Getting the Most from Your Diversity Dollars

By Jennifer Garcia-AlonsoMatt KrentzClaire Tracey, and Miki Tsusaka


Getting the Most from Your Diversity Dollars: The Podcast

Listen to the podcast

Needed: A More Rigorous Approach

For senior executives, the main question about gender diversity is not whether but how. How can the company identify specific initiatives that offer the greatest payoff? How should the company allocate its resources—in time, managerial focus, and capital—to generate the biggest diversity gains? Unfortunately, many companies simply don’t know the answers to these questions, so they use a trial-and-error approach, trying a lot of things and hoping to see good results.

Getting the Most from Your Diversity Dollars: The Podcast

Listen to the podcast

Women@BCG Podcast

Getting the Most from Your Diversity Dollars: The Podcast

Listen to the podcast

That approach leads to wasted resources, and it may even be counterproductive. Although more than 90% of companies have some sort of gender-diversity program in place, our research findings indicate that only one in four women feel that they have personally benefited from such programs. When companies generate a lot of activity but little progress toward gender diversity, leaders begin to feel that interventions don’t work. Meanwhile, women grow frustrated, morale sinks, and company performance suffers.

Getting the Most from Your Diversity Dollars: The Podcast

Listen to the podcast

  • Companies need to understand which initiatives are worth pursuing and which are not. With that objective in mind, we surveyed roughly 17,500 employees and interviewed more than 200 senior executives at companies in various industries across 21 countries, seeking their perspective on 39 specific diversity initiatives. (See “Key Findings.” For a detailed description of our methodology, see the appendix.) This report discusses global findings; BCG is also publishing a series of market-specific reports that discuss the findings for individual countries in more detail.


Getting the Most from Your Diversity Dollars: The Podcast

Listen to the podcast

For companies seeking to stretch their diversity spending and generate a higher ROI, our findings point to three clear imperatives:

  • Companies need to understand their particular underlying issues. Do middle managers not support company-wide measures to increase gender diversity? Do aspects of the company culture implicitly favor men? In many cases, senior leaders (particularly men) and the women who work for them have dramatically different perspectives on these issues.
  • Companies need to know which measures are the most effective in addressing these problems. Our findings show that some common measures—such as one-time training sessions —are less effective than many people imagine. Conversely, other measures—such as interventions that target key points in women’s career paths (when they take parental leave or consider an international posting, for example)—tend to be underestimated but often lead to significant results. We call these measures hidden gems.
  • Companies should thoughtfully and deliberately implement the programs they introduce, testing the ROI of interventions and imposing a minimum bar for new ideas. Implementation is critical. Issuing a new policy on paper is all too easy. It is much harder to follow through and continually refine programs over time based on real-world experience. This report is a call to action for companies to run diversity programs with the same rigor that they would apply to any other major commercial priority.

Correcting the Misperceptions

Many companies are investing heavily in efforts to improve the gender balance of their workplaces. On average, our survey respondents reported being aware of 11 gender-diversity initiatives underway at their company, out of 39 total in our analysis. In the countries we studied, societal attitudes toward working women have improved, and the aggregate share of women in the workforce is growing in most markets.

Yet women remain underrepresented—often woefully so—on most leadership teams and boards. (See Exhibit 1.) And gender diversity programs aren’t helping as much as they should.

For company leaders, the first step in improving gender diversity involves accurately understanding the current situation at their organization. This may entail overcoming some initial misperceptions. Often, senior leaders—typically men—don’t grasp the scope of the problem or recognize the circumstances in which women experience the effects of gender imbalance most directly. For example, when asked if their company offers support for women from executives and middle managers, 72% of male respondents say yes, compared with only 54% of women.

Similarly, men often have different ideas about where things go wrong. They tend to identify recruitment as the biggest challenge in gender diversity. Women, by contrast, think that advancement and retention are the biggest challenges. (See Exhibit 2.) These misperceptions matter because they shape companies’ investments in diversity initiatives. Our data shows that companies tend to spread their diversity investments evenly across the five categories we tracked: recruitment, culture, leadership, retention, and advancement. That amounts to taking an all-of-the-above approach to the issue, rather than allocating investments in a way that emphasizes the areas that women see as the biggest problems: retention and advancement.

What Works

Companies need to focus on efforts that generate the highest ROI. Our analysis indicates that gender diversity initiatives fall into four main clusters. (See Exhibit 3.)

Hidden Gems

These initiatives are highly effective, yet many companies don’t pursue them actively enough. The ROI on these measures tends to be very high—much higher than leaders typically believe.

Targeting Interventions to Moments of Truth in Women’s Careers: At several critical junctures in a woman’s career, companies can take measures that have a disproportionate effect on that career’s trajectory. These moments of truth include returning from parental leave, applying for an overseas position, and aiming for a promotion with much greater responsibility. (See “GSK Provides Extra Support at Critical Career Junctures.”)


