Measuring customer experience

Most customer experience research seems to fail to capture the totality of customer experience with its different components. Traditional methods mostly measure cognitive and conative experience components by investigating self-reported pleasure from past experience as a metric of customers’ current feelings rather than the experience in totality. Several experience scales were inspired by Pine and Gilmore’s (1999) four realms of experience: educational, escapist, esthetic, and entertainment, which are differentiated by the level of activity and customer involvement. As an example from the tourism and hospitality research, Ali, Hussain, and Ragavan (2014) classified experience in these four dimensions and developed the 16-item customer experience scale in the context of resort hotels. Another experience scale was initiated in a retail setting and incorporated other experience dimensions: joy, mood, leisure, and distinctive (Bagdare & Jain, 2013). Klaus and Maklan (2013) introduced an alternative measure of customer experience, the customer experience quality (EXQ) scale, which consisted of 19 items and was validated across different services and service channels.

With a slightly differentiated term, memorable experience, Kim et al. (2012) suggested using the Memorable Tourism Experience scale (MES) with seven constructs: hedonism, refreshment, meaningfulness, local culture, knowledge, novelty, and involvement. However, their second-order model with affective, cognitive, and behavioral components as the latent variables demonstrated that affective and cognitive components shared common variances and that the MTE scale fits better using the first-order structure. Another study by Otto and Ritchie (1996) uses hedonics, peace of mind, involvement, and recognition to develop a tourism service experience scale, which was tested in different tourism settings, including hotels, airlines, and attractions. The most cited experience scale study by Brakus et al. (2009) describes sensory, affective, intellectual, and behavioral components of brand experience and suggests measuring these components by using a 12-item brand experience scale. This scale was further replicated by Barnes, Mattsson, and Sorensen (2014) in a destination setting. However, applying the sensorial and affective scale items evoked by the design, identity, communication, packaging, and environment of the famous conventional brands (Apple, Coca-Cola, BMW, etc.) into the destination context causes several doubts. For instance, such survey questions as “the destination is interesting in a sensory way”, or “destination gives me bodily experience” developed for the conventional product brand studies might be incomprehensible for tourists in the destination context.

Additionally, such self-reported scales pose several limitations. First, study results can be influenced by respondents’ biases such as social desirability, resulting from people’s desire to under-report socially undesirable activities in favor of more attractive ones (Ganster, Hennessey, & Luthans, 1983). Second, people’s responses can be biased by the availability heuristic as they tend to overestimate events that come to mind (Kahneman & Tversky, 1979). Third, people’s memories can also be influenced and reorganized by their knowledge and beliefs (Ross, 1989). Fourth, people’s answers can be biased by their mood at the time of answering questions (Cantor & Kihlstrom, 1987). Therefore, a comprehensive understanding of tourist experience requires research methods that allow measuring components of experience moment-by-moment as it unfolds.

Recent literature reviews and critical reviews shed some light on the difficulty of measuring experience; however, they have not offered innovative measurement methods or techniques. Neither Adhikari and Bhattacharya (2016) nor Packer and Ballantyne (2016) offered any new methodological insights to capture experience. After discussing the challenges of measuring consumer experience, Palmer (2010) acknowledged the difficulty of measuring nonlinearity of customer experience using any single method. He discussed in detail the changing, dynamic, and transitory nature of experience and thus the difficulties of measuring this elusive concept. Highlighting the challenges springing from situational factors, non-linearity of experience, and the difficulty of identifying the optimal experience level, he deemed the standard survey design inadequate, especially in measuring the changing nature of affective dimensions of experience.

Hwang and Seo (2016) also acknowledged the multidimensional sequential nature of experience and listed a number of measurement challenges including the variable definitions and components, changing nature of affective attitude, difficulty of measuring the experience itself, subjective and irrational emotional dimensions, the questionable validity in self-reported emotions, and situational and context-driven nature of experiences. Thus, they called for innovative approaches to the measurement of customer experiences and suggested using “experience sampling method, grid technique, netnography, and structured content analysis”, as well as more of a cultural lens in studying customer experience (Hwang & Seo, 2016, p. 2238).

The experience sampling method makes it possible to study people’s subjective experience while interacting in natural settings by asking respondents to report on their feelings, behaviors, and thoughts on random occasions over time (Bolger & Laurenceau, 2013). Several recent tourism studies investigated visitors’ experiences during different activities by using the experience sampling method. For example, Jones, Hollenhorst, and Perna (2003) compared models of optimal experience of whitewater kayakers by applying the Experience Sampling Method. Birenboim, Reinau, Shoval, and Harder (2015) used experience sampling data collection technique to explore subjective experiences of zoo visitors and concluded that the quality of experiences varies in time and space. Borrie and Roggenbuck (2001) collected experience sampling data from Okefenokee National Wildlife Refuge visitors to analyze the nature of the wilderness experience and its temporal dimensions. Experience sampling could be useful in tourist experience studies since it helps to reveal impacts of a particular event by comparing responses before and after the experience; compare consumers’ experience of events; and examine duration, magnitudes, and sequences of visitor emotional states (Csikszentmihalyi & Larson, 2014). Nonetheless, the experience sampling method is also dependent on participants’ self-reports and has several limitations related to self-report measures, which can be eliminated by applying more objective measures of tourist experience.

