Working with what they have

Editor's note: Based in Singapore, Robin de Rooij is director Asia Pacific at Netherlands-based research firm SKIM. Based in the U.K., Robert Dossin is the firm’s client solutions director EU.

More than 70 million people with lower incomes are crossing the threshold into the middle class each year; virtually all of them reside in emerging economies. By the end of the decade, roughly 40 percent of the world’s population will have achieved middle-class status by global standards, up from less than 20 percent today. The rapidly growing ranks of middle-class consumers span a dozen emerging nations, not just the fast-growing BRIC countries, and include almost 2 billion people who spend a total of $6.9 trillion annually. Research suggests that this figure will rise to $20 trillion during the next decade – about twice the current consumption in the United States.

Those in the lower socioeconomic classes (SECs) of developing countries in Asia, Africa and South America may not have a lot of purchasing power now but they represent a rapidly-growing market. For many multinational companies, these lower-income consumers present a key business opportunity. To ensure that companies can sell their products and services to this demanding group, it is important to use research to understand their preferences.

With limited Internet penetration in many markets, targeting lower SECs is typically a difficult task. Running research online can be close to impossible because of the difficulty in securing a representative sample, as the PC and laptop are much less prevalent than mobile phones. Hence, researchers often turn to face-to-face interviewing instead. This method of recruitment can be costly, time-consuming, labor intensive and often unrepresentative of the total market due to the logistical constraints of in-person interviewing.

Although desktop Internet reach is limited to reaching around 600 million people, current mobile Internet access is at 2 billion and growing. And the number of smartphone users is expected to reach 4 billion by 2020; that’s 80 percent of the adult population on Earth! As the gap between mobile and PC continues to widen, mobile is quickly replacing the laptop as the main channel to the Internet.

Now, with increasing mobile Internet penetration in emerging countries, there is finally an opportunity to access this target group through their smart phones – and even their low-tech feature phones.

Through the same development

In the past decades, technology has greatly impacted how we interact with the world around us. Where sending someone a message used to take a letter, an envelope, a stamp and a few days for delivery, we can now interact instantly by chat or video-call with anyone we know. Research has gone through the same development, with pen-and-paper research becoming quickly outdated. More and more research is designed to take advantage of modern-day computing power and the interface of choice: the smartphone screen.

The steady development toward research conducted on smartphones (devices with a capable processor and 3G or 4G accessibility) today enables the easy rendering of highly sophisticated mobile Web surveys. It is possible to replicate sites like Amazon and have respondents go through virtual shopping trips that perfectly mimic their real-life purchases. But the mobile reality in many developing markets is very different from what researchers in the developed world are accustomed to. Most people in these groups still use slow feature phones with small screens that are connected through much slower 2G networks, severely limiting what’s possible with a typical device in five to 10 minutes. This group is not used to shopping on Amazon on their phone, therefore a cool mobile survey that can replicate that experience doesn’t add value. Instead, it’s necessary to adapt to their reality and interact with the devices they have access to.

Forwarding-thinking researchers are exploring the limits of what is possible with mobile technologies in each market, using both basic and advanced methodologies. For developing countries, this means understanding how we can reach consumers using a low-tech phone, deploying advanced methods and applying trade-off techniques to forecast demand for specific products.

Case study: Painkiller usage and preferences in emerging markets
In the very competitive painkiller market, SKIM conducted a multi-country study using low-tech phones to assess market demand based on individual brand choice, package size, dosage, type (tablet or capsule) and price. The study output predicted significantly better than any previous forecasts and was very consistent across markets, including Brazil, Indonesia and South Africa.

Sample: It is clear that recruiting through traditional online methods yields a sample skewed towards higher SECs. The same is true when we recruit consumers who are smartphone users, which are more likely to be in a higher SEC. On the other hand, recruiting consumers who use only basic feature phones creates samples skewed towards lower SECs. To get a truly representative sample, one must recruit across all types of mobile devices. Because smartphone consumer penetration is not high in many emerging countries yet, it is important to ensure enough low-tech users will be recruited.

The mobile study in Indonesia, for example, was successfully completed with just two days of fieldwork in August 2015 and yielded 380 completes. Of those who responded, one-third used feature phones to complete the survey. The group consisted mainly of younger age groups, with around half of the respondents under 25 years old.

Design: Due to technical limitations, it is often assumed that more advanced research methodologies cannot be deployed on feature phones. However, if one strips the survey of all unnecessary components to focus on the essentials, it is possible to deploy advanced conjoint techniques on feature phones by enabling consumers to do simple trade-offs using their small screens.

For the Indonesian component of the painkiller study, researchers set out to determine how consumers choose a painkiller and which are the most appealing attributes. There were two or three concepts shown on-screen at a time. Consumers were asked to complete seven different choice tasks, each task varying the attributes of the product (brand, units, dosage, price and format). In each task, respondents chose one concept that they preferred most. This study was engineered to be completed with only 13 taps on a mobile phone.

Survey length: Related to the point above, respondents typically have a shorter attention span when answering a survey on a mobile phone compared to a laptop or a desktop or when being interviewed face-to-face. When setting up research to be conducted on a mobile device, the survey needs to be limited to the questions that truly contribute to answering the business question and avoid respondent fatigue. When a longer list of questions is necessary, it is best to split the research into multiple parts and ask respondents to participate in the second (or third) part at their own convenience for an additional incentive. This approach typically leads to higher involvement in phased surveys. In recent studies using this approach, up to 80 percent of feature-phone respondents who had already participated in the first part of the survey continued to participate in the second part when offered an incentive.

The challenge of reengineering conjoint from tablet to phone is not trivial: Just try to fit this the conjoint task shown in the table onto the display of a smartphone. Respondents were told to expect the survey to take 10 minutes; it it took on average only eight minutes.

The Indonesian portion of the survey quite successfully showed that brand is the most important attribute among consumers in the region and that Indonesians use painkillers for headaches and toothaches. Researchers also learned that respondents prefer tablets over capsules and – very importantly – that they are very price-sensitive.

Using this survey design, researchers could predict future buying behavior with statistical rigor based on just five trade-off choices made by 380 consumers. When compared with their actual purchase choices, researchers were able to successfully predict with accuracy comparable to that of traditional online studies.

Half the time

Traditional face-to-face interviews are usually costly due to their labor-intensive nature and usually take a lot of time because interviewers have to travel to the respondent, or vice versa. With mobile research, respondents can easily access the survey from their phones and insights can be gained in half the time. The cost of hiring staff is also significantly reduced, thereby making this form of research relatively more affordable when compared to traditional methods.

With its small screen sizes and shorter interview length, research on mobile devices is now a solid alternative to in-depth research in many countries. However, when targeting lower-income SEC consumers, tailoring the survey for low-tech phones is a requirement.

Finally, as business requirements of multinationals require increased access to lower-income consumers, the demand for quick insights is growing. Mobile phones – both smart and low-tech – are great tools for breaking down logistical and socioeconomic barriers for better access to consumers in every corner of the world.