For example, women returning from maternity leave can benefit dramatically from measures such as flexible working models, additional administrative support, and more direct access to leadership. (See “Why Paid Family Leave Is Good Business,” BCG article, February 2017.) Similarly, when a woman considers taking an overseas position, clear communication and targeted logistical support—such as advice on housing, child care, schools, and taxes—can materially shift the likelihood of her moving, and thus her future career trajectory. (See “Women on the Move: Shaping Leaders Through Overseas Postings,” BCG article, May 2017.)

Addressing Unconscious Biases in Decisions About Recruitment, Evaluation, and Pro­motion: Periodically, companies make far-­reaching decisions about employee recruitment, evaluation, and promotion. At these moments, managers can yield to unconscious biases that cause them to favor people from similar backgrounds. In addition, organizations’ assessment processes (for example, automated tests for job applicants, or mechanisms to provide feedback about superiors, subordinates, and peers) can incorporate systemic and hidden skews.

There are specific ways to avoid these biases, such as by rephrasing job descriptions to make them gender-neutral, requiring diverse short-lists of candidates for open positions, adjusting the parameters and wording of automated testing, and refining the language or questions asked on performance evaluation forms. Although such changes are inexpensive to make, they can have a significant positive impact, so their ROI is strong. For example, Kimberly-Clark in the US revamped its hiring processes by, among other things, changing job descriptions to include transferable skills and demanding diverse lists of candidates for open positions. As a result, in four years, the company increased its share of women in leadership ranks from 17% to 30%.

Increasing the Visibility of Role Models in the Organization: Survey respondents ranked role-model visibility third among the 39 initi­atives we looked at: 44% of respondents cited it as being among the most effective measures. (See Exhibit 4.) Moreover, companies that already have role models in place—such as women in senior positions and men who have taken unconventional paths to the top—have found that giving these people a higher profile requires very little additional investment. The hard work is already done. By profiling such role models in employee communications, putting them on panel discussions, and celebrating their accomplishments, organizations can inspire women at lower levels of the company.

Supporting Internal and External Networks for Women: Fostering professional networks for women can be a very inexpensive proposition. Yet they create a valuable alternative channel of information and support. They are particularly effective in male-dominated organizations and industries. (Notably, only 27% of men in senior management positions rank these networks among the top five most effective interventions, compared with 38% of all women respondents.)

Offering Professional Development, Such as Talent Management Programs and Executive Coaching: Training and professional development can improve specific capabilities and allow women to access the tools they need to build long-term careers at a company. These measures work best when they are tailored to an individual’s needs, rather than taking the form of a one-size-fits-all program for all women at a particular level. They are particularly effective if used to help women negotiate transition points in their career. Coaching and development programs can be expensive, but not when compared with the cost of replacing a promising woman on the executive track.

Engaging Male Employees: A strong correlation exists between male support for gender diversity and reported progress over the past three years. (See Exhibit 5.) After all, many of the interventions described here benefit both men and women: improving diversity is in everyone’s interest. Moreover, when men participate in gender-diversity policies and programs, the programs become normalized and any stigma that might be associated with them fades. Companies can celebrate men who actively model key behaviors (such as taking their full paternity leave) and who actively sponsor women; involve men in designing and implementing key diversity policies; and set up a network of men interested in acting as champions.

Proven Measures

Proven measures deliver a high ROI, and leaders rightly believe in them.

Offering Flexible Work Models: Companies worldwide recognize flexible work policies as an effective way to improve gender diversity: 51% of respondents rank it among the most effective measures (more than any other). This is true across countries, genders, and leadership levels. Yet implementing flex-work policies poses some clear challenges. For example, women are more likely than men to fear that adopting such a model will limit their future career opportunities.

Leadership teams should therefore keep a few things in mind. First, flex-work should apply to both men and women and be visibly used by successful role models of both genders. Otherwise, there is a risk that it will lead to negative stereotypes (for example, “Only moms use those policies”). Companies must also make sure that the policy works on a day-to-day basis—and improves rather than diminishes productivity—by thinking through the details in advance, leveraging technology, continually assessing how and where to get work done most effectively, and fine-tuning the program over time. (See “Commonwealth Bank of Australia Creates a Customer-Centric Model for Flexible Work.”)


Committing Publicly to Gender Diversity: Public commitment to improving an organization’s gender diversity is crucial. The CEO and the entire leadership team must explicitly identify diversity as being among the company’s top priorities because of its importance to the firm’s strategic commercial agenda—and they need to back up that position with action, including in all of their interactions with middle management. (See “Unilever Launches a Multifront Campaign to Boost Gender Diversity.”)


