Impact of Brand Equity on Purchase Intentions of Customers in Insurance Industry of Pakistan

Insurance firms play an important role in an economy’s growth and development and are key financial institutions that are critical for the success of other companies. They provide stability to both individuals and businesses by agreeing to bear the financial liability of insured parties. However, in Pakistan, the share of insurance towards GDP is significantly lower than in similar developing countries. Thus, we have developed a model that can help the insurance sector improve its image and sustainability. We collected the data from five major cities of Pakistan based on pre-developed questionnaires. The study has a useable sample size of 424. We found that Brand Attitude (BA) and Customer Satisfaction (CS) promote Brand Equity (BE) and Purchase Intention (PI). Further, BE is a determinant of PI. However, brand image has contradicting effects. On the one hand, Brand Image (BI) does not affect BE, but on the other hand, BI has a significant association with PI. The results also suggest that BI, BE, and CS mediates BA and PI. Also, BI and CS affect BE. However, BE does not mediate BI and PI, but BI mediates CS and purchase intentions


Introduction
Since the inception of the brand equity (BE) concept, many researchers have extensively used it in different domains (Keller, Apéria, & Georgeson, 2008).Chow, Ling, Yen, and Hwang (2017) suggest that BE is an effective framework for understanding the possible consequences of multiple brand strategies.BE has a customer and firm-based perspective (Datta, Ailawadi, Van-Heerde, 2017).The customer perspective focuses on the "consumer mindset, which includes structures such as attitude, awareness, associations, attachments, and loyalties" (Keller & Lehmann, 2003).Simultaneously, the firm-based perspective stresses "employing product-market outcomes such as price premium, market share, relative price, and financial-market outcomes such as brand purchase price and subsidized cash flow of licenses and royalties" (Liu, Wong, Tseng, Chang & Phau, 2017).
Given the prevailing competitive environment, firms and companies spend considerable resources on BI and BE.They understand that a strong BI is necessary for sustainable growth, product differentiation, and competitive advantage (Sinclair & Keller, 2017).Many researchers suggest there is more need for studies on BE in developing countries, especially in the domains of insurance (Hosseini & Moezzi, 2015;Theurer, Tumasjan, Welpe, & Lievens, 2018).Similarly, others recommend more studies on the customerbased BE model in different domains and societies (Chatzipanagiotou, Christodoulides, & Veloutsou, 2019;Machado et al., 2019;Algharabat, Rana, Alalwan, Baabdullah, & Gupta, 2020).Such new studies may bring further insight into the association between customer-based BE and PI.This study has responded to the calls of earlier studies and has examined Pakistani insurance consumers' attitudes and behavior towards insurance products.It specifically identifies the impact of BI, BA, BE, BI, and CS.

Insurance Industry
Insurance firms play an important role in an economy's growth and development.Insurance has two purposes (Waseem-Ul-Hameed, Ali, Nadeem, & Amjad, 2017).First, insurance functions as an economic instrument that is critical to the success of other companies.Insurance gives stability to individuals and businesses by agreeing to bear the financial liability of the insured parties.Second, insurance accumulates assets in the economy, community, and privately held sectors as a financial entity.Khan et al. (2018) suggest that insurance educates and inspires people to save for the future.Wang, Asghar, Zaidi, Nawaz, Wang, Zhao, and Xu (2020) suggest that insurance, apart from its traditional role, can significantly contribute to the fight against underdevelopment, insecurity, illness, illiteracy, unemployment, negative trade balances, and other economic features prevalent in underdeveloped countries.Similarly, Hussein (2019) also suggests that the insurance industry plays an important role in economic and social development.
Pakistan is reshaping the economy to face the demands of the global marketplace.The government has implemented several reforms to promote and consolidate Pakistan's status as an emerging regional market.As a consequence of the reforms, the financial industry has been deregulated and liberalized (Takaful, 2019).Consumers' perception of insurance products in Pakistan is low.Therefore, the insurance sectors' contribution towards GDP is 0.9%.However, other regional countries' insurance sectors' contribution to GDP is 2.2%.At the same time, the global average is 6.6%.The major reason for the low insurance ratio in Pakistan is the poor service and bottlenecks in the claim process (Pasha, Hamid, & Shahzad, 2017).The insurance sector in Pakistan offers two types of insurance, i.e., life and general insurance.Life insurance focuses on life and health.On the other hand, general insurance mostly covers motor vehicles, homes, travel, and phone (Ahmed, Arshad, Mahmood, & Akhtar, 2019).

