In-situ research in Latin America

An enriching process

Editor’s note: Martha Rivera is senior research associate - Hispanic research at Doyle Research Associates, Chicago.

As a qualitative research method, in-situ research (in-home, in-store, in-car, etc.) was not intended to simply shift in-depth interviews from a traditional interview facility to a more natural setting. Questioning a consumer in his or her home does not automatically cancel the artificiality of the situation. To ask a housewife, for instance, how she prepares a particular dish while interviewer and interviewee sit on her couch may not be significantly less artificial than asking her the same question in a nicely furnished facility. She does not regularly prepare the dish in either setting. Even if the interview were conducted in her kitchen, artificiality may not be 100 percent eliminated.

The synergy produced by observation, participation and questioning is the real value of in-situ research. This goal is why anthropologists got out from behind their desks in the early twentieth century. We modern consumer researchers should do the same. With in-situ research in Latin America we encounter and must deal with not only cultural differences expressed by consumers but also by local research teams. Recognizing these differences leads to a more accurate and comprehensive understanding of the Latin American consumer.

Being on the same page

Ethnography is a coherent research approach encompassing a solid conceptual framework. It is expected that the ethnographer will perform three different types of activities while interacting with the interviewee during any in-situ research: observation, participation and questioning.

Observation

Observation of a behavior within its habitual context by direct observation or videotaping and photos is one of the most powerful sources of information about consumers. It reveals details that would never be reported by the consumer. Some of the insights resulting from the use of observations as a key element of consumer insights research are:

• No person can fully report his or her own unconscious behavior. Some behaviors that became habitual long ago are just ignored. You need to carefully think through how you brush your teeth before you are able to describe it, and despite your best effort you always miss some component of a behavior. Due to the fact that consumers do not recognize those minimal or automatic components of their behavior, they are not reported to the interviewer. Hence, ethnographic observation can bridge these gaps.

During an in-situ research study conducted in Venezuela  , in an ordinary bathroom I saw a large multicolored, perfectly round sphere made of small bits of body soap. The interviewee explained that he usually compressed together these bits of body soap when they had become too small. None of the previous interviewees had reported something like that, but all subsequent participants were asked what they regularly did with the last pieces of their body soap. Most of them automatically placed the old small piece to the new soap bar without another thought. Supplementary questions revealed that the flat form of the client’s soap bar was thought to facilitate this, and that these consumers actually looked for that particular brand for that reason. Hence the client abandoned its plan to change the shape of its soap.

• People’s reports of their own behavior always contain at least a small portion of biases and inaccuracies. Subconscious aspirations - and the human being’s natural need for resolving the gap between what they know about their own behavior and what they do not - and other factors explain the inconsistencies between what people report and what is found out through observation.

While testing a new type of package for pre-prepared meat, Mexican housewives claimed that they would readily buy any pre-prepared food item because it offered the convenience that they reportedly expected from all types of products. However, it was observed that their refrigerators were filled exclusively with fresh food items, thus suggesting that these housewives habitually prepared their meals from scratch and did not consume any form of pre-prepared food. Further questions about such apparent incongruity confirmed this inference.

• Rarely is any person able to recognize and accurately describe the contextual triggers of their behavior. Because of repetitive occurrences within a specific context, there are subconscious associations and interdependent or isolated components of a particular environment - all of which can ignite behaviors. Often the person is hardly aware of the cause(s) of these behaviors.

Contextual triggers of the behavior rarely act alone. Several types of them converge to exert influence on people’s behavior. They might be physical (i.e., the actual place), time-related (routines exerted repetitively at specific hours or dates), social (people who are involved in the behavior or are present while it is performed), and behavioral (previous and simultaneous behaviors, such as chain behaviors, whether they are directly related to it or not).

