The first model that grounds this study is Schmitt’s (1999) experiential branding model. This comprises a five-dimensional framework for experiential branding, which covers sensory experience, affective experience, intellectual experience, behavioural experience and social experience. According to Schmitt (1999), these dimensions are evoked by various stimuli, including the brand’s design, packaging, environment and communications. Firstly, the sensory experience refers to the visual, auditory, tactile, atmospheric and gastronomic experiences that a customer has when they interact with the marketing environment. The affective dimension is the emotions, which are evoked when immersed in the environment. Further, the intellectual (or cognitive) aspect of the experiential branding model refers to the customer’s reasoning about the marketing environment with which they engage (Schmitt 1999). An environment that provides a strong cognitive experience is able to effectively communicate messages to the client about different positive elements in relation to visiting and making purchases in the store (Brakus, Schmitt & Zarantonello 2014). The fourth experience that the model comprises is the dimension of behaviour. The latter dimension is concerned with customers’ experiences with the dynamics of the marketing environment. Lastly, the social experience concerns the nature and quality of relationships formed in the store environment, both between customers and employees, and among the different customers in the store. The basis of this model correlates with this study’s conceptual model in the following ways: both are grounded in the concept of sensory branding and the way in which the servicescape (store design and layout, interaction with other customers, the service process and customer service) influences consumers’ affective and cognitive behaviour.

Hultén (2011) developed a multi-sensory brand experience model based on the argument that stores should make use of diverse sensorial strategies in order to differentiate its brand and connect with its target markets. According to Hultén (2011), the five senses (smell, sound, sight, taste and touch) play important roles in customer experience when visiting a store. The importance of each of the senses may differ, depending on the product or service that a store is selling (Hultén 2011; Hulten, Broweus & Van Dijk 2009). For example, a bakery may focus on the smell of bread, which is a strong tangible means to compel visitors to enter the store. Likewise, a coffee brewery will benefit from the attraction of the smell of coffee, whereas a clothing store might consider a fragrance or an earthy smell. Further, the importance of sound in terms of type of music, as well as the volume, will differ according to the kind of product being sold and the relevant target market. However, by utilising sensory branding, retailers can gain tremendously in terms of sales and profits. The application of this model to this study is as follows: in a similar way to the aforementioned model, the majority of these sensations are incorporated into the ‘store environment’ construct (atmospherics, visual and auditory sensations), while the gastronomic sensation (taste) is less relevant to fashion and has therefore been omitted.

Store environment is a critical aspect of marketing and refers to the physical environment of a store, such as the music, lighting, store layout and arrangement of merchandise (Hu & Jasper 2006; Wang & Ha 2011). Aspects of the store environment have been found to have a significant influence on a store’s financial performance (Lam 2001). Likewise, the store environment can elicit a wide range of behaviours from people within the store (Geetha, Bharadhwaj & Piyush 2013), where certain elements can influence customers to stay in the store longer, increasing the probability that they may buy an item. Conversely, store elements can also have a negative effect on consumer’s behaviour by influencing them to leave the premises prematurely.

Brand attitude refers to the direction and strength of the perception that consumers have in relation to a brand (Yoon & Park 2012). By gaining insight into brand attitude, marketers are able to determine customers’ brand perceptions and ascertain their willingness to make a purchase. A customer’s brand attitude also tends to reveal his or her predisposition to recommending the brand to others (Tong & Hawley 2009). Yoon and Park (2012) found that sensory advertisements have significant impacts on brand attitudes. In particular, some ads allow consumers to associate with the brand and therefore to view the brand positively (Yoon & Park 2012). On the contrary, consumers who find some ads offensive or ludicrous tend to transfer these perceptions onto the brand itself (Yoon & Park 2012). Brand attitudes are therefore an important measure in determining possible buying behaviour among current and potential customers.

Brand experience is another important category in the fashion retail industry and is defined by Brakus, Schmitt and Zarantonello (2009) as the totality of subjective, internal consumer responses, such as sensations, feelings and cognitions that are evoked by brand-related stimuli. It is further defined as the overall impact that a brand has on a customer, which is derived from the combination of the different marketing elements that are put together in order to communicate the brand (Shamim & Butt 2013). The brand experience is often regarded as the sensory embodiment of the brand (Hultén 2011) and refers to how the brand connect with the five different senses of consumers, from how the logo of the brand appears to people, to how appealing the ambience is to people who visit the stores where products of the brand are sold. A number of studies have found that brand experience has a significant influence on brand loyalty (Brakus et al. 2009; Gentile, Spiller & Noci 2007; Ghodeswar 2008; Ponsonby-McCabe & Boyle 2006; Schmitt 2009).

