The role of brand equity in a new rebranding strategy of a private label brand

Abstract The purpose of this study is to understand the effects of the rebranding process on private label performance output, namely, on brand equity. More specifically, the study aims to investigate the performance of brand equity constructs (brand awareness, brand associations, perceived quality and brand loyalty) before and after the rebranding process. A questionnaire was administered to 466 shoppers, who put forth their perceptions, of the brand’s image, before and after the rebranding. When analyzed altogether, brand equity constructs have not suffered significant changes, meaning that previously existing brand equity had successfully been transferred onto the new brand; however, new brand equity was not created in the process. Consumers do still associate the private label brand image with its previous brand identity. Nevertheless, their ability to easily identify the various brand tiers was improved with the new image. Consumers associate the rebranded image with being “innovative” and “original” and describe the previous image as “inexpensive” and “trustworthy”. Brand awareness and loyalty are the factors that relate the most to consumers’ perceptions of the brand before its rebranding. This study contributes to the brand management literature by providing a new look into the under-researched problem of rebranding and brand equity, empirically validating the real-life market case.


1
The role of the brand equity in a new rebranding strategy of a private label brand 1. Introduction Over time, private label brands (PLB) have witnessed an impressive evolution in terms of geographical proliferation, product category expansion and market share growth (Cuneo et al., 2012;Dawes & Nenycz-Thiel, 2013). The diffusion of PLB has been so widespread, that brands, once known as low-price/low-quality, commodities, now require special branding attention (Nielsen, 2014;Sutton-Brady et al., 2017). PLBs have thus identified a need to update their brand image, which has in turn led to the implementation of rebranding strategies.
Rebranding has become a way of ensuring that market relevance is kept, through name, visual identity or positioning changes to accommodate varying requirements in business contexts (Miller et al., 2014). However, rebranding requires extensive financial effort and organizational resources with no guarantees of success forthcoming.
Brand monetization and valuation of the brand portfolio have undergone substantive scrutiny in both the literature as well as marketing practice (Davcik et al., 2015). Managers are facing rapid market changes and competitive pressure has ensued, with the proliferation of brand portfolios and the spreading of brand equity. The sources of these changes and pressures are to be found in the volatility of consumer demand (Collange & Bonache, 2015), external stakeholder co-creation influences (Markovic & Bagherzadeh, 2018) and inequality of resources distribution within the firm and across its brand portfolio (Sharma et al., 2016, Davcik & Grigoriou, 2020. However, the literature is scarcer on how the rebranding process and brand value repositioning may affect consumer demand and internal resource distribution. For instance, Lidl in Sweden has struggled over the years with negative consumer perceptions as a cheap and poor-quality retailer label. This all changed in 2012. They came with a new rebranding strategy, with a deliberate repositioning focus, aimed at changing cheap into a highquality retailer. This meant substantial investment in employee training and store layout and atmosphere. This rebranding strategy resulted in the attraction of 95.000 new households to Lidl´s custom, customers that had not been available before (Lindberg, 2015).
The literature on rebranding in the main is anchored on case studies from several managerial or geographical contexts and situations, relying on limited or no empirical evidence as to its overall effects on the organization and its brand portfolio (Miller et al., 2014;Roy & Sarkar, 2015). A dominant body of rebranding literature focuses on corporate rebranding (Miller et al., 2014;Muzellec & Lambkin, 2006;Merrilees & Miller, 2008) or institutional rebranding (Stuart, 2018) strategies. However, another stream of rebranding research, consumer-centric rebranding literature (Pepe et al., 2011;Collange & Bonache, 2015) remains limited and scarce. Roy & Sarkar´s (2015) study is a rare example in the literature, one which looks precisely into the effects of consumer-based brand equity of a brand before and after rebranding. In establishing an academic and practical case for the current paper, it is acknowledged that there are works on customer-centric rebranding, but not much has been found on the contrasting of pre-and post-rebranding. For instance, Ibeh et al. (2005) on internet branding provide mainly anecdotal evidence of rebranding intending to suggest best-practices that point to collaborative and customer-centric e-branding. Saraniemi (2011) andKonecnik andGo (2013) criticize destination branding studies as being too customer-centric whilst, Ahonen (2008) acknowledges that branding in SMEs has adopted a customer-centric product branding view, and Berry (2000) asserts that branding service is a guarantee of a consistent, uniform level of service provision and that a customer-centric brand image is thus a basis for building long-term trust.
In line with this research paradigm, the current study aims to expand the domain of brand equity, by studying rebranding and consumer-centric aspects of these business processes, beyond the limited findings available in the current marketing literature. The brand equity research domain in marketing literature is well known for its two dominant research streamsthe consumer-and financial-based brand equity research paradigm (Davcik et al., 2015). In the current study, the authors are focusing on a consumer-based research stream because the consumer-centric rebranding literature is scarce and inconclusive, and addressing the research gaps presented requires focus from the consumer side. Considering the managerial context of this study, its overall objective, is thus, to evaluate the effect on the brand equity of rebranding a private label, herein termed for the purposes of the current research, "Major Portuguese Retailer" (MPR) 1 through a consumer-centric perspective. MPR is used as the case under scrutiny, as it is a major retail brand in Portugal, encompassing a wide range of products and holding a significant market share in Portugal. The purpose of this study is then, to understand the effects of the rebranding strategy on PLB performance, namely, on brand equity. More specifically, the study investigates the performance of brand equity constructs before and after rebranding. The study uses a consumer-centric approach in its investigation and applies well-known consumer-based brand equity factors, such as brand awareness, brand associations, perceived quality and brand loyalty (e.g., Aaker, 1991;Yoo & Donthu, 2001) in the rebranding context of the existing retailer and its brands. This study aims to contribute to the brand management literature, by providing a new look into the under-researched problem of rebranding and brand equity and doing so by using real-life data and empirically scrutinizing a relevant retailer context. None of the works 1 Due to the privacy requirements, we will use the name Major Portuguese Retailer (MPR) instead of the true brand name we have studied. stated above, look at rebranding from a brand equity perspective, as to what happens to brand equity following a rebranding exercise and that constitutes our academic and practical contribution.
The manuscript is organized in several sections. The next section deals with conceptual development and the literature review. The third section delves into brand equity dimensions, rebranding processes. The research hypotheses are derived in this section of the study. The methodology is presented in section four, followed by data analysis and results in section five.
Confirmatory factor analysis of longitudinal data is used to test the hypothesized relationships.
The final section discusses the findings and conclusions.

