Impact of SMEs strategy on loyalty: the hairdresser case

Purpose – This study aims to propose a theoretical framework and provide empirical evidence on the most successful marketing strategiesfor obtainingbehavioural loyalty in smallandmediumenterprises(SMEs). Design/methodology/approach – The data are based on 475 telephone surveys conducted among Spanish hairdressers.The authors have used ordinaryleast squaresto estimate theempirical model. Findings – Pricing, services and communication (Web page and in-store communication) are the main drivers of customer loyalty. SMEs have to be cautious with the use of social networks to avoid damaging loyalty. In addition, those positioned at high-price segments should pay more attention to communication on the Web, and all companies should ﬁ nd a balance between in-store communication and the sale of products for use at home. Research limitations/implications – Further research should try to replicate the ﬁ ndings with data from consumers and ﬁ rms. Practical implications – Service managers need to understand the optimal strategy to succeed in the market. The key insights of this study could also apply to other sectors, such as health, personal care and wellnessservices. Originality/value – Previous research focussed mainly on large companies, while the role of loyalty in the success of SMEs has been poorly studied, with focus only on the antecedents and the measurement of loyalty. This study contributes to the previous research by analysing the effect of the strategy (price, range of services, communication, size and location)inthe achievementof loyalty in SMEs. we of attitudinal the worker attitudinal further price number


Introduction
Customer loyalty is one of the most debated subjects in marketing (Kamran-Disfani et al., 2017;Serra-Cantallops et al., 2018;Watson et al., 2015) and has interest among academics and practitioners (The Economist, 2016) who want to understand the effect of strategy on customer loyalty and performance, given that achieving customer loyalty is a primary marketing goal (Watson et al., 2015).
Building and maintaining a loyal customer base creates multiple potential benefits for firms and brands. Loyal consumers remain with a company because they are pleased with the service and therefore are more likely to buy additional services and spread favourable word-of-mouth communication (Zeithaml et al., 1996). The literature of services state that customer loyalty drives firm performance (Hogreve et al., 2017;Coviello et al., 2006). First, loyal customers amplify sales through frequent repurchases and customer referrals (Kamakura et al., 2002). Second, loyal customers reduce service costs and marketing expenditures because they are familiar with service provider's processes and are reluctant to switch (Anderson, 1998).
In the rich history of loyalty, the empirical research has mainly been undertaken on large organizations, and there is a gap concerning the role of loyalty in the management and success of small businesses (Mostajer Haghighi et al., 2014). The scarce literature on loyalty in the small business services are focussed mainly on the antecedents and measurement of loyalty Johnson, 2006, 2009;Mostajer Haghighi et al., 2014;Russell-Bennett et al., 2007;Wiid et al., 2016;Yim et al., 2008), whereas marketing strategy remains less studied. While some aspects of the strategy on loyalty have been studied as loyalty card programme adoption , brand management decision and brand experiences (Mitchell et al., 2015;Ong et al., 2018), changes in different attributes of a service (Lam et al., 2009) and other components have been left out. Price, store format, product quality, human interactions, store image, among others, are key elements of the whole strategy that have not been investigated on SMEs and affect to retail patronage and loyalty (Gauri et al., 2008;Pan and Zinkhan, 2006;Kamran-Disfani et al., 2017). Johnson (2006, 2009) and (Yim et al., 2008) analyse the achievement of loyalty, from a consumer perspective, in a sector mainly composed by SMEs, the hairdressing salon market. These authors found that trust, commitment and interactions (loyalty to service workers) are key drivers to engage customer loyalty. However, we propose that other factor to explain loyalty, the marketing strategy. Could managers choose a strategy to engender loyalty? This research question remains unanswered because other elements of the strategy, such as pricing, assortment of services and communications, are out of the scope of these studies, at least in their potential role of the components of the strategy such as moderators to generate loyalty. Therefore, this paper adopts a managerial perspective trying to fill the gap and contributing to the previous literature through the analysis of the effect that the SMEs marketing strategy exerts on customer loyalty.
The hairdressing salon market is a business-to-business-to-consumer (B2B2C) model whose core stakeholders are manufacturers, hairdressers and consumers. The hairdressing salon market represents more than 3.5 million salons worldwide. More than 800 million customers make use of hairdressing services annually, making 6 billion visits to hairdressing salons (Machek, 2010). Worldwide, seven out of ten women regularly attend hairdressing salons, visiting them, on average, five times per year. Overall, these figures show the potential of this market (GFK, 2016).
According to Key Stone (2016), in Spain, the hairdressing market employs 200,000 people directly and 35,000 people indirectly. It is highly fragmented with 50 K salons representing the highest concentration within Western Europeone salon per 1,000 inhabitants on average. It is estimated that this business generates around e3,300m per year, 0.3 per cent of the Spanish GDP. In this market, more than 80 manufacturers and importers operate with 140 different brands. From a manufacturer standpoint, hairdressing salon market in Spain accounts for e520m.
We contribute to the previous literature by adopting a managerial view, and with this aim, we ask hairdressers about the strategies they implement and about the consumer loyalty in their salons. Thus, the study has two main purposes: (1) to establish a suitable theoretical framework; and (2) to provide empirical evidence related to the most successful marketing plan in the achievement of behavioural loyalty in SMEs in the hairdressing salon market.

