Printable versionSend by emailPDF version

A large professional services firm waned a way to understand who was leaving the firm, why they were leaving and what could be done to reduce turnover and improve hiring efficiency.

The firm had relevant information in multiple data silos across the firm with no way to design, build or implement a solution that was predictive and prescriptive for their needs.

Our solution included:

  1. Definining the problem: We developed a set of testable hypotheses with key stakeholders focused on human capital issues such as retention and recruitment.
  2. Acquiring and preparing data: We worked with IT, HR, and Finance units to synchronize and clean data sources as well as evaluate data usability. We also performed statistical testing and built visualizations to study the relationship amongst variables and identify features for our model build.
  3. Developing and deploying model: Given the business problem and data available, we selected the most appropriate algorithm with care not to overfit. We deployed predictive model into production, re-implementation data transformations and the model into their existing data platforms.

Results:

  • We directed HR department to implement the predictive model including requirements for technical infrastructure, training and suggested oversight.
  • We prescribed solutions to reduce the cost of turnover as well as increase productivity through an increase in workforce quality and engagement.