Private Label Brands
Store brands, own labels, own brands, retailer brands, private brands, and home brands are some of the many designations that private label brands (PLB) have been known for overtime.
Much like its designations, PLBs have had several different definitions over the years, however, they are commonly known as super and hypermarkets' brands and products, sold exclusively on their stores, alongside other brands (Sutton-Brady et al., 2017).
As a result of continuous PLB growth, retailers have been betting on different strategies to maintain PLB consumption, focusing on quality, design and marketing efforts, thus acquiring a competitive advantage, which has gradually allowed them to market their products better, with incisive impacts on profitability (Cuneo et al., 2012). PLBs have registered a noteworthy growth over the last few years, reaching "a large number of countries, sectors, and product categories" (Cuneo et al., 2012: 428). According to Nielsen's (2014) report, PLBs show better outcomes in developed countries, specifically in Europe, North America, and Australia. At the time, Portugal had the 5th highest PLB share, with 33% of the retail market, whilst the leading country was Switzerland with a 45% share.
Most PLB literature points out that whilst PLBs were initially attractive because of their lower price points, nowadays, these brands are seen as equivalent (or, in some cases, better) than the national brand (NB) (Dawes & Nenycz-Thiel, 2013). 70% of Nielsen's (2014) sample of respondents indicated that they believe PLBs are a good alternative to NBs and 71% said store brands have significantly improved their quality over time. PLBs have proven that they are no longer a low-price low-quality option and can successfully compete with NB. In many instances, PLBs may indeed offer premium quality products in demanding markets (Cuneo et al., 2012).
Despite an unequivocal positive evolution of PLBs, some scholars believe that there will always be a gap between PLB and NB, specifically with regards to differentials in quality and perceived risk between the two alternatives (Dawes & Nenycz-Thiel, 2013;Pepe et al., 2011).
Even though consumer trust has been conquered over time, consumers may still be reluctant when it comes to buying PLB when a NB is available for the same functionality (Batra & Sinha, 2000). It is precisely because of this why branding strategies have become particularly important for PLBs.

Rebranding
Over the years, there have been increasing numbers of examples of organizations using rebranding as a way of ensuring that they remain relevant in the market (Miller et al., 2014;Lindberg, 2015). Most rebranding literature is based on case studies from various managerial or geographical contexts and situations (Miller et al., 2014), but are rather flimsy when it comes to empirical evidence of their effects and outcome. The term "rebranding" was much more common in the business press than in the academic literature (Muzellec & Lambkin, 2006), so for a while, arriving at a conceptual definition of rebranding was proving to be difficult. Many other terms are used to designate rebranding, often according to the author's definition of the concept. For instance, brand revision, brand repositioning, renewal, makeover or even reinvention, are some of the terms used (Hankinson et al., 2007;Miller et al., 2014). Other existing definitions allude to the need to change some or all of the tangible and intangible attributes of a brand (Daly & Moloney, 2004) or even the need to internally change the organization and its employees rather than just its visual identity (Hankinson & Lomax, 2006). A significant amount of rebranding literature is based on an analysis of case studies in diverse settings and contexts. Some authors have used case studies to develop a better understanding of the rebranding process, which can be applied in different industries and rebranding categories (Muzellec & Lambkin, 2006;Finney & Scherrebeck-Hansen, 2010).
Most of these case studies focus on the drivers of rebranding, emphasizing which contexts have led to the need to change an already developed brand (Lomax et al., 2002;Lindstrom & Andersen, 1999;Merrilees & Miller, 2008). Others, however, go further and also analyze rebranding enablers, barriers, and outcomes (Miller et al., 2014;Vallaster & de Chernatony, 2006;Davis & Dunn, 2002;Daly & Moloney, 2004), attitudes and staff (Hankinson et al., 2007). Understanding the whole process of rebranding helps in comprehending, and possibly even forecasting, which rebranding strategies will be successful, thus enabling for significant managerial implications to be derived (Miller et al., 2014).

