The franchising industry plays an important role in the South African economy and contributed over R302 billion revenue to the overall gross domestic product (GDP) in 2013 (Franchising Association South Africa [FASA] 2014:14–17). Within this industry, the fast food franchising sector leads by commanding approximately 22% of the overall franchise sector (Stats SA 2014:140–198). South Africa experienced a 160% increase (R516.3 million to R1342.9 million) in the income generated by the fast food industry between 2006 and 2012 (Stats SA 2014:140–190). The growth of the South African fast food industry has been attributed to factors such as rising household incomes, growth in the black middle class segment, participation of women in the labour force and the increasing value of household time (Mackay et al. 2013:68; Roberts-Lombard 2009:236).

Governmental, political, social and economic change in South Africa between 2001 and 2011 has shown a drastic shift in the consumer profile; the biggest shift has been the decline of 48% of the lower consumer profile as segmented by the LSM 1 to 4. A major shift in the LSM 5 to 7 has been identified by a 57% increase in this segment, as well as a 45% increase in LSM 8 to 10 (South African Audience Research Foundation [SAARF] 2012). Therefore, the aggregated projected food sales in South Africa, for the period 2011 to 2016, are forecasted to grow at around 3.75% (FASA 2014; SAARF 2012). This growing middle market leaves the fast food market open to both local and international players, where the key to success is primarily based on positioning in the market through site location and brand performance (North & Kotze 2004:30).

Extant theory assumes that both franchisors and franchisees choose franchising in order to improve performance over time (Hua & Dalbor 2013:724) and consequently revenue growth performance is operationalised as the study dependent variable. Traditionally, research on franchising has focused on agency costs (Michael & Combs 2008:76) to ensure that the responsibility lies with the franchisor to maintain high quality franchisees across their network distribution. In this regard, business format franchising is a widespread form of operation (Nijmeijer et al. 2013); under this arrangement, advertising, brand building, product-service mix and overall quality monitoring of the franchised outlets remain the obligation of the franchisor to ensure that revenue growth and sustainability of both parties is maintained. While acknowledging the importance of such an arrangement and its derived benefits, several studies have further identified various factors that influence the performance and profitability of both the franchisor and franchisee (Kacker et al. 2013:25). Factors such as the franchise site location, convenience, quality of the physical environment and service of the restaurant, including its image and perceived customer value, have all been associated with improved performance (So et al. 2013:33).

Moreover, successful franchisors and retailers are customer centric and their strategic and tactical decisions revolve around their present and potential customers (Urban & Kongo 2015:2). Store choice decisions in the food retailing industry have been widely discussed in the literature. A consumers’ decision to shop at a store is in general based on location or convenience, assortment and quality of merchandise, service quality, price image, ambience conditions and promotions as well as brand selection (Jaravaza & Chitando 2013:303).

Building on in this research direction several key variables are operationalised for the purpose of this study which have been identified in the literature as important factors to franchise growth. The selection of variables is by no means exhaustive. It is acknowledged that the actual process of franchisee sales growth is far more complex and that no single factor can determine the outcome of this process. Notwithstanding the complexity of the phenomenon and the reciprocal nature of relationships between different factors and franchise growth, hypotheses are formulated and restricted to a number of variables and links. These variables, which inform the hypotheses, are discussed below.

Various models of site selection and location exist within the fast food and general retail environment, with the most prominent model, the location-allocation model (Goodchild 1984:87), analysing the relationship between the facilities available within the location as well as the demand density within the area (Frasquet, Gil & Molla 2001:25). Additionally, the Shopping Centre Attractiveness (SCATTR) scale is based on 21 attributes and has been used successfully to assess which factors influence the attractiveness of shopping centres from a customer perspective (Wong, Lu & Yuan 2001:26). The SCATTR scale includes factors such as family shopping environment, banking facilities, retail merchandising, food court, entertainment and parking availability. Additional attributes within the South African environment have been noted such as the need for security (North & Kotze 2004:33). Building on previous research findings, a number of these site location factors are discussed below to inform our hypotheses.

