Features

XXI.1 January - February 2014
Page: 36
Digital Citation

Are you sure your software is gender-neutral?


Authors:
Gayna Williams

Women account for 50 percent of people on payroll [1] and account for even greater purchasing power. Some products (all too often pink) are deliberately built for a female audience, and some are built for a male audience. However, many products are intended to be gender-neutral. Here, I address the latter and ask, exactly how well designed are these supposedly gender-neutral products?

Many papers focus on ways to increase the participation of women in computer science and the challenge of channeling more female candidates into IT positions. There is endless talk about the optimal representation of women in groups and how to achieve it. These are worthwhile topics for discussion, but I want to take the conversation in a different direction. I don’t want to see women missing out on great technologies while we wait for the pipeline of female candidates to fill, and I don’t want businesses missing out on being successful in the marketplace by not successfully creating products and services for 51 percent of the population [2].

Frequently, products are built with the assumption that they are gender-neutral; no gender is intentionally assigned to their users. But this does not mean they are gender-neutral. The Gates Foundation realized this during a crop-development project, when they discovered that an unintentional bias favored the perspective of male farmers and overlooked the requirements of female farmers, who were the primary family caregivers and tended the crops daily. This led to them proactively requiring gender awareness in project research, rather than assuming gender-neutrality [3]. Such awareness is often missing in software development. When gender is assumed not to be relevant to the users’ tasks, it is not discussed. For example, Margaret Burnett’s research on gender in writing and debugging software code has demonstrated gender differences in approaches. However, the features of commercial software tools are usually optimized around the preferences of male developers [4].

We need to consider the changes in women’s influence in the world of consumer and business spending and behavior. Women account for a large portion of consumer spending and are the decision makers in all major product categories: healthcare, finance, and education. Women are the most frequent users of social networks and online shopping, and even when they are not the early adopters of a technology, they provide a stable mass market. In the workplace, women are in more influential line-management positions than before and make up a large number of new small businesses. Technology appears in situations and environments where previously it did not and therefore is more relevant to and accessible by women, making women a significant, lucrative audience that businesses should not ignore or fumble over due to gender blindness.

So how do we remove gender blindness when we know we can’t change the IT workforce to reflect the gender balance in the customer base?

We could look to the methods of user-centered design (UCD) to address the challenge. However, those of us who have worked in technology companies and applied user-centered design methods know that to be successful, one must be strategic in applying a UCD approach and prioritize where to invest time and effort to persuade others. We are also aware that many critical product decisions are made outside documented processes; they are made in hallway conversations, pre-meeting discussions, and through differing interpretations of ambiguous rules and criteria.

So rather than looking to UCD as the only solution to address gender blindness, I took a broader look at the product-development cycle through a gender lens to discern where blindness can occur.

Insights

ins01.gif

Gender Distribution

Looking at the gender distribution within software development teams, we see that they are predominantly male, approximately 80 percent [5]. This is no surprise given the number of women graduating from computer science programs. For simplicity, imagine a development team that is 20 percent women, with fewer than 10 percent of the management roles filled by women. Now ask yourself, with 20 percent women on the team, what’s the likelihood that this team will develop a product free of gender blindness for a user base that is 50 percent women and 50 percent men?

Many decisions in the design and development of a complex product or service will be made under time pressure by individuals with limited access to information. This translates to gut-based decision making. It’s not uncommon for data to be subject to confirmation bias by individuals making quick decisions or compromises. The consequence is that this spontaneous decision making is made by four times as many men as women, and nine times as many male managers as female managers. There is no intention to not design for women, but the truth is that speedy decision making contributes to gender bias (see Figure 1).

Bias in Who Speaks up

Every team member brings a set of insights and experiences. Their perspectives are shaped by job discipline, age, family status, education, work experience, global experience, and other factors.

I conducted a series of interviews with individuals at a software company that yielded a perspective on how people allow personal opinions to influence their work on products and the potential impact on designing a product for a particular gender. The interviews were with both men and women, at different levels of seniority, in different job disciplines.

When first asked, men often initially assumed the products they were building were gender-neutral. However, when asked to reflect on whether a workforce that is approximately 80 percent male could take into account the female perspective to deliver gender-neutral products, their uniform answer was “probably not.” Women’s response to the same question was that their products definitely had a male bias. The question and the topic did not seem as novel to the women as it did to the men.