Theoretical framework
A retail strategy is the overall plan guiding a retail firm. It influences the firm's business activities and its response to market forces, such as competition and the economy. The implementation of the strategy involves the choice of factors such as the store location, the range of services, pricing and communications to achieve objectives of the firm (Berman et al., 2018). The fundamental assumption of our set-up is that the goal of hairdressing salons seeks to build loyalty through the combination of different elements of the marketing strategy such as price, customer service such as breadth of services, unisex services and sales of professional products to use at home, loyalty to leader supplier, communication (Web page, social media and in-salon support material), number of employees and location. In Figure 1, we present the conceptual framework proposed in this paper.

Salon loyalty
To measure loyalty, three different perspectives have been used in the literature: behavioural, attitudinal and mixed (Watson et al., 2015). The behavioural perspective focusses on measuring behaviours such as repeated purchasing, which has obvious benefits for a firm's financial performance. Because we adopt a managerial perspective, we measure loyalty from a behavioural perspective (Hult et al., 2017;Larsson and Viitaoja, 2017;Wiid et al., 2016). In the current study, we measure salon loyalty from a hairdresser perspective as the share of frequency of visits of the clients to the salon (Ailawadi et al., 2008).

Pricing positioning
Pricing positioning is a central strategic decision for all companies. Pricing constitutes a critical marketing strategy because it is the only variable that ultimately generates income (Monroe, 1979). The retailer's price positioning is an important factor in store choice and loyalty (González-Benito and Martos-Partal, 2012) and an important element of the retailer strategy (Gauri et al., 2008). Therefore, we expect that it will constitute an important factor for salon strategy to drive loyalty.
According to the classical economic theory, a higher price increases perception of monetary sacrifice, implying a negative relationship between price and the purchase probability. At the same time, for many services a higher price increases perception of quality, and thus it positively affects the purchase probability (Rao and Monroe, 1988). Monetary savings often represent a key motivation for purchase, especially for consumers sensitive to price and promotions. Store loyalty might diminish monetary savings though because customers must waive the advantages of offers and promotions provided by different retailers. Price-focussed consumers who perceive significant benefits from comparing prices across brands and stores likely spread their purchases among various stores to obtain the benefits of the best deal. Therefore, it is common in the retailing industry to find a negative relationship between consumer  price sensitivity and store loyalty (Ailawadi et al., 2008;Martos-Partal and González-Benito, 2013;Pan and Zinkhan, 2006). Moreover, hairdressing services differ from other retailing sector services, in that the former are perceived as a greater involvement in the purchase decision, both in experience quality and intangibility. The prices of hairdressing services are difficult to comprehend as these services cannot be kept, demonstrated, returned or resold. Furthermore, the experience of consumers with the service they receive in the hairdressing salon is intangible, and, as such, it cannot be sold or transferred to a third party. Therefore, we expect higher prices in salon to increase the perception of service quality and thus to have a positive effect on loyalty. Hairdressing salons that are positioned in the high-price segment can generate more confidence and attract a clientele that is less sensitive to prices and, consequently, less willing to change their hairdressing salon. Previous literature also finds empirical evidence that higher prices generate loyalty (Dawes, 2009;Sayman and Hoch, 2014). If customers trust a particular stylist with a high standard quality, they will be willing to pay the price it costs. Hence, we propose as hypothesis that higher price positioning provides greater customer loyalty:

H1.
A salon with higher price positioning engenders more customer loyalty.