Feeling the impulse to wash clothes while passing by a full dirty-clothes bag is an example of a physical trigger; to feel the same impulse on Tuesday early in the afternoon because that’s when such activity is usually performed is an example of time-related triggers. Lighting a cigar as soon as you see a particular friend who is often present while you are smoking exemplifies the action of social triggers. That you automatically pick up a beer after preparing popcorn and turning on the TV to watch the game is a good example of a chain-behavior trigger.

In-situ observation of the investigated consumption behavior will reveal these contextual triggers, thus generating otherwise unattainable insights. While visiting Colombian housewives to learn about their habits related to the preparation of empanadas (which are fried, filled pastries, similar to burritos but fully enclosed and small) I noticed that most of them stored cooking oil right next to a solid version of the same product which is considerably cheaper and performs the same function but lessens the flavor of the food. (Both varieties are regularly consumed in low-income households, but housewives tend to save the liquid oil for special plates.) Many reported frying the empanadas with liquid oil; however, as they began taking out the ingredients from their pantries, some interviewees glanced at the solid oil and decided to use it instead of the liquid. When asked about this change in their habits, they related that seeing the solid oil was a reminder of the higher price of the liquid and led them to prefer the solid.

Participation

Participation is, so far, the least frequent of the activities conducted in-situ by consumer researchers. Long ago anthropologists learned about the immense value of experiencing the investigated behavior to gain insight about it. We marketing researchers often limit our in-situ activities to plain questioning, pictures or videotaping and some observation. However, no immersion into the investigated consumption behavior is complete unless the researcher participates in it.

In-situ participation means that the researcher gets involved with the consumption behavior while it is performed. Rather than solely observing, the researcher would develop sensitivity related to the behaviors by means of, among other things, assisting the consumer to perform the investigated behavior, being part of the social situation when it is conducted, and repeating some of the consumer’s actions to learn from them. As a result, insights will be produced by a visceral apprehension of the actions and reactions related to the behavior and by a psychological connection to it.

Not all consumer behaviors are suitable for researcher involvement. Ethical and practical issues restrict the participation of the researcher in some activities, while participating in some others is not a real need. However, in-situ participative research is indicated when respondents belong to an ethnic or social group that is very different from the researcher’s, as is the case for U.S.  companies interested in Latin American or U.S. Hispanic consumers. In fact, participative methods were originally developed as a practical way to better understand unfamiliar ethnic groups.

Some specific business situations in which participative in-situ research was helpful include:

• Learning about motor or sensorial responses involved in consumption behavior. We once tested three different ball sizes for roll-on deodorant bottles. Female Ecuadorian consumers were provided with three choices and were requested to use them and to loudly describe their sensations. We, the researchers, were given the same three and sought to imitate all the movements performed by respondents, thus obtaining a deep sensory and motor comprehension of their verbal response.
• When researchers’ unfamiliarity with the product makes it hard to interpret verbal reports from consumers. We were commissioned to investigate the general mindset of Venezuelan truck drivers in regards to air filters, so short in-truck trips were performed. The truck drivers were requested to teach the ethnographers about the replacement of an external air filter on the truck’s exhaust system. As a result, the researchers were able to remove the air filters from their cases and to identify the right moment to replace them. Most drivers used their air filters far longer than recommended by the manufacturer.  The truckers waited until the filters got to a particular tone of gray, which we were able to reproduce for the written report. The manufacturer was able to prepare a simple sheet correlating the different degrees of gray with increasingly adverse consequences for the truck’s performance.
• When new ideas to improve or develop products are sought. We conducted an ethnographic study to understand decisions of mothers related to the food items they gave their children for lunch at school. It was learned that mothers regularly gave them bags of potato chips, and we learned that the children struggled to open them. We also experienced difficulties with them. Apparently children did not regularly consume the chips but did not return them home. In response a different sealing system for the bags was created.
• The desired answer to a particular consumer-related question has proved to be hard to find. I was once sent to visit a little town of fishermen who historically had been purchasers of large numbers of batteries. Without apparent reason, the fishermen markedly decreased their purchases. During my visits to their homes, I learned that they had recently devised a method to check if the battery still had power. It was a complicated process that included, among other things, burying the batteries underground and storing them in a freezer for 24 hours. I found it difficult to believe that this process led them to detect power but one of the fishermen placed a tiny metallic sheet at the bottom of the battery and told me to put my tongue on the top. “You will feel a flow of energy similar to what you feel when your boyfriend kisses you,” he assured. It was true.