In the interviews I asked whether personal opinions arose in their work and their decision making. People in every discipline except market research and user research said they did surface their personal opinions or personal agendas in decision making.

Several factors that lead to voicing opinions were shared across gender and discipline. Common motivating factors included enthusiasm for getting something they personally wanted to use in the product and to have the product solve a problem they once experienced. The more a product was something used in their own lives, the more they were aware of bringing personal experiences to decision making. One man noted that he expressed a lot of opinions when working on a gaming platform and fewer when he worked on software for medium-size businesses, with which he didn’t have personal experience. People also reported enthusiasm for increasing exposure to something in a product, or adding options for more features when they really wanted a product to be liked. A third category of personal opinion was based on the person’s professional discipline; a team member may contribute to a discussion with a hidden agenda, such as a desire to benefit by trying a new approach or tool. A quality assurance manager said she had angled for a particular outcome because it would permit her team to try a new engineering approach.

Market and user researchers claimed they did not bring personal opinions and agendas to discussions. Their roles depend on their credibility for providing insight from unbiased interpretation of results. But even when researchers stay true to data and results, gender bias can creep in based on the research methods they employ.

The one type of opinion not voiced in the workplace during product development is one that exposes gender differences. Of the women who said that yes, they do bring opinions to the table, all said explicitly that they do not raise perspectives related to gender. Here are their three reasons for not sharing gender-based opinions:

  • Women do not want to offer ideas or feedback that would require identifying with women’s needs because often they do not want to draw attention in the room to the fact that they are women.
  • It is easier to generate support for a gender-neutral requirement, so if battles are to be picked, it is better not to propose a gender-specific case.
  • For women who have learned to act like “white male geeks” at work, considering a female perspective doesn’t cross their minds.

The interviews revealed that personal opinions influence decision making, but the fact that women are not comfortable in voicing their opinions when they are rooted in a female perspective suggests that including some women on a team will not correct for gender bias in the product.


Women are playing a more substantial role in the workforce and in how money is spent; therefore, it behooves businesses to ensure they are not blindsided by an erroneous belief that gender-neutral products happen by chance.


These findings are similar to those of Mendelberg and Karpowitz, who looked at how the positions of group participants vary with the percentage of women in the group [6]. Women are reluctant to raise certain issues in a face-to-face situation until the group consists of 60 to 80 percent women; yet in an anonymous survey situation they have no reservations in raising the same issues. This behavior is not observed with men.

Clearly it will require more than 20 percent representation of women on engineering teams for the power of their views to be leveraged unless we think differently about our processes, organizations, and approaches to communication.

Hidden Gender Bias in Product Cycle

Building successful products and services that do not exhibit gender bias need not await a more gender-balanced workforce, although when that is achieved it will help. In considering the product cycle from idea generation and funding through development and deployment, a close look shows many places where unintended gender bias affects decision making. Many product teams can point to a systematic process that drives their development cycle, yet every day hundreds of decisions are made by individuals on the teams. These decisions are made quickly, influenced by work and personal experience. For example, in the early stages of idea incubation, brainstorming sessions are often held and ideas are voted on to make the initial product feature list. Brainstorming is well known to be loaded with challenges for diverse representation, and voting tends to eliminate minority views. More than once in the interviews, women mentioned they wouldn’t suggest “feminine” ideas or vote on what appealed to them as a woman because they knew they would be outvoted; they didn’t want to be associated with a losing idea or cast a wasted vote. The next stage requires securing funding from predominantly male executives or investors, and teams will do whatever it takes to appeal to them, such as customizing their prototype pitches without realizing they are heading down a gender-biased path. During development, gut decisions affect the final product in numerous ways. Finally, beta feedback is captured from mostly male users. This was not necessarily an issue when feedback focused on bugs, but today’s more sophisticated tools provide qualitative feedback that influences future versions of products. Google was scrutinized on this issue by websites such as White Men Wearing Google Glass [7] and TechCrunch, which analyzed gender feedback in Twitter feeds on the hashtag #ifihadglass [8]. Google responded positively by sponsoring women-only hackathons with Google Glass developer kits to increase feedback on product direction. (More details on how gender bias appears in the product cycle are at http://ifshecanican.com.)

