The nature of consumer decisions ranges from simple to very complex, and the time devoted to the decision process will vary accordingly. Consumers’ risk perception, the information required to conclude the decision and consumers’ confidence to conclude the decision therefore differ considerably from one decision to the next (Zander & Hamm 2012). Authors have tried to compile a complexity continuum to differentiate products in terms of the complexity associated with particular decisions to explain consumers’ behaviour in the marketplace when dealing with different product categories (Erasmus, Donoghue & Dobbelstein 2014). When discussing differences in the complexity that are associated with different purchase decisions, impulsive decisions are however not accounted for (Erasmus et al. 2014). Per definition, impulsive purchase decisions are made on the spot, without prior meditation or deliberation (Bayley & Nancarrow 1998; Rook & Fisher 1995; Saad & Metawie 2015; Sharma, Sivakumaran & Marshall 2010).

Impulsive purchases are not necessarily limited to certain product categories, or to low risk, low complexity purchases. In a specific context, a consumer’s decision to purchase a commodity on the spot may go beyond rational thought. Depending on the product, a consumer might argue that if an opportunity is lost, it might be impossible to get hold of a particular product again, to enjoy the advantage of the discount, or to pride oneself on owning a unique product that will be admired by others (Baumeister 2002; Hausman 2000; Rook & Fisher 1995). Impulsive purchases are unique in that they are based on stimulated interest and emotion that may even become uncontrollable and overpowering. The subsequent hedonistic desires may then override rational thought and stimulate action, such as to purchase products or participate in activities that a consumer has not considered prior to entering the retailer or venue (Kalla & Arora 2011). For example, an urge to have fun and to spoil oneself could supersede the functional benefits of the products, denoting internal antecedents as important contributors (Ungerer 2014; Virvilaite, Saladiene & Žvinklyte 2011).

Impulsive behaviour does not involve any pre-planning or prior information search: it is an immediate behavioural response to an individual’s desires (Beatty & Ferrell 1998; Ungerer 2014) when the individual almost instantaneously reverts to underlying emotional reasoning to act (Amos, Holmes & Keneson 2014; Silvera, Lavack & Kropp 2008). Impulsive behaviour occurs after consumers experience an urge which is often irresistible (Beatty & Ferrel 1998). Rational thought, on the other hand, implies calculative, informed decisions with some understanding of the value of the decisions (Babin & Harris 2016; Solomon 2007). These decisions include the use of product information, even if it is experience based (Jackson 2005), to enable an evaluation of the possible outcomes and advantages or disadvantages of a purchase decision (Solomon 2007; Van Hamersveld & De Bont 2007). Only when faced with an overwhelming array of options, a consumer would revert to emotions as a heuristic to facilitate choices (Van Hamersveld & De Bont 2007).

Because consumers are tempted beyond logical reasoning, one may argue that attempts to evoke impulsive purchases are not ethical. In defence, retailers have come to the realisation that consumers have become more critical of traditional brick-and-mortar shopping environments (Marsland 2015; Stops 2013), arguing that consumers appreciate stimulation, and a service scape that excites and entices (Chang 2016). For modern-day consumers, shopping is no longer an activity to solely acquire goods and services. Retail experiences have now become pivotal in terms of purchase decisions (Stops 2013). It is even argued that brick-and-mortar shopping will only be sustainable if it succeeds in providing an environment where consumers’ longing for pleasure, fantasy and fun can merge into a memorable experience (Portas 2010; Stops 2013). Experiential retail therefore seems to be the ideal approach to engage consumers in a memorable way (Pine & Gilmore 2011) through non-instrumental benefits and pleasure unlike traditional retail that predominantly focuses on utilitarian benefits and monetary gains (Fiore & Kim 2007). Consumer behaviour in experiential retail environments differs from that in traditional retail environments as instead of focusing on the purchase process, experiential retail aims at entertaining consumers. This means that the whole environment instead of merely product-related aspects will contribute in shaping the experience.