Brand Equity
Increasing brand equity is generally the goal of organizations that choose to undergo a rebranding transformation (Goi & Goi, 2011). Brand equity is a vital concept in marketing, management and branding studies (Baalbaki & Guzmán, 2016) since higher levels of brand equity are often translated into higher cash flows and increased competitiveness (Yoo et al., 2000;Pepe et al., 2011).
Two focal brand equity perspectives have been recognized in the literature -financialbased and consumer-based brand equity (Davcik et al., 2015). Financial-based brand equity focuses on studying and measuring financial value as the measure of brand success (Davcik et al., 2015;Veloutsou & Guzman 2017). In contrast, the consumer-based brand equity paradigm represents the extent to which a consumer response is affected by the brand (Keller, 1993). The current study investigates the consumer-centric rebranding literature, and, therefore, the authors are focusing on the consumer-based brand equity research paradigm. The branding literature shows that brand equity has a positive influence on stock market responses (Lane & Jacobson, 1995), brand reputation (Liu et al., 2017), market share (Sharma et al., 2016), consumer preference, brand choice, purchase intention as well as consumer willingness to pay (Yoo et al., 2000;Lee & Leh, 2011). The main purpose of managing brand equity is in attracting new customers, maintaining existing ones and using it as a mechanism for effecting an emotional connection between the customer and the brand (Lemon et al., 2001).
Measuring the value of a brand however continues to be a challenge for organizations (Muzellec & Lambkin, 2006;de Oliveira et al., 2015). Numerous brand equity models and scales have been developed and published in the literature, but, research shows there to be no consensus as to which model better estimates brand equity (Davcik et al., 2015). For instance, Biedenbach and Marell (2010) in the B-to-B context show positive effects of customer experience on brand awareness, associations, perceived quality, and brand loyalty. Iglesias et al. (2019) provide evidence in service context on how sensory brand experience has positive effects on brand equity through customer satisfaction and commitment. Models, such as Aaker's (1991) and Keller's (1993) can be applied to most cases, contexts, and industries, although scholars admit they do not capture the complexity of every case and lack empirical verification (de Oliveira et al., 2015). Muzellec & Lambkin (2006) proposed the rebranding paradox concept, where they explore the likelihood of losing equity when significant components of a brand are altered. The authors have focused mostly on brand name changes though. Marketing and brand equity literature often endorses the idea, that only through continuous investment on the existing brand can one create a strong brand (Sharma et al., 2016). This paradox seems to indicate that the concept of rebranding may in effect be counterproductive and inconsistent with the concept of brand equity (Muzellec & Lambkin, 2006). Brands are organizational assets that may have taken years to be created and developed, and if changes are made, this means, abandoning existing brand values. By abandoning these assets, firms risk losing the positive association between the brand and its key stakeholders.
The rebranding phenomenon is, therefore, a risky exercise with no guarantee of success (Miller et al., 2014).
When defining brand equity, Aaker (1991: 24) stated that "If the brand's name or symbol should change, some or all of the assets or liabilities could be affected and even lost, although some may be shifted to a new name and symbol". Muzellec & Lambkin (2006) suggest that the main goal and challenge of a rebranding strategy is to successfully create or transfer brand equity by transferring positive associations to the revised brand, whilst creating new ones.

Brand equity dimensions, rebranding, and research hypotheses
Brands have been conceptualized as the mental perception of the consumer -"brands are what people perceived them to be" (Baalbaki & Guzmán, 2016: 230), which emphasizes consumers' role and power in developing brand equity (Keller, 1993). For this reason, it is necessary to analyze the concept of rebranding from a consumer perspective. Thus, in order to evaluate MPR's own-brand rebranding results, the consumer-based brand equity model proposed by Aaker (1991) and later added to by Yoo et al.'s (2000) was used. Aaker's (1991) four brand equity dimensions will serve as the basis for the research model and hypotheses development, as the goal of the research is to understand and evaluate which factors affect consumer behavior, following the rebranding process.

