Earlier this year, the Wall Street Journal reported on Unilever’s novel approach to diversify its candidate pool for entry level positions.
Unilever’s strategy, implemented in 2016, saw the company move away from on-campus recruiting and the submission of resumes. Unilever had traditionally focused on recruiting from a small number of colleges, using recruiters for the process.
Instead, Unilever placed ads on social media and job search sites, then invited candidates to apply directly. The application software uploaded information directly from the candidates’ LinkedIn profiles.
In a further departure from the usual recruitment process, downselected candidates were then asked to play a series of online games which assessed issues such as concentration and short-term memory. The next step was to submit a video interview; the software assessed response times, facial expressions and vocabulary.
After these steps were completed, candidates participated in their first and last interview with HR and management personnel.
Unilever stated that this process increased the percentage of overall candidates who received and accepted job offers (as well as increasing efficiency in the hiring process). The number of colleges in the applicant pool increased significantly and the anecdotal experience of Unilever management was that the successful candidates were as strong as, or stronger than, previous intakes.
Unilever’s initial view is that this process has great potential for reducing bias in the hiring process. As the WSJ article points out, human input into the software naturally involves bias in the choice and weighting of desired characteristics (facial expressions and vocabulary would be particularly prone to issues here); and software is not capable of recognising and compensating for the bias of its inputs or programmers. However, even if the process only increases the number of colleges represented, it would have a positive effect in opening up the talent pool.
In addition, the conscious effort involved in identifying and encoding desired characteristics has great potential to help organisations understand their own culture.
Next week we will look at recent reporting on what happens when organisations try to predict whether new leaders will fit within their culture.