This study aims to identify and explain internal driving forces that might prompt consumers’ impulsive behaviour in experiential retail settings. Theoretically, this will underpin and enhance an understanding of the antecedents and driving forces that underlie consumers’ impulsive behaviour in these contexts. Findings will contribute to the body of theoretical knowledge on consumer behaviour by amalgamating literature on impulsive behaviour with that of experiential retail. Despite the acknowledgement that impulsive behaviour relates to the experiential approach of shopping (Hausman 2000; Mowen 1988), these concepts have not been explored together in the context of experiential retail. To date, several researchers have attended to experiential retail as a differentiation tool to identify principles of the experiential servicescape that could be incorporated by traditional retailers (Gentile, Spiller & Noci 2007; Plevoets, Petermans & Van Cleempoel 2010; Senthil, Chandrasekar & Selvabaskar 2012; Verhoef et al. 2009). Popular theories grounding experiential retail include Pine and Gilmore’s (1998) four realms of experiences (entertainment, educational, aesthetic and escapist) that could be used by retailers to design memorable experiences. Although this theory provides an overall guideline that provokes thoughts towards creating certain experience realms within any type of business, there is a lack of specific factors related to each realm. Schmitt (1999) also developed a theory based on strategic experiential modules (SEMs) wherein he explains how attending to consumers’ senses, actions, feelings, thoughts and relatedness can lead to the development of different experiences (sensory, affective, cognitive, physical and social). Experiential retail will inevitably result in the creation of an experience in the same way that every incident a person is involved in creates an experience. However, the challenge lies in creating a memorable experience which, when emerged in it, urges consumers to spontaneously react to offers to not only satisfy consumers’ desires but also promise desirable outcomes for retailers. In creating such experiences, an understanding of the internal motives that drive consumers in actual experiential venues is crucial.

Investigation of consumers’ behaviour in experiential retail contexts is limited. In 2007, Fiore and Kim developed a framework capturing experiential and utilitarian shopping experiences through synthesising variables deduced from previous studies without empirically testing the framework. The study also did not focus on consumers’ impulsive behaviour, which is central to consumers’ behaviour in experiential retail contexts. Extant literature that addresses consumers’ behaviour, specifically their impulsive behaviour, is still sparse. This study addresses the shortcomings depicted in existing theories and literature by empirically identifying the internal driving forces that predict consumers’ impulsive behaviour in experiential retail settings.

The theoretical background for the study integrates existing research on impulsive behaviour and prominent factors identified within experiential retail literature. Two prominent theoretical approaches, one representing impulsive behaviour (Muruganantham & Bhakat 2013) and the other experiential retail (Arnold & Reynolds 2003), were merged. This led to the identification of the most prominent variables that might explain consumers’ impulsive behaviour in an experiential retail context. Figure 1 indicates the respective variables that were deduced for the study as well as additional literature in support of this.

Muruganantham and Bhakat (2013) studied impulse behaviour in traditional retail settings by extracting secondary information from over 50 years’ publications. Nine internal factors that seem prominent to prompt consumers’ impulsive behaviour were identified, namely emotions, hedonism, fashion, normative evaluation, variety seeking, enjoyment, impulsiveness, self-identity as well as product involvement.

Empirical studies on experiential retail indicate that mood and emotions, hedonism and recreation (enjoyment) are related to consumers’ behaviour in an experiential retail setting. Regarding hedonism, Arnold and Reynolds (2003) further identified six different hedonic desires, namely adventure, gratification, value, idea, social and role shopping, to represent the various pleasurable motives that might drive consumers’ shopping. These hedonic desires concur with the factors that were identified by Muruganantham and Bhakat (2013). Idea shopping represents fashion, while gratification shopping represents normative evaluation and adventure shopping represents variety seeking. The different hedonic desires as termed by Arnold and Reynolds (2003) are used to describe the hedonic constructs appropriate to experiential retail.

In this study, three of the constructs identified by Muruganantham and Bhakat (2013) were excluded, namely (1) impulsivity as a lifestyle trait because the nature of experiential retail, often temporary, novel and scarce, arouses impulsivity even in visitors who are not typically impulsive, (2) self-identity, arguing that impulsive behaviour in an experiential context does not transpire as an attempt to express or reflect the inner self but merely to enhance the experience or to acquire memorabilia and (3) Product involvement as an influence on consumers’ impulsive behaviour in an experiential retail context, arguing that in an experiential retail setting, the environment rather than the products gets consumers immersed.

Based on the amalgamation of factors, the hypothesised relationships between the independent factors and consumers’ impulsive behaviour as the dependent variable are subsequently explicated.