Brand awareness
Brand awareness consists of a person's ability to recognize or recall a brand, its name or symbol (Aaker, 1991). Walsh et al. (2008) show that the ability of an individual to remember a brand, when in contact with it, amidst other brands, is seen as "recognition", whereas, "recall", or unaided awareness, refers to the ability to remember a brand without any mention of it.
Brand recognition, brand recall, top-of-mind and brand dominance are all brand awareness factors, that influence brand equity, irrespective of whether or not, the brand is part of the consumers' pool of possible choices (Aaker, 1991;Liu et al., 2017). Brand awareness shows different levels of impact according to the brand: the objectives for recent brands may be to increase recognition, whilst the goals for popular brands may have more to do with recall (Aaker, 1996). Awareness impacts consumer choice, perceptions, attitudes and loyalty (Aaker, 1996;Huang & Sarigollu, 2014). Consumer brand awareness influences decision-making processes, as brand notoriety is linked with a higher likelihood of a brand being part of a set of options, which in turn ultimately has an impact on market outcome (Huang & Sarigollu, 2014).
Without this construct, none of the others can exist, since there needs to be a node in an individual's memory to which all other perceptions can attach to (Aaker, 1991). Aaker (1996) emphasizes the gap in the brand awareness literature, by concluding that most brand awareness measures refer only to brand name, but other attributes are key in understanding the value of a brand. For instance, a brand's visual identity and symbols are very significant brand awareness influencers. Hence, it makes sense to conclude that the transformation of the private label brand image may induce a change in the brand's awareness.
Thus, we postulate that: H1: Private label brand awareness is positively affected by its new image

Brand associations
Brand associations are typically linked with a specific memory about a brand and are often referred to as the consumers' notion of the brand (Aaker 1991). Associations are an important concept since they often influence the consumers' decisions (Romaniuk & Nenycz-Thiel, 2013;Swoboda, 2016). Consumers' memory stores information in nodes that can represent several concepts such as attributes, brands or products (Krishnan, 1996;Huang & Sarigollu, 2014).
These nodes are organized in a network form and the connections that define this network are associations. The number of associations and their strength depends on the consumers' perceptions of all the stimuli they are exposed to (Swoboda et al., 2016).
For the purpose of this study, the focus of associations has been on the brand and productrelated attributes. In subsequent sections and analyses, the attributes chosen to measure brand associations are attributes that can be associated not only to the brand but also to the brand's products and packaging. Since MPR's own-brand rebranding, has focused mainly on the brand's products' packaging, it was essential to also consider attributes that link with the visual imagery of the brand.
Colors, shape, text, logos, symbols, names and images are all stimuli categorized within product packaging and represent a direct communication channel between the brand and the consumer. These stimuli create associations on the consumer's minds, which refers to the quality and value of a product (Ampuero & Vila, 2006;Wu et al., 2011;Beneke & Carter, 2015). Due to the relevance of brand image in brand associations, it is possible to infer that: H2: Private label brand associations are positively affected by their new image

Perceived quality
Perception can be conceptualized as an interpretation, arrangement, and selection of stimuli which then becomes part of the individual's view of the world (Solomon et al., 2006). Quality perception has been defined as the consumers' judgment about a product's overall perceived performance (Zeithaml, 1998). This construct is positively related to the consumers' experiences with the brand. The higher the perceived quality, the better the consumer's interactions with the brand (Yoo et al., 2000). It is also considered a component of brand value since consumers with positive experiences are more likely to prefer the brand over other brands (Zeithaml, 1998). Brand image is a key antecedent of brand equity that may positively affect the perceived quality of the brand and its value (Kim et al., 2012;Iglesias et al., 2017).
Perceived quality varies amongst subjects, as each individual has its own criteria for quality (Aaker, 1991). Experiences with a brand and, consequently, perceived quality, are the assessments of all interactions with both product/service and the brand. Consumers' judgment of quality is usually not "rational", and is based on superficial associations, including appearance, color, taste or functionality (Baalbaki & Guzmán, 2016). According to Beneke & Carter (2015), brand image and packaging attract attention and present cues and information about the product that aid consumer choice, which then results in a positive influence on the perceived quality of PLB. Thus, it is proposed that: H3: The perceived quality of private label brand is positively affected by its new image

