The systematic instability of consumer preferences

by Stijn van Osselaer

Contrary to conventional thinking, research has shown that consumer choice is a motivational and dynamic process based on goals and ‘circumstances’.

Traditionally, it is assumed that consumers purchase products based on their attributes, notably brand name and what the product contains, and that these preferences are “stable”. That may be true when predicting the choice between similar brands (such as Coca-Cola and Pepsi) within a single product category. However, this traditional model breaks down when it comes to making more subtle predictions such as choices between products from different categories (for example, coffee versus cola versus mineral water), or even subcategories (such as Belgian Wit beer versus Tripel).

The reality is that consumers are not interested in the characteristics of a product, but in what it does for them. In the case of a soft drink like cola, consumers are more concerned with its functional aspects, such as whether it gives them a buzz or helps them stay awake, rather than the fact that it contains caffeine.

Meeting goals

In other words, consumers mainly pick a product in anticipation of the goals it will fulfil. However, these goals and their importance are highly situation-specific. The importance of the goal to stay awake, for instance, is greater when driving late at night, than, say, in the morning. This increases a preference for coffee over mineral water regardless of a consumer’s ‘general’ preference for coffee or mineral water. This also makes preferences inherently “unstable” and situation-specific.

Nevertheless, this is no reason to stop trying to predict consumers’ choices. Crucially, the variability in consumer choice is not random, but systematic. We just need to determine which goals will be most important in which buying situation.

To determine this, we need to know what drives this importance. One obvious driver is the consumer’s level of goal-related deprivation or satiation. Once a goal has been satisfied, its importance level falls. In contrast, if a goal has not been fulfilled in a while, consumers feel deprived and the importance level rises.

For example, if we have taken a satisfying drink, the thirst-quenching goal is fulfilled and the importance of whether a drink can slake our thirst or not is lessened. However, if we become thirsty later, or if it is very hot outside, our hydration levels will drop and the desire to satisfy our thirst becomes more important, giving a big edge to products that we think will satisfy this goal.

Thinking systematically about satiation and deprivation does not completely solve the problem in many situations, however. Even when satiation and deprivation are taken into account, there are often many goals that could potentially determine choice. For example, drinks differ in their ability to: warm you up or cool you down; give you a pleasant taste experience; promote healthy living; make you look cool to others, or give you a buzz.

Which of these many goals is most the important also depends on which one is “activated” or highlighted in a consumer’s memory. Of course, the most obvious way of activation is by directly bringing the goal to the consumer’s attention – though an advertisement, for instance. This type of “goal priming” can even take place subconsciously. For example, consumers surreptitiously primed with words related to the goal to save money (for example, “frugal”) are, in a later unrelated task, more likely to choose cheaper products.

Other ways in which goals can be activated are through being a specific buying context or by the presence of other products in the buying situation. In these cases, goals that have historically been salient in that situation will be activated automatically and consequently have a greater impact on choices made.

“The reality is that consumers are not interested in the characteristics of a product, but in what it does for them. ”

Know your consumers

This model of goal-based selection offers a structured way of thinking  about consumers and their choices. Although this requires study and practice, several guidelines should apply broadly:

  1. In predicting consumers’ choices, do not simply measure supposedly general preferences out of context. It is better to either measure preferences in multiple situations that mimic different buying situations as closely as possible, or determine which goals are most salient in the relevant buying situations and then measure the extent to which consumers feel your product and competing products will satisfy those goals.
  2. Highlight those goals, relevant to the buying situation, where your product is more appealing than the competition.
  3. Explicitly tie your product to important consumer goals. Do not assume that if consumers know your brand and its attributes or make-up, they will automatically link them to the related goals.
  4. Remember that you compete for goals: it is often the case that a goal can be satisfied by several quite different products (for example, cola, coffee and fresh air will wake you up or keep you awake), and different goals are important in different situations. Furthermore, the alternatives on display in the buying situation will highlight their own goals to the consumer. Therefore, think about the competition for goals and the specific buying situation.
  5. Finally, understand your consumers – see the buying situation from their perspective: what goals matter to them at that moment; how well do they think your product satisfies those goals, and which alternatives could achieve the same goals?


Stijn van Osselaer is Professor of Marketing, Department of Marketing Management, Rotterdam School of Management, Erasmus University. His main research interests are in branding and the influences of learning, memory, and cognition in consumers’ decisions.

This article draws its inspiration from A Goal-Based Model of Product Evaluation and Choice, which was written by Stijn van Osselaer and Chris Janiszewski and published in Journal of Consumer Research, Vol. 39, No. 2 (August 2012), pp. 260-292.

This article was published in RSM Insight 12. More information about and back copies of RSM Insight can be found here.

Download Original Article (.pdf)

Share this article: