Using Data Driven Hiring to Predict Employee Turnover

Quality of Hire

Implementing assessments into the recruitment process can be the catalyst for a data driven hiring strategy. Collecting key pieces of data from the candidate before and after they are hired allows organizations the ability to build predictive hiring models. Organizations can then validate their hiring decisions based on hard facts versus initial gut feelings or impressions. These hiring decisions allow organizations to improve the quality of their workforce by ensuring only the best fit candidates are being on boarded.

Hiring the wrong individuals into your organization greatly impacts finances, recruitment efficiencies and employee productivity. There is significant financial gains associated with screening out candidates with the characteristics associated to turning over within the first 90 days of employment. Being able to identified these characteristics and screen them out with a pre hire assessment allows companies to start making better data driven hiring decisions.

To start, candidate assessments need to be rolled out into a recruitment process, validated based on historical data and then continually improved to drive better hires. You start with the best based on screening for the characteristics and attitudes that fit your organization and then continually get better by validating your results.

Questions are layered into the assessment that are associated with likelihood of employee turnover. Both the candidate’s score and their actual length of employment are then validated to make sure the prediction tool aligns with actuality. The two results allow us to uncover any correlations that exist and then more accurately weight assessment questions based off of actual turnover results. This improves a company’s risk of hiring employees with the likelihood of turning over.

We used this methodology on one of our largest retail partners. We took a sample of 20,000 candidates who applied for a position within the organization and completed a pre hire assessment over a one year period. The assessment was built with five key turnover predictive questions that were then validated with 1,000 hires this organization made.

The initial assessment laid the foundation for the turnover predictor model. By measuring key characteristics and then cross­ validating the assessment results with actual turnover numbers, we were able to create a scoring key that better measured likelihood of turnover. Initially, this client was seeing upwards of a 40% turnover rate within the first 90 days of an employee’s lifecycle.

Building in the turnover predictive model, we were able to create a multiple hurdle approach. Not only were candidates being screened for best fit characteristics and attitudes but additionally they could now screen for their likelihood of turning over. When we used this model to predict what improvements we would have seen in better pre screening these same hires, the data showed that within the first 60 days of employment a 18% decrease in turnover would have occurred and within the first 90 days we would have seen a drastic 27% decrease in turnover.

These results validated the turnover predictor questions and now allow us to better screen and remove any of these high risk candidates from the recruitment process with data driven hiring. Our client now is making better hires, reducing the financial burden of non­-committed employees and improving their market competitiveness with a stronger workforce.

Related Case Studies

Prairie Lube

Automotive

Fresh St. Farms (H.Y. Louie)

Grocery