The Challenge
A well-known local recruitment firm was engaged to find the next Managing Director of a large Kenyan utilities company. The applicant pool was broad — diverse backgrounds, demographics, and levels of seniority — and the process ran across several weeks, involving over 15 interview panels and hundreds of candidates.
The requirement was clear: the outcome had to be transparent, defensible, and free from bias.
Some Context
Anyone familiar with hiring in Kenya understands the pressure of kujua mtu — the expectation that personal connections influence outcomes, particularly at large institutions. Removing that influence without creating friction or embarrassment for panellists is a genuine challenge. This is precisely where structured data analytics adds value: it surfaces the best candidates through the numbers, not the relationships.
Our Solution
At the conclusion of the interview process, we facilitated a structured data review session with all panellists present. We began by examining outliers — candidates with unusually high or low scores — to validate whether those scores reflected genuine performance or panel-level bias.
From there, we applied an abstraction framework that normalised scores across panels, controlling for differences in individual panellist scoring tendencies. The result was a clean, comparable ranking that could be presented to the board with full confidence in its integrity.
The Result
The process delivered a successful hire from the strongest available pool of candidates — selected through a method that was transparent, auditable, and free from undue influence. The client reported that the board accepted the recommendation without challenge, citing the clarity of the supporting analysis.
