Don’t tell me stories – the narratives of retirement and their relation with brand associations

Brand positioning based on sociocultural issues was defined by Rossatolos [2013] as a brand narrative. Narrative refers to “spoken or written account of connected events” [Sarpong et al., 2019], often including emotional appeals that shape how consumers use products to create meaning [Brechman and Purvis, 2015]. Brand narratives are also defined as stories, accounts, or tales [Ardelet et al., 2015].

Creation of meaning by employing a particular brand narrative is done by using different cultural codes, which are a collection of signs [Oswald, 2012], a system of words, symbols, or behaviors or “a hidden systems of meaning” used to carry messages set in a particular context [Hyatt and Simons, 1999]. Codes may be expressed at an observable level of culture, through verbal and nonverbal signs. In the following part of the paper, we explore the codes related to retirement and how they can be applied for creating brand narratives.

There are different cultural categories that could be investigated within the context of cultural branding, with retirement being mentioned as one of them [Holt, 2004]. The aging population has become one of the most crucial social antecedents of various shifts observed and it has implications for nearly all sectors of the economy [United Nations, 2019]. The number of older persons – aged 60 years or above – is expected to more than double by 2050 and to more than triple by 2100, rising from 962 million globally in 2017 to 2.1 billion in 2050 and 3.1 billion in 2100 [United Nations, 2019].

Due to the aging of the population, retirement patterns have gained considerable attention in many societies. Retirement has been investigated as a psychological concept [Laberon et al., 2019], where researchers have highlighted behavioral and psychological antecedents, outcomes, and psychological mechanisms underlying the retirement process [Shulz and Wang, 2011].

The end of productive life is inevitably connected with the cultural perception of “being old”. But old age seems nowadays more like a personal decision rather than a demographic metric, according to the study entitled “Predictors of Attitudes to Age Across Europe” by Euromonitor International [2011]. Greater longevity is changing the narrative of both aging and retirement.

According to marketing experts from companies serving older customers, there are two ways of approaching the issue: The first approach is based on the belief that, as they age, people get isolated and lose their interest in the world around them. The second approach, which is more modern, is based on the view that the oldest should keep themselves active for as long as possible and that they should seek out substitute activities to replace the ones they can no longer carry out [Euromonitor International, 2011]. Several surveys claim that a decent percentage (between 71% and 83%) of American baby boomers intends to keep working after retirement [Euromonitor International, 2011]. There is even a new language to describe the phenomenon of people who refuse to retire – they are called “nevertirees” [Foxon, 2014]. The evolving landscape surrounding retirement has changed its meaning and nature [Henkens and van Solinge, 2021] and as a result must be reflected in the communication built around brands targeting older customers.

One of the concepts to investigate the effect of brand narrative built around the product are brand associations [Boisvert and Burton, 2011] that provide information about either positive or negative image rooted in the brain memory recalled by the consumer [Emari et al., 2012]. The research on brand associations has been already developed by many scholars who investigated how they affect the consumer response [Río et al., 2001], the positive relationship between the brand perception and willingness to buy the product [Chan et al., 2013], as well as the relation between the brand image and the willingness to pay the premium price [Anselmsson et al., 2014 ]. Moreover, the brand association serves as a consumer's support to process and retrieve information in order to differentiate or position the brand [Till et al., 2011]. Finally, brand association provides a purchasing reason, as it is related to brand attributes, the target consumer market, and the benefits brought by possessing the product [Len et al., 2007].

As already discussed, the concept of brand associations provides more insights into brand valuation as presented by Buil et al. [2008], who introduced the detailed view of measuring the brand association by including items such as perceived value (PV), brand personality (BP), and organizational associations (OA). That is the approach we follow in our study when investigating the relations between various brand narratives of retirement and brand associations.