Becoming Gender-Aware

I believe that technology teams want to make the best products possible for their customer base, with maximum reach, and do not want to ignore gender bias. Below are recommendations designed to help teams remove unintended bias and to enable the views of women on a team to be heard. Removing bias doesn’t require costly efforts, laborious processes, or more hurdles; the key is raising awareness and the understanding that bottom-line profitability can be affected if habits don’t change.

It costs little or nothing to follow most of these recommendations for removing gender bias:

Costs nothing

  • Always use she as the default pronoun in your company.
  • Executives and investors (approximately 90 percent male) should push back on customization in demos and pitches that is intended to appeal to them personally; they should ask for a female customer to be represented in demos when the product is meant to be gender-neutral.
  • Data in presentations should be broken down by gender in footnotes (e.g., “beta feedback: 92% male, 8% female”).
  • Men should not assume they know the experiences, motivations, and behaviors of their spouses, daughters, or mothers; they should ask for their perspectives.

Small effort

  • Male investors and executives should have a trusted female reviewer for a balanced insight.
  • When brainstorming is important, assemble one group that is at least 60 percent female.
  • When using informal voting systems, consider equal vote distribution by gender.
  • Have a representative who is comfortable channeling ideas and feedback on products from women on the team, including: feedback on designs, quality criteria discussions, any other product- and marketing-specific issues.
  • Allow women to post anonymous product feedback related to “improving the product from a female perspective.” Explicitly consider gender suggestions and differences.
  • If a team believes there is no gender difference in behaviors or preferences, then dare them to do female-only research to balance the high male participation in building the product.

More than a small effort

  • Convene panels of early women adopters to balance the technical early adopters.

Conclusion

The purpose of this article is to highlight where the effects of gender-bias blindness can affect business success. Women are playing a more substantial role in the workforce and in how money is spent; therefore, it behooves businesses to ensure they are not blindsided by an erroneous belief that gender-neutral products happen by chance. Businesses need to be gender-aware in order to be deliberate in their decisions. This article is not about driving change to increase the number of women in software development, as that remains a complex systemic challenge; rather, it is about raising awareness of bias in processes where there are immediate opportunities for change. The reason to change is business opportunity.

References

1. English, A., Hartmann, H., and Hayes, J. Are women now half the labor force? The truth about women and equal participation in the labor force. Institute for Women’s Policy Research. Mar. 2010; http://www.iwpr.org/publications/pubs/are-women-now-half-the-labor-force-the-truth-about-women-and-equal-participation-in-the-labor-force

2. United States Census Bureau. Age and Sex Composition in the United States: 2011 – Population Table 1; http://www.census.gov/population/age/data/2011comp.html

3. Bill & Melinda Gates Foundation. Creating gender-responsive agricultural development programs. Feb. 2012; http://docs.gatesfoundation.org/learning/documents/gender-responsive-orientation-document.pdf

4. Beckwith, L., Burnett, M., Wiedenbeck, S., Cook, C., Sorte, S., and Hastings, M. Effectiveness of end-user debugging software features: Are there gender issues? Proc. of ACM Conference on Human Factors in Computing Systems ACM, New York 2005, 869–878.

5. National Center for Women in Information Technology. NCWIT scorecard: A report on the status of women in information technology 2010; http://www.ncwit.org/sites/default/files/resources/scorecard2010_printversion_web.pdf

6. Karpowitz, C.F., Mendelberg, T., and Shaker, L. Gender inequality in deliberative participation. American Political Science Review 106, 3 (2012), 533–547; http://dx.doi.org/10.1017/S0003055412000329

7. http://whitemenwearinggoogleglass.tumblr.com/

8. http://techcrunch.com/2013/03/03/google-glass-women-problems/

Author

Gayna Williams is founder of If She Can I Can, which focuses on career coaching and leadership training for women and gender research in industry settings. She also consults on creating positive people experiences through organizational change and customer-focused engagements, leveraging her years as director of user experience teams in high-tech industry. gayna@ifshecanican.com

Figures

F1Figure 1. Can the product team on the left build for the user population on the right?

COPYRIGHT HELD BY AUTHOR.

The Digital Library is published by the Association for Computing Machinery. Copyright © 2014 ACM, Inc.

Post Comment


@Sheila L Deskins (2015 01 27)

Nice article.
There is still a very large need of awareness for the issue, and glad to see this article written. The harder part is how to get theses issues to be considered important and acted on by males. Males are still the predominate leaders and decision makers.