Questioning

As discussed, in-situ questioning is only one of the three key elements interacting to produce synergistic outcomes from in-situ research. While consumer insights obtained by means of more traditional methods (such as focus groups and in-depth interviews) rely on respondents’ answers to either direct questions or projective exercises, insights from in-situ studies depend on observation and participation as much as on those answers. Hence, in-situ questioning has some distinctive functions:

• Interfacing with observation and participation to interpret what is watched or experienced. A permanent exercise of “thinking on your feet” is needed to achieve genuine synergy between questioning, observation and participation. “Can you show me?”- and “Can you teach me?”-type questions are constantly inserted in the interview guide. Hard conclusions are not to be drawn while the in-situ visit is being conducted. Sensitivity is needed to detect when and where the researcher should participate in the consumption behavior or should conduct more detailed observation.

To support the introduction in the marketplace of a new type of salt with a fluoride additive, I visited Colombian housewives. While one of the interviewees was talking about the new product, I noticed that pronouncing the product name was difficult for her. I also tried to pronounce it but had trouble myself. After two or three more interviews my tongue was still blocked, as were those of the interviewees. Launching the product with that name would be a serious deficiency in Latin America because the vast portion of grocery shopping is done at small stores where goods are obtained from attendants behind the counter. Hence, after the initial four in-homes we asked respondents for a pronounceable name for the fluoride compound. One of the names they suggested still forms part of the brand name.

• To guide the respondents through an unusual research experience that is fun for them and provides useful information about their behaviors. Being visited by a research team that takes photos, observes everything and participates in daily activities is trying for any consumer. Discussions on ethical issues and rapport techniques are mandatory for researchers, but they are not the focus of this article.

Instead, I want to point out that proper in-situ questioning can and should lead the consumer through a pleasant research experience in which he feels he or she wants to show more, say more, and share more with the researcher. During an in-situ project about lingerie, young female executives showed us some of the items they wore daily and others they reserved for special occasions. At some point the question “Is it complicated to care for these items?” was asked. As a response, several of the interviewees spontaneously offered to show us how they washed them, and some actually did it in front of us.

• To address inconsistencies between an interviewee’s responses and a researcher’s observations or experiences. This is one of the hardest moments of the in-situ research. By making it apparent that some contradictions have been identified, the researcher may make the Latin American consumer feel shy and stop them from participating candidly. However, inconsistencies between what was said and observed must be addressed. In fact, that is one of the critical reasons why in-situ research is conducted.

In order to learn how Peruvian housewives cooked a dehydrated mix-of-grains soup, whose preparation instructions specifically included keeping the whole package of grains boiling in a tightly covered pot, several in-homes were conducted. Most respondents rattled off by heart the cooking instructions when they were requested to teach us their methods to prepare it. However it was observed that many of them would interrupt the boiling process in order to take out the peas, and then heat it up again to boiling. Rather than saying, “Why do you not stick to cooking instructions?” we asked respondents if the peas were cooked already. In addition to learning their reasons for taking the peas out earlier, several other possible changes to the instructions were revealed.

Latin America is on a different page

Differences related to marketing research practices in Latin America and the U.S.  may discourage large international companies from conducting consumer insight studies for their business in Mexico or Central or South America . However, the size of the market - among other factors - makes Latin America particularly interesting for most business sectors, hence making consumer research mandatory. Understanding and accepting those differences will make the research process flow smoother and will produce more useful consumer insights, in turn providing better support for business in the region. Some of the differences that we at Doyle Research Associates have found while conducting in-situ research projects in Mexico and Brazil are described below.