This cross-sectional quantitative study followed a deducto-hypothetico approach (Walliman 2005), which implies that the hypotheses were deduced from existing literature. The hypotheses aimed to confirm relationships among independent and dependent variables (Aaker et al. 2011; Hopkins 2000). The electronic survey was economical, enabled a rapid turnaround in data collection and permitted the recruitment of respondents from a broad geographical area based on the large database of an established research company, Consulta Pty Ltd. (Creswell 2014).

A pre-test aimed at a minimum of 100 responses was launched on 10 July, 2015. A total of 225 responses were obtained during 3 days. The internal consistency of items measuring the same construct was assessed by means of Cronbach’s alpha to detect problematic items (items that did not correlate sufficiently with others measuring the particular construct). Problematic items were reworded to improve reliability, face validity and clarity. The pre-test further included 11, instead of the 8 retailers respondents could refer to. However, results indicated that because some of the 11 retailers only incorporate certain elements of the experiential approach in their service delivery such as having a theme and thematic products on offer while predominantly operating a traditional retail format, respondents got confused. All retailers included were therefore reselected during a focus group session by an expert panel and limited to only eight, which could be classified as fully experiential. This necessitated the need for a pilot test, also aimed at 100 responses, which was launched on 25 August, 2015. The study was closed on retrieval of 101 responses. After consultation with a statistician, the questionnaire was deemed reliable and valid and the final questionnaire was fielded.

The final data collection that took place from 18 September to 5 October, 2015, delivered a total of 402 usable questionnaires. Worth mentioning is that 1460 interested candidates did not meet the requirements to participate as they did not visit one of the eight identified experiential retailers. Even though the participation rate decreased dramatically after limiting the retailers to fully experiential ones, the inclusion of fully experiential retailers enabled more accurate measurement within an experiential retail context and improved the validity of data. In order to calculate the minimum sample size for the bigger study, the recommendations of Tabachnick and Fidell (2013) were followed, namely, N > 50 + 8m (m indicates the number of independent variables). Considering that this formed part of a broader study including six independent variables, 98 questionnaires were required. However, for stepwise multiple regression analysis, it is recommended that a ratio of 40 cases is required for every independent variable (Pallant 2013), which amounts to 240 questionnaires. This target was easily exceeded.

A heterogeneous sample of South African citizens that included males and females, 18 years of age or older, from various ethnic backgrounds, with divergent educational levels and annual earnings, was recruited electronically. Non-probability, purposive sampling was employed to recruit willing respondents, who have previously visited one of the identified experiential retailers. The investigation required consumers’ retrospective evaluation of their actual experience at one of the selected eight experiential retailers. These retailers represent different types of experiential retail formats such as themed restaurants, experiential retail stores, experiential destinations and Pop-up stores. Table 1 provides a concise description of each of these experiential retailers.

These stores are further representative of fully experiential retailers. This implies that the retailer incorporates an experiential approach in every aspect related to the business and not only in the designing of the store or products. A pre-requisite for respondents to participate in the study was having visited at least one of the listed establishments within the 3-year period preceding the data collection. Three years were considered appropriate bearing in mind that at the time of data collection, only a limited number of fully experiential retailers were operating in South Africa and that some experiential retailers have a temporary nature such as Pretville, Pestaurant and the Magnum Pop-up store. Consumers more easily recall positive experiences than negative ones (Kim, Ritchie & McCormick 2012) and tend to remember their best experiences, as opposed to materialistic purchases where consumers tend to best recall the worst ones (Martin, Reimann & Norton 2015). Respondents who had visited more than one experiential retail setting before completed the survey based on their most memorable experience to ensure the recall of specific behavioural reactions encountered while there. Limiting the point of reference to a single experiential retail trip enabled effective comprehension of impulsive behaviour within a specific experiential retail environment. The electronic survey enabled automatically substituting the selected retailer’s name in relevant questions to constantly remind respondents to refer to the same experience.

A total of 17 questions and 43 items included in the electronic questionnaire are relevant to report on in this study. Two questions in the questionnaire identified the experiential retailers that were patronised as the most memorable visit. Two screening items were used to restrict access to items pertaining to purchases and activities to only respondents who had actually purchased items or taken part in experiences in the respective experiential retail settings during their visit (Table 2). Seven questions related to the dependent and independent variables; five questions gathered demographic information using nominal scales. Income and education level were measured in alignment with the predetermined classification criteria.

