Created model to predict leaving risk

Top 3 global business process outsourcer. We worked with the UK team and a Strategic Client. We combined our practitioner understanding of the factors affecting colleague attrition, with a machine learning to give a leaving prediction dashboard, presented in an enhanced MI dashboard that significantly reduced team leader effort to implement ‘interventions’.

Challenge

• Attrition was the most significant business cost affecting margin.
• Clients were beginning to take notice and enforce penalties.
• All historical efforts had failed to curtail leavers, which is a common problem in busy contact centres.

Solution

• We ran a dual approach of colleague journey & solution design that addressed root causes of attrition & activity to mitigate it.
• Data Science modelling ran in tandem using over 220 data points across people, process & performance used to build the model.
• An iterative Minimum Viable Product approach was used to drive pace and business engagement in co-creating the solution.

Outcome

• A minimum viable product that predicted the likelihood of leaving within 6 weeks, that had up to an 85% confidence level.
• The creation of the MI dashboard effectively reduced team leader one to one preparation time from 45 to 5 minutes.
• Out colleague entered approach combining our direct experience and best of iterative Data Science - all pushed into a reworked MI dashboard created a real world AI win.

“Nicola’s drive, vision and meticulous planning set her apart as a business leader and solution provider. In addition, Nicola’s approach with a classroom of customers was engaging, professional & persuasive. I would recommend Nicola and her team to anyone looking to innovate and improve their operation.” Senior Project Lead, Concentric