Recruiting

• Developing screeners for U.S. Hispanics is complex work because they must include particular variables such as: preferred language at home and work; immigration generation (to obtain even samples of first- and second-generation respondents); and years attending Spanish or English school as an indicator of the participant’s language ability. Developing screeners for Latin American consumers is a different story. We have not often seen in Latin America our long, closed-ended question screeners but have come across either open questions or short, closed-ended questionnaires. Hence, Latin American fieldwork personnel doing interviews to identify potential respondents may find it difficult to handle extensive screeners.
• We have also detected important cultural differences related to how the screener questions and instructions are assimilated. While it is true that translation is a factor, wording also plays a key role. For example, you can indicate to a U.S.  interviewer “accept all boxed checks” and he will understand. However, this expression translated to either Spanish or Portuguese will not make sense within the context of the purpose for which it was stated. Mexican or Brazilian interviewers would expect a clearer expression such as “accept all respondents who said yes to questions two, three or four.”

Incentives

• Giving cash incentives to participants in qualitative studies is not a popular practice in most countries in Latin America . Mexico and Puerto Rico are the exceptions. Surely consumers in other Latin American countries would not reject cash incentives, but they are not used. Rather, most marketing research agencies give them either an “expensive” present or a store card. Presents may include small home appliances, fine perfumes or liquors or other foreign goods. Big-box or department store cards are the most popular. Not longer than one decade ago or so client companies still used to give respondents baskets of their products. They were highly appreciated because they were seen as a demonstration of genuine, personal gratitude of the client company toward consumers.

Scheduling

• Differences in business hours are a factor while scheduling in-situ projects in Latin America . In Mexico City , for instance, lunchtime is seldom earlier than 2 p.m. and takes up to two hours; hence, most potential respondents do not leave their work before 6:00 p.m. Sao Paulo ’s working hours are similar to those in the U.S.   While recruiting suppliers would find available respondents at literally any time, those who could participate during working hours hardly would be genuine representatives of average consumers. Very likely they would be unemployed people with a very reduced buying power (if any), thus their purchasing criteria are not necessarily typical.

Participants

• Shyness and modesty in front of foreign people are distinctive characteristics of the Latin American culture. While clients want to be present during in-situ studies, their presence could inhibit Latin American consumers or could result in responses founded on their subconscious desire to be agreeable to the foreign person. Very often the visited consumer will ask their visitors if they want something to drink. If the visitor accepts that offering and actually takes some water, soda, coffee or juice, the respondent can relax and feel more comfortable with the entire situation.
• Several members of the family will usually be present during in-home visits. Regularly, the research team will be first seated in the living room and often most present members of the family will come to greet the team and stay. Making them participate in all activities performed during the visit is not only an invaluable additional source of insights, but it also contributes to the comfort of both consumers and research teams.

Videotaping, filming, taking notes

• Security is a real issue in Latin America . People are constantly afraid of being robbed at their own homes. That is one of the reasons why research teams may unexpectedly be denied permission to videotape or take pictures. Even if the participant had previously agreed, he/she could simply change his/her mind at the last minute. Because of this, research teams must be quite skillful at observing. In order to be prepared for this eventuality and also because our care for the quality of our work, we regularly include an observation guide within the in-home guide.
• We have found that taking notes while conducting qualitative research with Hispanic respondents (either U.S. Hispanics or Latin Americans) can hamper interviewee spontaneity and participation. To the respondent, the qualitative sessions can resemble interviews before authority figures such as physicians, law enforcement officials and others, thus increasing the cultural shyness already mentioned.

Very enriching

In-situ research in Latin America can be a very enriching experience in both the professional and the personal spheres of life. It is an invaluable opportunity to learn from Latin American consumers rather than merely acquiring knowledge about them. Consumer insights obtained through in-situ studies are priceless keys to open client companies’ minds towards a region of the world whose consumers have proved to be loyal and highly profitable for most business sectors.