Positioning in service and assortment
Product/service selections (or assortment) is defined as the number of different items in a category (Levy et al., 2012). The breadth (number of brands) and depth (number of stockkeeping units) of an assortment offered in a shopping centre help retailer to meet the heterogeneous tastes of their customers (Dhar et al., 2001).
Store formats refer to competing categories of retailers that match varying customer needs and shopping situations, and it is related to store assortment (González-Benito et al., 2005). A format with a higher assortment (such as hypermarkets) can provide a one-stop shopping convenience for their customers and could meet more heterogeneous needs (González-Benito et al., 2005). For example, Finn and Louviere (1990) found that different apparel shopper segments tend to choose shopping centres that they associate with different combinations of features. Shopping centres that provide good services and a wide selection, but with less emphasis on low prices, are more likely to fall into the shoppers' consideration set. Moreover, previous literature in retailing finds evidence of a positive relationship between greater assortment or store format and loyalty (Pan and Zinkhan, 2006;Kamran-Disfani et al., 2017).
Therefore, we expect that salons with greater assortment and more services (such as unisex salons or/and salons with major breadth of services) could meet more heterogeneous needs and show higher loyalty levels. A hairdressing salon has two different sources of income: services and professional products to use at home. Professional products have higher prices than retail alternatives because they have an exclusive distribution and high-tech and high-quality products, and they are highly intensive in capability training and education of hairdressers (Machek, 2010). Therefore, we consider as part of salon assortment both the professional products sold at the salon, as well as the hairdressing services.
Loyal customers are more likely to buy additional services (Zeithaml et al., 1996). Therefore, the number of products or services a customer buys are critical factors that impact on the depth of the customer-company or brand relationship. One way to intensify the depth of this relationship is to increase the amount of retained products (Bolton et al., 2004).
Previous arguments are consistent with the social exchange theory (Blau, 1964); firms which offer more services to customers provide concrete benefits to them and explain why customers engage in repeated exchanges and behave in response to being bestowed with benefits by an exchange partner. Based on the previous reasoning, we propose as hypothesis (H2) that higher service positioning such as unisex services or salon with higher breadth of services and professional products to use at home deliver higher customer loyalty to the hairdressing salon: H2. The number of services offered and customer loyalty have a positive relationship.
H2.2. Salons with a major breadth of services reveal higher customer loyalty.
H2.3. Salons with high sales of professional products to use at home show greater customer loyalty.

Loyalty to leading providers
Peculiarities of these professional products sold at the salon made it difficult for the hair stylist to switch to another provider because of the high switching costs, which will drive higher relational exchanges between the manufacturer and the hairdressers (Zeithaml, 1981). According to Trout and Ries (1986) and the double jeopardy theory (Martin, 1973), history shows that the first brand in the brain gets on average twice the long-term market share of the second brand and, again, twice as much as the third brand. Hence, the leader gets almost all the material advantages. This thought matches one law-line generalization in marketing; the double jeopardy effect. According to double jeopardy effect, larger brands win twice; they have more buyers and these buyers tend to buy more often (Ehrenberg et al., 1990). Consequently, these brands generate brand loyalty, which is one of the key elements of brand equity (Aaker, 1991).
The professional products sold in the salon are important elements in the hairdresser strategy, and the hairdresser makes the decision about the brand to include in the assortment. Consumers, however, visit the salon mainly for the service and stylist expertise and not for the product Johnson, 2006, 2009). Brands with higher equity could drive more customer loyalty for hairdressers more concerned about building image and reputation (Aaker, 1991). If we rely on the image transfer theory (Park et al., 1996), hairdressers must consider that an image transfer could occur between the brand sold and their salons, which would help building salon image and trusting, as it happens with co-branding. Consequently, we expect that salons loyal to leading brands generate higher loyalty.
H3. Loyal salons to leading brands show higher customer loyalty.

Communication strategy: Web page, social media and in-salon communication activities
According to the economy of information theory (Nelson, 1970(Nelson, , 1974, consumers have uncertainties about the attributes and benefits of the products and services they aim to purchase because of imperfect or asymmetric information, which characterises most markets. Consumers can search for information about products and services to compensate for this lack of knowledge. Communication is an informative tool and a key element in the strategy of a firm (Kotler and Armstrong, 2010), and then the hairdresser has to choose right communication policies to inform consumers. Communication causes consumer trust, which stimulates commitment to the relationship and loyalty to the firm (Morgan and Hunt, 1994;Ball et al., 2004).
Nowadays, the first-place search for consumers is the internet. They look for either information about businesses to shop online or simply primary information such as the address or phone number of a hairdresser salon. Therefore, building an online presence is critical for hairdressing salons, not only to share information with consumers, but also to gain credibility and trust. As outlined by Verisign (2013), 56 per cent of consumers declare that they do not trust a business without a website. In addition, today consumers like to connect and communicate with companies more than ever. In this sense, establishing an online presence through a website or social networks is an excellent platform to achieve this goal. Maintaining ongoing communication and providing updates on company initiatives, news and events through digital channels helps companies reach a broader audience to create and strengthen relationships and to build trust, which is critical for the loyalty to the company.
Communication in the store is a key retail mix element (Berman et al., 2018). Moreover, the physical environment is influential to communicate the firm's image and its purposes to customers. In fact, physical setting may also influence customer's ultimate level of satisfaction with the service (Bitner, 1992) and provoke loyalty. In this paper, we focus on insalon communication activities, specifically display and advertising in the salon. Display activity improves in-store visibility, gaining consumer attention to the category and, therefore, positively influences in-store purchase decisions (Dhar et al., 2001).
According to the previous arguments, we focus on the most used communication tools in the hairdressing salon (Web page, social networks and in-store communication), and we propose that the use of those tools (online, offline) ensures higher customer loyalty.
H4. More communication activities and consumer loyalty have a positive relationship.
H4.1. Salons with Web page engender more customer loyalty.
H4.2. Salons with social media generate more customer loyalty.
H4.3. In-salon communication activities lead to more customer loyalty.