Previous studies have shown that brand narratives foster building particular brand image [Dahlen et al., 2010]. However the question of whether it is true, despite the cultural content of these narratives, remains open. As cultural ideas and codes change over time, so can brand narratives, and the relation between these dynamic changes and brand associations remain for further investigation. Williams [1981] has offered the concept allowing to capture how cultural codes evolve over time, indicating three types of codes: residual codes – past cultural ideas and their representations, dominant – current, contemporary cultural ideas and their representations, and emergent – new, fresh cultural ideas and how they are presented [Mahtabi and Eslamieh, 2015]. This notion provides a framework that allows to investigate and to understand how sociocultural paradigms evolve over time [Mahtabi and Eslamieh, 2015].

We may observe such evolution of cultural codes regarding retirement, which was earlier conceptualized as an extensive unhappiness caused by the feeling of having not much to live for, and now evolved to the stage where retirement is treated as a new beginning where the life beyond work is discovered [Kloep and Hendry, 2006], with the shift toward a positive view of retirement often defined even as a period of freedom [Apouey, 2020]. Therefore, the growing complexity of the cultural ideas of retirement created the need for seeking a renewed way of thinking about aging [Coleman and O’Hanlon., 2017]. Thus, the objectives of the narrative are built around shaping the decisions of aging individuals transitioning into the new meanings of retirement created through contemporary discourses of “positive” aging [Rudman, 2015].

The remaining question is whether consumers tend to reject emergent cultural ideas in branding and opt for old-fashioned, traditional narratives? Or, alternatively, do they choose well-known contemporary stories? Previous research has shown that older and mature consumers remain attached to product options that they have been using for a long time and may be less prone to try or buy new products [Lambert-Pandraud and Laurent, 2010]. But, what if an established brand offered an innovative, progressive narrative, based on the emergent cultural trend? Would it be less acceptable for consumers than the traditional narrative based on the residual cultural trend? To understand more completely the key tenets of using brand narratives and their impact on brand-related effects the following hypothesis is proposed:

H₁: The specific cultural brand narrative of retirement has an impact on brand associations.

Age is reported to be one of the drivers of consumer behaviors and therefore many scholars have already mentioned the importance of addressing the needs of older adults in marketing strategies [Nataraajan and Bagozzi, 1999]. Such an approach is rooted in using the aging process as one of the explanations for changing preferences [Gregoire, 2003], as the retirement experience may be reflected in postretirement lifestyle approaches and consumption changes [Hopkins et al., 2006]. As mentioned by Lambert-Pandraud et al. [2005], we may distinguish four main mechanisms that are the drivers of changes in consumer behavior over time. These are biological aging (responsible for decline of physical capabilities), cognitive decline (responsible for limiting the number of information analyzed in purchase process), socioemotional selectivity (growing as the time perspective is getting shorter which is manifested by using the emotions and feelings as the purchase drivers), and change aversion (represented by lower willingness to choose a new product, even if the old one is not satisfactory). Furthermore, the preferences exhibited by an individual are regarded as a part of that person's consumption history – which is longer in an elderly individual – and not as just a behavior at a given point in time [Frytak et al., 2003]. Hence, the analysis should focus on causal pathways of consumption activities and their determinants across the entire life span of the individual [Moschis, 2007].

The research on differences in the decision-making process on brand choice has revealed that elderly customers are more likely to repurchase and actively resist switching brands once a favorite has been established [Karani and Fraccastoro, 2010]. Cole et al. [2008] have highlighted some drivers that affect the brand preferences that vary with age: a habit mechanism that develops with age and decreases in willingness to innovate. Similar results were obtained by Sikkel [2013], who showed that among elder customers a decreased tolerance to ambiguity, consistent with low cognitive–affective complexity results in a higher need for longer, well-established relationship with the brand and affects its associations, as the consumers were seeking for brands with a long history and presence on the market.

The patterns observed across previous studies allow to draw inferences regarding the differences related to aging: in consumer preferences [Gregoire, 2003], in brand switching and willingness to repurchase the brand [Karani and Fraccastoro, 2010], in decreasing the willingness to innovate [Cole et al., 2008], and in the increasing preference for older, well-established brands [Yoon et al., 2009].