An 11-point Likert-type agreement scale was considered most suitable to measure the constructs of the study (mood and emotions, adventure, gratification, value, idea, social and role shopping, recreation, irresistible desire creation, impulse purchases and impulse participation), where 0 = strongly disagree and 10 = strongly agree. Considering that this formed part of a bigger study which required the use of more advanced statistical analyses, including Covariance-Based Structural Equation Modeling (CB-SEM), 11 points were considered more suitable for the analyses requirements as well as to enhance construct validity (Awang, Afthanorhan & Mamat 2016). The 25 items measuring the independent variables originated from previous research (Arnold & Reynolds 2003; Kim 2003; Mathwick, Malhotra & Rigdon 2001) but were adapted to allow retrospective assessment of the visit at a specific experiential retailer instead of relating to shopping behaviour in general. The scales from which items were adapted were proven reliable and valid in the studies from which they originate (Arnold & Reynolds 2003; Kim 2003; Mathwick et al. 2001). Irresistible desire creation was investigated by means of three items that were self-formulated based on the literature. The three items related to impulse purchases and four items related to impulse participation were adapted from Rook and Fisher (1995) (Table 2).

One item pertaining to impulse purchases as well as one pertaining to impulse participation was removed as they were reverse coded items, causing consistency concerns. The Cronbach’s alphas subsequently increased from 0.57 to 0.62 for impulse purchases and from 0.63 to 0.70 for impulse participation. Considering Hinton et al.’s (2004) guidelines for assessing reliability results, wherein Cronbach’s alpha values between 0.50 and 0.69 are regarded as moderately reliable and values of 0.70 and above as highly reliable, internal consistency reliability is confirmed.

Table 3 indicates that the male and female distribution was almost equal with most respondents aged between 26 and 55 years. White people and Africans dominated the ethnic representation. Respondents were well educated, while lower to middle income categories were strongly represented.

The statistics program IBM©SPSS©23 was used to analyse the data. Internal consistency reliability was assessed by evaluating Cronbach’s alpha values (De Vellis 2012). Stepwise multiple regression analysis was subsequently conducted to test the hypothesised relationships between the independent and dependent variables. The cut-off value for significance is p ≤ 0.05 (Malhotra 2010). The purpose of regression analysis is to determine whether the independent variables explain a significant variation in the dependent variable and confirm the existence of a relationship (Malhotra 2010). This study specifically aimed at exploring the best predictor variables.

Cronbach’s alpha values were calculated for each of the constructs to ensure adequate internal consistency (see Table 2). All independent variables portrayed high Cronbach’s alpha values (mood and emotions = 0.96; adventure shopping = 0.91; gratification shopping = 0.83; value shopping = 0.83; idea shopping = 0.86; social shopping = 0.91; role shopping = 0.90; recreation = 0.88). The same was true for irresistible desire creation, with a Cronbach’s alpha value of 0.91. Impulse purchases portrayed moderate internal consistency reliability with a Cronbach’s alpha value of 0.62, while impulse participation portrayed high internal consistency reliability with a Cronbach’s alpha value of 0.70 (Hinton et al. 2004).

Stepwise multiple regression analysis was conducted to explain the variability by means of the significant independent variables. Based on the assumptions of multiple regression analysis (Hair et al. 2014), the following confirm regression analyses as appropriate to analyse the data. Pearson’s correlation coefficient identified adequate linear association between variables, exceeding 0.3 in most instances (Pallant 2013). The linearity of residuals was confirmed by means of a scatter plot. The data were retrieved cross-sectionally to eliminate independency concerns. Homoscedasticity of plotted residuals and normality assumption based on histogram outcome were visually assessed as reasonable.

Stepwise regression analysis results identified mood and emotions (B = 0.309; p < 0.0005) as the best significant positive predictor of irresistible desire creation, followed by recreation (B = 0.307; p < 0.0005) and gratification shopping (B = 0.232; p < 0.0005). The explanatory value R² indicates that these three variables cumulatively account for 61% of the variance explained in irresistible desire creation.

