Our Method

Our method of market research involves gathering the most accurate information possible about consumer preferences. Typically, market research surveys about products only try to find out how much people like or dislike various product features. For example, in the case of mobile phones you could gather data on the product features shown in column 1 of the table below. Let’s imagine that we survey four people, and we ask them to click buttons on a computer to show how much they like or dislike the different product features. They click a plus sign button as much as they want to show how much they like a feature. Or they can click minus signs to show how much they dislike a feature. They can click a zero button if they neither like nor dislike it.

Product Feature:	Person 1	Person 2	Person 3	Person 4
Front facing camera	+++	++	++++	+++
Over 6-inch diagonal screen	++++	+	++	+++
Under 6-inch diagonal screen	++	+	+	+
Battery Life under 24 hours	—	0	++	+
Battery Life 24-48 hours	++	+	++	+++
Battery Life over 48 hours	+++++	++++	+++	+++

From the table above you can the like/dislike scores for four hypothetical people. The normal way to work out which product each person would buy is to add up their total scores for all the features of a given product. For example, if the iPhone has a front facing camera the “iPhone score” is +3 for person 1. If its screen size is over 6 inches we add +4. If its battery lasts over 48 hours we add +5 for that feature. Our hypothetical iPhone gets a grand total of +12. Let’s suppose we work out total scores for five other phones in the same way. When we analyze the data the product that get the highest total score is hypothetically “bought” by person 1. We then move on to person 2 on our computer and work out which product that person buys. Let’s imagine we have a thousand people in our database. If 30% of those people give the highest combined product-feature score to the iPhone then we say it has a “virtual market share” of 30%. If you have set up your survey well then this percentage should be close to the iPhone’s actual market share in the real world. Now that you have a database of desirability scores you can play “what if” games with a potential new product design. Let’s suppose that no phone has a big screen and a battery life over 48 hours. You can analyze the 1000 people again and work out what percentage of them would give the highest combined score to this new potential product. This is the typical way that market research work using the most commonly used methodology called “Conjoint”.

A different methodology was proposed in the brilliantly innovative book by the late Eric Marder entitled “The Laws of Choice: Predicting Consumer Behavior” [1]. Marder argued that it is not a product’s actual combination of features that drives consumer behavior. Rather it is what any consumer believes to be true whether that is right or wrong. For example, if person 1 in the example above believes the iPhone has 48+ hours of battery life that is what will decide person 1’s buying decision. For an Eric Marder type survey there is a second phase of data collection that gathers data on what consumers believe to be true about a given product’s features. Person 1 believes the iPhone has 48+ of battery life then we add the +5 score to person 1’s total whether or not it is true.

Philip Truscott carried out research to determine which scoring method is more accurate and published an article about it in 2019 [2]. This showed that the belief-based scoring method is significantly more accurate than traditional conjoint scoring method. What is more important corporate clients gain much more value from the belief-based surveys. If you only know how much value consumers place on different features the only way to use the data is to design a new product with a new combination of features. With belief-based surveys you can find out what good features your product already has but the public don’t know about. Let’s suppose your phone already has 48+ hours of battery life but the data shows that only 20% of consumers know this. You can now calculate how much your market share will go up if your advertising increases that percentage from 20% to 50%. For brand managers you now have a tool to fine tune your advertising process.

Marder, E. (1997). The Laws of Choice: Predicting Customer Behavior. Free Press.
Truscott, P., & Chiam, M. (2019). Belief – Based Marketing vs. Conjoint: An Illustration Using the Indian Mobile Phone Market. Indian Journal of Marketing, 49(4), 7-19–19. https://doi.org/10.17010/ijom/2019/v49/i4/142973