Bridging Brand Parity with Insights Regarding Consumer Behavior

2.1 BP

In today’s world, finding any kind of information is getting easier, and from a brand management perspective, competing brands can easily get information about their competition and try to offer similar products or services that satisfy the needs and wants of the consumers. When the differences between brands vanish, BP arises. BP is defined as “the overall perception held by the consumer that the differences between major brand alternatives in a product category are small” (Iyer & Muncy, 2005, p. 222). It is also defined as consumers’ perception of similarity among brands (Henderson et al., 1998). According to Keller (2013), points of parity associations are important, but they should be backed up with points of difference associations in order to create a reason to buy in the mind of the consumer. From a microeconomic perspective, the decision-making process of the consumer is quite complicated, and sometimes, the similarity between brands demotivates the consumer from buying a specific brand. For this reason, it is important for brands to have a points of parity association, while also creating points of difference associations that will attract the consumer.

2.2 BI

BI is one of the most important aspects of building a strong brand. According to Keller (1993), together with brand awareness, BI is one of the dimensions of brand equity under the brand knowledge framework. His conceptualization of BI consists of brand associations. Similarly, Aaker’s (1991) approach to brand equity also includes brand associations. Therefore, the key to a positive BI is creating positive brand associations. Brand associations are related to consumers’ perceptions associated with a specific brand (Ali et al., 2018). For instance, when seeing a specific brand’s logo, if the consumer feels happy, this is a positive brand association, leading to a positive BI. BI can be a distinguishing factor regarding brand preference and may create a brand-loyal consumer base. Keller (2013) categorizes BI into four sub-dimensions, where the first is about the types of brand associations. There are attributes consisting of product-related (product’s physical material or service characteristics) and non-product-related association sources, such as usage imagery, user imagery, packaging, and price. Usually, when a consumer sees a high price, there will be a positive BI signaling high quality. There are also functional (main advantage of the product or service), experiential (the feeling when using the product), and symbolic (such as social acceptance or self-respect) benefits as types of associations. The last type is attitude, which is a general evaluation by consumers regarding a brand. The other sub-dimensions are favorability (whether a brand possesses attributes and benefits that satisfy consumer needs and wants), strength (related to the relevance and consistency of the information to the consumer), and uniqueness (whether the consumer’s brand-related associations in mind are brand specific and unique) of brand associations that all together creates the image of the brand.

2.3 BS

As already stated, the decision-making process of consumers is quite complicated. However, in general, it can be said that consumers are willing to purchase a product or service that satisfies their specific needs or wants. When this happens, the consumer is usually satisfied with the brand. Chinomona et al. (2013) define BS as “overall consumer’s evaluation based on the consumer’s total purchase and experience with a brand of product or service.” From this definition, it can also be seen that the actual purchase, including the decision-making process of the consumer, is related to BS. In another approach, BS is directly affected by brand performance, where if the consumer’s cognitive and affective evaluation of perceived brand performance exceeds the expected performance, it can be stated that there is BS (Hwang et al., 2021). Moreover, if a consumer buys regularly from a brand, this can also show that there is BS. In order to preserve the value of a brand, BS is one of the key elements in brand management (Thai et al., 2020).

2.4 BL

From a brand equity perspective, the BL of a specific brand’s customer base is usually the core of the brand’s equity (Aaker, 1991). Oliver (1999, p. 34) defines BL as “a deeply held commitment to rebuy or repatronize a preferred product/service consistently in the future, thereby causing repetitive same-brand or same brand-set purchasing, despite situational influences and marketing efforts having the potential to cause switching behavior.” Accordingly, repeat purchase behavior is an important component of BL. Keller (1993) also emphasized that BL happens when favorable beliefs and attitudes for a brand end with repeat purchase behavior. The re-purchasing is more about behavioral loyalty. Dick and Basu (1994) also mentioned that regarding BL research, several behavioral measures were used. On the other hand, there is attitudinal BL, which is defined by the dispositional commitment of the consumers regarding unique value associations with the brand (Chaudhuri & Holbrook, 2001). In summary, BL can be conceptualized in two main categories: behavioral perspective and attitudinal perspective (Buil et al., 2008). Keller (2013) considers BL the top of his brand resonance pyramid model, where there are four sub-dimensions. The first sub-dimension is behavioral loyalty, which comprises repeat purchases and share of category volume. The second one is attitudinal loyalty, which occurs when a consumer has a strong bond with the brand. The third sub-dimension is the sense of community, which is defined by the kinship or affiliation with other people associated with the brand. The last sub-dimension is active engagement, which happens when consumers are willing to invest time, money, energy, or other resources in the brand. This article considers BL holistically, where the explanations of the dimensions above are intended to define BL in a descriptive manner and not from an operational perspective.

