Four Guiding Principles for User Research with Low-Income Women in Emerging Markets

September 16, 2020

By Sophie Theis, Qualitative Research Specialist at Women’s World Banking

Qualitative user research (also known as UX or user experience research) is a suite of strategies and methods that are essential in our work to design financial services for low-income women in emerging markets. User research infuses the design process with direct input from end-users, helping us define the problem space, discover unanticipated needs, and gather feedback from users to continuously refine a product or service. Put simply, there is no substitute for user research to gain a deeper understanding of customers’ behaviors and preferences to inform development of new products and services that help solve problems that they face.

The partners we work with—technology and financial institutions in Bangladesh, Egypt, India, Indonesia, Mexico, and Nigeria—recognize that they cannot design effective products based on their assumptions about customers. And they know that without making an intentional effort to investigate women’s needs, they will default to designing for only half of the population.

User research’s deep and continuous focus on the end-user is powerful. But a narrow focus on the individual user can also lead to blind spots. User research treats people as experts in their own lives and decisions, but we know that people cannot always explain why they act the way they do. Sociologist and user researcher Alba Villamil writes, there are limits to users’ ability to explain to researchers why they face certain barriers, especially when those barriers are complex and systemic.

In our work at Women’s World Banking and with partners, we design products that must respond to the complex and multi-level barriers to women’s access to finance. Our research needs a broader lens to capture this. Without attention to the social context and power dynamics in which users (and researchers) are embedded, research can easily misinterpret user behavior, bias results, or even inadvertently place research subjects at risk—risks that are more pronounced when conducting research with women.

To that end, here are four principles we’ve developed to frame our user research with women in low- and middle-income countries, plus tips for putting each principle into practice during research design, implementation, and interpretation.

  1. We need to view women’s individual behaviors within the context of their relationships.

In all stages of research, if we only study individual users’ engagement with a product or service in isolation, we will overlook systemic issues that influence their preferences and actions.

In many households, a “gatekeeper” has final say over what women can and cannot do. This gatekeeper may be her husband, or if she lives with her husband’s family, her in-laws.

Understanding a woman’s relationship with the gatekeeper can help explain behavior that at first does not seem economically rational. For example, when a woman says she is not interested in expanding her business, does this mean she lacks entrepreneurial ambition? Instead of making this assumption, researchers should explore from her perspective the risks she may anticipate if she expands her business. For many women, growing her business may raise concerns about losing control of her earnings to the gatekeeper, being unable to complete expected care work and household responsibilities, or even threaten the gatekeeper’s identity as breadwinner.

It would be a mistake to conclude that she is not entrepreneurial. In fact, financial products and technologies can play a role in shifting the risk-benefit calculation of business expansion by addressing these concerns, offering greater privacy, or helping women save time.

Relationships can constrain behavior, but they can also be powerful catalysts for behavior change. Many of the products we design draw on social support to build new habits, such as a savings product embedded in peer groups within Mexico and India or learning to use bank accounts provided through a social protection program in Indonesia Yet, in many cases, women need the permission of the gatekeeper of her household to adopt a new financial product.

If we want to empower women to use a product or service, we need to look at how her social context may restrict or support that usage.

Practical tips

  • As a starting point, review existing literature and learn from local community leaders, frontline workers, and researchers about gender norms and financial decision making in the specific context where you are working; partner with local researchers in research design; and conduct early stage discovery research with women focused on gender norms and household constraints to build an understanding of gender norms in context.
  • In customer journey mapping, take note of which decisions and stages in the customer journey are influenced directly or indirectly by different gatekeepers, and whether that influence restricts or supports her usage of financial services or involvement in financial decision-making.
  • Separate from interviews with women users, consider gathering feedback on prototypes directly from gatekeepers. What attitudes and mental models do they hold that influence their support of their partners’ use of a service or product? How do they need to be engaged?


  1. We need to recognize and respect women’s need for privacy.

Several decades of academic research in economics and sociology have dismantled the assumption of the “unitary household,” wherein all household members pool their resources and share the same preferences. When designing research with low-income women, we should not assume that there is full sharing of either resources or information within a household. In fact, a woman’s ability to keep her own assets private and separate from her household can be critical for preserving her economic autonomy.

In our research exploring the relationship between women’s financial inclusion and empowerment in Indonesia, women e-commerce entrepreneurs noted how keeping their savings private enables them to prioritize certain spending decisions. Women explained that with their own money they can send money to their parents—a choice husbands would not allow with the family’s joint budget. Privacy is also important to mitigate risks of a family member appropriating funds, as happened to one woman in our research who had no choice but to “lend” her husband $200—the equivalent of a month’s business profits. Privacy was so valued by participants in our research that some women described buying jewelry, taking home the certificate of purchase—and leaving the jewelry at the store as a strategy to save privately.

If we want to accurately understand women’s behaviors and motivations, we should never assume a unitary household model. We need to create a safe research environment where participants are able to speak privately about their finances and seek to understand the innovative strategies women use to ensure their privacy and autonomy in daily life.

