Do you know how your company’s DEI efforts are going? What data does your company collect to track and shape these efforts?
In this episode, DEI strategist and consultant Lily Zheng explains why data-driven efforts are everything. The way people make lasting progress on diversity, equity, and inclusion is to measure outcomes.
During this year’s Women at Work Live event, Lily explained the opportunities that data can create for DEI. They give examples from different companies, including one that was able to discern where exactly their recruiting efforts became inequitable and how the company fixed it. Lily also has advice for making a difference with data even when your company is tiny or you’re starting from scratch or there’s no budget.
Guest experts:
Lily Zheng is a diversity, equity, and inclusion strategist, consultant, and speaker who works with organizations to achieve the DEI impact and outcomes they need. They are the author of DEI Deconstructed: Your No-Nonsense Guide to Doing the Work and Doing it Right.
Resources:
- “What Needs to Change About DEI — and What Doesn’t,” by Lily Zheng
- “To Make Lasting Progress on DEI, Measure Outcomes,” by Lily Zheng
- “The Failure of the DEI-Industrial Complex,” by Lily Zheng
- “To Avoid DEI Backlash, Focus on Changing Systems — Not People,” by Lily Zheng
- “To Build a DEI Program That Works, You Need Metrics,” by Joan C. Williams et al.
- “How to Push for Policy Changes at Your Company,” from Women at Work
Sign up for the Women at Work newsletter.
Email us: womenatwork@hbr.org
AMY BERNSTEIN: You’re listening to Women at Work from Harvard Business Review. I’m Amy Bernstein.
I have a couple of questions for you. How are your company’s DEI efforts going? How do you know? What data does your company collect and track that shapes those efforts? To strategist Lily Zheng, data-driven efforts are everything. The way people make lasting progress on diversity, equity, and inclusion is to measure outcomes, and I couldn’t agree more. During this year’s Women at Work live event, Lily explained the opportunities that data – when used ethically, of course – can create for DEI. Lily will give us examples from their consulting with different companies like the one that found out where exactly it’s recruiting efforts, which started out fair, took a turn, and how the company fixed the problem. Lily also has advice for making a difference even when the company is tiny, even when you’re starting from scratch, even when there’s no budget. Lily is someone who always makes me and Amy G. think and laugh, and we’re delighted to share this conversation with you.
Hey, Lily.
AMY GALLO: Hi, Lily.
LILY ZHENG: Hey, folks. Great to see y’all.
AMY BERNSTEIN: So, we’re going to get to data in a sec, but first I want to hear about the decisions you’ve observed business leaders making in response to the backlash against DEI. Can you take us through one of them?
LILY ZHENG: Yeah. Yeah. So, just to provide some context for the folks listening in, we are currently experiencing a backlash against diversity, equity, and inclusion work where perhaps folks are being misled by misinformation, or are otherwise not as enthusiastic about it as they have been perhaps compared to five years ago. And in response to it, I think we’re starting to see companies diverge pretty substantially when it comes to their approaches to DEI in the sense that whereas in 2020 or 2021, you saw a very consistent approach to DEI. Many companies started employee resource groups, organized voluntary, DEI committees and councils, made commitments, made donations, published a whole bunch of DEI positive articles on their website. Now we’re starting to see that diverge where some companies are continuing to stay the course or double down even and say, “No, this is actually what we care about. This is our brand. These are our values. We are not going to stop whether or not it’s popular.”
And we’re seeing other companies start to withdraw that support and say, “Okay, now that it’s not popular, we’re going to cut our DEI staff. We’re going to withdraw our DEI funding. We’re going to quietly be less vocal about our support or call it something else, or quietly put it into another department or otherwise take actions to deprioritize DEI, both in language and in commitment.” And so, I think that divergence is the big thing that we’re seeing right now.
AMY BERNSTEIN: Yeah.
AMY GALLO: And are you seeing companies having to dial back because of this, or choosing to dial back? Give us an example of a conversation you’ve had with a leader, with a group of leaders about this backlash and how they’re navigating it.
LILY ZHENG: Absolutely. So, I think what’s consistent across most companies that I’ve worked with and most leaders that I’ve spoken with in the last year or so is that everyone is anxious. Now, what people do in response to that anxiety is very different. Some folks say, “I’m anxious, but I have a lot of confidence that we’re doing things that are having an impact, and so I’m going to push through that anxiety. We’re going to keep doing what matters because I know that this is effective work.” Other leaders on the other hand, are perhaps leaning in, I would say, a little too far into that anxiety and saying, “You know what? Suddenly this feels risky. Suddenly even something like starting an employee resource group feels risky. Suddenly even saying the word diversity or inclusion or equity feels risky, and so maybe we just don’t do that.”
