Using Data for Better Employee Retention: 5 Things to Know
September 21, 2020
One of the most challenging aspects of offering equity to employees is knowing how much equity to grant. The NASPP Conference session “Data & Analytics: Telling the Equity and Mobility Story” demonstrates how companies can use more holistic data and statistical modeling to better answer this question.
The session featured Jennifer Link and Eric Scaringe from KPMG along with Kathi Green from Phillips 66. The panel explained how KPMG used data analysis to help Phillips 66 develop a strategy for employee retention, including a retention grant program. Here are five things I learned from this session.
1. Employee Turnover Is Costly.
Jennifer noted that the cost of replacing entry level positions is approximately 50% of the employee’s base salary. For mid-level employees, this cost jumps to 125% of base salary and for senior executives, it is 200% of base salary.
2. You Can Make More Informed Decisions About Equity Grants
Your company likely has a plethora of data available to help you determine what your retention grants should look like, including which employees should receive them and how large they should be. Accessing this data will allow you to make more informed equity granting decisions that are based on statistical modeling rather than a gut feeling.
Some of the key data points that result in a more holistic analysis might include supervisor performance (this turned out to be key for Phillips 66), age, years to retirement, work location, outstanding equity awards, and more.
3. You Can Predict the Future
Jennifer explained that using this data in a statistical analysis of employees who have already left allows you to calculate a retention risk for your current employees. Essentially you are using the past to predict the future by determining how similar your active employees are to those that voluntarily left within, say, the last three years. This allows you to identify those that might be thinking of leaving and those who aren’t likely to leave. This, in turn, helps you determine how to compensate your employees, including whether they should receive retention grants.
4. You Have Four Categories of Employees
By marrying this retention risk analysis with employee performance data, you can bucket employees into four categories:
- High performer, high retention risk: These employees have high value to the company and there is a high risk that they will leave. They are candidates for retention awards.
- High performer, low retention risk: These employees are also high value to the company but likely will not leave. You still want to grant equity to them (otherwise they might end up in the prior category), but perhaps not at retention levels.
- Low performer, high retention risk: The company might want to start succession planning for these employees.
- Low performer, low retention risk: This company might want to consider how to improve these employees’ performance or sever their employment.
5. Equity Is a Powerful Tool for Retention
Kathi explained that Phillips 66 initiated this project because they were experiencing high turnover within a couple of key organizations within the company. The analysis revealed that supervisor performance was a critical factor in employee resignations. This likely can’t be solved by equity awards alone, but equity can help. Phillips 66 implemented a retention grant program for their high-risk high performers. They are now nearing the end of the vesting period for those grants and have achieved almost 100% retention.
Phillips 66 also implemented additional training for supervisors and other programs, which surely contributed to their success in retaining employees. But supervisor training isn’t an instant fix. The equity awards may have been an important incentive that gave the other programs time to take hold.
This was a very interesting session; the potential to use employee data for better decision-making around equity awards is intriguing. It certainly seems like something that many companies could benefit from.
The panel did not discuss how gender, race, and similar factors might be incorporated into the analysis. I think including this data could help companies ensure that they aren’t inadvertently creating inequities among employees. For example, if men are more likely to leave than women (and note that I have no evidence of this, I’m just hypothesizing), there could be unintended consequences to a retention program focused primarily on employees who are likely to leave. I still think that this analysis has a lot of potential, however.