Accessibility: Accessibility is the ease with which a customer can get into and out of a site. To assess a site’s accessibility, a fast food outlet or retailer simultaneously evaluates several factors such as road patterns, road conditions and barriers (Jaravaza & Chitando 2013:305). Considering that in South Africa many consumers rely heavily on public transport, accessibility to franchise fast food outlets is pivotal. Indeed, if the accessibility of the shopping centre does not allow for ease of access to retail and fast food outlets, this will decrease the chance of consumers patronising the retail location and negatively influence growth (Strydom 2014:3). Building on past empirical findings (North & Kotze 2004:32) and by conceptualising accessibility broadly in terms of parking availability, level of security, access to the main transport route as well as ease of access to the retail mall itself, in the first instance it is predicted that:

Competitive intensity: Competitive intensity is conceptualised as the level of direct competition that the focal firm faces within its immediate business domain (Doherty 2009:529; Emerson 2010:191). Competitive intensity can increase competitive advantage and influence consumer purchasing behaviour over time because competition erodes customers’ perceptions of differential advantage along unsustainable dimensions (Doherty 2009:533). In South Africa, the increased competition from international brands such as Domino’s, Pizza Hut and Burger King entering the local market place have put pressure on local franchise fast food outlets in recent years (FASA 2014:14). Emerson (2010:193) finds that increased competition from international competitors and saturation of local brands tend to drive both prices and profitability down as competing brands look to get a bigger piece of the market share. Consequently, saturation and cannibalisation of existing stores encroaching on existing trade areas can lead to reduction of the probability of growth (Emerson 2010:195). Thus:

Convenience: Convenience is a multidimensional construct and research shows that convenience is significantly related to customer satisfaction and behavioural intentions, consumer switching behaviour and customer perceptions and retention (Strydom 2014:4). Convenience has also been found to act as a positive moderator of satisfaction’s effects on repurchase visits and spending. Location and convenience are well established key success factors within the retail environment, which interact with other factors to influence consumers’ store choice and include price, selection of products and service levels of competitors (Jaravaza & Chitando 2013). Convenience and ease of transporting goods to the household are important considerations which ultimately drive down the cost for the consumer of travelling to a major retail mall as opposed to selecting a closer local mall (Strydom 2014:4; Terblanche & Boshoff 2010). Convenience plays a vital role as many consumers visit a pre-selected retail shopping centre that is located within walking distance of their home and or along a main transport route (Jaravaza & Chitando 2013). Consequently it is expected that:

Central facilities: The optimal retail location mix of both tenants and offerings is crucial to franchise success in order to attract the maximum number of consumers within a given trade area. As each retail location is unique with its own demands, store outlets must take cognisance of their trade area and ensure their offerings are in line with the market requirements (Jaravaza & Chitando 2013:307). Central facilities of a retail establishment are comprised of many different offerings as noted by Wong et al. (2001:27) and include: banking facilities, major retailers and supermarkets, various specialised retail shops and entertainment facilities, which must not only fit with the surrounding trade area but also act as a pull factor for consumers who patronise the retail mall. Research finds that clusters of retail stores within a shopping centre generate increased consumer traffic (Jaravaza & Chitando 2013:307). The location of fast food outlets within retail establishments is often primarily located within a food court or strip mall area where many competitive offerings are available (Shaw 2009). This strategy coincides with the central placement theory of location (Grosh & Craig 1983), which supports the generalised assumption that consumers will patronise the closest available store within the retail environment; thus, the stores located closer to the entrance of the mall or in an emerging market closer to a main transport route should perform at higher growth rates (Tayman & Pol 2011:46). Building on this empirical evidence it is predicted that:

Building strong brand equity is vital for franchising, based on the relationship between that of franchisor and franchisee where it is essential to co-promote brand equity (Nyadzayo, Matanda & Ewing 2011:1105) and customer satisfaction (Terblanche & Boshoff 2010:3). Brand awareness, brand image and brand attributes are important for the franchising industry as consumers differentiate across the spectrum of competing brands, resulting in decisions that ultimately relate to profits, sales growth and survival (Gillis & Combs 2009:553).