So, I’ve talked to a couple leaders who said, “What if we just call it other things? What if we find other terminology so that no one sees that we’re talking about DEI?” or, “What if we just say inclusion and drop the diversity and drop the equity?” or, “What if we just not only don’t talk about it, but we just stop doing it? I’m sure no one will notice.” And now we’re in this twilight zone of backlash where not only do all of those problems that I just talked about exist, now, some leaders are being swayed by folks who think that there is concrete proof that DEI is causing harm. And I’m like, where is that? There’s not concrete proof of anything you’re doing, let alone harm. So, I think we’re so far behind.
AMY GALLO: Right.
LILY ZHENG: We’re so far behind that I don’t even know how they can conclude that DEI is somehow making things worse. Maybe perhaps you collect some data, and then we can see if it’s actually causing harm because I think it’s just essentially going nowhere with a lot of companies, right? And this backlash, it could be… Okay.
AMY GALLO: Do it, Lily.
LILY ZHENG: A wild hot take.
AMY GALLO: I love a Lily hot take.
LILY ZHENG: Oh yeah, I’m full of them. A hot take could be that part of this backlash could be maybe a distraction from the possibility of collecting data. And so, people are so scared of the possibility that they could be held accountable that they’re buying into this narrative that suddenly everything is too risky, right? Suddenly DEI is evil. Suddenly DEI is bad. And so, that’s why we’re just never going to collect data ever, and we’re going to move on and hope this all goes away.
AMY GALLO: I do want to know, do you work with any companies that are based in a state where government has banned DEI initiatives or offices? And if so, how are you advising them to keep up the work? You mentioned some of the things that people are considering. Let’s drop the diversity and equity, let’s stop altogether, let’s call it something else. How are you advising them to keep going despite the, I would call a hostile environment?
LILY ZHENG: Yeah. Yeah. So, I am currently working with a few companies that are based in states where DEI is risky, maybe where there’s been proposed legislation to ban it, which I believe hasn’t passed yet for the companies that I’m working with. But I’m definitely working with several global companies that are operating in these states. And those companies have to be careful, because of course, they’re not going to change their entire global policy to align with the laws of one state, but they are tiptoeing, I’d say. They’re walking on eggshells. For those companies, a lot of them are saying things like, “Well, maybe we should just stop doing it.” And to them, I say, “Look, what matters beyond all else is that you’re actually achieving the outcomes that you say you are.”
And so, what I care about is, are you achieving diversity? Are you achieving an inclusive workplace? Are you achieving workplace equity for everyone involved? And if you have to, if for some reason someone’s made it illegal to say the word equity, that’s rough. What can we do to work with that and to make sure that we’re still achieving those outcomes? Because those outcomes matter more than anything else. I’ve been working with some of those companies on… For example, I gave someone the exercise the other day to design a workplace equity initiative around pay without ever using the word equity. It’s not actually that difficult, right? So, let me come up with one off the top of my head. We are taking efforts to correct disparities in pay by gender to ensure that the processes we’re using to pay people are fair for everyone. Done.
AMY GALLO: Right.
LILY ZHENG: I didn’t even say the word equity, I didn’t say diversity, I didn’t say inclusion, but if you actually follow through on that, you should be achieving something like greater equity at scale.
AMY GALLO: Yeah.
AMY BERNSTEIN: So, I have a question for you, Lily. If you notice that your company is pulling back on its DEI effort or whatever it’s calling those efforts and you believe the work was making a measurable difference, there’s data, what can you do? Have you seen employees actually reverse that pullback, or at least in some way affect a change in that policy?
LILY ZHENG: So, your if is very interesting because the number of companies I’ve seen that have extremely good measurement and are pulling back on their DEI efforts is very low. I’m trying to –
AMY BERNSTEIN: Right.
