According to an article on the cfo.com website, it is easy to predict who will stay. On the other hand, predicting who is about to leave is much more difficult. This is one recurring theme when efforts are made to design a model for predicting staff departures.
Key predictor: Is performance fairly evaluated?
There was a bank in California which set up a bonus scheme for sales of a new credit card. All the top salespeople were in one region. That led to an initial assumption that all the best sales talents were located there. However, when a model was constructed to predict performance, it turned out this was not the case at all.
Foot-traffic levels related to branches, credit rankings and income levels were all included in the model, along with percentages of servicing visits and sales opportunities. The best predictor for sales success was the average credit rating in each region. Thus if a top salesperson from that most successful region moved to a different location, they would sell fewer credit cards.
Now imagine the bank had acted on its initial assumption. It would very likely have upset a number of good salespeople by telling them their performance was poor and they should show more effort.
Lacking control of the working day is also harmful
People also tend to leave their job if they feel they don’t have control over their working day. If people are given demanding tasks with tight deadlines when their diary is already full of meetings, they can become really frustrated. This then affects their performance and in the long run also their chances of being promoted.
On the other hand, a moderate level of pressure is healthy. Too much pressure is harmful but so is no pressure at all as it makes people wonder whether anything they do actually matters.
-jk-