Number of employees
Consistent with the social exchange theory, we expect that the higher the number of employees, the greater is the loyalty. Moreover, given that in this sector the average number of employees is low, more employees could serve more customers, which, as time goes on, could get interactions with indirect effects on loyalty. Previous research found that in personal service contexts, loyalty to service workers are key drivers to engage customer loyalty (Bove and Johnson, 2006;Yim et al., 2008;Ganesh et al., 2000). This means that loyalty to the employee translates into behavioural loyalty to the firm, a channel explained by the closeness and interaction of service workers and customers. Thus, we postulate that: H5. The number of employees and customer loyalty are positively related.

Location
One of the more critical decisions that a small retail establishment makes is the choice of its location (Kuo et al., 2002). It is not possible to determine how many businesses with positive potential fail because of the entrepreneur's failure to find an adequate location for the business survival (Scarborough and Zimmerer, 2006). Even minor differences in location can have a significant impact on market share and profitability (Craig et al., 1984). Therefore, the choice of the location of the salon can help making or breaking the business. In fact, location decisions are more difficult to change (in comparison to pricing or services) because retailers make important investments to purchase and develop real state or commit to long-term leases with developers (Berman et al., 2018). These authors distinguish three types of locations: the isolated store, the unplanned business district and the planned shopping centre.
Consumers go to shopping centres to have an experience of leisure, family time and enjoyment (Berry, 1996). The act of purchase is only one of the aspects to consider and the border between commerce and leisure has been diluted. Shopping centres are no longer a place just to buy but also to put the desire for "fun" into the context of successful retail value propositions (Howard, 2007;Wakefield and Baker, 1998). This evidence matches with the motivation theory, which proposes that the value of the shopping/service experience has not only a utilitarian but also a hedonic dimension (Babin et al., 1994). Hedonic benefits are experimental and affective, i.e. consumers appreciate them for their own qualities, not as the means to an end (Hirschman and Holbrook, 1982).
Based on the previous premises, we propose that consumers in shopping centres are less loyal than consumers in any other location. It can be justified by two reasons. One, the motivation of consumption is different because they seek leisure or relaxation rather than a pure functional cutting and colour service in a street salon. Second, their main goal or plan is not to visit the shopping centres salons, but because they are there, they take advantage of the time and enjoy the experience that this type of salons can offer.
H6. Location at shopping centres has a negative relationship with customer loyalty.

Moderating effects
Worldwide, there is a current premiumization trend at beauty markets (Euromonitor, 2017). Thus, the evolution to upper pricing positioning over different salons could have consequences for consumers' motivations to repeat the visit. Hence, the price itself can be one of the factors that moderates the effect of other marketing strategies on customer loyalty. Specifically, we propose that price positioning could moderate the effect on consumer loyalty of the size of the hairdressing salon measured by the number of employees and the effect of Web page communication. Bove and Johnson (2006) find that the number of workers moderates the transference of positive attitudinal loyalty for the service worker to attitudinal loyalty to the service business. The transfer is stronger when the service business has only one or two workers, as these personnel represent the service business to the customer. They argue that in a company with one or two workers it is likely that they may be partners in the business. Even if they are not partners, each one is a larger portion of the "business" than when there is a medium or high number of service workers. We go further and propose that price positioning moderates the relationship between the number of employees and customer loyalty. At price-oriented salons, customers' motivations to keep loyal relate more to low prices than to quality perceptions. In contrast, at high pricedoriented salons, customers' motivations focus more on quality and trust, so the importance of perceived quality may be higher, and the heterogeneity of service could become a problem when the number of employees increase because a higher number of employees could also imply a lack of personalization or lower emotional link with consumers. Each day more, consumers like to be treated in a special and personalized way and their desire is stronger to the extent that they pay more.
Heterogeneity is one of the characteristics of services and concerns the potential for high variability in the performance of services. The quality and essence of a service can vary between producers, customers and time (Zeithaml et al., 1985). Therefore, even if more employees could serve more customers, time can make it translate, first, to employee loyalty and, then, to customer loyalty. In salons with high-price positioning, customers expect high service quality, high employee capability and more experience of the stylist. Customers pay the price for the trust to a particular stylist with a highquality standard. However, as the number of employees increases, it becomes more difficult to keep the same standard across employees, which could damage the loyalty to the firm. This is explained because customers are not going to be willing to pay the same price for stylists with lower quality levels. Then, we expect that: H7. Salons' price positioning moderates the relationship between the number of employees and consumer loyalty, so the relationship between loyalty and size in high-price salons is negative.
On the other hand, because salons must align their price positioning with their marketing actions, we propose that price positioning of the salon moderates the relationship between the most important communication activities to build salon image (Web page) and customer loyalty. Communication is key to build image and engender trust and loyalty to the firm (Morgan and Hunt, 1994;Ball et al., 2004) and Web page is the key tool to do it (Verisign, 2013). Given that customers' motivations focus more on quality and trust in high-price salons, the effect of Web communication to generate loyalty ought to be more important for these types of salons. However, firms and, more specifically small firms, have difficulties to develop integrated marketing communication between online and offline strategies. Therefore, high-price salons are going to have more difficulties to satisfy the needs of consumers that are paying more for services, i.e. consumers that expect the same quality and trust in the salon than in the available information on the Web: H8. Salons' price positioning moderates the relationship between Web page communication and consumer loyalty, so the relationship between Web page communication and consumer loyalty in high-price salons is negative.
Moreover, we propose that the proportion of customers who bought professional products to use at home moderate the effect of in-salon communication on customer loyalty. As previously argued, display activity improves in-store visibility, drawing consumer attention to the category and thereby positively influencing in-store purchase decisions (Dhar et al., 2001). According to Valenzuela and Raghubir (2009), shoppers are more likely to purchase an item when they can easily see from the aisle, demonstrating a need for more visible and eye-catching displays. Therefore, display of professional products to use at home could damage the loyalty to the salon because if more customers buy these products, they are going to use them at home and, therefore, the number of visits to salons will be reduced: H9. Proportion of customers that bought professional products to use at home moderates the relationship between in-salon communications and customer loyalty, so the higher the proportion of customers using products to use at home, the lower the consumer loyalty through in-salon communication activities.