The majority of previous studies described the prevalence of affective components of visitor experience (Aho, 2001Gentile et al., 2007Meyer & Schwager, 2007Palmer, 2010). However, as was also acknowledged by Mauss and Robinson (2009), measuring emotions is one of the most difficult tasks, and both the existing methods and those newly introduced techniques fail to capture emotional states before, during, and after the visit. The emotional response starts from the appraisal of the situation and results in subjective experience, physiological reactions, and behavior, while there is a debate about the precedence of emotional reactions versus cognitive appraisals in the social and cognitive psychology literature (e.g. Parkinson & Manstead, 1992). Consequently, different measurement techniques might be applied to investigate emotional responses as part of the customer experience.

Although emotions are traditionally measured retrospectively in tourism and hospitality research (e.g., Soodan & Pandey, 2016), there is an opportunity to measure them by applying the moment-based psychophysiological techniques, which reflect electrochemical changes in neurons, muscles, and gland cells (Stern, Ray, & Quigley, 2001). Different emotions involve different patterns of activation of both branches of the human autonomic nervous system (sympathetic and parasympathetic), which can be useful in analyzing people’s emotional states (Kreibig, 2010). The sympathetic system is associated with activation and mobilizing, while the parasympathetic system is related to relaxation and dampening. Arousal has been reported as the best way to reflect responses of the autonomic nervous system (Cacioppo et al., 2000). Arousal can be measured by using electrodermal activity, electrocardiography, pupillometry, and some other techniques (Stern et al., 2001). At the same time, different branches of autonomic nervous system activity can work independently or against each other (Bradley & Lang, 2000), hence the second dimension of valence could be applied to reflect these differences (Russell & Mehrabian, 1977). Valence can be measured by applying facial electromyography, electroencephalography and other methods. Several studies suggest that multiple measures of the autonomic nervous system may provide a better degree of autonomic specificity (Stern et al., 2001). Psychophysiological responses are not controlled by people and can eliminate the limitations of self-report and behavioral measures (Larsen & Fredrickson, 1999). Moreover, psychophysiological recordings provide moment-by-moment values of respondents’ emotional reactions (Wilhelm & Grossman, 2010) and can be used before, during, and after the experience.

Some researchers also suggest measures to capture the sensory components of consumptions, which include hearing, seeing, tasting, smelling, and feeling (Agapito, Valle, & Mendes, 2014Holbrook & Hirschman, 1982). Bech-Larsen and Nielsen (1999) applied elicitation interview techniques to compare different sensory attributes of a product. Gretzel and Fesenmaier (2010) introduced the Sensory Experience Elicitation Protocol (SEEP), which consists of open-ended questions to elicit sensory association networks in consumers’ minds. They categorized tourists’ sensory experiences in the Midwest United States to apply in destination marketing. Another example of sensory impressions research is the study by Agapito et al. (2014), who applied survey questions related to five human senses (sight, smell, hearing, touch, smell) to analyze sensory tourist experiences of Southwest Portugal and segment visitor experiences into four sensory-informed themes: rural; nature-based; beach-related; and balanced experience. A number of researchers suggested using other methods of capturing people’s sensory experiences (e.g., Dunn, 1997Hayes, 2015Wendin, Allesen-Holm, & Bredie, 2011).

However, recent evidence suggests the mutual influence of emotional and sensory brain cortices. For example, Vuilleumier (2005) described the role of the amygdala in providing sensory signals, which influence the representation of emotions, while Sacco and Sacchetti (2010) explained the effects of sensory cortices on emotional responses. Hence, it seems possible that measuring the affective (feelings, emotions) component of experience could be similar to those for measuring the sensory component.

Each measurement technique has its advantages and limitations. In spite of evident advantages of self-report measures related to their simplicity, low-cost, and measuring multiple concepts in one setting (Paulhus & Vazire, 2007), their limitations do not allow capturing unconscious, affective, and sensorial components of experience or the temporal changes in them (Poels & Dewitte, 2006). The experience sampling method aimed at providing the temporal profile of experience is also influenced by self-report biases (Csikszentmihalyi & Larson, 2014). Several psychophysiological techniques that could provide measures of affective and sensorial responses are also limited by susceptibility to stimuli, individual differences, and interpretation of results (Stern et al., 2001). Therefore, only applying a combination of several methods such as self-report scales, experience sampling, laboratory experiments, and psychophysiological techniques could help to capture the totality of customer experience with its different components at pre-visit, on-site, and post-visit stages.