Tips for Market Research in Emerging Markets
For W.E.I.R.D. researchers working in contexts different from their own, humility and curiosity are key
Our team packed into a modest but neatly organized house in Cagayan de Oro, Philippines. The walls were decorated with Catholic iconography, motivational posters, and cutout pictures of NBA stars. A teenage girl fried sisig a few feet away, and a gaggle of children looked on. Our team of a designer, researcher, content strategist, and front-end engineer watched a participant, “Carla” (not her real name), testing a new prototype.
After struggling with one step in the product flow, Carla started to tell us that in this situation she would ask her husband to help her, because they share a phone. Then Carla stopped short. Although she needed help, she realized she might not want her husband to see this specific content on her phone. She wondered how she’d be able to hide some of it before asking him for help.
This research revealed that people who share phones want a different experience than those who view their phone as completely personal and private. Those who share a phone may prefer to be asked for their password more frequently, for example. Such differences are important because sharing phones with family members and friends is common around the world. There is also evidence from Kenya that rural populations, lower-income populations, and women are more likely to share a phone
W.E.I.R.D. researchers and the connectivity frontier
Increasingly, people coming to Facebook products for the first time will also be coming to the internet for the first time. As a researcher at internet.org, Facebook’s effort to make internet access more affordable around the world, I used mixed methods research to understand the experiences of people who are new to the internet.
When connecting to the internet, people in much of the world optimize over a different set of constraints than we do in the U.S. and in other places where Wi-Fi and LTE data are nearly ubiquitous. They must reckon with slow and spotty network coverage, power outages, limited phone storage space and battery life, and other factors, in addition to the phone sharing our team observed in the Philippines.
To help with some of these challenges, a large and growing population is now using lighter, more affordable tools to access the internet. Over 200 million people around the world now use Facebook Lite, a streamlined version of the Facebook app that takes up less space on a phone.
The quality of these initial experiences is essential to setting people on a path to accessing information that is relevant to their daily lives. At the same time, the bulk of existing psychology and user experience research has been carried out with W.E.I.R.D. populations: people from Western, Educated, Industrialized, Rich, and Democratic contexts.
Many insights from the market research we do here in Silicon Valley do not apply to other countries and populations.
For me and other researchers with a W.E.I.R.D. background, understanding the needs of people who don’t come from the same context requires deliberate thought and changes in approach. Here are some tips I’ve learned doing research at internet.org and in my career in international development research:
1. Constantly interrogate your cultural biases.
As researchers who are different from the populations we study, we must be ever attentive to how our involvement sways results. It’s especially important to be attentive to our own biases, both in designing research and interpreting results.
Is it about you or them?
International researchers must balance being physically present in the interview, which allows for a more immersive experience, and protecting the authenticity of the conversation. When we visited Carla’s family, we were sensitive to bringing five foreigners into her home. While exposing product team members to the realities on the ground is important, it can hinder your ability to observe participants acting naturally. In research with populations that are different from you, consider how your role as an outsider might impact responses. Keep an eye out for these two scenarios in particular:
- Social desirability bias, or participants telling the researchers what they think they want to hear, is common in research. To address this, we explain to research participants that all information is confidential and that we are interested in their honest opinion. There is some evidence that using technology that removes the human interviewer from the survey, such as interactive voice response computer calling, results in more accurate reporting when researching sensitive topics.
- Participants may overstate needs when foreigners are present if they perceive that the research will result in increased benefits, such as benefits from an NGO or foundation. You can read more about this kind of measurement error here and here.
One of the most effective ways to address such biases is to work closely with researchers from the country or location where the research is being done at each step of the planning and execution process. This is always best practice as their input and insights are invaluable.
2. Be mindful of gender, class, language, and power dynamics.
In focus group discussions, especially when covering topics that may differ by income or gender, it’s ideal to have a homogeneous group of people who come from about the same socio-economic level. In a focus group discussion in rural Mexico, for example, one woman who was much wealthier than the others made the other group members feel bad, creating a dynamic that was hard to repair in the session.
Similarly, when conducting a face-to-face survey in Pakistan, we hired interviewers of both genders and assigned one-on-one interviews with an interviewer of the same gender so participants would feel comfortable.