Brand loyalty
Brand loyalty generally refers to a consumer's commitment to a product, service or brand (Yoo et al., 2000;Iglesias et al., 2019). Oliver (1997) defined brand loyalty as a strong consistent commitment to rebuying and patronizing a brand, regardless of efforts on the part of competitors to the contrary, that is, switching-inducing behaviors. Brand loyalty strongly affects brand equity, as loyal consumers show a stronger willingness to pay steeper prices and are less likely to switch than non-loyal consumers (Liu et al., 2017).  show that consumer loyalty is indirectly affected by consumer perceived ethicality because of the mediators such as consumer commitment and perceived quality.
This construct can be divided into three forms: behavioral loyaltyrepeated purchase; attitudinal loyaltyintention to buy; or composite loyaltyrepeated purchase and intention to buy (Smutkupt;. For the purposes of this study, the attitudinal approach was adopted, which regards brand loyalty as an intention to buy a branded product or service as the main choice (Yoo & Donthu, 2001;Smutkupt et al., 2016). In line with the concepts of brand associations and perceived quality, brand loyalty depends on an individuals' perception of the brand. Tangible and intangible features are the variables that allow consumers to decide if they want to be loyal to a brand, based on whether the brand and its products have the characteristics that they are searching for (Liu et al., 2017). According to Wu et al. (2011), a strong image attracts, differentiates and creates positive associations with the brand, thus increasing purchase intention. Hence, PLB's changes should influence loyalty towards the brand: H4: Brand loyalty towards private labeled brand is positively affected by its new image The conceptual model below (Figure 1) illustrates the hypotheses and authors whose literature acts as foundations to them. The current model seeks to understand the effects of rebranding strategy on the brand equity of MPR's own PLB as well as to answer the proposed research hypotheses (H1-H4). Most of the respondents were female (65.8%). 36.4% of the respondents' ranged between the ages of 45 and 54. The second (24.7%) and third (12.9%) biggest groups were those whose sample members were aged between 35 and 44 and 55 and 64, respectively. Concerning the respondents' annual income, the two biggest groups are on opposite ends of the scale. 31.5% of the respondents have an annual income of more than 35,000 €, whereas 19% state that their annual income as being less than 10,000 €. Most respondents have a bachelor's degree (56%), whereas 26.1% and 15.4% have masters' degrees and secondary education, respectively.
It is not surprising that our sample is skewed toward female purchasers because women are dominant retail shoppers. A recent Nielsen global report reveals that women continue to assume most of the household responsibilities, and 90% of them state that they have shared or total responsibility for daily retail purchases (Nielsen, 2019).

Survey pre-testing
Before proceeding with the distribution of the survey, a pre-test was conducted in order to identify possible issues or difficulties in interpretation. It occurred between the 10 th and the 14 th of August 2018 and was conducted with 20 respondents. Some interpretation issues were, indeed, identified. The purpose of the questionnaire was to apply the same questions and constructs to the brand's image before the rebranding and after the rebranding. However, respondents found that to be unclear and often thought that there was something wrong with the questionnaire itself. In order to clarify possible doubts, an additional "structure and operation" section was included in the questionnaire. Other minor changes in semantics were also applied, in line with other issues put forth by the respondents.

Questionnaire structure and measures
The questionnaire was composed of several sections: (1) the first section of questions had the purpose of assessing the consumers' shopping behaviors. Respondents were asked to assess how frequently they shopped, how frequently they shopped at MPR and how frequently they bought MPR's private label brand products. Respondents were also asked whether they knew and purchased MPR's own-branded products. Those who said they did not buy MPR's ownbranded products were asked why, whilst those who said they did not know MPR's own-brands skipped directly to the demographics, the final part of the questionnaire; (2) The aim of the second section is to assess each research model constructs for both pre-rebranding image and post-rebranding image. Pictures of the pre-rebranding and post-rebranding MPR's private label brand images were presented 2 . They were not told which of the images was the current one, nor the one used before the re-branding. Subjecting consumers to the visual imagery of a brand has proven to be a successful tool in evaluating brand image (Aaker, 1996); (3) The next section in the questionnaire asked respondents to assess which of the images in their opinion related more to the idea of MPR's own-brand and whether they could identify (how many) sub-brands and brand tiers were there in any of the pictures shown. This served the purpose of ascertaining which visual connection to the brand was the strongest (pre-rebranding or post-rebranding) and whether the change had influenced sub-brand awareness; (4) The final section of the questionnaire sought to identify individual demographic characteristics of the sample.
To assess the four brand equity constructs, measurements were adopted from relevant authors' items and adapted to this specific case. Table 1 presents a list of items per construct.
All items were measured on a 5-point Likert-scale, where 1 meant completely disagree and 5 completely agree.
Due to the fact that MPR's own-brand rebranding process, focused mainly on the brand's products' packaging, it was essential to consider sufficiently generalized attributes, that did not relate only to the brand and its products but also it's packaging, thus representing overall consumer perceptions of the brands' image and character.
The respondents' anonymity was guaranteed and this procedure aimed to reduce the common method bias (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). However, in order to assess the potential for common method variance, the Harman's single-factor test was conducted (Podsakoff et al., 2003). The exploratory factor analysis with all items demonstrated that a single factor did not account for the majority of variance, revealing that the risk of common method variance was unlikely to be a problem.