It calls for further investigation to better understand how these consumers act on the basis of various brand narratives built around different cultural codes. We assume that age would moderate the relation between particular retirement brand narratives and brand associations. Therefore the following specific hypothesis has been proposed:

H₂: Age moderates the impact of retirement brand narrative on brand associations.

Diversity in modes of retirement has already been investigated by some scholars providing the exploration of distinct ways of conceptualizing and experiencing retirement [i.e. Hornstein and Wapner, 1985]. These conceptualizations were recently examined to reveal the impact on personal wellbeing during retirement [Bačová and Halama, 2020]. However, there are several factors having an influence on the individual perspective, with the cultural representations of aging and retirement being dominant. Diversified cultural codes of retirement were put together by Oswald [2012], who defined three types of codes: residual (retirement as “the clinic”), dominant (“the cruise”), and emergent (“the road trip”) that are being used in different contexts by narratives employed by the brands. The details of the codes are presented in Table 1.

These codes create a cultural landscape [Holt, 2004] in which companies can search for inspirations for their brands’ narratives. Companies offering products related to retirement should reconsider its changing meaning in terms of their decisions of how to communicate brands, products, or related services. The changing codes of retirement serve as the foundation for various brand narratives. Based on our previous research [Hajdas, 2019], in this study we investigate the conservative narrative (based on residual “clinic” code), contemporary narrative (based on dominant “cruise” code) and progressive narrative (based on “road trip” code). That is the approach we further explore.

Our research model is presented in Figure 1.

Figure 1

To investigate the model proposed, we have performed an experimental study with different age groups, based on the financial product (voluntary pension saving account). The experiment design, procedure, and results are described in the following section.

To test our research hypothesis H₁ and H₂ the experiment has been used based on investigating the separate and joint effect of two factors: age (with three groups distinguished −30 to 39, 40–49, 50–60) and three brand narratives represented by three types of codes (residual: “the clinic”, dominant: “the cruise”, and emergent: “the road trip”). The proposed model was tested using a survey-based quantitative method. We have decided to use the financial product – a voluntary pension saving account, as it has already been reported that retirement financial services providers have been very active in the marketing communication tools usage [Koehler and Mercer, 2009]. Moreover, as revealed by Estelami [2009] and Huhmann and McQuitty [2009] the financial services advertising employs some communication approaches with limited financial information, providing instead some biased perceptions of financial offers. Such a perception may be rooted in the cultural code used to build the particular narrative.

Four versions of a print ad were prepared for the purpose of this experiment. The main layout of all four versions of the ad was the same; including the same description of a saving account: A voluntary pension saving account is a solution for you if you want to voluntarily accumulate capital for the future. It is an interest rate account, where you accumulate additional money for retirement. When you reach the age of 65, you can withdraw the money anyway you like – in whole or partially. Differentiated paragraphs of the body text were used to describe the particular narrative.

In the control group the following text was used: Retirement can vary. It will depend on your condition and money.

For the experimental Group 1, the body text was supplemented by the content representing the residual code: Retirement can be difficult. Lower income and more worries – shopping, medications, bills. It's a time when we still want to be physically and financially independent, but that doesn’t always work out.

For the experimental Group 2 the body text was supplemented by the content representing the dominant code: Retirement is a time of well-deserved rest. It's the longest vacation when you don’t have to do anything anymore. It's time for yourself, for your relatives. Books? Gardening? A walk in the woods? Retirement is a time to relax.

For the experimental Group 3 the body text was supplemented by the content representing the emergent code: Retirement can be an adventure. It's a time when you don’t have to do anything anymore and can take full control of your life. Perhaps it's a time for the trip of a lifetime? Or pursuing a long-postponed passion? A new sport? A new home? A new business? Retirement is a time that belongs only to you.

All participants, after being randomly assigned to a respective stimulus, were given time to read the advertisement which was displayed by the whole time of fulfilling the questionnaire, and then were asked to answer series of 16 questions on brand association measures and in the second section – 7 questions related to the characteristics of the sample.

