than individuals, yet raw performance metrics at the individual level ignore how an employee interacts with teammates and shift employee focus from collective to individual goals, harming team morale and performance.

Read More