Statistical analysis
Different types of statistical analyses were conducted, including descriptive statistics, means comparisons with t-test and confirmatory factor analysis.
Confirmatory factor analysis of longitudinal data, more precisely paired data, was used in order to test whether the customers' perceptions of the new rebranded image are influenced by their perceptions of the previous image. This method fits the intended purpose because the same set of items were measured at two different points in time, before and after rebranding.
Seven models were estimated. They were organized in a hierarchical ordering with increasing constraints from one model to the next. The first model (M1) estimated the correlation between the brand equity constructs of the old and the new brands and had no correlated measurement errors. In the second model (M2), the measurement errors of an indicator before and after rebranding were also correlated. The third model (M3) aimed to test whether the factor loadings are the same before and after rebranding, i.e. whether the loadings are invariant. Therefore, equal loading constraints are added. If the loadings are the same, then it can be argued that the construct is the same before and after rebranding. The fourth model (M4) requires that error variance of the same measures, before and after rebranding, are equal.
Pairwise comparisons between nested models were calculated to choose the best fit model. The model chosen will then be used to study the effect of rebranding on the means of the latent variables, i.e. to know if the perception of the new brand is higher or lower than the previous one for each of the brand equity dimensions. Therefore, three more models were estimated. In model M5 intercepts were forced to be equal whereas the factor means were free. The sixth model (M6) aimed to test the equality of factor means and the last model (M7) tested the equality of factor variances. The chi-square difference (χ2diff) test statistic was used to compare the model fit of pairs of nested models. If the χ2diff value is significant at 5%, the nested model with a smaller number of degrees of freedom fits the data better than the least restrictive model.

Buying behavior and overall associations
The majority of respondents stated that they go shopping at least once a week: 44% of respondents answered that they go shopping more than once a week and 42% said they go about once a week. Regarding shopping at MPR, more than 30% of the respondents shop about once a week, 23% shop at MPR more than once a month and only 13% shop more than once a month. Almost all respondents (96%) stated they are familiar with MPR's own-brand. From those, 22% of respondents buy MPR's own-branded products every time they shop at MPR, half buy most of the time at MPR, and only 3% never buy products from MPR's brand. These respondents do not buy MPR products mainly because they are used to buying other brands (46%) or do not like MPR's brand (31%).
Respondents' were asked which one of the packaging images (presented in the prerebranding and post-rebranding pictures) they associated more with the MPR's brand. Most respondents (62%) identify the pre-rebranding image as the one they associate more with MPR's own-brand. However, respondents who shop at MPR more than once a week, associate more with the post-rebranding image than the previous one.

Sub-brands and brand tiers
Respondents were also asked whether they could identify different sub-brands or brand tiers in each one of the pictures that were shown to them corresponding to the pre-rebranding and post-rebranding images. Overall, respondents were better able to distinguish between different sub-brands in the newly revised image, although the difference was not deemed significant. 56% of the respondents stated they could tell different brand tiers apart in the postrebranding picture, whereas 51% could tell different brand tiers apart in the pre-rebranding figure. This is an interesting finding since for pre-rebranding, four sub-brand pictures were shown, whilst in the post-rebranding picture, only three sub-brands were shown. Regarding the pre-rebranding picture, only 12% of the respondents chose the correct number of sub-brands and 32% thought only two sub-brands were shown. In the post-rebranding picture, 26% of the respondents chose the correct number of sub-brands, yet 36% of the respondents thought only two sub-brands were shown. Frequent shoppers showed higher percentages when it came to being able to identify brand tiers, whereas sporadic shoppers showed higher percentages of not being able to identify brand tiers. When looking at the group of respondents that shop at MPR more than once a week, 65.9% of them were able to pinpoint sub-brands in the pre-rebranding picture, whilst 62.5% could identify sub-brands in the post-rebranding image. On the contrary, from the group of respondents who shop at MPR less than once a week, 69.8% could not identify brand tiers in the pre-rebranding picture, whilst 58.1% could not identify brand tiers in the post-rebranding image. It is also possible to note that the percentage of respondents that can identify different brands is higher in post-rebranding (56.5%) than in pre-rebranding (51%).

Rebranding effects on brand equity
Paired sample t-tests were conducted in order to assess the theoretical hypotheses.
Significant negative differences mean that the MPR's own-brand equity was affected by its new image except for reversed statements. The t-test results for all brand equity variables are presented in Table 2.   TABLE 2 There are several differences in the means of the brands' awareness items suggesting that respondents answered more positively to the pre-rebranding image than to the post-rebranding image, meaning that consumers associate MPR's private label brand image with the previous image and not with the new rebranded one. Similarly, at the moment, respondents have a harder time recognizing MPR's own-branded logo, identifying products in-store and identifying MPR's own-branded products amidst other brands' products, with the new, rebranded version of the brand.
Regarding the brand associations construct, respondents seem to associate the terms "inexpensive" and "trustworthy" to the pre-rebranded image, while they associate the terms "innovative" and "original" to the rebranded version of the brand. The consumers' perception of quality only showed differences in one of the four variables. When asked to evaluate their concurrence with the sentence "this image represents a reliable brand", respondents responded more positively to the previous image, than to the rebranded version.
Lastly, regarding brand loyalty, results showed that customers were prone to switch, the rebranded brand for other brands, more readily than they did before the rebranding. However, and as expected, they were less likely to buy other brands when there were discounts coupled with the products, irrespective of the rebranding.

Effects of pre-rebranding perceptions on a new image
Confirmatory factor analysis for longitudinal data was conducted in order to understand to what extent the post-rebranding perceptions of the familiar respondents with the MPR's ownbrand (N=349) are correlated to their pre-rebranding perceptions. As above mentioned, seven models were estimated. Table 3 presents a summary of their model fit measures and the nested pairwise comparisons. All the models presented have an acceptable fit. Comparisons of the first four models suggest the choice of M2 or M3; this means that the correlation of measurement errors for each item is needed and, although worsening the model fit, forcing equal loadings is also reasonable. As can be seen in Table 4, the models M2 and M3 have similar correlations values between constructs, so M2 is chosen.  3   TABLE 4 Results of M2, M5, M6 and M7 comparisons (see Table 3) show that there are no differences between the construct means before and after rebranding (M2, M5, and M6 fit equally well) but the factor variances are different (M2 and M6 are preferable to M7, the model which test the equality of factor variances). Therefore, model M6 is chosen. This model is characterized by having equal factor means and different factor variances, meaning that although there is variability in the brand equity of the two brands, the rebranding did not affect the brand equity of the own-brand as it was supposed to do, and was examined in the previous section. Table 5 presents the estimates for the items intercepts and factor means of the model M6, as well as the factor loadings for brand equity before and after rebranding.

TABLE 5
Although the fit is at the limit of acceptable for the most of models (cf. Hu & Bentler, 1999;Davcik, 2014), M1 and M4 have no excellent fit (RMSEA values higher than 0.08 and CFI values lower than 0.9); meaning these models do not reproduce the observed data very well because of various constraints on the hierarchy of models. This result is not surprising because (1) M1 has no correlated errors between before and after rebranding, assumptions are not realistic for paired data, and (2) M4 presents the same measurement model before and after rebranding; M2 or M3 are preferred than M4, thus equal error variances should not be considered.
All the pre-rebranding versus post-rebranding variables are correlated (see the correlation values for the model M6 in Table 4), meaning that consumers' perceptions of the current brand image were related to their previous perceptions. Brand recognition and loyalty are the variables with the higher correlations. This means that customer loyalty to the brand and recognition thereof, before the rebranding, strongly impact on brand recognition and loyalty after the rebranding. Consumer's modern association of the post-rebranding brand image was weakly related to the associations of the pre-rebranding brand image, as it was considered more inexpensive and trustworthy.