Purchase intention refers to a consumer’s plan to buy a product or engage in a service while it involves the particular circumstances and timing in which the consumer will buy an item (Lu, Chang & Chang 2014). Such buying behaviour is usually driven by a certain need or compulsion that arises, or if the item’s prices reaches a certain desirable level (Lu et al. 2014). Purchase intention is commonly measured using three dimensions, such as priority, periodicity and spontaneity (Anderson et al. 2014; Ko & Megehee 2012; Moore 2014). Priority items entail goods that are highly prioritised such as food and petrol, whereas periodicity refers to how often the consumer intends to purchase the product. Some products have very high periodicity, which means that consumers purchase them at predictable intervals of time (Moore 2014). Lastly, spontaneity refers to the length of time between the customer’s decision to buy a product and the actual execution of the intention (Moore 2014).

The data collection was conducted among 501 black women in South Africa, considered to be part of the middle-class. By means of systematic sampling, every fourth individual from a pre-screened panel of 2000 elements were selected for questionnaire completion. This sample size was calculated in the following way: from an estimate population of 3.4 million black middle-class women in Johannesburg, an allowance of a 95% confidence level was allowed, using a 5% confidence interval. By means of the Raosoft_® sample size calculator, the researcher has estimated a sample size of at least 384. To decrease possible sampling errors, a sample of 501 participants was surveyed.

Research scales were operationalised on the basis of past research in the same field of study, and adequate modifications were applied in order to fit the current research context and purpose. All scales were measured using a 7-point Likert scale. For the first variable, store environment, a modified version of Baker, Grerral and Parasuraman’s (1994) scale was used. Brand experience was measured using a scale adapted from Sahin, Zehir and Kitapçi (2011). To measure brand attitude, Martinez and Chernatony’s (2004) scale was used, and lastly, purchase intention was measured using Shukla’s (2011) purchase intention scale.

The profile of the participants is presented in Table 1 followed by a discussion.

The findings in Table 1 indicate that the majority of the respondents (40%) have a diploma, whereas 26% and 24% have a graduate degree and a professional certificate, respectively. The remainder, which accounts for 10%, has a postgraduate degree. Fifty-two per cent of the respondents are between the ages of 25 and 34 years, with 26% between the ages of 17 and 24 years and 22% being older than 34 years of age. The analysis further indicates that 53% have a monthly household income of between R8001 and R40 000, with 19% earning less than R8000 per month and 14% earning above R40 000. More specifically, from the portion of respondents earning between R8001 and R40 000, three relatively equal clusters are evident, with incomes ranging between R8001 and R16 000 (21%), followed by those between R16 001 and R25 000 (16%), with 16% earning between R25 001 and R40 000. In terms of occupation, almost 60% are full-time employed and working for a company. Fourteen per cent of the respondents are students or trainees, whereas 10% are seeking employment. The remainder, which accounts for a relatively small portion, either works part-time or is self-employed.

The scale accuracy analysis is presented in Table 2 followed by a discussion of the measurement scale reliability and validity.

As indicated in Table 3, the inter-factor correlation values for all paired latent variables were lower than 1.0; this suggested the existence of discriminant validity according to Pooe, Mafini and Loury-Okoumba (2015). Therefore, indicating that each construct is distinct from the other constructs proving existence of heterogeneity (Malhotra 1996). Furthermore, a correlation value between constructs of less than 0.7 is recommended in the empirical literature to confirm the presence of discriminant validity (Nunnally & Bernstein 1994).

This section presents the analysis of model fit by assessing a number of different indices, so as to ascertain the degree of acceptable or good model fit (Table 4). The following indices are examined: CMIN, the normed fit index (NFI), incremental fit index (IFI), Tucker–Lewis Index (TLI), goodness of fit index (GFI), and the root mean square error of approximation (RMSEA). Because of standardised regression weights being lower than 0.5, STE3, STE5, STE 9, STE12 and BE2 were removed from analysis (Hair et al. 1998). Table 4 presents the results of the model fit indices.

It is evident from Table 4 that all the model fit indices reach the acceptable thresholds. RMSEA is indicated by a value of 0.04 that is below 0.08 as recommended by Hooper, Coughlan and Mullen (2008). CMIN, NFI, RFI, IFI, TLI and CFI all meet the recommended threshold of 9 (McDonald & Ho 2002).

Table 5 presents the results of the hypotheses and path coefficients followed by a discussion.

The results in Table 5 provide clear evidence that four of the five hypotheses are supported. In other words, store environment has a positive influence on both brand attitude (H1) and brand experience (H3). Furthermore, brand attitude has a positive influence on purchase intention (H4), and lastly, brand experience positively influences purchase intention (H5). The relationship between store environment and purchase intention was found to be insignificant and the hypothesis therefore unsupported (H2). Upon examining the estimates, the strongest relationship was found between store environment and brand experience (0.79), followed by store environment and brand attitude (0.69). The weakest relationship was found between brand experience and purchase intention (0.16). Brand attitude had a relatively strong influence on purchase intention (0.42).