2.5 Development of the Hypotheses

To make it more convenient and concrete for the readers before explaining the rationale of the development of the hypotheses, the conceptual framework of the study can be seen in Figure 1.

Figure 1

Conceptual framework.

The hypotheses were developed based on information processing theory. Information processing theory has also been used in established brand-related studies (e.g., Leong, 1993), and it is explanatory regarding consumer behavior in the context of cognitive processes (Tybout et al., 1981). A person needs only a piece of information to choose one of two similar alternatives (Miller, 1956). Consumers receive information from different marketing communication options that, in turn, create BI, BS, and BP, which affect several outcomes. The first hypothesis of the study is considering the effect of BI on BL. As described before, consumers have some brand associations in mind regarding their perceptions of a specific brand. As brand associations, along with the overall BI, become more positive, there will be a tendency to become more loyal to the brand. For instance, even if a competing brand offers a discount, if the BI of the regularly purchased brand is very positive, there will be no switching behavior because the BL level would also be high. This logic is supported by previous research findings (Alhaddad, 2015; Büyükdağ, 2021; Greve, 2014; Kim et al., 2018; Thai et al., 2020). Accordingly, the first hypothesis of this study is:

H1. BI has a significant effect on BL.

Having a positive BI will also influence consumer satisfaction. For example, when you eat food from a brand that you have a positive association with, such as good flavor and high-quality service, there will be a higher chance of high BS. When looking at the literature, Ali et al. (2018) found that BI had a positive effect on brand purchase intention, and it can be inferred that a satisfied consumer will have more intention to purchase. Moreover, Alfakih et al. (2022) hypothesized and found support regarding the positive effect of BI on BS. Last but not least, in their study regarding chocolate BI and satisfaction, Puska et al. (2018) found a positive impact of BI on BS. Therefore, the second hypothesis is:

H2. BI has a significant effect on BS.

The third hypothesis of the study considers BS and its effect on BL. In general, the more consumers are satisfied with a brand, it is expected that they will become more loyal to that specific brand. As mentioned before, BS is related to the purchase of the brand and the experience with the brand. As consumers buy more frequently from a brand and have positive experiences with the brand, such as with experiential marketing activities, BS will increase and will turn into stronger BL. Chinomona et al. (2013) found the effect of BS on BL through brand trust. Although this is not showing a direct effect, it is an important insight. Moreover, previous literature also found a positive, direct effect of BS on BL (Hwang et al., 2021; Mokha, 2021; Susanti et al., 2021). Based on the logical relevance of these concepts and previous findings in current literature, the third hypothesis is:

H3. BS has a significant effect on BL.

As mentioned before, BP happens when consumers do not perceive differences among brands in a product category. It can be stated that when this happens, there are only points of parity associations for these brands regarding consumers’ perceptions. However, as stated by Keller (2013), points of difference associations may create reasons to buy a specific brand and differentiate the brand from its competitors. From this perspective, it is expected that BP would be negatively correlated with BL. The reason is that if consumers do not observe considerable differences between competing brands, they can easily switch to another brand, for instance, if there is an effective marketing communication activity. In the literature, there are few contradictory results regarding the relationship between BP and BL. Iyer and Muncy (2005) found that as BP increases, developing a loyal consumer base decreases. However, Li (2010) found something slightly different than the Iyer and Muncy findings. As stated before, there are generally two types of BL, which are behavioral and attitudinal loyalty. Li (2010) observed that BP does not impact attitudinal loyalty, in fact, as BP increases, it has a positive impact on consumers’ behavioral loyalty. In summary, there are few contradictory findings in the literature, but based on the previous literature, the fourth hypothesis of this study is:

H4. BP has a significant effect on BL.

