Shedding new light on an old flame
Editor’s note: Ruth Sando is senior associate at the Washington, D.C., office of Barbara Perry Associates, an ethnographic research firm. Donna Sweeney is associate research manager for the Consumer & Product Insights Group at SC Johnson, Racine, Wis.
Sherlock Holmes’ enduring popularity with readers arises from the excitement of watching him build important inferences from close observation of the details around him. While his alter ego, Watson, looks at things and sees nothing, Holmes extracts the relevant information that leads to the solution of the crime. Readers recognize that there is nothing magical about this process, and with good powers of observation and analysis, they could appear to be just as smart. Such is the inductive research method at work.
Ethnography, a research tool that is based on an inductive research approach, can produce exciting new findings for product developers, marketers, market researchers and advertisers. The current generation of market researchers has expanded the qualitative toolkit to include ethnography in consumer research. The process of observation and interviewing in context, often with video and audio taping as well, has become fairly standard. It does this in the same way Holmes produced his successes - deep attention to what is happening, without imposing assumptions, until the data reveals the meaning.
Ethnographic research needs to be approached in a way that maximizes the benefit of its inductive process and helps participants feel comfortable with the process. Otherwise, its open, interpretive and seemingly unstructured means of analysis can become a liability. In such cases, when the data has been collected, there are two types of problems that may develop. Some people will have already heard enough to have decided what the “solution” is, and then become disengaged. Others may feel overwhelmed by the quantity and diversity of the data and develop “analysis paralysis.”
So how can you ensure that projects not only conclude successfully, but also produce maximum value for the organization? If you look to the literature, the amount published on qualitative research is vast, but much of it appears to focus more on the data collection and project management than on how to handle the data after it is collected. The analysis stage of the project, particularly in a fast-paced business environment, is as crucial as collecting the right data. But it is perhaps not surprising that fewer writers address the challenges of analysis. Good analysis is a thoughtful and interpretive process, but one that is systematic, iterative and improved by structure even though the opposite appears to be true because of its initially unstructured nature.
Our experience, as shown in a recent project, points out the importance of an analytical process that resembles a funnel, one that has a very wide top and a narrow spout. In other words, one needs to make sure that the approach is very broad at the beginning, but becomes very tight and focused at the end.
To illustrate this approach, we will use a recent ethnographic project at SC Johnson. The objective was to produce new ideas for an old, commoditized product line. As will be shown, key elements of this process included putting together a diverse team from the organization, talking to consumers on the fringe of some key trends, iterative debriefings which allowed for themes to emerge, making use of state-of-the-art analytical tools, and keeping the focus on an actionable outcome.
Setting the stage - making something old new again
While there is no historical record of the first candles used by man, the earliest evidence of candle use dates from the fourth century B.C. Candles have undergone many technological changes in the centuries since their early use, but they still consist of a wick of some sort and an energy source such as the wax used for candles today. The current candle market is highly competitive, saturated and becoming increasingly commoditized. So, how does a company with a product that has been around for centuries gain ideas for developing the next generation of that product in such a challenging market?
The candles research team at SC Johnson tackled this very problem. They decided to utilize the ethnographic approach to gain a deeper understanding of candle usage and meaning directly from consumers. The ultimate goal of the research was to identify and describe consumer need states surrounding the experiential aspects of this product in several key areas. These needs would then be used to drive ideas for new products and next-generation candles. An additional goal, however, was to structure the project so that it would provide the context for aligning the organization around a future vision by collectively living through the research process. This second benefit of having a team approach to analysis creates the buy-in of the team members. Many times things like this are done by a few people and the results are just archived. This was a different story.
Following the funnel analogy, a broad, a cross-functional project team was formed that was led by SC Johnson’s Research, Development & Engineering Department, with the assistance of Barbara Perry Associates. The project team also included members from the marketing and market research departments and the company’s advertising agency. Inclusion of team members from these particular functions was designed to allow the project to be as inclusive as possible during the collection and analysis phases of the research. The objective was to create ownership of the findings among all team members, especially key stakeholders and decision-makers.
The data - building blocks of creativity
The first steps in the project were carefully planned to ensure both effective use of the team and collection of a maximum amount of data. The recruitment of team members was followed by a brief training session to produce alignment on the research process, create a focused research guide and articulate commonly held assumptions within the company on candles and candle uses. This last activity produced team awareness of the “lens” through which they would be filtering the data and helped them put aside pre-existing biases.
Respondents recruited for this study had to be heavily involved in candle use and also be considered “leading edge” on several key lifestyle, behavioral and attitudinal trends. Respondents were interviewed in their homes for an extended period of time and completed homework beforehand. The flow of the interview was to first understand their lives and values and then understand how the product fit into that context. The interview guide mainly served to ensure that we eventually covered all the categories and that all the interviews would have a basic set of data in common. However, by opening the control of the interview to the respondent, we obtained a variety of information that was broader and led the research in new and unexpected directions. Creativity in this project rested on the collection and organization of a wide variety of data that would allow the team to expand its thinking about the product category and pull in fresh insights about the way people had adapted it to their needs and attitudes. Our most important data consisted of notes, audiotapes, videotapes and pictures from in-home interviews as well as respondents’ homework assignments (logs and collages).
