Introduction

The Technology Acceptance Model, TAM (Davis, Bagozzi & Warshaw, 1989), has become the bona fide model of IT acceptance, forecasting the extent to which a new IT is used. The basic premise in TAM is that two behavioral outcome beliefs about a new IT, namely its perceived usefulness (PU) and its perceived ease of use (PEOU), are significant predictors of its intended future use. Over three- dozen studies show that this premise holds across IT types and user experience levels, including user intentions to use B2C websites (Gefen, Karahanna & Straub, 2003; Gefen & Straub, 2000; Gefen & Straub, 2003). While it is undeniable that perceived beliefs about behavioral outcomes , such as the two beliefs identified by TAM, are primary predictors of behavioral intentions (Ajzen & Fishbein, 1980; Fishbein & Ajzen, 1975), including the intention to adopt a new IT (Davis et al., 1989; Karahanna & Straub, 1999), it is also a fact of life that force of habit still dictates many behavioral intentions once people have gained experience. ^[1] In other words, repeated previous behavior often dictates current behavior independently of any rational assessments (Chaudhuri, 1999; Triandis, 1971), such as those described by the Theory of Reasoned Action (Ajzen & Fishbein, 1980; Fishbein & Ajzen, 1975) on which TAM is based.

While research dealing with TAM did examine additional antecedents of ITuse, based among other things on the diffusion of innovation theory (e.g., Karahanna et al., 1999), it has done so by focusing on other behavioral outcome beliefs, such as positive image, compatibility, trailability, visibility, and result demonstrability (Moore & Benbasat, 1991, 1996). TAM and related research has not, however, examined habit as a predictor of future IT usage, possibly because TAM was originally geared toward the acceptance of a new IT, rather than examining its continued use. TAM research has not examined this aspect of IT usage, although the predictions of TAM have been shown to apply also to experienced users (Gefen & Straub, 2000; Karahanna et al., 1999), and although habit is a major predictor of consumer behavior (Chaudhuri, 1999; Tucker, 1964) and of behavioral intentions in general (Ouellette & Wood, 1998; Trafimow, 2000; Trafimow & Borrie, 1999; Triandis, 1971). It is important to note in this regard that habit and TAM complement each other: TAM dealing with a rational assessment, habit dealing with the lack of one.

How then does previous usage preference of an IT—habit—apply to IT adoption? And what is its relative weight compared with the verified intended use-antecedents identified by TAM? Answering this is the objective of this study.

The data, collected from experienced online shoppers with business-to-consumer (B2C) e-commerce websites, show that, as hypothesized, habit was a significant predictor of the future use of the IT, in this case continued purchase and other services through the specific website, in addition to its PU and PEOU. Moreover, the data also show that habit alone is a major predictor of the continued use of the website, explaining 40% of the variance, and that habit is a significant predictor of both PU and PEOU. The exceptionally high values of explained variance suggest that, in the particular case of B2C users, habit and behavioral beliefs are major predictors of continued activity at the given website where the user has previously purchased.