Brand awareness: Brand knowledge comprises two elements which form part of the memory node that consumers hold when identifying a particular brand; these are brand awareness and brand image. The conceptualisation of brand knowledge through memory recall is important, considering that brand awareness comprises brand recognition and brand recall performance (Keller 1993). Brand recognition relates to the consumers’ ability to ratify a prior experience with a product utilising the brand as a clue, whereas brand recall refers to a consumer’s ability to recall the brand when given a product category. Brand awareness within the fast food industry has been found to have a positive influence and significantly strengthen the brand relationships that customers display towards their preferred fast food offering (Mackay et al. 2013).

Brand image: Brand image, according to Keller (1993), is the perception of the brand as reflected by associations, which are embedded in the consumer’s memory. Brand associations are the additional information nodes that are connected to the brand node in the consumer’s memory and embrace the meaning of the brand. In the fast food industry, brand image typically refers to the sum of emotional perceptions, ideas or symbolic attitudes that customers associate with restaurants (Mackay et al. 2013:68).

Brand attributes: Product-related attributes refer to the physical composition of the product such as the ingredients required to produce the product, while non-product-related attributes refer to price information, packaging and the product’s visual appearance (Keller 1993). Product-related attributes such as taste, texture and quality are important in understanding how consumers differentiate between brands. Differences between the brands are communicated overtly by making comparisons with competing brands’ attributes or tacitly without stating their obvious competitive advantage. A brand’s success is weighted heavily on both direct and indirect factors of the brand attributes and its positioning in the market, which is key to ensure the continuous consumption of the product (Odin, Odin & Valette-Florence 2001:75).

Price and promotion: Related to brands, price is an important attribute as consumers have beliefs regarding the association between price and value of a brand. Therefore, price promotions on higher quality brands have shown to take market share away from similarly priced quality tier brands and lower tier brands. However, lower tier brands with price promotion will not take any market share from the higher tier brands but rather cannibalise similarly priced brands (Roberts-Lombard 2009:236).

In summary, brand awareness, brand image, brand attributes and pricing and promotion are important considerations for fast food franchise outlets as consumers rely on these factors to differentiate across the spectrum of competing brands, which ultimately influences sales growth (Gillis & Combs 2009:556; Terblanche & Boshoff 2010). Moreover, in acknowledging the importance of how the consumer interacts with a product or service of a certain brand through brand awareness, firms recognise the need to consistently deliver a superior quality product in order to increase customer loyalty, which will result in greater revenue streams and the cost of acquiring new customers will be reduced (Roberts-Lombard 2009:236). Consequently, it is hypothesised that:

To statistically test the hypotheses, the focus of the study was on the franchise fast food industry in South Africa and the sample was drawn from customers at two leading chains from the Famous Brands Group, across the greater Gauteng region. Two leading fast food brands who are both prominent market leaders were selected and thus the sampling frame represents customers purchasing categories of franchised fast food, in terms of hamburger and pizza offerings. The respondents, as well as the participating fast food outlets, will remain anonymous.

A convenience sampling technique was used at the store or outlet level of analysis, with a random sample selection of the sub-population taking place at several franchise outlets. All the targeted franchised fast food outlets were located in Gauteng and operated in either regional retail centres, small regional malls or community centre malls. The site locations were spread across the greater Johannesburg area and included the townships of Soweto and Alexandra, which allowed for sufficient segmentation of consumers from LSM 4 to LSM 10 to be surveyed.

Demographic characteristics of respondents reveal that the majority of respondents (58%) were in the 18–25 age category, while 65% were female. Close to half (48%) of the respondents indicated that they purchased fast foods twice a week and 36% revealed that the most common form of transport used to visit the fast food outlet was public transportation.

