In a sense, there is no such thing as an irrational decision. The choices we make are based on deep-seated ideas about how the world works – which researchers call “mental models”. We could be working with the best possible information, but if our mental models are out of step with reality, our decisions will be similarly skewed.

In life as in business, if we want to understand – let alone change – behaviour, we first must excavate the mental models that drive it. But because mental models are often buried far below our conscious awareness, fully apprehending them can be challenging. Standard workplace tools such as customer surveys and evaluation forms are poorly designed for the task.

Practitioners looking for new approaches to understanding stakeholders can adapt a well-established research approach known as theories-in-use (TIU). TIU employs structured interviews to unearth mental models and express them in simple “if-then” propositions. The logical connections and associations thus discovered have sparked completely new theories in academic literature. The body of TIU research is relatively small, but highly impactful. As my new article in the Journal of Marketing[1] points out, three of the ten most cited JM articles of all time use a TIU approach. The article provides practical guidelines for adopting TIU in professional contexts.

Examples of TIU

First, let’s identify key areas of application. TIU examines mental models of various stakeholders, such as customers, managers or policymakers, to gain better insights about business phenomena that often are “new to the world” or simply not well understood. Business disciplines have no shortage of such situations. Think of notions such as “digital maturity” or “customer centricity” that usually point at a construct (or a set of interrelated constructs) in need of deeper exploration.

As an example, among the three above-mentioned top Journal of Marketing articles are landmark TIU-based investigations into the constructs of service quality and market orientation. The service quality article triggered a decade-long research stream, which in turn produced novel conceptual frameworks and practical insights on service quality within many industries and companies.

Other productive targets for TIU may include high-priority topics such as corporate social responsibility, sustainable growth or new organisational functions (think of a Chief Growth Officer or Customer Success Management, etc.). Comparing various mental models within stakeholder sub-groups can bring new theories to light as well. For example, what drives choices in large firms versus start-ups when hiring a Vice President of Customer Success? What motivates senior executives to take such a decision? What for? Under what circumstances?

The stages of TIU

TIU investigations follow a special process. At first, you may find this interview flow unfamiliar, even difficult. Like any other new skill, it requires practice to master. Keep in mind, you don’t have to be an experienced interviewer to get started with TIU. In some ways, novices may be well suited to the approach, because they have fewer assumptions and biases to crimp their curiosity.

Construct development and definition.

Once you have chosen the topic you want to learn more about, you may find it requires further elucidation. One interview won’t be sufficient to achieve a firm grasp of an especially elusive construct, e.g. market orientation. You will need to aggregate the answers of several interviewees to questions such as “Can you tell me a bit about your approach to [construct]?” and “Do you have a common definition of what [construct] means in your organisation?” Each interviewee will provide the answer as they see it. Through active listening, you can locate areas where multiple answers overlap and begin to craft a higher-level definition. From there, you can continue to test and refine over successive rounds of interviews by using prior responses as a jumping-off point, e.g. “Based on interviews to date, [construct] may be defined as […]. What are your thoughts?”

Proposition development.

With a solidly defined construct, you can start drawing connections to real-world outcomes, i.e. the above-mentioned if-then propositions. These can include propositions both “upstream” and “downstream” of the core construct. For example, you can ask questions designed to determine antecedents (“In your view, what are the main drivers of market orientation?”) or barriers (“What blocks market orientation in your organisation?”). You can also begin asking about novel consequences (direct and indirect) of the construct, e.g. “What are the key benefits of market orientation?” or “Were there any surprising outcomes when your organisation implemented [change initiative]?” Finally, you should try to identify moderators, or factors that affect the degree to which X leads to Y. Again, it is critical to conduct several rounds of interviews, iterating/consolidating responses and fine-tuning your propositions by showing them to late-stage interviewees.

Argument development.

The goal of this stage is to add plausible causation and nuance to the novel theorising you have drawn out of the interviews. Once you are confident in your propositions, you can start investigating the reasons behind them, e.g. “Why do you think X will cause Y to occur?” On the principle that there are exceptions to every rule, you should also hunt for limitations to your if-then logic, building upon previous results to create “if-then-except-when” theoretical propositions.

Iteration and abstraction

These iterative and recombinative techniques can be used across, as well as within, studies. Overlapping theories developed in sequence can be combined to form a more broadly relevant theory. Let’s say you discover, over separate interview rounds, that increasing R&D spending amid recession improves firm performance during recovery and that increasing operations investments at the same juncture achieves similar results. Since both interventions arguably come under the heading of “capacity building”, you could extrapolate the proposition that “Greater investment in capacity building during a recession leads to higher post-recession firm performance.”

Limitations

We enthusiastically recommend TIU to managers who want to learn more about how customers, colleagues and even policymakers see the world. Across business disciplines, TIU can help generate new insights on diverse questions, such as “What drives advertisers’ decisions to allocate spending across social media channels?”, “What factors drive venture capitalists’ decisions to invest in fintech start-ups?”, or “Which mental models guide hiring procedures of wealth managers in banking?”

However, every approach has its limitations and pitfalls. For example, TIU should be used to construct theories, not to test them. That said, properly conducted TIU interviews yield information that can help prepare for theory testing, thereby reducing associated costs significantly.

It takes a skilled interviewer not only to unearth mental models, but to convert other people’s (very possibly flawed) decision-making processes into potentially viable theories. Thus, TIU is much more than a mere transcription of decision-makers' beliefs, ideas, or thoughts. The approach requires the investigator to take a step back and thoroughly weigh insights gained with an open mind.

A good rule of thumb: The newer the phenomenon or construct under investigation is, the more caution is warranted when exploring mental models during the interview process. TIU’s effectiveness therefore depends upon a certain amount of skill and experience on the part of both interviewer and interviewee. Otherwise, the latter’s uninformed guesses may hamper the theorising of the former. Or, to put it another way, ‘garbage in garbage out’.

Wolfgang Ulaga is a Senior Affiliate Professor of Marketing at INSEAD.

Don’t miss our latest content. Download the free INSEAD Knowledge app today.

Follow INSEAD Knowledge on Twitter and Facebook.

[1] Co-authors for this paper are Valarie A. Zeithaml of University of North Carolina-Chapel Hill, Bernard J. Jaworski of Claremont Graduate University, Ajay K. Kohli of Georgia Institute of Technology, Kapil R. Tuli of Singapore Management University, and Gerald Zaltman, professor emeritus at Harvard Business School.