Brand Attitude (BA) and Brand Image (BI)
Consumers develop an attitude towards a brand as they have limited cognitive processing capability.Zhang, Li, Ye, Qin, and Zhong (2020) suggest that consumers' attitude towards a brand enhances its image (Kim, Jang, & Kim, 2021).An attitude represents an effect on an object.The BE model examines the effect of attitudes on market share.The study found that "the market share of two-third of surveyed brands increased" due to consumers' positive attitude (Baldinger, Rubinson, 1996).
Brand strength refers to the characteristics of a brand, due to which consumers develop a positive attitude towards it, leading towards enhanced BI and loyalty.BA comprises three components which are cognitive, affective, and conative.All three individually and collectively affect BI (Byun, 2020).Lin, Lin, and Wang (2021) and others suggest that BA is a pre-defined aspect of brand evaluation.This pre-defined evaluation enhances BI and customer loyalty (Zhang, Zheng, & Zhang, 2020).Marketers use the brand association as a tool for product differentiation and stimulating positive feelings towards a brand.Favorable attitudes towards a brand motivate consumers to develop loyalty and sustainable relationship.Extant literature has documented that attitude towards a brand directly correlates with their behavioral intentions (Yodpram & Intalar, 2020;Liu, Dzyabura, & Mizik, 2020).

Brand Attitude (BA) and Brand Equity (BE)
BA is a key term in customer behavior (Arghashi, Bozbay, & Karami, 2021).Sadrabadi, Saraji, and MonshiZadeh (2018) have referred to BA as "a relatively long-lasting, onedimensional overview appraisal of a brand that presumably energizes behavior." Customers with a positive BA are more inclined to pay a higher price for it (Agmeka, Wathoni, & Santoso, 2019;Paul & Bhakar, 2018;Aaker, 1992).Many researchers suggest that BA is a precursor to a firm's BE and sustainable growth (Ramesh, Saha, Goswami & Dahiya, 2019;Sadrabadi, Saraji, & MonshiZadeh, 2018).BA reflects a brand's likeability and how favorably a consumer perceives it (Kim, Jang, & Kim, 2021).Agmeka, Wathoni, and Santoso (2019) argue that consumers' behavior towards a brand depends on their attitude.If they have a positive attitude towards a brand, their attitude towards BE would also be positive.On the other hand, a negative attitude towards a brand would adversely affect BE.Extant literature also supports the correlation between BA and BE (Arghashi, Bozbay, & Karami, 2021;Sadrabadi, Saraji, & MonshiZadeh, 2018).Hedonic behaviors (e.g., enthusiasm, delight, and enjoyment) are considered important for promoting BE and purchase intentions (Liao et al., 2017;Sadrabadi, Saraji, & MonshiZadeh, 2018).Given the importance of emotions, many brands use emotional appeal in their advertising strategies.

Brand Attitude (BA) and Customer Satisfaction (CS)
BA and CS are distinct constructs.However, both of them have a causal relationship.BA enhances CS, and CS positively impacts BA (Oliver, 1980;Lee, Han, Radic & Tariq, 2020;Rivera, Bigne, & Curras-Perez, 2019).Hwang and Mattila (2019) suggest that CS is consumption-specific, and it is transitory.On the other hand, BA is relatively enduring.Consumers with a strong BA may ignore random bad experiences and hence may not complain.Oliver (1980) found that satisfaction significantly depends on consumers' past experience and may influence consumers' post-purchase attitudes.However, attitude towards a brand or product is not dependent on experience.Advertisements, BI, and corporate image are precursors of attitude (Byun, 2020;Augusto & Torres, 2018).Past studies have documented that BA has a direct and indirect association with customer satisfaction.For example, researchers have found that BI and corporate image mediate CS (Ko, & Chiu, 2008, Bozbay Karami & Arghashi, 2018;Evardsson & Gustavsson, 2013).

Brand Image (BI) and Brand Equity (BE)
Many researchers believe that BI is an integral part of BE, while others think BI and BE are distinct constructs, but the former is a precursor of the latter (Keller & Brexendorf, 2019).Liu (2020) stresses that high equity brands appear to have a more favorable brand association (BI) than low equity brands.Marques, da-Silva, Davcik, and Faria (2020) also report that premium rates and higher BE attribute to higher product quality and image.Similarly, Sürücü, Öztürk, Okumus, and Bilgihan (2019) observed that "positive BI" affects BE "positively, " while negative BI affects BE "negatively." Consumers' evaluation and decision-making process to buy a brand significantly depend on BI.Thus, a brand with a strong image and equity reduces consumers' search in the purchasing process (Tran, Nguyen, Do & Nguyen, 2020;Erkan, Gokerik, & Acikgoz, 2019;Han, 2017).Brand association depends on marketing mix strategies.At the same time, the brand association is a significant determinant of BI and BE (Keller & Brexendorf, 2019).Thus, one can infer that marketing mix strategies (i.e., product, price, promotion, and place) mediate brand association and BE.Strong and favorable BE also allows marketers to differentiate their brands from competitors leading towards increased market share and sustainable growth (Iglesias, Markovic, Singh, & Sierra, 2019;Alam & Khan, 2019).
H4: BI and BE are positively associated.

Brand Equity (BE) and Purchase Intentions (PI)
BE refers to the premium value it generates for its brands.A firm can create BE by delivering valuable, memorable, and recognizable propositions for its target audience (Sanny, Arina Maulidya, & Pertiwi, 2020).Premium brands command consumers' PI and promote sustainable relationships with customers.Sawaftah, Calıcıoglu, & Awadallah (2020) suggest that consumers do not hesitate to pay a premium price for brands with high equity.On the contrary, consumers do not pay premium prices for low BE (Kim & Park, 2013).PI depends on attitude towards products and services.At the same time, BA affects PI and BE (Kim, Chun, & Ko, 2017;Kala & Chaube, 2018).Consumers are risk avoiders, generally.While buying products and services, they spend considerable time reducing the risks associated with their purchases.Thus, a few customers collect information from the internet and other available sources (Agmeka, Wathoni, & Santoso, 2019).
In comparison, others make their purchase decisions on the experience of their friends, peers, and families (Hien, Phuong, Tran, & Thang, 2020).In addition, many researchers suggest that consumers have a perception that purchasing brands with strong equity would reduce the risk associated with buying.Therefore, they have a high inclination towards the brand with strong BE than brands with weak BE (Hermanda, Sumarwan, & Tinaprillia, 2019).
H5: BE and PI are positively associated.

Customer Satisfaction (CS) and Brand Equity (BE)
Researchers have discussed CS from different perspectives (Dimitriades, 2006).However, most researchers agree on the definition derived from expectancy disconfirmation theory (McQuitty et al., 2000).The theory presumes that CS is the difference between customer expectation and the perceived performance of a brand (Oliver, 1999).Extant literature acknowledges that CS, directly and indirectly, affects sustainability, competitiveness, and BE.The service industry, including insurance companies' growth, image, and equity, significantly depend on CS.With the growth and development of the service sector, researchers have revisited and redefined satisfaction from the customer's perspective (Aaker, 1992).Mc-Kenna (1991) suggests that the service sector should pay less attention to advising and promotional strategies and more to building infrastructure that helps firms developing products and services aligned with customers' needs and wants.These efforts may enhance CS and positively affect BE (González-Mansilla, Berenguer-Contri, & Serra-Cantallops, 2019).
CS directly and through other marketing-related variables affects CS (Ailawadi et al., 2003).Nassar (2017) and Blackston (2000) suggest that market-oriented firms focus on CS to stimulate a positive attitude towards a brand and motivate consumers to pay premium prices.These factors positively affect CS and BE (Iglesias, Markovic, & Rialp, 2019;Kala & Chaube, 2018).Many past studies found a positive association between CS and elements of BE, while others found a correlation between satisfaction and overall BE (Tran, Vo, & Dinh, 2020;Zameer, Wang, Yasmeen, & Ahmed, 2019).
H6: CS and BE are positively associated.

Customer Satisfaction (CS) and Purchase Intentions (PI)
CS depends on customer value, which determines loyalty, retention, and PI (Dash, Kiefer, & Paul, 2021).At the same time, extant literature also suggests that customers' perceived value also affects PI (Hermanda, Sumarwan, & Tinaprillia, 2019).Given the importance of CS, most businesses ensure that their customers are fully satisfied with their goods and services as they promote PI.Peng and Basit (2018) stress that CS directly affects the PI and indirectly affects corporate profits, market share, and sustainability (Hossain & Zhou, 2018).
Satisfaction is an ongoing process that can change from time to time (Watanabe, Torres, & Alfinito, 2019).Many organizations tend to increase automation to improve efficiency.However, it reduces human interaction with customers.Human interaction is necessary for increasing customer satisfaction (Moslehpour, Wong, Lin, & Nguyen, 2018).Past studies have also stressed that automation, on the one hand, enhances efficiency and reduces CS.Besides satisfaction, past studies have widely studied CS and PI together since PI is a significant predictor of actual behavior.Jauhari, Kusumawati, and Nuralam (2019) suggest that PI mediates customer satisfaction and actual buying behavior.While studying the association between CS, marketers should also measure the percentage of PI that translates into actual buying behavior (Hossain & Zhou, 2018).In most cases, a higher percentage of PI converts into actual behavior, but there are exceptions.For example, in green marketing and high involvement products, the conversion ratio between PI and actual behavior is low (Peng & Basit, 2018).