Empirical model 3.1 Research method and measures
We provide empirical evidence of the conceptual framework proposed in Figure 1 in a sector characterized by SMEs, the hairdressing industry. The population of interest was defined as hairdresser's owners or salon managing directors. We conducted a survey in Spain in 2013, whose respondents are a sample of the population of interest. The data were gathered by a marketing research company with high reputation and specialization in the industry. The data collection was Computer-assisted telephone interviewing (CATI). We used a stratified sampling method across Spanish regions. We obtained 475 completed questionnaires and validated 463 questionnaires after dropping observations with missing information across multiple items. Sample size represents approximately 1 per cent of total Spanish salons (53,061) according to Key Stone (2016) and are representative of the salons market population across Spanish regions. It was found that 97 per cent of the Spanish salons were not affiliated at a national chain and the average sample salon age is 10 years.
In the survey we use single item measures for two reasons. First, we try to minimize respondent refusal given the difficulty to get responses by owners and directors of firms. Second, because we measure concrete singular objects and attributes, it is recommended to use single item measures instead of multiple items measures for this kind of construct (Bergkvist and Rossiter, 2007;Bergkvist, 2015). All items and sources used to build the variables and their descriptive statistics are reported in Table I.

Loyalty to the salon measure
We adopt a behavioural perspective on service loyalty. Specifically, we follow Ailawadi et al. (2008) and focus on the share of shopping trips to the salon. The item used was: "Considering all your customers, what percentage would have weekly frequency?" Therefore, the variable "loyalty" represents the percentage of weekly visits (over total visits) to the hairdressing salon by an end-consumer as declared by the hairdresser. This variable is thus bound between 0 and 1, and the mean value is 0.23.

Pricing strategy
The "pricing" variable was seized from the following item: "What is approximately the average value (in e) of the ticket in your salon for hairdressing services and products?" We imputed missing cases (30) of this variable by calculating the weighted average ticket at the Spanish province where the salon belongs. The average price per salon is e23.5.

Services strategy
"Unisex" measures hairdressing salons with unisex services using the survey question: "What kind of clients do you attend to in your salon?" The variable takes three possible answers (100 per cent men dedicated; 100 per cent female dedicated or unisex), which were transformed to a scale (1, 0) being 1 unisex and 0 "not unisex". Of the sample, 87 per cent are unisex salons.
The "service" variable measures breath of services and it was built using the survey question: "Do you have the following services at your salon?" The hair-care services asked in the survey were: hair colour, hair beauty treatments including nutrition, hydration, volume, repair, specific treatments for hair care (fall, dandruff, grease, etc.), haircut and styling and straightening/permanent. The aesthetic services include waxing, manicure, pedicure, massages, nail reconstruction, anti-aging treatments, make-up services and other aesthetic treatments. Answers to previous questions were transformed to (1, 0) dummies being 1 Yes/0 No. Once transformed to a dummy, we compute the variable as the sum of hairdressing salon services multiplied by x , (x being the percentage of hair services invoice sales over total salon sales as declared by respondents to the survey) plus the sum of all the aesthetic services multiplied by 1-x . The variable has an average sample value of 4.48. Sales of professional products to use at home were measured by "product purchase" variable. The item was: "Could you tell me what percentage of customers who come to use salon services also buy the products?" On average, 12 per cent of consumers purchase professional products to use at home.
Hairdresser loyalty to leading supplier is measured by the variable named "leader loyalty". One common measure of behavioural loyalty considers the expenditure Adapted from Bove and Johnson (2006) distributions across stores/brands and define the preferred store as the store that gets the greatest expenditure from the consumer (Bustos-Reyes and González-Benito, 2008;González-Benito and Martos-Partal, 2014). Therefore, we compute the variable considering the percentage of purchases in professional products that hairdressers acquire for their salon to measure behavioural loyalty to the supplier. It was computed considering the item: "According to volume of purchase: who is your main supplier?" There is a high market concentration from suppliers as the premium segment accounts for 41 per cent of the market ($219m). This is the major battlefield to compete where L'Oréal Company and Coty Company together account for 51 per cent of the total premium segment. Overall L'Oréal is the market leader with 34 per cent share, while Coty is the second player (17 per cent) (Kline, 2016). Therefore, we consider L'Oréal as the leader supplier brand and, then, we distinguish between salon loyal to the leader brand, then when the salon points out as main supplier to L'Oréal, and in this case the leader loyalty variable takes a 1 value and 0 otherwise. 23 per cent of the sample has L'Oréal as their main supplier.

