Consumers engage in both internal and external search for product information (Beales, Mazis, Salop, & Staelin, 1981; Bettman, 1979; DeSarbo & Choi, 1999; Moore & Lehman, 1980; Newman, 1977). Internal information search involves consumer retrieval of memory that stores product knowledge. External information search refers to activities other than memory, such as consulting with salespeople, friends, reading other sources, looking at ads, direct observation and so on. Consumers who make purchases often employ both types of search in a sequential and iterative fashion. First, consumers consult information stored in memory. If they do not have sufficient information, then they search for additional information from external sources. This information is then accumulated in memory for later use. The amount of information search exerted for a specific purchase is therefore a function of both internal and external resources available to the consumer. When both internal and external search are carried out, results that are conflicting will need to be resolved properly.
External information search encompasses both goal-directed, pre-purchase activities and ongoing search activities (Lee & Hogarth, 2000b; Peterson & Merino, 2003). Most research on information search has focused on pre-purchase information search which involves the consumer’s cognitive effort to reduce uncertainty for specific purchases (Beatty & Smith, 1987; Punj & Staelin, 1983; Srinivasan, 1990). Meanwhile, other researchers assert that ongoing search is not specifically related to imminent purchases and is more related to nonfunctional motives including entertainment (Bellenger & Korgoankar, 1980; Holbrook & Hirschman, 1982; Tauber, 1972) and product interests (Bloch & Richins, 1983; Bloch, Sherrell, & Ridgway, 1986). For example, Bloch et al. (1986) found that the perceived enjoyment of shopping and enduring involvement are related to ongoing search and that heavy ongoing searchers tend to be heavy spenders within the product class.
Consumers can gather information from various external sources. Many researchers have reported different types of information sources (Andeasen, 1968; Beatty & Smith, 1987; Freiden & Goldsmith, 1989; Lee & Hogarth, 2000b; Newman & Staein, 1973; Schmidt & Spreng, 1996). Often used information sources can be categorized into several types: seller provided information (e.g., stores, catalogs, salespeople), media (e.g., television, newspaper, magazine, radio, the Internet), interpersonal sources (e.g., friends, family, relatives), third party sources (e.g., Consumer Reports, J. D. Powers), and direct inspection (e.g., direct observation, product trial).
There has been a stream of research on the categorization of different modes of information search or browsing based on both empirical evidence and theoretical propositions (Claxton, Fry, & Portis, 1974; Furse, Punj, & Stewart, 1984; Kiel & Layton, 1981; Newman & Staelin, 1972, 1973; Westbrook & Fornell, 1979; Wilkie & Dickson, 1985). These researchers have sought to identify the pattern of information search by focusing on distinct groups of consumers. There appear to be several factors (situational, environmental, and individual) based on which distinct patterns of information search are categorized. However, few researchers (Maute & Forrester, 1991; Moore & Lehmann, 1980; Punj & Staelin, 1983; Schmidt & Spreng, 1996; Srinivasan & Ratchford, 1991) have attempted to develop a causal model to reveal the relationships among these factors for information search behavior.
In theory, there are three approaches to modeling information search behavior (Schmidt & Spreng, 1996; Srinivasan, 1990). Some researchers applied the economics of information theory to behavioral aspects of search (Avery, 1996; Goldman & Johansson, 1978; Miller, 1993; Urbany, 1986). According to this approach, consumers acquire and process information until the marginal cost of information search exceeds the expected marginal benefit of information search (Ratchford, 1982; Stigler, 1961). In other words, perceiving increased benefits will lead to more information search and perceiving increased costs will lead to less search effort. Thus, investigating variables affecting the perception of costs and benefits of search would be necessary for understanding external search behavior.
A psychological or motivational approach, on the other hand, assumes motivation to be the driving force for external information search (Howard & Sheth, 1969). Motivation refers to the desire to exert effort for a certain task (Bettman, 1979) or the arousal for achieving goals (Park & Mittal, 1985). Different search goals (i.e., optimizing vs. satisficing) and different levels of product involvement (i.e., relevance or interests) may affect search direction and intensity. Therefore, any variables that increase the motivation to search will yield more search effort and vice versa.
Finally, an information processing approach emphasizes consumer memory and cognitive ability (Bettman, 1979). This approach has led to extensive research on search behavior affected by prior knowledge, familiarity, experience (Brucks, 1985; Johnson & Russo, 1984; Punj & Staelin, 1983), expertise (Alba & Hutchinson, 1987), and prior belief (Duncan & Olshavsky, 1982; Urbany, 1986).
Most of the empirical studies have examined direct effects of various antecedents on information search in bivariate situations (Guo, 2001; Lee & Hogarth, 2000b). While direct relationship between external search and its determinants is in and of itself important to understanding consumer information search behavior, the complex nature of information search behavior will require investigating the relationships between various factors and information search in a multivariate setting. External information search is influenced by a number of determinants and, in a multivariate setting, the magnitude and direction of the relationships between search and antecedents will be different. Several researchers have argued that ability, motivation, and cost and benefit mediate the effects of various antecedent factors on information search activities and suggested models for testing (Maute & Forrester, 1991; Moore & Lehmann, 1980; Punj & Staelin, 1983; Schmidt & Spreng, 1996; Srinivasan & Ratchford, 1991).
In summary, empirical studies on consumer information search behavior to date have provided some general implications and guidelines for further research in this area. First, past research has found four different relationships between antecedent variables and information search: positive, negative, inverted U, and null relationship (see Guo, 2001, for a detailed discussion). Second, different studies have adopted different measures of information search from single measures to multiple measures to aggregate (weighted/ unweighted) indices of information search effort. Third, most of the relationships have been studied in bivariate settings with few exceptions.
These general models from traditional information search studies provide a good starting point for investigating information search behavior in an online environment. However, several factors particularly relevant to the Internet may need to be developed. Factors such as characteristics of the Internet media (e.g., accessibility, interactivity, flow, customization) and Internet skill and experience (Liang & Huang, 1998; Ratchford, Talukdar, & Lee, 2001) will need to be incorporated as we move into an increasingly online shopping environment.