Brand Image (BI) and Purchase Intentions (PI)
Besides communicating the value proposition, brands have brand personality, commonly known as brand personification (Agmeka, Wathoni, & Santoso, 2019).Consumers find that alignment between their personality and BI motivates them to buy.BI reduces a brand's search cost and motivates consumers to purchase brands with a strong image.BI also helps firms to communicate their value proposition to the target market (Kala & Chaubey, 2018).For example, BI is associated with upward stretched products (allied) while others with lower stretched products (economic and low cost) (Han, 2017).Thus, consumers of the lower-income strata know which brands are more appropriate for their needs and wants.
Similarly, consumers in the upper strata know which brand is good for them.BI strongly affects PI and does not frequently change (Martín-Consuegra, Faraoni, Díaz, & Ranfagni, 2018).Thus, a brand with an image of being an economical brand targeted to lowerincome should not target upper-income strata.If a firm wants to target upper-income strata with the same brand, it will be advisable to launch the brand with a new brand name (Temaja & Yasa, 2019).

Mediating Relationships
We have proposed eight mediating relationships in the study.The rationale for articulating the mediating relationship is the above discussions suggesting that each proposed mediating relationship's independent variable affects the mediating variable and impacts the dependent variable.

Conceptual Framework
Based on the preceding discussion, we establish a conceptual framework that includes several direct and indirect relationships.Figure 1 illustrates the conceptual framework of the study.

Data Collection Procedure
The study examines the effects of BI, BA, and CS on BE and PI.It also measures the mediating role of BE, BI, and CS.The study focused on the current insurance policyholders for data collection.Since the sampling frame was unavailable, the study used a non-probability sampling technique.We outsourced the data collection part to a local professional research firm.The enumerators distributed 500 questionnaires and received 424 useable responses.

Respondents Profile
The respondents comprise of insurance policyholders from Pakistan's metropolitan cities, i.e.Karachi, Lahore, Quetta, Islamabad, and Peshawar.72% of the respondents were males and 28% were females.Age stratification shows that 21% of the respondents were in the age group of 24-29 years, 29% in the age group of 30-35 years, 30% in the age group of 36-40 years, and the remaining 20% were over 40 years.Concerning income, 48% of the respondents were in the income group of Rs. 50,000 to 75,000, 30% were in the income group of Rs. 76,000 to 100,000, 20% were in the income group of Rs. 101,000 to 125,000, and the remaining 2% were in the income group above Rs.125,000.In terms of education, we found that 3% of the respondents had only matric level education; 17% had intermediate education; 57% had bachelor degrees, and the remaining 23% had master degrees or higher qualifications.

Measurement of Constructs
The survey questionnaire used in the study has two parts.The first part includes demographic-related questions, and the second part includes questions on five constructs adapted from the previous literature.The study measured BI through five items adapted from Low and Lamb (2000), BA has five items adapted from Spears and Singh (2004), customer satisfaction has three items adapted from Hellier et al. (2003), BE has five items adapted from Foroudi et al. (2018), and PI has five items adapted from Foroudi et al. (2018).All the variables were measured on a five-point Likert scale, where one represents highly disagree, and five represents highly agree.

Data Analysis
The study has used Smart PLS software for preliminary analysis and analytical testing of the proposed hypotheses using the Partial Least Squares-Structural Equation Modeling method (PLS-SEM).For estimating complex statistical relationships between latent variables, Hair et al. (2012) and other researchers believe that the PLS-SEM method is superior to the CB-SEM approach.

