McDonald’s now works with A.I. that predicts which applicants will succeed in the long run based on 55 characteristics. It turns out that this even differs considerably from restaurant to restaurant. How does that work?
Prediction is not that simple
When you talk about assessments, it is often thought that you can estimate someone’s personality quite well with one test, and that you can also predict how well someone will perform in a certain position. Research by the fast food chain at the American McDonald’s restaurants showed that it is not that simple. Based on A.I. is clear that it differs per candidate and per restaurant. Every location has different factors that determine the success of an employee. And vice versa: a candidate will differ in performance depending on where he or she starts work.
55 characteristics
Assessment provider Sprockets provides the technology that has recently been integrated into McHire, the recruitment platform specially built by Paradox for the fast food chain. Sprockets’ algorithms looked at 55 needs, values and personality traits of the employees of 4 participating McDonald’s stores in the study. First, the local top performers were asked for each location. By analyzing them on all 55 characteristics, a ‘success profile’ was then created for a particular position in a particular location. Subsequently, the applicants were also presented with a similar questionnaire, after which the best candidates were selected against the yardstick of the success profile.
No ‘standard profile’ at McDonald’s
What makes Sprockets’ story so distinctive, however, is that it shows how important even that one McDonald’s location is over another. No ‘standard profile’ of a successful McDonald’s employee is made, but the difference is examined per location. And that turns out to be bigger than expected. ‘There was hardly any overlap between the locations. Each branch turned out to have quite a different success profile.’
Reduced turnover by 22%
Particularly relevant information for McDonald’s, with about 2 million employees 1 of the 4 largest employers in the world, but at the same time an employer that struggles to find enough candidates, and so far has an annual turnover of more than 130% . And that is precisely where Sprockets claims that his A.I. has value. For example, the algorithms would have reduced annual staff turnover by 22%, and 90-day staff retention would have improved by 43%. That saves a drink on a drink in terms of introduction and recruitment costs.