Based on the literature review, suitable measures for the purpose of this study were identified where theoretical and empirical support for each construct was evident. Primary data were collected for the various items representing the independent variables (IV) and objective secondary data for the items representing the dependent variable (DV), all of which were based on the conceptualisation of the constructs as discussed in the literature review. Multiple items were used to measure each of the constructs representing the IVs and the DV and these were operationalised as follows.

Franchise site location factors: The following factors were hypothesised to influence consumers’ choice of franchise fast food outlets: accessibility, intensity of competition, convenience and central facilities available. Past research (Frasquet et al. 2001; North & Kotze 2004; Wong et al. 2001) has used elements of the SCATTR instrument to assess retail attractiveness factors and these were adapted in terms of the fast food franchising for use in our study: (1) Accessibility was measured with items such as ‘the shopping centre is easy to access’, ‘the shopping centre has good parking’ and ‘the shopping centre has good security’; (2) Competitive intensity was measured with items such as ‘the shopping centre has many fast food outlets to choose from’ and ‘there are many fast food outlets located within one kilometre’; (3) Convenience of the retail location was measured with items such as ‘the shopping centre is located within one kilometre of work or home’, ‘I choose the closest shop when I enter the shopping centre’ and ‘the shopping centre is well established’; (4) Central facilities within the retail location was measured with items such as ‘the shopping centre has a bank that I use’ and ‘the shopping centre has many shops that I use’.

Franchise brand factors: A combination of brand awareness, brand image and brand attributes scales that have been used in past studies were adapted in context of the fast food franchising for use in our study (Mackay et al. 2013; Terblanche & Boshoff 2010). (1) Brand awareness was operationalised with variables such as ‘my restaurant brand is located within one kilometre of my workplace’, ‘my restaurant brand is easy to locate within the shopping centre’ and ‘my restaurant brand has good visible signage’. (2) Brand image was based on variables such as ‘the brand logo allows me to find my restaurant easily’ and ‘I know what to expect when I see the logo’. (3) Brand attributes was based on variables such as ‘the food is served hot’ and ‘the quality of the food is fresh’. (4) Price and promotion was operationalised with variables such as ‘the food is good value for money’ and ‘prices are competitive’. (The price point and promotional offering that was available at the store level was based on a national level price and promotion basis. Additionally, all national promotions were run within the stores for the same period of time.)

Revenue growth: Although there is no consensus on the appropriate measure of firm performance, researchers have pointed to growth as the crucial indicator of performance (Steffens, Davidsson & Fitzsimmons 2009). It has also been argued that growth is a more accurate and easily accessible performance indicator than accounting measures and hence superior to indicators of financial performance (Steffens et al. 2009). Five-point scales were used for the revenue growth measures, anchored by ‘1 = much more than’ and ‘5 = much less than’. The sales revenue or turnover revenue was collected from audited sales reports submitted from the selected store outlets to Famous Brands on a monthly basis. The period used for the sales information analysis included a year-on-year comparison from 01 March 2014 to 30 November 2014 and 01 March 2015 to 30 November 2015. This allowed for an average trading period of 10 months to be assessed, including both peak and off-peak trading months. The year-on-year growth percentage was calculated through the difference between average revenue for 2014, less the average revenue for 2013. Absolute growth was simply computed as the size at one year minus the size of the previous year.

All items for the IVs were measured along a five-point Likert-type scale, ranging from ‘1 = mostly agree’ to ‘5 = mostly disagree’, where respondents were required to indicate the extent of their agreement with each statement. In some instances items were reverse coded in the scale analyses and the wording was adjusted to reflect a franchise fast food context. Since the study used a self-report questionnaire to capture the individual-level measures at one point in time, common method bias may affect empirical results and conclusions. A number of procedural steps were taken to minimise the risk. In order to reduce socially desirable responses and item ambiguity, the questionnaire featured specific, clear, concise items, with a ‘counter-balanced’ question order and the respondents could choose to remain completely anonymous (Podsakoff et al. 2003:888).