Descriptive Statistics
The study in descriptive statistics has computed mean, standard deviation, skewness, kurtosis, and Cronbach's alpha.Table 1 illustrates the descriptive statistics.  1 further demonstrates that the skewness values range between -1.10 and -0.70.At the same time, the kurtosis coefficients range from 0.60 to 1.25.Since the skewness and kurtosis values ranged between ± 2.5, we have inferred that the study's constructs do not deviate from the requirement of univariate normality (Hair et al., 1998).

Convergent and Discriminant Validity
Convergent validity reflects the logical association between the constructs.At the same time, discriminant validity shows the uniqueness and distinctiveness of the used variables.The study has assessed the convergent validity based on composite reliability and AVE.For discriminant validity, the study has used the Fornell and Larcker (1981) criteria.We have illustrated the results in Table 2.The results show that the composite reliability values ranged from 0.863 to 0.919, and AVE highest value is for BA (AVE=0.741),and the lowest for BE (AVE=0.617),suggesting the constructs do not deviate from the requirements of convergent validity (Cunningham, Preacher, & Banaji, 2001).The study found that the square roots of AVE are greater than the Pearson correlation values, suggesting that the constructs used in the study are unique and distinct (Fornell & Larcker, 1981).

Confirmatory Factor Analysis (CFA)
We used CFA to test how well the measured variables are associated with the latent variables.The results indicate that all the factor loadings are greater than 0.60, suggesting an association between the indicators variables and respective latent variables.

R-Squared
We have used bootstrapping for generating the results.Table 4 shows that the lowest adjusted R-squared value is for BI (Adj.R 2 =0.312) and the highest adjusted R-squared value is for PI (Adj.R 2 =0.432), suggesting adequate predictive power of the model

SEM Results
We have tested eight direct hypotheses and seven mediating hypotheses.Our results support all the hypotheses except one direct and one indirect hypothesis.The summary of the results is presented in Table 5.Also, refer to Figure 2 for the measurement model and Figure 3 for the structural model.Our results support all the hypotheses except the association between BI and BE and the mediating role of BE on BI and PI.

Discussion and Conclusion
Due to the highly competitive environment and consumers' low inclination towards insurance-related products, insurance companies in Pakistan are operating below their potential.As a result, their contribution towards GDP is lower than other developing countries.Thus, the insurance sector needs to build BE.Thus, we have proposed a model that contains five variables (i.e., BA, BI, BE, CS and PI).The proposed model has eight direct and seven indirect relationships.Of the 15 hypotheses, our result supports all the hypotheses except one direct and one mediating relationship.We found that BA and customer satisfaction promote BE and PI.And BE is a predictor of PI.However, BI has contradicting effects.On the one hand, it does not affect BE, but on the other hand, BI has a significant association with PI.The results also suggest that BI, BE, and CS mediates BA and PI.Also, BI and CS affect BE.However, BE does not mediate BI and PI, but BI mediates CS and PI.

Implications
Some researchers assert that since BE lacks managerial usefulness, marketers' efforts may not enhance BE.Many marketers have enhanced BE through the trial and error approach, which is dangerous for a brand (Faircloth, Capella & Alford, 2016).Based on empirical evidence, this research supports how a firm can enhance its BE.The study has the following advice for marketers.First, marketers can enhance BE by independently working on it.Second, the managers should realize that its utility is beyond measuring the value of a brand.Many conventional marketers still use BE as a balance sheet asset, which provides valuable information but not strategies for creating and enhancing BE.The study found that BA and CS affect BE "positively." Also, the literature suggests that consumers have a strong willingness to pay premium prices for the brand with strong equity.Therefore, firms should spend considerable resources in building and maintaining BE.
Although our results do not support the association between BI and BE, still their alignment is necessary for sustainable growth.The insurance companies in Pakistan still focus on conventional personal selling.Hence, we found no association between BI and BE.Thus, the study advises insurance companies to manage BA and BI, enhancing sustainability and BE.Insurance companies should focus on BA as it promotes brand association and BI.Service sector success depends on customer satisfaction; therefore, we suggest that insurance companies launch innovative products that may satisfy consumers' needs.

Limitations and Future Studies
The research on insurance firms is restricted to five metropolitan cities, indicating a need to explore customer behavior and attitude in other cities of Pakistan.Due to the unavailability of the sample frame, we have used non-random sampling.Other researchers can make efforts to obtain a sample frame and use random sampling.
Researchers can use the developed conceptual framework in other service and none service sectors.A comparative study between the two sectors may also bring more insight into the phenomenon of BE.Pakistan has a diversified culture.Therefore, incorporating culture as a construct in Asian studies is important.

Figure
Figure 2: Measurement Model