What do you Mean - A Human Centred Approach?
Since the release of our report: Understanding the Experience of Farm Women the most common question we’ve had is not about specific data points, but what it means to use a human centered approach to customer research. After all, doesn’t the act of collecting information about farmers’ opinions, behaviors and characteristics almost always entail some form of discussion or interaction with a human?
The reality is that the vast majority of market research in agriculture is about products. It’s about what products farmers bought. It’s about which farmers bought your product, more than why they bought it. Conventional research only considers the farmer through the lens of the product. This most often results in what we call a “Phantom Segment.” A segment that exists only on the assumption that the past behaviour of buying the same product somehow makes these farmers similar in other ways. Human centred research flips the script. It seeks to define and explore groups of farmers who are similar in motivations, attitudes and behaviours that help to predict both what they will purchase and how we might influence that purchase behaviour.
Doing human centred research means recognizing differences at three different steps in the process. The first step in human centered research is framing the research problem correctly – with the explicit recognition that your subjects are humans first and farmers second. This means that many of the topics that we as Agri marketers are interested in, such as how our customers learn about new products, how they process data and information, how they feel about brands, are human issues more than they are farmer issues. As such, we can draw on a large body of knowledge and experience in behavioral economics, consumer and B2B marketing. The most important of which is to remember that what may appear to be an irrational behavior is almost never such in the customers mind. It’s why we almost always conduct qualitative research first. The primary purpose of the qualitative is to understand the range of attitudes and behaviours. This step, when conducted in an empathetic manner, also provides insight into language and terminology. In a recent study, our client (prior to the research being conducted) told us that farmers at all scales of operation were alike in their desire for “independence” since previous surveys had shown no difference in the ratings when different scales of operation were considered. Our qualitative research discovered 5 discrete definitions of “independence”, and subsequent survey work showed that when exposed to these different definitions, farmers were not at all unified in the definition as it applied to them!
The next step in human centred research is to recognize and confront the “farmers, farms and farming problem”. As Agri marketers we are very good at understanding farms – acreage, commodities produced, products used and farming – direct seeding, no till, harvest methods and so on. But we don’t seem to spend much time trying to understand farmers! In human centered research we deal with this issue by examining and utilizing an understanding of farmers’ goals, motivations and relationships to the world around them. For example, in previous research we found that farmers’ observance (or not) of rural cultural norms was one predictor when it came to farm management and decision making. Similarly, the Farm Women study had a major focus on the relationship between our respondents and the larger rural culture. Those relationships became the defining variables in understanding a farm woman’s “experience” and sense of identity as a “farmer”.
Adding confusion to this situation, is a tendency in ag market research to believe that a priori segments are predictive of behavior. For example, the fact that someone is a dairy farmer will certainly impact the nature of the products and services that will be purchased, but it is a poor predictor of the strategy that the farmer will employ in purchasing those products and services. Similarly, there is a deep-seated belief that being a “large scale farmer” is somehow predictive of a certain pattern of decision making. As humans, our motivations, personal preferences and methods of living are simply too diverse for that to be true.
That brings us to the third step – correctly framingand conducting the data analysis. As the Noble Prize-winning economist Daniel Kahnemann said: “the human mind is a machine for jumping to conclusions.” Too many market research reports are simply tabulation exercises, enabling and facilitating the reader’s ability to jump to potentially unfounded conclusions by categorizing responses only in demographic and farmographic terms. We believe that human centered research means driving relentlessly to understand the root cause of variation in the dataset. In the Farm Women research, categorizing responses regarding everything from roles on the farm, to attitudes regarding farm safety according to conventional demographics and/or farmographics provided little if any insight into the big picture. Standard measures such as farm size, age of respondent, type of commodity produced, education etc. had no systematic correlation with the level of involvement in farm roles or the nature and severity of the barriers that respondents feel they face. The a priori segments simply did not help and in fact they obscured what was really going on.
Our approach took the route of simply understanding that humans (in this case farm women) are naturally diverse and we began to look for patterns in that diversity. This entailed using principal component analysis to understand themes within the responses, cluster analysis to group respondents into segments that were coherent in their similar attitudes regarding challenges and barriers, and discriminant analysis to model the relationships between these respondent segments in order to more fully understand the totality of their experience as a farm woman. With this analysis, a coherent and compelling picture of 5 different models of experiences of farm women emerged, providing deep insights that are helping us replace the stereotype of a farm woman with a more realistic, factual and multifaceted understanding.
Remember that any research respondent is subject to all those human characteristics that are explained by behavioral economics – recall error, risk aversion, aspirational thinking and so on. Among other issues in the Farm Women study, we paid particular attention to terminology and the use of gendered terms as well as question order to avoid bias. We used every opportunity to let the respondent define the terms of the response – using open ended questions wherever possible and letting the qualitative research define the range of responses to closed questions.
Humans make decisions. Farms don’t. In a world of exploding information and technology where farm decision makers are faced with ever more subtle differences between options, understanding customers as humans first will be the single most critical factor for success.