LILY ZHENG: – struggling to think of one right now. Usually what companies are doing is they’re doing a lot of movement around DEI. They’re doing a lot of actions, a lot of initiatives, but very little impact tracking, and they’re proposing reducing some of those initiatives. And so, we’re, as practitioners and proponents of DEI, in the tricky position of saying, “Well, simultaneously, I don’t think you should be doing initiatives just for the sake of doing initiatives, but also you shouldn’t be taking them away just for the sake of taking them away.” So, to those organizations, I would say, “Why are you moving these? Give me a good reason.” And we can use this also to say, “Why are you doing them? Why do they exist in the first place?”
And some of the folks who I talked to behind closed doors will say things like, “Honestly, Lily, we have no idea why we’re doing these. We’re only doing these things because some people ask for them in 2020 and we don’t really know. And now people are asking us to take them away, so I guess we’ll take them away.”
AMY BERNSTEIN: Wow.
LILY ZHENG: And I’m like, “Wow, that is… Thank you for being so honest, but also that is the least rigorous thing I’ve ever heard.” This isn’t how we run workplaces, right? It’s like, “Oh, why did you hire this person?” “I don’t know, because someone told me that they wanted it.” Right? No way. No way. We run organizations because there is a need, because we’re trying to do something to create some sort of impact. That should be why we organize any initiative, any intervention. And if we want to take one away, it’s because it’s not having the desired impact that we want it to have. So, frankly, I don’t mind if some companies get rid of some DEI initiatives that are not working, that they have data to show aren’t working. If they don’t have data in the first place, I think that’s the problem, right? You need to be able to show the value, the impact of every DEI thing you’re trying to do.
And then if something is working right in the scenario that you said, then there better be a really good reason why you’re getting rid of something that’s working. Are you going to replace it with something that’s going to work better, or are you just being pressured by external sources? If it’s the latter, then I don’t know what to tell them, right? You’re ignoring the data to cape to political pressure, so maybe don’t do that, right?
AMY GALLO: That’s good advice.
LILY ZHENG: Then that becomes something a consultant can’t fix. And I’m just like, “Well, I sure hope you’re ready for folks to be very mad at you for a long period of time.”
AMY GALLO: That’s right.
LILY ZHENG: But that situation is pretty rare for all the reasons I named, right? The lack of measurement is often the major problem there
AMY GALLO: Yeah. We have from a participant named Bethany. She asked, “Having data requires self-disclosure. How do you get more people to opt into self-disclosure efforts?”
LILY ZHENG: I think it’s because people view data as… What is it? People are scared that data will force them to be uncomfortable, and I think that’s accurate, right? In the same way where if you never take a COVID test, you never have COVID, in the way that if you never do an audit of your cybersecurity, you never have any cybersecurity issues, right? I didn’t coin this, but I say it a lot, it’s FOFO, fear of finding out, right?
AMY BERNSTEIN: Wow.
LILY ZHENG: People are scared. They’re scared of learning something that they might know intuitively but can deny. And then when they see the data, then they’re like, “Oh man, I can’t deny this anymore.” So, I think what I tell people is that fear is very normal, it’s very human, but we can view the possibility of having data, having transparency as an opportunity to grow and as an opportunity to improve. Progress is quite literally impossible unless you’re able to measure your present state. And so, for those organizations that are doing 20 DEI initiatives, I actually tell them, you are honestly wasting your time with these 20 initiatives unless you can show what it is they’re trying to achieve, right? The potential of these 20 initiatives only comes about if you’re actually measuring how impactful they are. Otherwise they’re just there because they’re popular, or they’re just there because someone wanted them to be there.
But wouldn’t it be cool if you had your 20 initiatives and you could say, “These 10 initiatives are increasing the belonging of these marginalized groups by 25% year over year?” That’d be incredible. Do you know how many folks you could brag to if you could say that? Do you know how incredible that would be if you could use that data to attract candidates? But instead, all you’re saying is, “We have an ERG.” Right? And all of your competitors are saying, “Well, we also have an ERG.” Wouldn’t it be cool if you could say, “Our ERG is different than our competitors because ours actually meaningfully increases people’s chances of career progression”? “Can our competitors say that? I don’t think so. They’re not collecting data.” Data gives you that opportunity.
AMY GALLO: Yeah. You wrote in an article for us about how, imagine if we ran other business initiatives with metrics like, “We participated in a sales webinar.” Right? No, you measure sales by dollars. Why would we not measure DEI initiatives with the same rigor?
LILY ZHENG: Right.
AMY GALLO: In that article, you go through sort of different areas of DEI and talk specifically about the outcomes that you could measure, which I think is so helpful, and I think our audience will find that really useful.
AMY BERNSTEIN: Yeah. Actually, I want to pull the camera back and sort of… You just spoke about the persuasive power of data. What are the other reasons that you advise your clients to ground their DEI efforts in data
LILY ZHENG: Accountability. That’s perhaps the most powerful one, being that every leader wants to feel like when they make a commitment, that they can be celebrated for it, right? Leaders want to feel good. Everyone wants to feel good. You can’t feel good unless you can feel like your promises are being kept and that you are delivering on what you have told people you’re going to deliver on. And so, sure, leaders can make a promise like, “In 2025, I will commit to racial equity.” So, I guess sometime in 2025, you could say, like, “Hey everyone, I committed to it.” And then people can say, “Yay. That’s great. Wonderful.” But if you say, “In 2025, we are going to meaningfully close the racial pay gap or the racial gap in promotion rate, or the racial gap in access to opportunity, we’re going to commit to closing that by 50%.” And then by the end of 2025, you can say, “Okay, we gathered the data. It turns out we closed it by 20%, which isn’t exactly what we promised, so I didn’t quite meet that promise, but we’re at 20. Next year, we’re going to make sure to keep on closing it.” There’s a very different kind of feeling around that, where even if you don’t quite meet your goal, the feeling in the workplace, “Well, there’s another empty promise that I can completely ignore.” It’s “Wow. I actually think we’re doing something. Wow, I actually think that this leader that made a commitment isn’t just spewing hot air. They’re doing something. I’m working in a company that’s doing something.” Most companies don’t do that, and that creates an incredible sense of loyalty, of commitment, of engagement, of satisfaction. It’s running a good organization, right? It’s running a healthy organization. You do what you say you do. Collecting data gives you that potential to be truly accountable.
AMY GALLO: Yeah. Lily, in that response, there’s so many reasons to have data, accountability, retention, persuasion, bragging rights, and I think actually even buy-in. There’s an article we published called Data-Driven Diversity written by Joan Williams and Jamie Dolkas, who are both at California College of Law, and they talk about how sharing the data helps get buy-in from people inside the organization.
LILY ZHENG: Yes.
AMY GALLO: So, if you want to sort of decrease some of the internal backlash, share the data and be honest about it. I’m thinking about an organization I did some work with. They were measuring the pay gap, and unfortunately, their annual measurement showed that the pay gap for gender got greater, not smaller. And so, they had a big question of, how do we spin this to the organization?
AMY BERNSTEIN: Oh my gosh.
AMY GALLO: And it’s like, no, there’s no spin. You’re disappointed, and just be honest about it, right?
AMY BERNSTEIN: Well, in fact, I wonder if you can give us an example of an organization that did the measurement, and then in disappointment, went back and redesigned processes, if you can talk us through how that worked.
AMY GALLO: Yeah.
AMY BERNSTEIN: Yeah.
LILY ZHENG: Yeah, yeah, yeah. I have worked with some good organizations, as it turns out. They’re not all bad. Yeah, so I worked with one that… This was a couple years back. They found that their hiring processes had a lot of folks falling through, so they actually did a really good job recruiting. They had close to equitable gender representation in their initial stage of recruiting and decent, I won’t say perfect, racial representation in their initial recruiting, but we actually got data showing that the further along they got in the hiring process, the phone screen, the interview, second interview, they started to see women and people of color falling off very rapidly and there was a big disparity there. And so, they found that data, decided that it wasn’t enough data, and actually did a whole bunch of interviews with their hiring managers, and with their recruiters and their interviewers, and found that there were some, not necessarily individual biases, but sort of procedural biases in the process where, for example, they lacked hiring rubrics for their second stage interviews.
And so, they had developed this sort of informal process where they essentially said, “Okay, well, it’s not really in a rubric, but if the candidate is very confident about this specific thing, then that’s a good sign that we’re going to move them forward.” And it turns out confidence for that specific thing was very racialized and very gendered, and they ended up having a lot of their women candidates and their people of color candidates fall through that gap. And they didn’t even know it because they weren’t tracking and they didn’t have a rubric. And so, one of the ways they fixed it, obviously they created a rubric, but they also implemented hiring panels. They learned about ways in which folks were falling through the gaps, and said, “How can we correct for this?” They went through the resume screening and said, “Okay, actually, what are the criteria that are required for success in this role? And how can we pass through folks who meet these criteria?”
There’s also a problem where I think… They were doing the thing where they had a lot of applications, and so someone somewhere was like, “You know what we should do? We’re having a lot of difficulty parsing these. Let’s just make it so everyone with an Ivy league background just immediately goes through and we toss out the other ones.”
AMY BERNSTEIN: Oh, wow.
LILY ZHENG: That’s a very explicit bias, and that’s a bias –
AMY BERNSTEIN: Wow.
LILY ZHENG: – that dramatically impacts the demographics of who makes it through. So, all of these little fixes, apply all of these at once. And then we started to see, a year or two later, that it wasn’t entirely fixed, but they were actually improving the pass-through rate of these marginalized candidates. So, I still need to check back in with them. I don’t think they had fully fixed it by the time I stopped working, but that is an example of them identifying these challenges and doing something to address it and seeing some movement on that front.
AMY GALLO: Right. I love it. Thinking about the reason to collect data, there’s a great comment from Chanita Foster who says, “One thing that cooks my grits is the need –” which is such a good use of that term, “– is the need for us to show the data to justify the need for DEI. I think it’s wild we have to commodify DEI and show how it’s profitable to do something about structural and systemic oppression and exclusion, not because it’s the right thing to do. Meanwhile, companies have mottos and corporate values based on the right thing to do.” Any reactions to that comment?
LILY ZHENG: Yeah, I think it’s complex. So, I agree, right? We shouldn’t need to rationalize DEI and we shouldn’t need data for companies to feel like it’s the right thing to do. I think I see it less as finding data to show that DEI is profitable, which a lot of it exists, and frankly, I don’t like it. It turns out the so-called business case for diversity, this idea that hiring more people of color and more women is good for the business, actually has some pretty substantial backlash effects associated with it, where the more you say it, the more you actually turn away marginalized candidates because they feel like they’re going to be commodified within the organization.
AMY GALLO: That makes sense.
LILY ZHENG: Even this comment itself, I think reflects exactly what this data found, which is that we shouldn’t be using data to show that, I don’t know, if you hire one more Asian person or if you hire one more Black person, you’ll make 20 more bucks this year. That’s extremely dehumanizing and a terrible use of data. What I’m hoping we can get towards is less in justifying the need for DEI through data, but in recontextualizing DEI as whether or not it’s the right thing to do, it is something that organizations need to do full stop, and the data helps them hold themselves accountable to doing it.
And so, I don’t usually use data to try to prove some sort of why I use data to demonstrate the how and the what, right? So, I assume that leaders have their own reasons for doing DEI, but I say, “Look, I couldn’t care less whether you’re doing DEI because you think it’s the right thing to do or because it’s going to make you more money. I’m here to make sure that if you promise to do it, you’re going to do it and it’s going to work.” That is, I think, what we should be using data for.
AMY GALLO: Yeah, that makes sense.
AMY BERNSTEIN: So, let’s dig into that a little bit, Lily. Molly in our audience asks, “What type of data do you track to show whether DEI efforts are working?” And then she asks, “What if your company is smaller and the end of any sort of diversity is very small?” What are your thoughts?
LILY ZHENG: Oh, that’s so interesting. Okay. Well, two different questions, and I guess my answer will differ slightly for each question. So, first, how do you track the effectiveness of DEI interventions? AB testing is a really great way to do it. You can also do longitudinal measurement. It’s a little less precise compared to the AB testing, but you can measure how the outcomes that you’re interested in are changing over time for a target population after you’re applying these DEI initiatives. You can also do things like pre- and post-testing – so, for things like DEI training, which I think is people’s go-to intervention, and perhaps it’s not always as successful as we want it to be. We can, on a very basic level, pre-test people on their, let’s say, usage of particular skills on their awareness of certain concepts, on certain behaviors, and then test them on the exact same things after the training and then several weeks or months after the training as well to see what’s changed over time.
We can also collect great qualitative data from things like employee engagement surveys or employee surveys in general around the effectiveness of different DEI initiatives. So, that also gives you really useful data. Of course, it’s not quantitative data, it’s not the same, but qualitative data is just as valuable. It gives you different insights. So, a whole bunch of ways we can do that. The second question is, how do you do DEI work that’s impactful when you have small N within a small organization? You just widen the aperture of the group that you’re looking for. So, instead of looking at, for example, gender and race intersectionally, maybe you don’t have enough people to do that… Right? Maybe your entire company is seven people. And so, rather than saying, oh, what’s the belonging of your men of color versus white men versus women of color versus white women, there might only be one person in each category, so you can’t do that. You can say, “okay, how can we increase belonging for the entire organization of seven people?” Right? “And last year, we found that only two out of seven people felt belonging above 50%, and so we want to get that higher to four out of seven people.” Demographics get a little tricky when you have small n, right? That’s a whole other question, but you can at least make some pretty substantial progress if you just look at the entire group.
AMY GALLO: Yeah. And I love that you don’t have to be an n of seven. You can be an n of 70,000 to think about, how do we increase the belonging of everyone. I like that focus you take of these efforts are not targeted just for marginalized groups. They are also beneficial to the whole organization. You should be measuring that. You’ve also written for us about how demographic representation isn’t the only important outcome to measure. We’ve been talking about this, but people can measure employee career progression, for example, or social impact or conflict resolution, something dear to my heart or environmental impact. Within all of those options, have you found there to be outcomes that are better to prioritize before others?
LILY ZHENG: Better to prioritize?
AMY GALLO: Or more impactful?
LILY ZHENG: You’re asking some hard questions here.
AMY GALLO: Yes.
LILY ZHENG: So, if I’m thinking about… Okay, pay is a huge one. Everyone wants to be paid fairly. I think that that’s a good place to start. I think pay in some ways is even more important than satisfaction because I would much rather someone were paid fairly and dissatisfied than very satisfied and being paid horribly. Let’s see what else. Enablement, so people feeling like they can do their job, they’re given the resources to actually do their job. Respect, inclusion, these are similar things. So, feeling respected and valued by members of their team, I think that’s a very important outcome. It’s a good predictor of other ones. And psychological safety, that’s also really high up there. So, people’s feeling of comfort in taking risks, in making mistakes and doing so without feeling like they’ll be punished by the folks around them.
AMY GALLO: I can imagine how that one cascades to so many other –
LILY ZHENG: Yes.
AMY GALLO: – of the outcomes you would measure. Do you want to go to another…
AMY BERNSTEIN: Yeah. I want to ask Inaga’s question because it’s a good one for this conversation because starting from scratch from. She says, “I’m a DEI associate at a nonprofit where my supervisor and I are building our DEI foundations from scratch as a new department that was created due to a significant need for equity and anti-racism at our organization. What if you don’t have much data to go off of? Where do you start?”
LILY ZHENG: Okay. So, if you are starting off from nothing, it’s honestly a really exciting place to be because you have enormous opportunity to shape how things develop. I’d say first, you need to learn how your organization is functioning to begin with. So, if your culture is good, what makes it good? What processes are good? What aspects of your culture are powerful? What is helping people feel good within your organization? And then how can you operationalize those? How can you create norms, processes, sometimes policies, requirements, expectations to ensure that the things that are working really well continue to work well? This is something that I talk to a lot of startups about. The thing where a lot of startups say, “Well, we don’t need any formal structure because everything’s working really well already, and everyone’s great and we’re all buddy-buddy.” And suddenly they add another 50 people to their team, and then it’s a dumpster fire because they never took the time to operationalize what made their culture good to begin with until it stopped existing.
So, I think you can do the same thing, understand how your organization is functioning well and put in processes to sustain that and maintain that over time. Then also try to understand where your organization is not working well. So, this comment mentioned a very strong need for equity. What does that mean? Why did that happen? Who’s falling through the gaps? What are the disparate experiences? These are all things that you can do. Data helps. Data helps enormously. But even if you can’t collect quantitative data, I would argue already that you have data. If you said there’s a strong need for equity, you’re telling me that you have collected some data, maybe qualitative data, maybe comments, maybe feedback. You’re already using it, right? Qualitative data is just as valuable as quant data. And if you take actions based on that feedback, I would argue you’re already using a basic data-driven DEI approach.
AMY GALLO: Yeah. One of the things about collecting data is that it’s most helpful if you have consistent data for a long period of time, right? So, you can say, “We’ve improved this.” But I think about that question and, because they’re starting from scratch, they might collect certain data the first year, but then reconsider what they collect the second year, third year. Do you recommend that that people are constantly rethinking what data they collect or do you hope for that consistency over a long period of time?
LILY ZHENG: I think eventually, I want orgs to get to that consistency, right? Every big organization needs to have that. It’s a requirement. I think if you’re a very small organization, I’m not going to say that every startup of 10, 15 people needs to have a longitudinal employee engagement survey of a hundred items every year, right? That’s not something that you can do every year, and it’s not the right environment for it. But I do think that so long as you’re being intentional with how you use that data, and I mentioned qualitative, so long as you’re consistently collecting a lot of qualitative data, I think that’s good enough, right? So, maybe in lieu of that quantitative yearly survey, you instead have an open feedback form that you collect comments for. And every quarter or maybe every month, you review all of the comments you get, announce it during your team meetings and make changes based on those recommendations. I would say that that’s essentially longitudinal data-driven DEI work, even if it’s all qualitative. And so, the consistency is the most important regardless of what the actual form of the data is.
AMY BERNSTEIN: Yeah. So, you’ve talked about the importance of DEI as an accountability tool. I wonder if you can share one specific example that our audience can learn from – a company that is using it and using it well.
LILY ZHENG: So, I had a company that I worked with last year where they used their employee engagement survey to understand… It was belonging they were really looking into within their organization because they had gotten a lot of reports the year previous that lots of folks did not feel a strong sense of belonging within the organization, specifically disabled folks, women, and LGBTQ+ people, I believe. This organization used data to find out that the belonging gaps were because their managers had extreme variation in their ability to provide support for their direct reports. And so, they found that some departments and some managers were really good for pretty much everyone, and some departments and some managers were really bad for specifically women, disabled folks, and LGBTQ+ folks.
And so, what they did is they used that data to focus on improving those experiences. But because they were able to isolate it to managerial support within a few departments, they were able to focus on those departments. Imagine if you didn’t have the granularity of that data and just saw, “We have low belonging for LGBTQ+ people in our company. Let’s bring in a pride month speaker.” A pride month speaker… it’s not going to fix a department whose managers are all hom*ophobic, right? So, you see how unless you understand the challenge, your solution, your one-size-fits-all solution may not actually solve the problem.
And so, this organization was able to use that data to uncover that root cause, or at least some root causes. I’m sure there were other problems as well that weren’t captured by data, and we’re able to design a solution. In this case, it was targeted manager training, and specifically more guidance for the department manager, the department head of that department to address their problem. So, is that sort of what you’re looking for –
AMY GALLO: Yeah.
AMY BERNSTEIN: Yeah, exactly.
LILY ZHENG: – positive examples of how to use data this way?
AMY BERNSTEIN: Yeah.
AMY GALLO: Yeah, and also how to hold people accountable. I think in terms of what needs to change, it’s accountability and investigating the root causes.
AMY BERNSTEIN: But also the follow-through, I find so interesting – just understanding the nature of the problem before you leap to a solution.
AMY GALLO: Yes. Yeah.
LILY ZHENG: Right.
AMY BERNSTEIN: And data helped there.
LILY ZHENG: There’s a comment that I want to briefly make, because earlier on, you talked about building buy-in and using data to build buy-in, right? And I think this example that I shared is a good example of that because a lot of leaders that I talked to say, “DEI isn’t my problem.” Right? “It’s the HR leader’s problem, or it’s the DEI person’s problem.” And what you can do is you can actually say, “Well, let’s see, because you are responsible. So, you’re right, you’re not responsible for the entire organization, but you lead a department. So, when you can show me that your department’s DEI outcomes are peachy, doing great, then you can tell me that you’ve got it handled. But if they’re not looking good, then that is your problem. That is your responsibility. So, let’s see.” And I think giving people that data, not just everyone’s data, but the data that pertains to the area of the organization that they manage, is one of the most powerful ways that I’ve found to build buy-in because now suddenly it’s personal. No one wants to be leading a poor performing department.
AMY BERNSTEIN: Right.
AMY GALLO: And your point that a pride month speaker is not going to fix that department, it’s such a vivid example of why it’s important to get to the granular level and understand the root cause. I’m hoping for a Lily hot take here because one of the other questions I have is, what’s a popular practice that you see lots of companies using for DEI reasons that you’ve come to learn, either through research or personal experience, just is not effective and you wish everyone would stop?
LILY ZHENG: Okay, well, these aren’t hot takes. These are research-driven insights. Thank you very much.
AMY GALLO: There you go. Yes, there you go.
LILY ZHENG: Okay, so there’s a few. One of them is quite old, in fact. There was really interesting research on diversity statements, essentially showing that when you make a very public diversity statements as part of your hiring process, that actually results in substantially fewer members of marginalized groups hired, which is very unintuitive and very strange. The reason being that having diversity statements around, like, “Oh, we don’t discriminate, this is a fair process, we encourage diversity” actually encourages your candidates to hide less of themselves when they interview. So, they spend less time whitening their names, they spend more time being authentic, which sounds really good, except that then opens them up to more hiring discrimination during the hiring process themselves, which results in fewer of them being hired.
AMY GALLO: Right. Talk about unintended consequences, right?
LILY ZHENG: Exactly. Right?
AMY GALLO: Yeah. Yeah.
LILY ZHENG: So, the idea being that you cannot bootstrap an inherently racist or sexist hiring process just by having one comment saying, “We love diversity.” Right? You actually have to fix your hiring process. Another practice is de-identification of demographic characteristics, which I myself was calling it best practice back in 2015.
AMY GALLO: Explain what that is, Lily.
LILY ZHENG: Yeah. Yeah. So, you know how resumes have people’s names on them or their affiliations, and oftentimes those names are gendered and racialized, and so you can infer people’s name or their race or some other information about them just through the resume itself. The prevalence of studies finding bias against people with racialized names or feminine sounding names has led a lot of organizations to say, “Okay, got it. We’re going to strip all identifying information from these resumes and just review the facts.” Right? So, by taking out this information, we’re going to interrupt bias and that’s going to fix everything. It turns out it doesn’t. It doesn’t fix everything. In fact, sometimes it makes things worse because what happens is that people from different groups have different experiences in society, they experience discrimination, they experience marginalization. As a result, that has a substantial impact on the sorts of career trajectories, their career opportunities, their educational opportunities, and so on and so forth.
When you remove the context of their demographics, it makes those disparities stand out even more such that hiring managers will just say, “Oh, got it. Well, the only thing I have is candidate A has more experience here and candidate B has less experience here, so I’m just going to hire candidate A,” and it results in a dramatic, sometimes, in a dramatic drop of diversity and talent. And so, the idea here being we need to, instead of teaching people to literally not see race or not see gender, we need to teach people to be more intentional and more mindful to contextualize people’s demographics within their career experiences.
AMY BERNSTEIN: Yeah. I love that. Let me ask you one last question from our audience. Here it is. “How can we find the resources to do more with DEI? The HR training and evaluation efforts take resources we do not currently have.”
LILY ZHENG: To do more with DEI… really interesting. I would say DEI is not about doing more. It’s about achieving more. And so, I don’t care what you do. I care about what you achieve. And if this very expensive training doesn’t help you achieve some sort of change or help you achieve some sort of outcome, don’t do it. Power to you. Literally do anything else. Something I’ve said to folks before is that I would much rather you take the money you would spend on an expensive speaker that might not do anything for you long-term and spend it on pizza parties every month. Literally, if pizza parties every month would have a better positive impact on your team than a speaker, do that, right? And I think I would apply this in philosophy here. Look at the amount of resources that you have to spend and ask yourself, what is the greatest possible impact, lasting impact we can make with these resources on the DEI outcomes that we care about?
Go ahead and do that. There’s so many creative things you can do. There are programs you can invest in. There are community events you can sponsor. You could just take all that money and just pay people a little bit more in the organization. Maybe that’ll translate to better outcomes, right? But be creative about it. The goal is to shift those outcomes, not just to give the illusion of doing a whole bunch.
AMY GALLO: Thank you, Lily. Thanks so much for joining us today. This has been fantastic, as always.
LILY ZHENG: Thank you for having me.
AMY BERNSTEIN: Lily’s latest book is DEI Deconstructed: Your No-Nonsense Guide to Doing the Work and Doing It Right. The accompanying workbook is called Reconstructing DEI.
If you have ideas for policies that might move DEI forward where you work, but you’re not sure where to start, check out our 2022 episode, How To Push for Policy Changes At Your Company. In that one, Amy G. and I talk with Lily and a union leader about how to build a coalition around a cause, manage the risks involved in pushing for change, and ultimately how to get buy-in.
Women at Work’s editorial and production team is Amanda Kersey, Maureen Hoch, Tina Tobey Mack, Rob Eckhardt, Erica Truxler, Ian Fox, and Hannah Bates. We’re taking the summer to put together a solid season 10 for you. If there’s a particular topic you’d like us to cover, email us at womenatwork@hbr.org. I’m Amy Bernstein. Thanks for listening, and take care.