Tracking Performance Through Meaningful KPIs: There is nearly universal awareness that tracking performance through metrics and key performance indicators is highly effective. For example, General Mills USA has increased its focus on diversity KPIs, with senior leaders accountable to the CEO for specific metrics. Thanks to this and other initiatives, the share of women in the com­pany’s senior management team increased from 9% in 2013 to 33% in 2016. KPIs should be practical, meaningful, and limited to a critical few.

Sponsoring Women Through a Scalable Program: A good sponsor, with the appro­priate supporting infrastructure, can help women navigate inflection points in their careers and lobby for them to receive promo­tions, training, and key assignments. However, the program needs to be wide reaching and consistent. C-suite executives at many organi­zations consider sponsorship to be more effective than it actually is, primarily because they experience it first-hand when working on diversity issues. Yet a large cor­poration will have far more women who would benefit from sponsors than it has senior executives who can fill the role. It is important that the company identify and sponsor talented women at scale.

Crafting Antidiscrimination Policies: Around the world, respondents cited antidiscrimination policies as the second-most-effective measure. Most companies—especially in developed markets—have these policies in place and may consider this type of initiative a given. But our respondents told us that a smart antidiscrimination policy can make a significant difference, by giving executives a means to explicitly define and communicate the company’s values, and perhaps even differentiating it from competitors by using strong statements of zero tolerance.

Overrated Measures

The third category consists of overrated measures. These initiatives appear promising on paper, tend to be fairly easy to launch, and often involve short-term bouts of focused effort. Unfortunately, in many instances they don’t lead to real change.

Holding One-Time Training Sessions: Many companies invest heavily in professional training sessions—for example, to eliminate unconscious biases among managers. Yet even though many senior managers favor these initiatives, our female survey respondents tell us that the sessions are generally not effective. A single event—often led by expensive outside vendors—is not enough to change deeply ingrained behaviors.

Establishing Grievance Systems: Grievance systems are a necessary element at any company, and senior managers tend to feel they create a safety net, but women do not generally perceive them to be effective in changing an organizational culture. We believe that women tend to use grievance systems only for extreme situations and not for milder day-to-day negative interactions that our respondents tell us wear women down and affect their career paths.

Adding Gender Diversity to a Long List of Leadership Compensation Targets: Linking executive compensation to diversity sounds like an important step, but respondents did not rate it highly in our survey. Why? C-suite executives have so many measures in their scorecards these days that one more doesn’t carry enough weight to change their day-to-day behaviors.

Recruiting Senior Women Role Models from Outside the Organization: Senior role models hired externally can be effective—especially for companies that need to rapidly improve their gender diversity. But external hires are no substitute for an organic pipeline of talent, and they can give the C-suite a false impression of progress. Although such hires put women at the highest levels of the company, they do not show a clear pathway to the top, through the layers of middle management, which is the biggest challenge that many women face. Junior women replying to our survey could clearly recognize the practical difference between women hired externally at senior levels and internally successful role models who showed that it was possible to rise through the ranks.

Baseline Measures

The fourth category consists of basic measures that all companies should have. These initiatives do have merit, but they are either baseline factors that do not make a meaningful difference in women’s day-to-day experiences or insufficient measures when launched in isolation. In addition, too frequently, companies do a poor job of implementing them.

Mentoring Women: Mentoring is the first step many organizations take to promote gender diversity. The concept is valid: executives meet one-on-one with high-potential women to offer general career guidance and advice, but the arrangement is generally less formal and structured than a typical executive sponsorship. The impact of mentorship hinges on the relationship and commitment between the mentor and mentee, and unfortunately these variables are difficult to engineer. For most women, mentorship may create some helpful support but does not alter the course of their careers.

Addressing the Gender Pay Gap: A baseline measure of equity consists in paying men and women equally for the same work. Clearly, companies need to understand whether they have a gender gap—and if so, they should take urgent steps to reduce it. But unless the pay gap is extreme, it may not be noticeable to individuals, particularly since most people don’t know what their colleagues earn. Also, companies may have valid reasons (such as performance and tenure in a given role) for salary differences in some individual cases. Fixing the gender pay gap alone is unlikely to significantly improve the advancement situation for women.

Company Engagement in Public Debates and Membership in Female Advocacy Organizations: Giving the company a public platform from which to lobby for gender diversity is an outward-facing measure that rarely changes the way individual women on the inside look at their careers. Many women are often not even aware that their organization is involved in such efforts unless it heavily promotes its participation. The most suitable participants are companies that already rank among the global leaders in diversity and can use their public profile as a differentiator in trying to attract female talent.

Disconnected Cultural Initiatives: Isolated, superficial initiatives to change an organization’s culture—such as a policy to reduce the importance of face time in the office— often fail to yield real results. Although the ideas themselves have value, they need to be part of a wider, more concerted effort or employees will perceive them as mere lip service. For example, rather than simply issuing a policy about face time, companies might create a formal flexible work policy.