Based on the multi-item scales used, content validity from the customer’s point of view was tested in order to ensure that the instrument demonstrated sufficient face validity. A preliminary analysis via a pilot test was undertaken (n = 14) to ensure that the respondents had no difficulties in answering the questions and there was no problem in recording the data. Finally, the instrument was checked for internal reliability and consistency. Item statistics were calculated for each factor with the Cronbach’s alpha values reported in Table 1, all of which are deemed satisfactory (Cooper & Emory 1995:265).

In addition to the reliabilities already mentioned in the measures section, Table 1 shows the descriptive and correlational statistics. The overall mean scores showed high levels of agreement with each of the variables, where the mean is the midpoint average on the 1–5 Likert scale. Most scores were closely distributed along the ‘1 = mostly agree’ part of the scale and ranged from 1.254 for central facilities to 1.955 for revenue growth. To understand the nature of the relationship between the variables, a correlation analysis was conducted. Pearson product-moment correlation coefficients were calculated for each of the scales and are reported as aggregate scores with levels of significance denoted in Table 1. The correlation matrix shows several significant correlations between the variables that were significantly and positively correlated to each other: accessibility and convenience (r = 0.293; p < 0.01) and central facilities (r = 0.482; p < 0.01). Few significant correlations were detected between the variables and revenue growth, for instance competition (r = –0.403; p < 0.05).

Regression analysis was employed to test the hypotheses, which entailed using stepwise linear regression where initially the variables contributing the most to explaining any variance in the DV were entered, with subsequent variables included based on their incremental contribution over the first variable and based on the criterion that they were statistically significant. Based on this procedure the results show that the only factors that had a significant impact on the DV revenue growth were convenience and central facilities; see Table 2a-b. The regression results show model 1 with a R² = 0.116 and model 2 with a R² = 0.163, which means that these two predictors combined explain 16.3% of variance in the DV. The analysis of variance (ANOVA) section (not shown) shows an F-value of 7.106 which is statistically significant (p < 0.01). In terms of standardised coefficients the beta weights indicate negative t-values of which central facilities is significant at the 0.05 level. An examination of the collinearity diagnostics (not shown) reveals relatively average variance proportions for the variables. These diagnostics when read in conjunction with collinearity statistics show variable inflation factor (VIF) values of 1.084. These figures are deemed acceptable, indicating no major incidence of multicollinearity. When the values are 10.0 or more the regression coefficients can fluctuate widely from sample to sample, making it risky to interpret the coefficients as indicators of the predictors (Cooper & Emory 1995:598).

These regression results mean that only hypothesis 3 and hypothesis 4 are supported, where significant findings provide evidence of a positive relationship between convenience (H3) and central facilities of a retail location (H4) and revenue growth of franchise fast food outlets. Moreover, based on the limited amount of variance (16.3%) explained by the two IVs on the DV, the predictive and explanatory power of this model is limited and a fair amount of work needs to be conducted to further understand the influence of site location factors and brand factors on revenue growth of fast food franchises. Based on the results hypothesis 1, hypothesis 2 and hypothesis 5 cannot be supported.

By placing franchise site location factors and branding in the foreground and by relying on a sample of customers frequenting leading franchise fast food outlets, the findings of this study have important implications for both scholars and practitioners:

The study is limited in its concentration on a single industry – franchised fast foods. Every industry, however, has a different set of factors which might influence their revenue streams. Thus, the findings of this study may not generalise across franchising or industries. Future studies could expand the sample frame across multiple franchise sectors to improve generalizability. Moreover, the study’s measures of franchise performance focused solely on revenue growth as a single criterion. Future studies could consider additional performance measures as well as non-financial measures in evaluating franchise success, to counteract such deficiencies.