Hospitality Analytics: Use of Discrete Choice Analysis for Decision Support

Pedro Longart

Higher Colleges of Technology


Eating out of home has become an integral part of people’s lives. This is because people have changed their attitudes about food and also due to an increase in disposable income (Capstick, 2011). Eating out normally takes place in restaurants, which have become an important part of our everyday lifestyles, and offers ‘a place to relax and enjoy the company of family, friends, colleagues, and business associates’ (Walker, 2014; p. 160).

There are many types of restaurants. A fine dining restaurant is one where a good selection of menu items is offered with a high level of service. Mehta and Maniam (2002) claimed that fine dining (or gourmet) restaurants are the most formal, fine dining experiences. Casual dining is more relaxed and could be part of one or a combination of ethnic, family and mid-scale casual. Although most family restaurants are seen as casual dining, some operations are targeting a more upscale customer. Ethnic restaurants base their ethnicity on the type of food served: Mexican, Indian, Chinese, Spanish, Italian, etc. Casual dining may include restaurants that are mid-scale in many countries (or considered as upscale in others) such as chains like TGI Friday’s, Hard Rock Cafe, and Frankie and Benny’s.

Due to this variety, choosing a particular context is challenging. A snack on the way to work could be considered as eating out. However, that can be considered as a low involvement, repetitive, routine customer decision. Eating out for leisure purposes entails complex phenomena. Consumers give many reasons for eating out, and those reasons may be compatible with how they live their lives. For example, for some people, as Charles and Kerr (1988) found, a proper meal is an occasion in which table manners are adhered to.

From the restaurateurs’ perspective, they have a daunting task in working on a long list of aspects of their restaurants; therefore, a prioritisation of attributes by ascertaining attributes’ importance is necessary. Some studies have attempted studying restaurant attribute importance, and hence, the first objective of this literature review is to analyse the various attempts that have been made to find out what attributes are more important for certain consumer segments and for certain occasions for eating out. A second objective focuses on the critical analysis of methodologies used for these investigations and to suggest a more suitable method for the purpose of understanding consumers’ preferences in the context of choosing a restaurant.

Literature Review

Foundations of Consumer Research: Cognitive Approach

One of the cornerstones of the cognitive approaches of consumer research and most theories of choice is the presumption of a rational consumer. H a rgreaves- Hcap et al. (1992) started with the main conundrum for studying choice: What makes a choice rational? They explain that rationality is a matter of means, not ends. These rational actions can be represented as those of individuals seeking to maximise utility - or expected utility (Hargreaves-Heap et al., 1992). If a consumer is hungry, and he/she has two choices, he will most likely go for the choice that will better satisfy hunger. Utility can be defined as the measure of how the hunger is satisfied. If I have objective measures such as calories, and the information about the caloric content is available, then there is objective utility involved. However, in many cases, utility is subjective; then it depends - amongst other things - on preferences. This is the topic studied by the theory of subjective expected utility (SEU) that according to Simon (1986) is central to the body of prescriptive knowledge about decision making.

Behavioural Decision Theory

Theories of Choice

When investigating the behaviour of a large number of individuals making the particular decision of eating out and choosing a particular restaurant, the aggregate behaviour of restaurant patrons is the result of each individual decision. Ben-Akiva and Lerman (1985) explained that there is no universally accepted theory of choice that satisfies the requirements laid out above. A choice, they continue, is a sequential decision-making process that includes the following steps:

  • 1. Definition of the choice problem
  • 2. Generation of alternatives
  • 3- Evaluation of attributes of the alternatives
  • 4. Choice
  • 5. Implementation.

In this context, the choice problem is that of a consumer deciding where to eat out, not for convenience, that is, lunch between working hours. His place of residence or stay will define the alternatives (restaurants available). The next step is about evaluation of the alternatives, and a discussion of alternatives will be included later in this literature review. The consumer needs to collect information about relevant attributes of that restaurant. In order to do that, the consumer applies a decision rule to arrive at a choice. Then, implementing the choice is obviously the meal itself. Thus as Ben-Akiva and Lerman pointed out, any specific theory of choice is a collection of procedures that encompass the following elements: the decision maker, alternatives, attribution of alternatives, and decision rules.