Employee Surveys to Generate Feedback: Employee surveys are popular at many companies, but respondents give them low marks for effectiveness, primarily because companies rarely seem to do anything with the results. Among respondents who rank employee surveys among the five least-effective measures at their company, more than half point to poor implementation. Surveys can be useful, but only if leaders take visible action to address employee issues head-on.

Aside from these elements, one other initiative is worth mentioning: quotas for women in leadership positions. Quotas can be effective in generating momentum for companies that face large challenges and need to show rapid progress. But they are highly market-­specific; not surprisingly, they generated the most mixed reaction of all the initiatives we looked at. Some respondents hate them and believe they can be counterproductive (in that they stigmatize women as quota hires, unable to earn a position on their own merits). Others believe that they are a reasonable starting point for companies that need to generate immediate progress. Among survey respondents who are aware of the presence of such quotas at their organization, 26% rank them among the most effective measures, and approximately the same number rank them among the least effective. On the other hand, some companies have succeeded in building a gender-balanced workforce without instituting a quota system. (See “Bank Muscat Creates a Balanced Workforce Without Resorting to Quotas.”)


Implementation Is Crucial

What should CEOs do to create a step-change in gender diversity within their organization? Each company is at a different stage in its journey and will have its own challenges. Nevertheless, all companies can benefit from reviewing their current diversity initiatives and levels of investment and comparing these with our findings of what works. Companies at an advanced stage of gender diversity may already be pursuing some of our proven measures, but they can invest in additional hidden gems as well. Conversely, companies that are just starting their journey might choose to focus on a small number of proven measures to get those steps right and build momentum.

Leaders also need to pay close attention to how the company implements its diversity program. Mere activity—however well intentioned it may be—is no substitute for progress. In fact, CEOs need to apply the same rigor to diversity initiatives that they would to any large-scale change, setting strategic goals, measuring progress in terms of quantifiable outcomes, and refining over time.

Fundamentally, achieving success in gender diversity requires that leaders, managers, and employees change the way they behave, and this kind of change is hard. The process will likely take longer than many executives anticipate, and it will entail working in ways that feel unfamiliar and even uncomfortable. During the implementation phase, several principles are crucial:

  • Establish a clear business case at all levels. The argument in favor of gender diversity should be quantitative and unequivocal. Further, executives should communicate that case to midlevel managers, to ensure that everyone understands the rationale for diversity and is taking the necessary steps—large and small—to make it happen.
  • Get men involved. The group responsible for change should not consist of women alone. Otherwise, the program risks isolation from the company’s core operations as a special-interest “women’s program.” When men are active members of the diversity leadership team, companies make faster progress.
  • Think through the details. When companies fail to consider seemingly minor issues on the front end—and don’t continually monitor and refine policies after implementation—initiatives fail to deliver. Employees often see a wide gap between the formal policy and the way it works (or doesn’t) in the real world. Flex-time is a good example. Giving employees the flexibility to work on alternate schedules is easy—but it raises a host of follow-on questions. If a woman wants to work longer hours four days a week and take Wednesdays off, will she still have to come to the office for important meetings on her day off? Will someone be explicitly responsible for covering for her on Wednesdays? Will she ultimately be held responsible if her team makes a misstep when she isn‘t there? Is she alone responsible for explaining her schedule to clients and colleagues?
  • Tailor interventions to individual sites and markets. Even the most effective measures are not universally applicable. Pain points and cultural barriers can differ at the level of departments and business units in the same company. (See “PepsiCo Tailors a Global Diversity Program to the Needs of Local Markets.”) For example, in India and Brazil, close to 50% of respondents rated on-site childcare as among the most effective interventions, compared with fewer than 30% of respondents in the US and the UK.


  • Make sure that the first interventions succeed. Quick wins can generate momentum for a broader program, just as early failures can make people skeptical. For example, if a company rolls out a flexible working policy without devoting proper thought to the practical details, the first people to try it (likely women) may end up leaving—discouraging any further applicants and reinforcing stigmas regarding flex-work.
  • Be relentless in measuring outcomes, not just activity. Rather than measuring inputs, executives need to gauge performance based on changes in the career trajectories of women. This will require careful program design, to ensure the company has the right KPIs. (See Changing Change Management: A Blueprint That Takes Hold, BCG report, December 2012.)

Gender diversity is a critical opportunity for companies. Most leadership teams understand the stakes, and many have taken action to improve the balance in their workforces and executive suites. Now they need to make sure that those actions lead to real results. How? By treating gender diversity initiatives the same way they would any other strategically important business imperative.

Doing so entails accurately diagnosing the root causes of imbalance at the organization, identifying a small number of measures to address those problems head-on, tracking progress, ruthlessly culling initiatives that are not working, and focusing at a granular level on implementation details.