The last two hypotheses are about the moderating effect of BP. As stated before, as consumers start to perceive minimal differences among brands, the possibility of switching behavior rises. In that sense, if there are only points of parity associations regarding competing brands, when observing the effect of BI on BL, because of only points of parity associations building the BI, there might be a moderating effect of BP in that relationship. Similarly, considering the effect of BS on BL, BP may moderate this relationship as well. As mentioned above, BS is based on both the purchases and experiences with the brand. If there are no observable differences among brands, the purchase behavior and experiential relationship with the brands will be small, resulting in BP having a moderating effect on how BS affects BL. In conclusion, the last two hypotheses of this study are:

H5. BP moderates the effect of BI on BL.

H6. BP moderates the effect of BS on BL.

3.1 Sample and Procedure

Regular customers who shop at chain supermarkets in the province of Antalya were selected to collect data. In order to generalize the results, ten branches of five well-known chain supermarkets operating in the Konyaaltı district of Antalya province were identified. Permission was obtained by explaining the purpose of the study to the managers of the selected markets. The customers participating in the study were informed about the research topic through the survey description and convinced of the research’s reliability, and their consent was obtained.

The study employed a convenience sampling technique. The population of Konyaaltı district is approximately 205,000. With a 5% margin of error, the sample size was determined to be 384 (Sekaran, 1992). To ensure homogeneity, it was planned to administer 60 questionnaires for each market, totaling 600 data points.

The survey was initially developed in English and then translated back to Turkish by two linguistic experts, following the recommendation of Perrewe et al. (2002). A pilot survey involving 30 customers was conducted to ensure a full understanding of the questions by the participants and to address any potential issues.

In order to mitigate non-response bias, efforts were made to enhance response rates, while simultaneously addressing common method bias by providing participants with assurances of confidentiality to promote truthful and unbiased reporting. An attempt was made to reduce collinearity issues by collecting data at different time points and obtaining responses from different scales each time.

At Time 1, 600 questionnaires were distributed. Each customer received a survey packet containing a cover letter from the researchers, explaining the purpose of the study. The cover letter also provided assurances of confidentiality to mitigate the potential threat of common method bias, as suggested by Podsakoff et al. (2003). The customers were informed that they would receive another short survey approximately 3 weeks later. At Time 1, data on BI, BL, and demographic information were collected. A total of 480 valid questionnaires were returned, resulting in an 80% response rate.

Approximately 3 weeks later, the customers who completed the Time 1 survey (n = 480) were given a second questionnaire assessing customer BS, BP, and demographic information. A total of 408 questionnaires were returned, yielding an 85% response rate. Through assigned identification numbers, the researchers were able to match the questionnaires with each other. Due to missing data, only 392 responses were used for data analysis. Response error was examined by comparing the demographic data (gender, age, and work experience) of those who returned surveys at Time 1 with those who did not respond at Time 2 (Collier & Bienstock, 2007). No significant differences were found.

3.2 Demographic Variables

The average age of the participants was about 35, and 42% were women. Over 71% of employees had at least graduate degrees or higher. About 60% of the respondents were married, and 70% were workers. The reason for preference of 70% of the participants is about economic reasons.

3.3 Measurement

BS Scale: In order to measure the level of BS, the scale developed by Taylor and Baker (1994) and validated by Thuy and Hau (2010) was used. The result of the reliability analyses conducted by Thuy and Hau (2010) of the scale, which consists of a total of 5 statements, is 0.93 (CFI: 0.91; AGFI: 0.88; RMSEA: 0.07).