The Alternatives

Wright (1975) found that decision makers try to simplify their decision making and when studying models of decision making, a clear example is that of how alternatives are elicited. Shocker et al. (1991) characterised decision making as based upon hierarchical alternatives. Ben-Akiva called them a choice set. The universal set refers to the totality of all possible alternatives, which in this case may be the restaurant in a certain location, that is, London. This set is just a starting point as it is impossible to consider thousands of alternatives such as in the London restaurant scene. Then, there is a set that springs to the customer’s mind; that means that they may remember them unaided. This has been called the evoked set (Howard, 1963), which comes from the awareness set, which is composed of evoked sets, inept sets and inert sets (Narayana and Martin, 1975). The model of Shocker et al. only considers the evoked set, which is also called the consideration set. They could also possibly be drawn from a list or restaurant guide (see External Alternatives). The definition from Shocker et al. appears to be more complete in separating awareness from consideration, indeed an important difference for marketers. In the context of restaurant decision making, it can be argued that the distinction is important and for that reason, Shocker et al.’s model will be preferred.

The consideration set is a reduction from the awareness set to a smaller set of alternatives (Gensch and Soofi, 1995)- Horowitz and Louviere (1995) warned that ‘using a consideration stage may lead to a misspecified model that would provide erroneous forecasts’ (p. 40). However, in the case of restaurants, it is sensible to assume that consumers engage in an extensive information search to arrive at a decision. This is linked with the information processing theory, which is an approach in which ‘the consideration set is formed and used by the consumer for subsequent purchase operations’ (Roberts and Nedungadi, 1995). The choice set has a very strong influence upon the individual intention to choose a particular restaurant.

This was evidenced by the study of Davis and Warshaw (1991)- Davis and Warshaw suggested that consumers employ screening procedures using decision rules, more particularly, non-compensatory rules to reduce the consideration set to a manageable size.

Decision Rules

The concept of utility is inextricably linked to decision rules. That means that consumers consciously or unconsciously assign values to the alternatives. That could mean the utility maximisation of satisfaction and minimisation of cost (or maximisation of value for money). Typically, consumers can only consider a small part of all the information available to them about a specific service. Heuristics are ‘rules of thumb’ (Shah and Oppenheimer, 2008) that individuals unconsciously apply to reduce the effort involved in decision making.

Decision rules have been traditionally seen as completely unrelated to impul- sivity. Yet, as pointed out by Hsee and Tsai (2008), they are closely intertwined as most decision rules are antidotes to impulsive behaviour and entail some sort of self-control mechanism.

As for the choice criteria, Devetag (1999) distinguished between two types of heuristics: compensatory and non-compensatory. A heuristic is said to be compensatory if good values on some attributes can compensate for poor values on other attributes. On the contrary, a heuristic rule can be defined as non-compensatory if that compensation does not have effect. In compensatory behaviour, according to Statt (1997), the consumer uses more than one criterion to evaluate a product or service and there are two versions of this type of rules. The first and simpler version has the consumer as unconsciously adding the pluses and minuses of each alternative and the one with the most pluses wins. In the more complex version, the relevant attributes are weighted according to their importance for the consumer. Solomon (2007) described non-compensatory rules as ‘choice shortcuts’ in which people eliminate all options that do not meet the consumer’s basic standards. Understanding these shortcuts is essential for a restaurateur as his/her restaurant may not even enter in a consideration set because of the nature of non-compensatory behaviour with regard to certain attributes and more particular features of that attribute, which are called levels of the attribute. That is, the restaurant may not reach the minimum threshold of that attribute (minimum level) to be considered by a decision maker. The concept of attribute level is discussed below.

< Prev   CONTENTS   Source   Next >