How to Get Buy-In—and Deliver—on DEI Goals
To creating lasting DEI change in your organization, leaders need a solid plan. Capturing metrics, using an evidence-based process and delivering on 1 percent solutions can move you closer to your DEI goals.
HR and Chief Diversity Officers (CDOs) will play the central role in ensuring sustainable companywide diversity, equity and inclusion (DEI) progress. Organizational change can take a decade and, not to put too fine a point on it, CEOs come and go, but HR is eternal. The first key to sustainable progress on DEI is to build bias interrupters into organizational systems instead of relying on companywide conversations about inclusion. Inclusion conversations are no substitute for interrupting bias in your company's business systems, using metrics to establish baselines and measure progress.
I will provide concrete guidance on how to construct an evidence-based case for your company's DEI goals and the structural change needed to achieve them. There are three basic steps: gather data to build a case, design evidence-based interventions and repeat.
Step 1: Gather metrics to determine if there is a problem.
Even if you have buy-in at the top, it's important to collect metrics that assess whether you have a problem, how big it is and where it shows up. No diversity effort can succeed without support from the middle, because it's midlevel managers who typically control access to opportunities.
Measuring demographic patterns in high-level jobs is important because they tell you whether your company has a problem getting women and people of color into leadership roles. But that is all they tell you. Such outcome metrics don't tell you why you have a problem or how to fix it. For that, you need process metrics.
It's easiest to explain process metrics in the context of hiring. To diagnose problems in hiring, you should be keeping track of who is in the original pool of applicants, who survives résumé review, who gets an interview, who survives the interviews, who gets offers, who accepts them, and (if salary is negotiable) what their starting salaries are. That's because the fix if you have a nondiverse applicant pool is totally different from the fix if no person of color makes it past the interview. Process metrics identify precisely what problem or problems you need to solve.
Another type of process metric provides a measurement of whether certain kinds of bias are playing out in a given system. For example, you can audit your performance evaluations to see if personal style—her "sharp elbows" or his "great smile"—gets equal airtime for different demographic groups. This will make it easy to see whether tightrope bias is affecting women and people of color. Studies suggest it usually is.
One comprehensive approach is to use the Workplace Experiences Survey, which provides a complete picture of whether employees report experiencing bias, whether they feel various business systems are fair and whether bias is affecting outcome measures such as belonging and intent to stay. The alternative is to collect the metrics needed for each business process to pinpoint whether bias is creeping in. Collecting the right metric can help CDOs and HR get the buy-in for an effective intervention.
For example, one creative HR employee at a professional services firm was trying to persuade her company to improve access to opportunities. So she did something simple. That firm already kept track of billable hours—and billable hours are the coin of the realm in professional services. She analyzed the hours in the two largest practice groups and found that, during the COVID-19 pandemic, white women were billing about one hundred fewer hours than white men, and people of color were billing two hundred to three hundred fewer hours. At her firm, professionals who did not bill enough hours did not thrive, so her findings were persuasive to leadership.
Another example comes from a company with a problem I call "magic jumpers." That's when white men jump up a level (or two) in a company where this rarely or never happens to women and people of color. It's very common.
Count the number of magic jumpers and keep track of who they are. A simple demographic breakdown can be telling—and persuasive.
Metrics are also important for other reasons. They are crucial for establishing baselines and measuring whether what you are doing is effective. They can help you celebrate small wins that build support for DEI efforts. Finally, given that eliminating bias is not a one-and-done proposition, they can keep guiding you toward your goal.
You may get pushback from in-house lawyers worried that keeping metrics will result in legal risk. You can introduce your legal department to some progressive outside counsel who have deep experience in keeping clients safe while they implement bold DEI initiatives. If your legal team's go-to outside counsel is more conservative, perhaps you could use a second opinion.
Keep in mind that businesses are willing to shoulder some level of legal risk for any business initiative they truly care about, so if your company insists on zero risk tolerance in the DEI context, that's an eloquent way to say that diversity is not really a priority for the company. But the key point is that the risk of negative publicity, lawsuits, and employee turnover from doing nothing, or too little, on DEI far outweigh the risks of collecting some data.
Step 2: Use evidence-based toolkits.
Small steps can produce big change—if they are evidence-based steps. Here's where you take the data you gathered in step 1 and try different interventions to interrupt the bias you've documented. Which interventions change the data?
While this step does involve a lot of experimenting, it's important to emphasize that you're not just throwing stuff at the wall to see what sticks; you can use social science to build a structured, strategic initiative.
Step 3: Keep at it.
You will not eliminate all the bias at your company with a single effort or in a single year—and that should not be the goal. The goal should be to take small steps consistently. If you build these kinds of evidence-based tweaks into your fundamental business systems, the organizational change you effect will be more resilient and long-lasting than a CEO-driven, conversation-based culture change. Use metrics to measure progress and go methodically—implementing evidence-based change after evidence-based change, deep in the day-to-day systems and educating managers at all levels.
Instead of trying to boil the ocean—and I know you know this—go for small wins at first to build up support for sustained organizational change. Do a pilot with some receptive managers and use your successes to pave the way for future steps.
The 1 Percent Solution
If you are an experienced HR or DEI professional, you know how many stakeholders there are in determining who gets hired at your organization, who gets access to valued opportunities, who gets a promotion and who gets a raise. You know only too well that adjusting any of the above will involve a complex organizational change process. And you know, only too well, how persistent and creative you need to be to effect organizational change. My advice: relentless pursuit of the 1 percent. Keep improving organizational systems in ways that are politically feasible right now, and then widely publicize the fruits of your labors to build political support for steps that will lead to other 1 percent improvements.
Reprinted by permission of Harvard Business Review Press. Excerpted from Bias Interrupted: Creating Inclusion For Real and For Good by Joan C. Williams. Copyright 2021 Joan Williams. All rights reserved.
Joan C. Williams is a Distinguished Professor of Law, Hastings Foundation Chair, and Director of the Center for WorkLife Law at the University of California, Hastings College of the Law. She is the author of Bias Interrupted and White Working Class and co-author of What Works for Women at Work.
An organization run by AI is not a futuristic concept. Such technology is already a part of many workplaces and will continue to shape the labor market and HR. Here's how employers and employees can successfully manage generative AI and other AI-powered systems.