Table 4 indicates that four variables emerged as significant predictors for impulse purchases. The relationship between mood and emotions (B = -0.302; p = 0.001) and impulse purchases is significantly negative. Gratification shopping (B = 0.297; p = 0.002) was identified as the best positive predictor of impulse purchases, followed by adventure shopping (B = 0.287; p = 0.018) and social shopping (B = 0.171; p = 0.022). Based on a cumulative explanatory value of 23%, impulse purchases are not strongly predicted by the variables that were attended to in this study.

In order of best predictors, results show that gratification shopping (B = 0.341; p < 0.0005), adventure shopping (B = 0.194; p = 0.042) and social shopping (B = 0.186; p = 0.007) are significant positive predictors of impulse participation in activities. These three factors altogether account for 42% of the variance explained in impulse participation.

The significant positive relationship between mood and emotions and irresistible desire creation supports H1_a. This suggests that the experience at an experiential venue is satisfactory as it leaves consumers with a craving for more. Support was also confirmed for H1_b in that a significant negative relationship exists between mood and emotions and impulse purchases. The relationship between mood and emotions and impulse participation is insignificant, thereby not supporting H1_c. These findings imply that the yearning for more will not necessarily culminate in purchases or participation in activities. This supports previous research that a negative mood state might be more conducive to impulse purchases than a positive mood state (Baumeister 2002; Verplanken et al. 2005). Hence, visitors who are already content might refrain from purchasing products or participating in activities to preserve their mood or because they do not need something to support their positive mood state. Furthermore, any purchase or participation in activities might spoil their experience as that entails some form of risk.

With reference to H2, only adventure shopping, gratification shopping and social shopping seem to positively predict impulsive behaviour. The adventurous nature of experiential retail was identified as a positive predictor for impulse purchases and impulse participation, supporting H2.1_b and H2.1_c but does not significantly predict consumers’ irresistible desire creation. H2.1_a was thus not supported.

Results further support H2.2_a, H2.2_b and H2.2_c. This confirms that the ability of experiential retail to relieve visitors’ stress and to serve as distraction from everyday obligations will fuel visitors’ urge to extend the experience and might even prompt them to purchase products or participate in activities on the spur of the moment.

The respective relationships between value shopping and idea shopping as well as impulse behaviour were not statistically significant (p > 0.05). Results did thus not support H2.3 and H2.4. Consumers visiting experiential retailers might therefore not do so with the aim of collecting novel ideas or purchasing exclusive, limited edition products but merely to enjoy the experience.

Social shopping positively predicts impulse purchases as well as impulse participation, supporting H2.5_b and H2.5_c. The relationship between social shopping and irresistible desire creation was not significant (p > 0.05), thereby not supporting H2.5_a. This is possibly owing to the perception of social risk. When visiting experiential retail venues with friends or family, consumers might feel obliged to purchase products similar to what some of their companions are doing, not to feel left out. When in a social milieu, consumers are usually more prone to conform to the behaviour of others and therefore participation in activities is more likely to occur when with significant others. Usually, friends do not want to dissociate themselves while the rest of the crowd are having fun (Hausman 2000). Irresistible desire is a personal state and unless it progresses into action, nobody else is exposed to, or experiences it. Therefore, others do not socially benefit from an individual’s behavioural state. Neither can they be disappointed by not experiencing it in a crowd, as it is solely internal and does not reflect externally at all. The absence of a significant relationship with irresistible desire despite the depicted significant relationship with impulse purchases and impulse participation indicates that an irresistible desire does not necessarily precede a consumer’s impulsive behavioural actions (purchases or participation). This is a valuable theoretical contribution to acknowledge different impulsive behavioural outcomes in the context of experiential retail.

Results did not support H2.6 as the relationship between role shopping and impulsive behaviour was not statistically significant (p > 0.05). Purchasing gifts usually involves some pre-planning, especially financial budgeting, and therefore consumers might have more control over their actions when purchasing gifts as they will not directly benefit from the product themselves. Impulsive behaviour is thus not relevant during role shopping.

