Did you know that Google’s hiring managers don’t decide whether a candidate should be hired? And McDonald’s restaurants with more employees over age 60 have on average 20% higher customer satisfaction? And did you know that within Xerox call centers recruiters don’t consider work experience, but personality since this is a better predictor for performance?
Organisations are increasingly using analytics within recruitment and HR to improve their decision making. But why should you follow their lead? During the 'data-driven recruitment-event' of Recruiters United I attempted to answer this question. Interested? Here are a few highlights.
Biased decision making
Are you capable of making good decisions based on data? If so, try to solve the following puzzle:
“You buy a baseball and a bat, combined they cost € 1.10 and the baseball is 1 euro more expensive than the ball, how expensive is the ball?”
I asked the audience of the Recruiters United event the same question, of which the majority answered 10 cent (figure below).Unfortunately 10 cent is not the correct answer, the answer should be 5 cent (find the explanation in full presentation). In 'Thinking, Fast and Slow' Daniel Kahneman explains why the majority will provide a wrong answer. It seems that we may think in two different ways: we can apply fast thinking where given a question we directly jump to the conclusion. Or we may use slow thinking, in which we solve the problem by analysing the problem step by step. Slow thinking provides the correct answer in this question, whereas fast thinking frequently results in the answer of 10 cents.Slow thinking is especially optimal compared to fast thinking when solving problems with much uncertainty, and when making decisions about people. Hence, since recruitment decisions are usually made in uncertain situations and involve decisions about people, analysis is vital in order to make better decisions.
A second important question one should ask is: does the improved decision making justify investments in recruitment analytics? This depends on a number of factors, including the current recruitment costs, quality of the candidates, and candidate volume. Unfortunately making a business case for recruitment analytics is a chicken or the egg problem: to obtain insight into the potential of recruitment analytics we have to use analytics, and as we will see in a minute this can already be challenging.
Consider for example that we want to predict employee churn next year, such that we can already identify possible candidates for replacement. To determine whether the potential cost savings compensates the cost of this analysis, we need insight into the cost of turnover. However, when we asked the audience to estimate the turnover cost compared to the annual salary of that position we obtained the following result:Interestingly, if we compare this figure with scientific studies, which estimate turnover cost between 100% and 200% of the annual salary, 51% of the recruiters underestimates turnover cost. If we would use this estimate to decide whether the analysis cost are worthy, we might wrongly conclude not to invest. In short, independent of whether recruitment analytics leads to improved decisions making, it is also required to determine the impact of your recruitment strategy.
Where to start?
So how do we put this theory into practice? Obviously there is no success formula, but you can avoid common traps by employing to following tips:
- Be aware of fast thinking when making decisions
- Use predictive methods when analysing data
- Use an experimental mindset
- Integrate data over the entire recruitment funnel
- Use common sense when interpreting data
Would you like to know how Endouble can help you with applying analytics in recruitment? Please contact us.
Check out the full presentation: