McKinsey & Co., the prestigious management-consulting firm, takes pride in giving advice to the world’s largest corporations about the smartest business practices. But a few years ago, McKinsey discovered it had a blind spot in its own workplace. At a time when companies are valuing diversity more than ever, the firm had a notable gender imbalance: far too few female hires and promotions.

McKinsey resolved not only to bring its own practices up to date, but also to become a leading authority on the issue. Since 2015, working with LeanIn.Org, the firm has published Women in the Workplace, a benchmark study. (For the 2018 report, 279 companies and 64,000 employees were surveyed.) A champion of McKinsey’s crusade for gender diversity happens to be a male executive, Kausik Rajgopal, the managing partner for the western U.S., whose office leads the firm’s work on the issue. He spoke with From Day One about the firm’s progress on the issue and the life experiences that made him an advocate. Excerpts:

FD1: When did the firm’s recent path to gender diversity begin?

Rajgopal:  We were having conversations about how do we grow the representation of women in the partnership.  And I remember many years before that, one of my mentees was a woman who came to me and told me that she was leaving the firm.  And she said something that stuck with me because she was about to become a mother.  She said, “Management consulting and motherhood are mutually exclusive.” I just remember being shocked by that. And that led me and several others to think about how we can build a profession, and certainly an institution, that not just attracts women but retains them at greater rates.

But then you fast forward to 2012-13, that’s when we started looking at the questions of representation systematically. And we said—quite logically, right?—if you want greater representation, you need to have it in our incoming recruiting classes. And we found that it was actually quite a bit below where we needed to be.  So that conversation on recruiting classes was probably the catalyst for a lot of the work that we’ve done on this.

Once you realized the situation, what were the next steps toward action?  

I think three big ones. One was to get our recruiting teams to be aware and make this a priority  reach out to women more systematically. So, for example, at Stanford, instead of just doing a single information session, which we had done historically, we said, “Why don’t we go do an event as well with the with the [dual-degree] women engineers?” So you reach out in a much more intimate, narrowcast way, as opposed to, “Here is a big company and show up to a single session and apply,” which has more of a factory feel to it.  And so that was one: using multiple venues to reach out to women who might be interested.

The second thing was to understand our own unconscious biases in recruiting. As an example, after interviewing candidates, we have a decision meeting where the partners who conduct interviews discuss all the candidates together. We began designating one of the partners in the room as the “debiasing advocate.” So everybody else in the room then expected them to push back on the questions. If somebody said, “Well, I found Jane to be very sharp and her resume’s impressive, but I’m not sure about her presence,” the debiasing advocate would then say, “Are you saying that because she’s a woman?”

And that led to a much more overt discussion, putting the bias on the table and being able to do that without judgment because we all have unconscious biases about lots of things. It doesn’t mean that we’re bad people.  But having the awareness and giving someone else the permission to name it and tease it out, and then have a discussion about what that means for what we’re trying to accomplish, was very powerful.  It was a real unlock for us.

The third factor was in reflecting on the face we present in recruiting, like who shows up from the firm to recruit. Getting that to be much more gender-balanced was a key part of that.  Do you look like a firm that welcomes gender diversity?  So we encouraged women to be ambassadors, to be more visible in the cultivation of candidates.

When did you start to see tangible results, to feel like, “Okay, we’re getting traction”? 

I would say it took a couple of years for all three of these pieces to come together. But once it did, the power of all three of these integrating created a multiplier effect, a sort of virtuous cycle. This year, the incoming class has reached virtual gender parity.

Along the way to that milestone, did your clients help push you in that direction?

Quite early on.  I would say culturally we are a very client-centric firm. We kind of wake up in the morning with benefits for them on our minds. I think culturally it’s very anticipant and appropriate, whenever we’re starting out to do something new and different, to get our clients’ advice and feedback. And for many of our clients, better diversity was already a priority.

So we were already starting to get feedback like, “In your team rooms, we’d like to see more women, like in the teams that you bring to serve us.” So our clients are typically not shy about giving out that feedback and when we posited the idea of doing the research with them, their advice to us was, “If you’re going to do it, just make sure that you are also institutionally committed to this as a priority and making that [representative] of your own commitments.”

 How did you decide to go about making this report—and making it a definitive one?

Three quick thoughts on that.  One is we quickly said we can’t do this on our own.  So we do this in partnership with LeanIn.Org and the Wall Street Journal. And Lean In, independently, obviously, had been thinking about this and advocating for gender diversity and parity in the workplace. So that was a natural notion.

The second thing we said was that we need to make this a multiyear commitment. This can’t be a one-and-done. So we expect it will be an indefinite commitment, really, which was our clients’ early feedback to us.  And then the third piece was we wanted to make sure that it was grounded in real data and not in anecdote.  So the methodology of the report is very much empirical.  We go out, we now survey more than 300 companies in Corporate America. We get data from them at every level, on everything from recruiting to performance evaluation to promotions, representation at different levels. It’s a pretty exhaustive look.

We’re very thoughtful about privacy of personal information, but on our blinded database we actually get a pretty detailed and rich dataset that we can then look at for patterns and trends year over year. And over time, as we build that methodology and approach, we’ve also overlaid on it specific questions in areas of interest. For example, the most recent report goes quite deep into women of color and their experience in the workplace, the particular challenges they face.  And in each report. I think we’ll do something like that, where we go deep on a particular question.

What trends have you found over several years now?

Longitudinally we see very little movement in the percentage of women who rise to the top. So when we look at representation of women in the C-suite, that number doesn’t really raise significantly over time.  The other thing that’s quite striking is there is, I think, a loose cultural assumption that some of the lack of representation is actually a choice by women.  That women, for example, after they have children, become less ambitious. Or they don’t want to make it to the top.  And we found quite recently in our research that that is not true. That they want to progress, they want to make it to the top. So those were striking assumptions.

The most recent report shows that the first promotion is actually quite significant. Women are falling behind early, because first promotions are most inequitable. Women are 21% less likely than men to be promoted to manager. Black women are particularly disadvantaged, as they are 40% less likely than men to be promoted to manager. Then you can imagine how it compounds from there.

What other kinds of connections that you were able to make?

One of the critical things that the report validated is the importance of sponsorship. And we often talk about the distinction between a mentor and a sponsor. A mentor is somebody who may occasionally give you advice, who sort of cares about you. A sponsor is somebody who really creates opportunity for you and is committed to your career and to your success. We saw quite frequently in the research that women who had been successful in rising to the top had a much greater and more consistent preponderance of sponsors—and not just mentors.  As somebody said to me, advice is cheap.  Consistent support is what matters.

The other really striking observation, which is a little sad, is that a stunningly large percentage of men believe that this is not a problem.  In a company where only one in ten senior leaders is a woman, nearly 45% of men think women are well represented in the leadership. Even more stunningly, a significant percentage of women, 28%, agree. To us, that suggests a little bit about the anchoring bias and the visibility bias, when you look at a leadership team and say, “Oh, there’s ten people there and there’s two women on there, so clearly women can make it.” So that was kind of an eye opener for us. The mindset was much more pervasive than we expected.

One other one: We’re doing a fair amount of research on what I call the “only” phenomenon, which is often in a room, or a team at any level, there’s only one woman in the mix. And that can be a lonely experience. So far, a lot of what we’ve talked about [in Corporate America] is diversity. The more powerful and more important bit over time is inclusion. And this dynamic where we’re going to have one woman at every level can actually a quite isolating experience for the woman.

Have experiences in your life inspired you to feel strongly about this issue?

It’s very personal for me. I’m the only child of an Indian mother who has a college degree.  She graduated from college in 1962 and she’s one of the smartest, most thoughtful, most proactive people I know. And for cultural reasons she stayed home to take care of me and raise me. It would be a mistake to say she never worked—she worked a lot—but she never worked in the workforce. And I always thought that that was a stunning waste of human capital for the world.

My wife, I’d say, is another role model for me.  She’s a software engineer, has worked in Silicon Valley for most of her career. And I remember when I shared with her some of the insights we had on unconscious bias, kind of in the early days of the journey that I was describing to you, she said to me, “It’s great that you are now admiring the problem. Are you actually going to solve it?” I think both of those are quite personal motivations for me. So I’ve personally been a pretty active sponsor and supporter of women.

Another part of your job is co-leading the firm’s global efforts in financial technology. What patterns do you see regarding the role of women in technology?

Around the world, there’s a stunning lack of adequate female representation in the workplace. In my travels globally, yeah, there are cultural differences, some nuance. But in general the patterns that I describe, which are based on research in an American context—I think they’re all true. This is a global issue.

Steve Koepp is a co-founder of From Day One. Previously, he was editorial director of Time Inc. Books, executive editor of Fortune and deputy managing editor of Time