The results further support H3_a in that recreational shopping significantly positively predicts the creation of irresistible desires. However, H3_b and H3_c are not supported and therefore recreation does not predict impulse purchases or impulse participation. This implies that consumers who enjoy shopping as a pastime do not necessarily have the intention to purchase products or participate in activities. Visitors to experiential retailers might subsequently merely visit the venue for enjoyment and to spend leisure time (Fiore et al. 2005). Shopping is nowadays regarded as a recreational activity and experiential retail has all the elements to fulfil this need. Pine and Gilmore (2011) propose that consumers pay an admission fee to visit experiential venues. This would provide retailers some financial support for their efforts to continually augment their offerings to provide recreational shoppers with a memorable experience and to remain sustainable.

The study extended former studies that propose that mood and emotions influence consumers’ impulsive behaviour (Amos et al. 2014; Bäckström & Johansson 2006; Verplanken & Sato 2011). However, unlike Bäckström and Johansson (2006) as well as Verplanken and Sato’s (2011) conclusion of a positive association between mood and emotions and impulse purchases, this study revealed that consumers in a positive mood state are not necessarily inclined to make spontaneous purchase decisions. Results indicated that it is highly likely that consumers, who are in a positive mood, would become emotionally excited and subsequently uplifted when visiting an experiential retailer, although they would not necessarily purchase anything on impulse. Rather, they would develop an irresistible craving for more, which often culminates as the total outcome of the experience. Consumers, who do not experience experiential retail as emotionally uplifting, might towards the end of their visit impulsively decide to purchase something or take part in an activity to change their mood state. This could be attended to in future studies.

Secondly, the study expands prior research that identified hedonism as a trigger for impulsive behaviour (Amos et al. 2014; Hausman 2000; Kalla & Arora 2011; Verplanken & Sato 2011). The results showed that three of the six hedonic motivations could prompt impulsive behaviour in an experiential retail setting, namely adventure, gratification and social shopping. Therefore, visitors to experiential retailers who purely go for the excitement and adventure often end up indulging in impulse offerings. For them, it is imperative that experiential retailers offer experiential products and activities as they are likely to participate and take advantage of it on the spur of the moment.

The study confirmed that experiential retail environments might be attractive for consumers who are looking for instant pleasure and distraction from their everyday obligations. These consumers might also be viable targets to purchase products on impulse and participate in activities as a form of distraction from their busy, stressful schedules. The results revealed that social shopping often leads to impulse purchases and impulse participation in an experiential retail setting, possibly because individuals do not want to feel excluded in a group. This validates results of Dhaundiyal and Coughlan (2016) that socialising increases impulse tendencies. Experiential retailers should therefore cater for groups, offering products that can be shared, as well as group activities, along with opportunity for individual participation.

The findings could not confirm value, idea and role shopping as antecedents for impulsive behaviour in an experiential retail setting. The purpose of visiting an experiential retailer might not be to purchase exclusive products or to gather innovative ideas, but rather to merely enjoy whatever the experience entails, as is typical of adventure shopping. It is recommended that experiential retailers offer special visitation packages that cover an entrance fee, and possibly a small souvenir in addition to the basic product or service offering. This will allow role shoppers to plan the visit and take advantage of offerings at the venue without feeling pressurised to commit to additional expenses that prompt regret owing to overspending. Similar to previous research, this study also indicated that a visit to experiential retailers as a recreational outing might be enjoyable in itself even without culminating in unplanned purchases or participation in activities (Fiore & Kim 2007). Mostly, these visitors merely attend to enjoy the experience without any purchase intentions.

This study makes a theoretical contribution through confirmation that brick-and-mortar stores can be modified to serve as recreational destinations through intentional manipulation of consumers’ mood and emotions, and through optimisation of consumers’ adventure shopping motives, desire for gratification and social shopping goals. It further suggests that impulsive behaviour within an experiential retail setting can transpire as one of three possible end states, namely an irresistible desire creation, impulse purchases and impulse participation in activities. An irresistible desire can be viewed as a satisfying inner feeling and a yearning for more. Unlike Beatty and Ferrel (1998) and Kim (2003), who identified an irresistible desire as a preceding mediator for further impulse actions, this study revealed it as an end state, which is not an additional behavioural action that might lead to increased sales or immersive participation. Such an experience nevertheless creates a foundation for good public relations, positive word-of-mouth and helps to attain a sustainable competitive advantage and customer loyalty, which are highly sought-after outcomes for prosperity in today’s competitive retail environments.