Communication strategy
The survey items related to Web page and social media activity use dichotomous answers: "Yes/No". For Web page communication, we compute the variable "Web page". This variable captures the results of the item: "Has the hairdressing salon a website?" and for social network communication we build the variable "rrss". This variable is computed using the survey question: "Has the hairdressing salon activity in social networks, such as Facebook, Twitter, etc.?" In both cases, the answers were transformed to a scale (1, 0), "Yes" to 1 and "No" to 0. In-salon communication activities are computed in the "animation" variable, which capture the animation effort in the salon. We measure the effort on a tenpoint Likert scale (from a minimum effort of 1 to a maximum of 10). We find that salons made more effort in-store communication (6.90 on average), whereas the internet presence is low (18 per cent of the salons have Web page and only 7 per cent have social network presence). The size of the hairdressing salon was measured by the variable named "employees". It was constructed considering the number of full and part time employees attending to answers to the following two questions: "How many full-time employees, owner or manager included, do you have in your salon?" and "How many part-time employees, owner or manager included, do you have in your salon?" We use the common rule in the sector (Key Stone, 2016) to have a comparable base on equivalent employees. A part-time employee represents one third of a full-time one. So, the final variable is computed as the weighted sum of full-and part-time employees declared by hairdressers. The average size of the salon in the sample is 2.36 employees.
Finally, location of the hairdressing salon is measured by "location", which is calculated from the answers to the following item: "Where is your hairdressing salon located?" Four answers were possible: street level, shopping centre/mall, apartment or other. Because we are interested in the effect of location at shopping centres, the previous answers were transformed to a (1, 0) dummy being 1 if the salon is located at shopping centre and 0 otherwise. Of the salons, 3 per cent are located in a shopping centre.

Analysis and findings discussion
Our dependent variable, p i , is defined as the share of weekly visit of customers to hairdressing salon i and, therefore, it measures the per cent as estimated by the hairdresser. The independent variables relate to marketing strategies. Because our loyalty measures range from 0 to 1 we transform the dependent variable to log-linearize the relationship between the dependent variable and the explanatory variables, where ln (p i /(1Àp i )) represents the logit transformation of the dependent variable and we estimate the model using ordinary least squares (OLS). In the 17 cases in which p i was equal to 0, we approximate the log by 0.01 to preserve the full sample. Thus, we assume a zero-weekly frequency of visits to a salon as an error in the variable. Then, our specification is expressed as: Where a is a parameter estimating the average share of weekly visits to hairdresser salons; X i is a vector of variables measuring the marketing strategies of hairdressing salon i; and b is a vector of the parameter to be estimated, capturing the effect of the marketing strategies on the proportion of weekly visits. As usual, we assume that, on average, the model does not make mistakes, i.e. the error has zero mean. In Table II we report the estimation results. We estimate two specifications. A first model (model 1 in Table II), where vector X i is composed by pricing, unisex, service, product purchase, leader loyalty, Web page, rrss, animation, employees and location. This model allows us to test H1-H6. Model 2 in Table II is proposed to additionally test the moderation hypothesis (H7 to H9). This second specification adds to the previous determinants of loyalty, the three moderator variables (pricing Â employees; pricing ÂWeb and animation Â products purchase). "Pricing Â employee", "pricing Â Web" and "animation Â products purchase" are the interactions of "pricing" and "employees", "pricing" and "Web page" and "animation" and "products purchase", respectively.
Before presenting the results, we analyse if collinearity is an issue in our model. We test for multicollinearity using the variance inflation factors (VIF). The VIF identify both correlation and its strength. The diagnostics in our estimated models does not reveal important multicollinearity problems (all VIFs are below 1.54 in model 1 and below 1.75 in model 2).
The direct effect of salon strategy through "pricing" on customer loyalty in the simplest model is positive and significant. Therefore, salons positioned at higher prices do generate more customer loyalty in support of H1. Hairdressing salons positioned at high prices could generate more confidence and attract a less price sensitive clientele, and, therefore, they will be less willing to change their salon. While superior product performance translated into higher price services must be a concern, consumers wishes would seem to be related more to final look or services satisfaction. Therefore, higher prices positioning could help to increase consumer confidence and become a powerful marketing strategy for retaining customers.
We find a significant and positive relationship between unisex and loyalty, providing support to H2.1. Nowadays, specialization is crucial and critical as a benefit to deliver prestige to the hairdressing salon. The raw data show that most Spanish salons -87 per centare unisex and the estimation results confirm this marketing specialization strategy to increase loyalty. We also find a positive and significant coefficient between customer loyalty and service variable, supporting H2.2. Moreover, the standardized coefficients in Table II show that the breath of service is the strongest force driving loyalty (0.193), pricing being the second (0.178) and unisex the third one (0.146).
Our results do not support that the service of sale of professional products to use at home generates customer loyalty (H2.3). These results could be explained by two core drivers explaining the cannibalization from professional market to retail market: (1) According to the recessionary environment during the past years, the price gap between professional products and products sell at the retailer have increased. (2) The development of online channels (e-commerce) has cannibalized part of the volume that consumers purchased at the salon (offline) and, therefore, they have now access and purchase these products at lower prices.
H4 proposes that hairdresser loyalty to leading supplier brands drives customer loyalty, but our results do not support it. Customer loyalty to the salon does not seem to be generated by hairdresser fidelity to leading manufacturers (L'Oréal). Therefore, consumers visit the salon for the service and hairdresser expertise, and the hairdresser decides the brands used in his salon Johnson, 2006, 2009). Another possible explanation could be that there are brands belonging to other companies that also are symbols of prestige and quality performance. Therefore, hairdressers could change to others market players without any cost to the loyalty of their consumers. Regarding communication strategies, our results support H4.1 (Web page) and H4.3 (in-salon communication), whereas H4.2 (social media) is not supported. We find a positive and significant relationship between loyalty and the use of Web page by the salon; this insight constitutes an opportunity for salons that do not currently use it or those underusing this marketing tool.
Regarding social networks communication, our finding shows opposite results to our proposal in H4.2 (results in line with Trainor et al., 2014). Thus, social media presence damages customer loyalty. Although digital communication is a reality that hairdressers cannot ignore in business, it looks like current social media digital strategies are not engendering customer loyalty. We can explain this result because despite thousands of hairdressing salons have opened profiles in social networks, most of them do not know how to manage them. They understand the need to have presence in social media, but it looks like they do not use it as an additional tool to the business. Therefore, we can summarize that the path to digital failure is to have presence without an integrated strategy, without providing value, with no coherence and integration with the business itself and with no link with the offline strategy.
The estimated coefficient of the variable animation is positive and significant, thus supporting H4.3. Therefore, the investment in store communication looks a good strategy to drive loyalty. Moreover, the standardized coefficients in Table II show similar importance across communication to engender/damage loyalty. The estimated coefficient of the number of employees is not significant, so, it does support H5. The number of employees is not related to behavioural customer loyalty. One possible explanation for this result is the small variation of the variable impeding a proper identification of its effect. According to descriptive statistics, 95 per cent of the salons have a number of employees in the range [1,4,5].
We do not find support for H6 in our data. Location of the salon at shopping centres does not have any effect on customer loyalty. This result is consistent with Jones et al. (2003), who provide empirical support that location is less important for small convenience stores offering less standardized and more personalized goods and services. Moreover, Bove and Johnson (2009) state that convenience, location and accessibility are the most frequent reasons for changing stylist and salon. Therefore, hairdressing salons at shopping centres do not generate less consumer loyalty. Consumers choose the salon considering a convenient location and for some consumers shopping centres could be a good one, as they make "shopping" while they find time for themselves.
To provide empirical evidence for the moderating effect of salon strategy on customer loyalty, we analyse the estimated results under the heading model 2 in Table II, which includes the interactions already mentioned. We propose that higher priced salon diminish the relationship between the number of employees and customer loyalty (H7). However, we find that the interaction of price with employees has a negative but insignificant coefficient and it does not support H7. Price does not moderate the relationship between number of employees and customer loyalty. This result implies that heterogeneity of the service and the lack of training in small firms are not important issues in this sector.
Regarding H8, we propose the relationship between Web page communication and customer loyalty is negative in high-price salons. The interaction variable has a negative and significant effect, thus supporting H8. Therefore, having online presence without an integrated strategy with the offline strategy damage loyalty for salons with higher prices because the website could not be signalling enough differentiation between higher price and lower prices salons, which could be damaging the trust towards the salon. Considering that an important reason of Web page communication strategy is to differentiate and build image vs core competitors, luxury or prestige hairdressing salons with no differences in communication hurt the loyalty of their consumers (Seringhaus, 2005).
Finally, we propose that higher proportion of customer using products to use at home, diminish the positive relationship between in-salon communication activities and customer loyalty (H9). In this case, the interaction variable has a negative and significant effect, so, our finding supports H9. Hence, hairdresser needs to be cautious with the sale of professional products to use at home. If they emphasize the communication on these products, they could lose the customer loyalty to salon as a result of having a lower frequency of visits. This is explained as these products could be used as substitutes for visits to the salon.