Conceptual contributions
The purpose of this study is to evaluate the effects of new rebranding strategies on consumer-based brand equity through the four constructs that make up consumer-based brand equity: brand awareness, brand associations, perceived quality, and brand loyalty, as well as to understand how the new brand image dimensions are related to the past ones.
Regarding overall associations, when respondents, were directly asked which image they associated with MPR's own-brand, most still picked the MPR brand image before the rebranding exercise. However, results show that shopping behavior has an impact on consumers´ ability to recognize the new rebranded image, as evidenced by the fact that the only group of consumers that associate MPR's own-branded goods with its new image being the ones that shop at MPR more than once a week. This may indicate that constant or recurring contact with the newly rebranded version of incumbent products is required for consumers to acknowledge them MPR's own-branded goods. The retailer brands were able to increase familiarity with its own-branded products thus promoting equity transfer (Ataman & Ulengin, 2003;Burt & Davies, 2010;Ngobo, 2011).
With regards to the brand tiers, there seems to have been an improvement with the new image, since consumers' ability to tell different sub-brands apart is higher with the rebranded version of the brand. The fraction of respondents that chose the correct number of sub-brands is also higher with the new image. This shows that even though the goal of the rebranding exercise was to create a uniform and coherent brand image, brand tier differentiation also proved to be successful. Additionally, results also indicated that frequent shoppers (who shop at MPR more than once a week) find it much easier to identify the different brand tiers, than sporadic shoppers, which once again corroborates the idea that consumers need to keep abreast of the brand and be in constant contact with it, for changes to be effective.
Regarding brand equity, results indicate that brand equity dimensions have not been altered by the image rebranding. This might seem undesirable, given that the main goal of most rebranding exercises is to improve brand equity (Goi & Goi, 2011). However, considering that the rebranding implementation is relatively recent, this outcome may change over time. More importantly, the lack of change in brand equity shows that the brand has not been affected by the "rebranding paradox". Rebranding exercises have no guarantee of success (Aaker, 1991;Miller et al., 2014) and may jeopardize current brands´ existing brand equity (Muzellec & Lambkin, 2006). In this case, results suggest that existing positive associations with the brand have not been lost. However, when each constructs' variables were analyzed, outcomes showed slight differences. Brand awareness and brand associations are the dimensions that show the most significant differences. It is interesting to note that within brand awareness, brand recognition was much more affected by the rebranding than brand recall, which makes sense considering that the literature establishes a difference between recognition and recall, by considering recognition as the ability to remember a brand by resorting to an aid (Walsh et al., 2008). In addition, consumers' ability to recall the brand was not substantially affected.
Regarding brand awareness, consumers' logo recognition, in-store recognition, and recognition amidst other brands were the most affected by the new image. Considering that for many years, MPR's brand shoppers were used to identifying and searching for the previous image, these results are not unexpected.
Results show that consumers see the new version of the brand as "innovative" and "original", whilst they consider the previous version to be "trustworthy" and "inexpensive". On the one hand, not associating the rebranded image with "inexpensive" is positive. On the other hand, considering the past image particularly, the word, trustworthy may indicate that the brand needs to work on its perceived risk -consumers´ trust of the previous brand was not completely transferred onto its new image. This argument is also backed by the fact that the only difference in consumers' perceptions of quality was with regards to seeing the brand as "reliable". As the literature suggests, perceived risk and subsequently, perceived quality are important influencing factors, especially for private label brands (Glynn & Chen, 2009;Cuneo et al., 2012).
Concerning brand loyalty, consumers are more likely to consider buying other brands after the rebranding, but they are less likely to purchase other brands if discounts are applied to products. This suggests that consumers perceive the new image as more on a par with NB.
Consumers' awareness of the current brand image is the most influenced by past experience. In addition, brand loyalty, followed by perceived quality, are also factors that transferred equity from the past to the present version of the brand. This supports the idea that equity is transferred, even when the brand changes and confirms that there was no significant loss of equity -therefore, no "rebranding paradox".

Managerial implications
In conclusion, the results emphasize two main managerial implications. First, no major losses of brand equity were identified. As the literature indicates, significant changes in a brand can compromise the existing brand equity. The challenge is to transfer the equity along with the transformation while focusing on creating it (Aaker, 1991;Muzellec & Lambkin, 2016).
Results indicate that MPR successfully transferred the majority of brand equity to the new brand image. Using the elements of the previous PLB image as the basis for the brand image in the rebranding exercise was an effective way of balancing the need for innovation whilst maintaining the essence and value of the brand. This suggests that in undertaking rebranding exercises, losses in brand equity are not necessarily a concern if proper care is taken with respect to symbolism in ensuring that the brand´s essence is kept. Second, time and continuity are key. Our study points out that consumers need to keep in contact with the brand for changes to be effective. Consumer interaction with the brand ensures familiarity with the new design and a propensity to lower strangeness levels with the new brand. Consumer-based brand equity constructs are very much based on consumer experience and interactions, which explains the identified differences between pre and post-rebranding variables. This contributes to the basis of branding and brand equity literature which states that creating and maintaining a positive reputation and equity of a brand is a continuous long-term project (Muzellec & Lambkin, 2006). Consumers have to be continuously subjected to a name, image or concept to associate it with a brand (Muzellec et al., 2003).

Limitations & future research
This research is not without its limitations. We could not include secondary data in our estimation models, such as marketing and advertising cost. MPR's rebranding strategy also focused solely on brand image, visual identity, and packaging. Another research limitation is the time span. MPR's rebranding was a long and continuous process, however, results seem to indicate that there was not enough time for consumers to fully grasp the new rebranding strategy.
From the methodological point of view, some of our models could show a more robust fit.
Several of specified models have fit at the limit of what could be considered acceptable.
Although cut-off points of 0.10 and 0.9 for RMSEA and CFI respectively were established in the literature, nowadays more restrictive cut-off points are suggested in order to prevent the acception of misspecified models (Hu and Bentler, 1999 Rebranding research that also evaluates the effects of marketing and advertising channels would offer a much more complete view of the rebranding effects on brand equity.
Understanding whether the brand used other communications efforts may help to empirically evaluate the importance of brand co-creation and stakeholder involvement, and what the ideal way of communicating the effected changes is. Figure 1 -Conceptual framework