BI Scale: A five-item BI scale developed by Keller (2003) and used by Kaplan (2007) was used to measure the tendency of customers’ perception regarding BI. The result of the reliability analyses conducted by Kaplan (2007) is 0.92 (CFI: 0.92; AGFI: 0.90; RMSEA: 0.06).

BP Scale: The five-item scale developed by Muncy (1996) and used by Ilıcalı (2016) was used to measure BP. The result of the reliability analyses conducted by Ilıcalı (2016) is 0.76 (CFI: 0.90; AGFI: 0.86; RMSEA: 0.07).

BL Scale: The five-item scale developed by Narayandas (1996) and used by Aydın and Özer (2005) was used to measure BL. The result of the reliability analyses conducted by Aydın and Özer (2005) is 0.88 (CFI: 0.88; AGFI: 0.84; RMSEA: 0.08).

Construct reliability and validity are shown in Table 1.

3.4 Results

SPSS and AMOS version 22.0 and Smart PLS 4.0 were used for analyses. All measures were subjected to confirmatory factor analysis to provide support for the issues of dimensionality, convergent, and discriminant validity. The retained item loadings exceeded 0.50; Cronbach’s alphas were all above the benchmark of 0.70; CR and AVE were also above the benchmark of 0.50 (Hair et al., 2006). This provides evidence of convergent validity among our measures. The estimated correlation between the variables is below 0.85, which provides evidence of discriminant validity, as recommended by Kline (2005).

Discriminant validity (Fornell–Larcker criterion) is presented in Table 2. The analysis of discriminant validity demonstrates significant attainment of discriminant validity. According to this finding, the following study has been able to achieve all the criteria since the reliability and validity of the measurement of the construct have been achieved.

Means, standard deviations, and correlations of the study variables are presented in Table 3.

As predicted, the relationship between BI and BL is positive and significant (r = 0.65, p < 0.01); the relationship between BS and BL is positive and significant (r = 0.77, p < 0.001). Due to the significant relationships observed, further analyses were made.

Hierarchical regression analysis was used to test the hypotheses. In addition, hypotheses regarding the moderating role were tested with the method proposed by Baron and Kenny (1986). Hierarchical regression analyses were applied to test the hypotheses. The results were checked through the Smart PLS path model and model fit.

Hierarchical regression analysis was applied to determine the effects of BI on BL and the moderating role of BP in this interaction and to test the hypotheses. The applied regression analysis results are given in Table 4. According to the results of the analysis performed by controlling age and tenure at the first stage, the relationship between BI and BL was found to be significant (β = 0.64; p ≤ 0.001). Hypothesis H1 was supported.

In the second step, it was determined that BP had no significant effect on BL (β = −0.06; p > 0.05). H4 was not supported. The results of this step are presented in Table 5, and the findings showed that BL did not have a moderating role in the relationship between BI and BL (β = −0.04; p > 0.05). Hypothesis H5 was not supported.

At this stage of the analysis, hierarchical regression analyses were applied to determine the effects of BS on BL and the moderating role of BP in this interaction and to test the hypotheses. The applied regression analysis results are given in Table 5. According to the results of the analysis performed by controlling age and tenure at the first stage, the relationship between BS and BL was found to be significant (β = 0.6; p ≤ 0.001). Hypothesis H3 was supported.

The results of this step are presented in Table 5, and the findings showed that BP did not have a moderating role in the relationship between BS and BL (β = −0.02; p > 0.05) either. Hypothesis H6 was not supported.

The relationships were retested by path analysis via Smart PLS to determine the model fit, indirect effects, and moderator effect. As already discussed, it was determined that there is a positive and significant effect of BI on BS (β = 0.751; p ≤ 0.001), supporting H2. The results of bootstrapped path analysis show that there were indirect effects between BI and BL via BS (β = 0.5; p ≤ 0.001). The structural model and indirect effects can be seen in Figure 2.

Figure 2

The structural model and indirect effects.

The models fits are presented in Table 6.