Conclusions and implications
This study proposes a theoretical framework and provides empirical evidence related to the most successful marketing strategies in obtaining behavioural loyalty in SMEs. The empirical results show that pricing, services and communication (Web page and in-store communication) are key elements of the strategy, which help in the consecution of customer loyalty. On the other hand, SMEs need to have caution in the way of using social networks to communicate with consumers to avoid damage to consumer loyalty. Moreover, SMEs positioned in higher price segments could damage their loyalty when they do not get enough attention to Web page communication. In addition, SMEs need to keep a balance between sales of products to use at home and in store communication to prevent hurting loyalty.
SMEs positioned at high prices can generate more confidence and attract less pricesensitive clientele and, therefore, consumers that are less willing to change. Second, service positioning is a key element to keep customer loyalty, firms with higher number of services allow serving more consumer needs, which implies higher loyalty. About communication and beyond the "word-of-mouth" offline strategy needed by SMEs, they must focus on communication strategies online and offline to retain customers and to attract new ones. Online presence is key to engendering loyalty while salons need to devote further efforts in the way they use social networks. The key for using social media communication for top-ofmind awareness is to deliver content that is consistent, relevant and helpful. Online communication should be linked to offline communication with special attention to in-salon communication. The effective combination of these two strategies, as well as best-in-class execution, could increase customer loyalty. Finally, the training to the staff is key to keep loyalty to the firm. Capability and capacity strategies to staff will ensure and increase crosssales via professional products sales to use at home, amplify the breath of services and optimize social media strategies to overall increase customer loyalty.

Managerial implications for manager of small and medium enterprises
Our findings drive interesting implications to hairdressers and to other sector characterized by SMEs. While loyal consumers go to the firm for the service and employee's expertise, also marketing strategies affect customer loyalty and customer loyalty build up is a sum of multiple effects. As a potential watch-out, we want to highlight that even though the high-price predisposes the consumer at first, repeated experience will lead her to dismiss her firm in the mid/long term, as the price is not in line with her service quality expectations. A high-quality service perceived at a high price accordingly would be the perfect combination. Therefore, quality service can afford to be marketed in a higher price segment. Moreover, the manager needs to make good decisions about the service offered in the firm because, as the number of services offered increase, the cost could also increase, and it could be covered from a wide base of loyal consumers. Moreover, it would also constitute a mistake not to offer customers everything they demand in relation to services and sales of products to use at home.
Managers also need to focus on engendering digital communication capabilities because the level of knowledge in digital marketing could make a clear difference within SMEs. Digital marketing is much more profitable than traditional marketing because it needs few resources to generate measurable results. Moreover, it is a convenient tool to attract the youngest audience with service proposals adapted to them. The cost of SMEs not being at social media is higher than that of being; the non-presence in these platforms/social networks could generate distrust to the consumer and it could suggest a lack of closeness. At the same time, the manager can benefit from the positive opinions and word-of-mouth. To do so, the manager needs good capabilities on digital communication, or he could externalize communication activities to satisfy consumer needs in the new digital era.
Another issue that managers should not forget is in-store communication, which is the last opportunity to communicate with customers to take a last-minute service consumption option. SMEs require agile communication media and capable of attracting attention: a good product or service deserves a good exposure. Firms should determine the location of the different communication actions chosen but they should also be concerned with easing orientation and increasing the average customer's ticket without damaging the brand-equity of the firm.
Regarding sale of professional products to use at home, firms should understand the changing and competitive economic environment where profits may decrease so they need to find sources of additional economic growth. However, they also must balance the impact on damage loyalty when the firm made a big effort in store communication about professional products to use at home. By applying direct marketing techniques, a proper training of the firm team and an increase of the promotion of the products intended to sale could easily facilitate the sale of these products without damage loyalty.

Limitations and future research
This study also suffers from limitations that suggest further research. First, one important difference between the hairdressing salon market and mass distribution market is the difficulty to obtain scanner data with real consumer purchases. This has been our main reason for conducting a survey. Moreover, we did not have access to objective income data from the firm. Future research containing this variable should analyse the effect of the strategy on firm's profitability. Second, in this paper we adopt a managerial perspective to analyse loyalty. The proposed model does not reflect the reasons why end-consumers motivate choosing a firm. Further research could adopt both the manager and the consumer perspectives to analyse all the drivers of customer loyalty (Hult et al., 2017). Then, it would be possible to analyse geodemographic, socioeconomic and psychographic characterizations of different loyalty profiles, for instance. Moreover, it would be possible to study alternative methods of exceeding consumers expectations at salon such as the physical setting and spatial layout of the store (Lee et al., 2005) or how "servicescape" (Bitner, 1992) could impact consumer loyalty. Third, in this paper we focus on the number of employees; however, it would also be interesting to analyse the level of training of employees. Fourth, to generalize our findings, future research could test the proposed model on data from other countries or data from other sectors with similar characteristics.