In addition to what was collected in the field, the team also identified two other sources of data during the planning meeting that proved extremely valuable during analysis. The first was a collection of a wide variety of candle products already on the market that had been gathered for a different project. The second was a pre-screened set of old product ideas that, for one reason or another, had not made it to market. Transcriptions of audiotapes became the basis for our thematic analysis during the next phase of the project, which was our final team debriefing and idea generation session.
Coding and organizing: working through the funnel
The data analysis can make or break a project, because the original data themselves provide us with rich text. It is the researcher’s challenge to take that data and create meaning, new directions and actionable implications for the organization. Continuing with our funnel analogy, the data analysis began in a way that was broad and inclusive.
It was a continuous process. We held lengthy team reviews and discussions after our interviews in each city. This “rolling debrief” process allowed themes, categories and new questions to evolve. In addition to sharing what had been seen and heard, team members visited stores in each city and brought back samples that conveyed some element of what was being learned. At the end of the last debriefing, we were ready to begin the process of moving toward the so-what challenge.
To begin this move, we needed to let the team work collaboratively through the data. Most of the team members realistically could only devote a very limited amount of time to this effort, due to the demands of their jobs. So a core team of researchers took the themes that emerged from the rolling debriefing sessions and, using NVivo (a qualitative software package from QSR International) created an online coding structure, with a tree logic, that incorporated all of the categories that had been discovered by the team plus any that emerged during this coding process. The software allowed us to create reports that sorted quotes across all interviews related to any one code. This “horizontal” look at the interviews made it possible to cross-check the categories we had developed. Did any look like they included material that was not really related? Were new sub-categories needed? We could also look for differences based on sorting the interviews by age or geography or any other basic segments that might seem useful. This was really the period of hypothesis-building. The rest of the team had the voices from the interviews in their minds as they went about their days and this became a period of fresh insights and deeper thinking about the data.
Team analysis and ideation: through the funnel we go
After two weeks of coding the data and generating the verbatims organized into themes, it was time to reassemble the full team and work through to the so-what at the end of the funnel. The team analysis took one day, which was made possible through careful planning and the use of materials that had already been organized and refined. To recreate the interview experience, we watched videotapes of the most important clips chosen for each interview. The room had posters made up of the collage materials created by consumers prior to our home visits. The software allowed us to create sets of consumer verbatims that team members could work on broken out by several coded categories. Using this material to determine the implications for new product ideas was a much easier task than working with the original interviews would have been. At the same time, because of their involvement in the field, work team members were very familiar with all the interviews and could evaluate the importance of concepts overall.
Each team was challenged to take its set of coded interview material and identify the underlying theme, the consumer need that was relevant, and product ideas that would meet the need. This was done by scanning verbatims, cutting out the most compelling, pasting them on flip chart paper along with their summary of the theme, need and product ideas. This exercise was the most challenging for the teams, requiring them to pull together their insights in a rigorous way. Each team generated as many as they could in an allotted time, and then reported back on what they had identified by presenting their information to the rest of the group. This is when the strongest themes began to emerge.
On the second day, to stimulate team creativity, we moved to reviewing and evaluating the product samples that had been collected. Each team member selected his or her three favorites, based on everything they had learned, and then explained to the group the particular physical product characteristics that were valuable or meaningful. The team then moved to an evaluation of old product concepts and applied a similar process of evaluation and discussion to these ideas. We found that some of these ideas were probably ahead of their time and may be worth reconsidering. By analyzing product samples and old product ideas in the context of what consumers were telling us, the team was able to make their thinking more tangible.
These two exercises took us to the bottom of the funnel - recommendations for new products. The team members moved smoothly to this stage, and true ideation started to flow naturally. They had many ideas for new products, all backed by the analysis that they had gone through. Altogether, the team produced numerous product ideas in a relatively short period at the end of this day. These were then reviewed for technical feasibility. Those determined to be technically feasible were moved on to the concept development and testing phase for new product ideas. The number of viable product ideas generated using this process (that is, ethnography followed by shared analysis and brainstorming) far exceeded that produced by typical brainstorming sessions. Again, due to the shared experiences of the team, the momentum is still there to move these initiatives forward.
Lessons learned
Analysis is often glossed over in discussions on research, and certainly in discussions on innovation. But analysis is the heart of what we are doing - taking information and squeezing out insights and implications. Project planning should focus particularly closely on this stage. Care should be taken early on to begin identifying existing assumptions and knowledge that may be challenged by new findings. In ethnographic work, the data is used for hypothesis-building - a very creative process which needs time to be effective. The researcher must obtain some distance from the data, after initially becoming familiar with it. Meanwhile, the brain is reworking the information and creating a new framework, new questions, relationships and insights. The creative conclusions are formed - they may come while driving to work, taking a shower or waking up in the morning. But they will emerge if space is created for this process.
It also helps to prepare carefully! Elements that can be arranged ahead of time all relate to the “broad” category. Putting together a diverse team from the organization, making use of analytical tools, leaving space in the schedule for contemplation, scheduling activities that broaden team thinking and keeping the focus on an actionable outcome all contribute to a successful project. The end result is alignment, commitment and energy, in addition to the new product, service or direction that was the project goal. The shared experience that team members have undergone and their ultimate convergence on a conclusion is irreplaceable in producing energy and commitment. Who said analysis was a tedious task?