Diversity programs not working? Cohort analysis can help

When financial services organizations want to increase diversity, they often turn to tactics like bias training in an attempt to make workplaces more inclusive. Studies show, however, that these broad strategies — which have been used in workplaces for decades — are rarely effective and can even be counterproductive.

This lack of progress in promoting inclusivity means that diverse recruiting initiatives often miss the mark. Hiring a large number of diverse employees won’t make a difference if an equal number leave the company each year in search of more welcoming environments. No wonder most major financial services organizations haven’t seen actual, concrete increases in diversity despite implementing programs for many years.

Truly moving the needle on diversity requires a more targeted approach that leverages nuanced data on employee experience, compensation, career pathing and other factors. This type of people analytics can help banks and credit unions understand what causes particular groups of diverse employees to leave — and how to entice them to stay.

Sign up for the free BAI Banking Strategies newsletter and get industry insights delivered to your inbox.

One problem with traditional diversity metrics is that they mask the diversity within minority groups. If you just look at the percentage of your workforce that belongs to a particular minority group, you’ll miss a wide variety of factors that could be influencing those employees’ experiences as well, such as job family, location and job level.

A senior manager’s reasons for leaving the company are probably different than those of an entry-level employee. Similarly, employees in different departments likely have different feelings toward the workplace. If you want to improve retention of diverse employees, you must address these individual situations, rather than rely on blanket initiatives that try to cover the entire minority group.

You can begin this effort by using people analytics to compare an employee to similar employees or employee groups, a process called cohort analysis. Cohort analysis can help you identify what is happening to diverse individuals, why it’s happening and what needs to be done to improve their experiences.

Financial services leaders nationwide are working to build diverse, equitable and inclusive organizations. Learn more by visiting BAI’s DEI Resource Center.

Diagnosing issues with cohort analysis

Start by identifying a group that has the potential to make a big impact on diversity at your company, such as members of a particular minority group who have recently been promoted to management. Make sure the group is time-bound — for example, you could choose to focus on a cohort that was promoted two years ago. Then, analyze two key factors:

1. Survivorship

How many of these employees are still with the company? If many of them have left, you have a retention problem to address. A retention problem isn’t always about compensation. Dive deeper to find any commonalities that may have spurred a decision to leave, such as missed promotions or interactions with a particular manager.

Remember that discrimination and microaggressions can reveal themselves in subtle ways that are hard to track with HR data. A lack of mentorship, being overlooked in meetings and similar issues can also drive an employee to look outside the company for their next career step.

2. Performance distribution

If the cohort you’re studying has mostly stayed with your company, the next question to ask is “Are they flourishing?” Compare compensation increases, engagement scores and other performance metrics for this group with peers in comparable positions across the company.

If those metrics seem low, again search for commonalities to discover which factors could be contributors. Subpar performance isn’t always due to lack of skill — the same factors that contribute to turnover can also sap motivation and morale. If the group is overperforming compared to their peers, that’s another important data point to examine. Look at which factors are likely contributing to positive performance, and see if you can replicate them for other target groups.

When designing diversity, equity and inclusion programs, it can be tempting to focus on high-level stats that appear to show the big picture at your company. But to really make a difference in those numbers, you must drill down deeper to understand the unique experiences of different cohorts within the wider group.

Once you’ve gotten started with cohort analysis, you can begin to incorporate other metrics, such as minority ratios by supervisor and promotion rates for different groups, to further refine your approach. By taking into account the varied situations of diverse employees, you can begin to design programs that truly make an impact, both for particular cohorts and underrepresented groups more broadly.

Ian Cook is VP, people analytics, at Visier.

Financial services leaders nationwide are working to build diverse, equitable and inclusive organizations. Learn more at BAI’s DEI Resource Center.