3. Be flexible and prepare low-tech UX research options.
Depending on the country or context, we can’t always expect tools like prototypes or video recordings to work well — or at all. I’ve learned not to plan anything too high tech, and to have lots of backups. We run tests at 2G speed, and bring pocket Wi-Fi connectors in case the network signal is weak. But no matter how prepared you are, you’ll likely have to be flexible and think on your feet:
- For research in Zimbabwe, we’d prepared slides to show to focus group participants, but the power went out early in the session. We had no internet, no projector, and a computer quickly losing charge. Less accustomed to power outages, my first thought was to print out materials. Um … of course printers also require electricity! In this situation, we had to revamp the focus group and come up with alternative analog activities on the fly, like asking participants to draw representations of their communication world.
- In our Philippines work, we learned that cell phone signal often does not penetrate cinderblock, a common building material in houses and buildings around the world.
- In conducting focus groups with women in Africa and India, we found that women may need to bring their children to the session. We found ways to accommodate mothers, bringing someone from the local research firm to play with children, and allowing mothers to nurse during a focus group if needed.
5. Value research participants’ time.
It is best practice to compensate research participants for the time they spend away from their jobs and other activities to participate in research. Specific compensation tips include:
- Make sure compensation is done discreetly so as not to cause tension in the household or neighborhood. If possible, have someone other than the moderator provide compensation in order to separate the research from the payment.
- In some cultures, non-monetary compensation may be more respectful and appropriate than cash. Groceries and other necessary household items can be a good alternative. Discuss this with local partners to find compensation that will be useful and appropriate.
- Don’t offer compensation that is out of touch with the needs of the study population, such as gift cards for online e-commerce for people who don’t have formal shipping addresses, or gift cards to a large urban store for people in rural areas.
6. Meet participants where they are, with language that is natural to them.
I see my colleagues in research employing amazingly creative research methods that often rely on making metaphors or ranking ideas according to a series of attributes. But with participants who may not have spent much time in a formal education system, I’ve found that asking concrete and direct questions works best. For example:
- For quantitative surveys, we’ve found it more effective to ask about an action over a recent, defined time frame than to ask about general behaviors or frequency. Recall is better over shorter periods of time.
- Carefully consider the units included in questions and use the phrasing that is most natural among participants. It may be difficult for people to estimate a distance in kilometers, for example, but they likely know how long it takes them to travel to a given location. Similarly, people may not use percentages often in their daily lives, but may be comfortable recalling amounts.
- Participants sometimes get emotional when the subject is personal, such as aspirations, how their business is doing, or interpersonal relationships. Make sure moderators are prepared to handle sensitive topics and show kindness to make the participant comfortable.
7. Get out of the office or lab — and out of the capital.
In emerging markets, the divide between urban and rural realities is stark. According to the World Bank, 32% of low-income countries’ populations live in the largest city, compared with 13% in upper-middle-income countries, on average. This implies that one large city attracts the majority of the economic opportunities and migration in lower-income countries. Visiting the largest cities in qualitative research, or having a large sample size from the cities in quantitative surveys, is essential in most cases. But failing to sample from peri-urban and rural areas can leave researchers with an incomplete picture of the country.
Make sure you’re ready for the special challenges rural research can present. For example:
- Be mindful of local happenings when recruiting or scheduling interviews. Farming schedules, market days, and religious and cultural celebrations can all affect people’s availability either at home or in town.
- In quantitative surveys, it may be wise to oversample rural populations in order to enable comparisons with urban ones. Having the statistical power to make comparisons between urban and rural populations is often important.
Rising to the challenge of building inclusive products
Internet.org products are designed for people who are doing a lot with limited data and a range of mobile devices. Despite these constraints, we want these products to be just as high-quality and well-suited to people’s needs as any “mainstream” product designed with W.E.I.R.D. users in mind. Great user research is a key step in that process — and it’s also a lot of fun! To anyone looking to get started with research in emerging markets, I encourage you to embrace, rather than be intimidated by, unfamiliar settings. Set your assumptions aside and approach people with curiosity, humility, and respect, and you will surely learn something unexpected.
I’d love to hear your tips and stories about research in emerging markets! Please share in the comments.
Author: Caitlin Sanford, Researcher at Facebook
Illustrator: Drew Bardana