People, People, Everywhere

We’ve often heard it said by many a CEO, that people (and data nowadays) are an organisation’s greatest asset. Why then do so few organisations apply modern analytics and data science techniques to enable their people to do their best work?

Matthew Bishop, the co-editor of Espresso writing in The Economist last year, bemoans the slow pace of big data adoption among the Human Resources (HR) community. There are examples of People Analytics (also called Talent Analytics) providing much needed insight into HR functions. Hewlett Packard for example, in a seminal data modelling exercise, came up with a flight risk score for each of its more than 330,000 employees. This score gave an indication of how likely an employee was to leave the organisation. Armed with this information, a manager could take proactive action to retain the employee - sometimes before the employee may have consciously thought of leaving themselves. The savings in staff replacement and productivity loss was by some accounts estimated to be $300m globally.

The Talent Analytics team at LinkedIn helped the organisation gain much needed visibility into hiring demand at a time when the company was growing at 40% a year and ‘couldn’t fill roles fast enough’. By building visibility and accountability into the hiring process and predicting hires to within 5% of actuals, the Talent Analytics team was able to quell the firefighting and save LinkedIn 15% of its recruiting budget in the first year.  

As 2018 gets off to a start, many employees will be looking forward to their bonuses. These awards are intended to reflect their performance over the prior year. Matthew Bishop again talks of the ‘anecdotal flimflam’ on which too many performance appraisals are based. There is a clear need here for objectivity and data to counter our unconscious biases.

Data is your friend

Many HR practitioners we’ve spoken to appear to have an innate aversion to science, saying it’s too complicated or over their heads. They do themselves an injustice. We suggest starting by encouraging their departments to be more data driven than they are today. Basic analytic charts and clear-headed visualisations can go a long way to help HR practitioners make better decisions. Do this before getting into more advanced and predictive applications that can help form a better view of an employee’s future performance or retention risk to name two possible use cases.

Of course, an algorithm can itself be biased and we’re not advocating replacing one set of biases with another. However you’ll be more confident in asking the right questions of your data scientists and challenging them when they can’t explain adequately how their model was built or under what conditions their assumptions break down.

People Analytics gives rise to legitimate concerns about employee privacy. However it remains a powerful complement to the range of tools at HR practitioners’ disposal. It would be a shame for organisations to ignore its potential.