Practical tips

  • Ensuring the privacy of research participants is essential. Arrange times to interview the participant when she is alone and, when doing remote research, explicitly confirm that she can speak privately before beginning the interview. Beware that many women may put the call on speaker phone when conducting phone interviews.
  • Use customer journey mapping with the participant to understand what information, assets, and actions women need to keep private and from whom.
  • Design of products and services should avoid undermining the strategies women use to maintain privacy and autonomy, as this could put them at risk.


  1. We need to study social norms, while taking steps to minimize their interference in research.

For both men and women, performing certain ideals associated with their gender can be strategic for numerous goals—including protection from violence, access to resources, and social status. The same dynamics can be at play when speaking with a researcher. Researchers should be alert to the risk of social desirability bias, when the research participant provides answers they think the researcher wants to hear.

A common example of social desirability bias is when women report that they are a housewife, even if they have an active business. They may downplay their economic contributions in order to emphasize their husband’s status as breadwinner and communicate their role as a good wife and mother according to local norms. Similarly, men may state that they involve their wives in all household decisions or give all their money to their wives to manage the household budget because this is what men assume the researcher wants to hear.

If we want to deeply understand women’s needs, behaviors, and motivations, we need to ensure our research unpacks social desirability bias in responses.

Practical tips

Researchers need to demonstrate that it is acceptable for participants to diverge from established social norms and make participants feel safe sharing how these dynamics play out in their lives. Techniques include:

  • Employ established informed consent practices throughout the research process. Not only is this essential for ethical research, it also safeguards data quality. When participants understand their participation is voluntary, they will be more comfortable providing truthful, fully nuanced responses to the researcher’s questions, which minimizes pressure to fabricate answers for questions they do not wish to discuss.
  • Ensure probing questions are included in interview guides to unpack socially desirable answers. For example, “What do you do for work?” should be followed by, “Are you involved in any side jobs or any other economic activities that generate income?”
  • Share vignettes about a decision and ask participants which stories are most like their lives; present scenarios and ask what women would do; or ask similar questions in different ways: what happens in your household, what happens in most households, and what are your attitudes towards certain normative statements?


  1. We need to distinguish between access to a product and ownership.

In UX research, it’s easy to just focus on how someone uses a product, rather than how or when they come to use it. Yet women in low- and middle-income countries are 8 percent less likely to own a mobile phone and 20 percent less likely to own a smartphone than men, and gender norms continue to play a role in discouraging women from owning phones.

Women are more likely to only have access to mobile phones and other technologies rather than ownership, with access having different implications for use. In our research on social commerce entrepreneurs in India and Indonesia, women often involve their partner to some degree in their business, using the husband’s phone, bank account, mobile wallet, or e-commerce account. But these products are not designed for joint use. In India, women social commerce entrepreneurs had to call their husband to confirm that their customer’s payment via GooglePay had registered in the husbands’ account. Several women grew so frustrated by the inconvenience that they decided to open their own bank account for the first time to get their own GooglePay account and stop relying on their husbands.

This example reflects the complexity of “joint” ownership. In some cases, joint ownership may be empowering. Family members or friends might serve as “proxies,” helping women learn how to use new technologies and completing tasks on her behalf to save her time. But in other cases, joint ownership may create friction or reduce women’s autonomy.

Because this varies by context, we should investigate, not assume, that our users prefer individual or joint ownership. If we want to understand how women use a product, our starting point should be to explore when she has access to the product, at what times, and for what purposes, rather than assume she has full ownership.

Practical tips

  • Always ask if women have access to a device, and if so, who the owner is. If a user does not own the device, when can she access it, what is she allowed to do with it, and how does this help or hinder her?
  • Investigate whether women prefer joint or individual ownership of the given technology or financial product, rather than assuming. Ask her to imagine how her life or business management would be different if her ownership was different.
  • In customer journey mapping, identify the main allies and proxies that support women in learning how to use a product and conduct participant observation.
  • Consider involving women’s allies and proxies alongside women users to see how they interact when conducting contextual inquiry, usability testing, and prototyping.


These norms, situational factors, and power dynamics vary in different contexts and for different groups of women, but they are key to understanding user behavior in a specific time and place. Our practical “tips” are not formulas, but rather critical areas of inquiry that help us avoid common researcher assumptions. When building new products that work for women, we need to actively explore how these themes play out in research participants’ lives so that design reflects the reality of their lived experience.

Many of these themes we can learn about from the end-users. Others we can gather from the people in their lives. Broader context from academics, community leaders, organizers, educators, frontline workers, and others are also critical, especially when trying to leverage product design to contribute to solving social and structural problems.

Using these themes to frame research is just one part of a broader approach to respectful, culturally appropriate research. We must also appropriately compensate participants for their time, reduce the burden on vulnerable communities, involve users in research design and interpretation, diversify our design teams, reflect on and actively work to combat our biases, and redress power asymmetries in research.

As researchers, we never tire of the refrain: it all depends. But we hope that through sharing these principles, what it depends on is a bit clearer—as is the need to study these dynamics in each particular context.

Are you working on these issues? Have suggestions for evolving this approach? Get in touch with us. We’d love to continue the conversation.


Thank you to Ker Thao, Whitney Mapes, Sonja Kelly, and Marina Dimova for the ongoing discussion and feedback that informed this blog.