We have collected a random group of 432 respondents, divided by age into 3 subsets with the following age intervals: 30–39, 40–49, and 50–60. Each subset had 1/3 of the total number of participants (144). All 144 respondents were randomly assigned to one of the groups: control group assessing brand associations with cultural-neutral narrative, Group 1 assessing brand associations with narrative based on “the clinic” code, Group 2 based on “the cruise” code and Group 3 based on “the road trip” code. Finally, each experimental group consisted of 36 respondents. The final sample of 432 randomly chosen participants consisted of 218 female and 214 male respondents. All participants had at least a bachelor's degree, and almost equally represented different regions of Poland, where 94% of the sample lived in towns with over 100,000 inhabitants; 76% of participants have an average monthly income below 1,168 EUR (1,287 USD), while 16% earn between 1,168 EUR (1,287 USD) and 1,752 EUR (1,930 USD), and the remaining 8% earn more than 1,752 EUR (1,930 USD). All participants have established an employment situation, and report having a fulltime job.

Based on the literature review, we generated the scale items and constructs to measure the brand associations. Following Churchill's paradigm [1979], we provided a multi-item scale for each construct analyzed. To ensure high reliability, we used the scales that have already been validated. In all cases, 7-point Likert-type questions were used with 1 – strongly disagree and 7 – strongly agree. A list of the items used to measure each construct is provided in Table 2.

Employed scale was tested for reliability using Cronbach's alpha and the results are presented in Table 3.

Cronbach's alpha for the entire scale equals 0.954, the coefficients for each construct equals the same value of around 0.9, and for the items it varies between 0.736 and 0.917. These coefficients meet the required criteria, that is, ranging between the recommended 0.70 and 0.95 [Taber, 2018]. Excluding any of the items from the PV and OA would lower the Cronbach's alpha. The only exception is observed for BP construct, where excluding the third item, I have a clear image of the type of person who would use brand X, would increase the Cronbach's alpha by 0.029 (i.e. 0.876 vs. 0.847); but as the difference is insignificant and it cannot alter the final result, we finally decided to include the item in the analysis. Based on the above analysis of Cronbach's alpha, we can assess the scale applied in this study as highly reliable.

To check the manipulation effect of our 3 versions of narratives, 4 groups of 48 respondents each were randomly chosen, and each group has seen one version of the ad. All participants answered the same series of the following question: Do you recognize the ad as conservative/contemporary/progressive. Next, respondents matched the answer on the 7-point Likert's scale (where one is strongly disagree, and seven is strongly agree).

To test the effect one-way ANOVA was applied. The means differed significantly on the level of 0.05 in the control group (F = 5.066, p-value = 0.010), Group 2 (F = 7.856, p-value = 0.001) and Group 3 (F = 7.892, p-value = 0.001). No difference between means was observed in Group 1 (F = 0.517, p-value = 0.599). The results meets our expectations partially, that is, narrative 3 (progressive) was recognized as more progressive and contemporary and statistically less conservative, narrative 2 (contemporary) was recognized more contemporary and progressive, and less conservative. However, narrative 1 (conservative) has no statistically significant effect on the respondents’ evaluations and a version presented to control group (with cultural-neutral narrative) was assessed as more contemporary and progressive and less conservative. The results of manipulation check are presented in Table 4.

We tested the potential sources of bias that can be caused by such factors as: gender, income and place of living. No differential effects were observed on gender (PV: F = 1.378, p-value = 0.241; BP: F = 3.171, p-value = 0.075; OA: F = 1.422, p-value = 0.233) and monthly average income (PV: F = 0.417, p-value = 0.796; BP: F = 0.915, p-value = 0.455; OA: F = 0.861, p-value = 0.487). An effect in PV and BP was observed on the town size (PV: F = 4.167, p-value = 0.0016; BP: F = 5.763, p-value = 0.003), however, between groups comparisons based on the Tukey's test did not allow to distinguish any apparent tendency in the evaluations, and therefore cannot result in any apparent tendency.

Additionally, we tested the equality of variances of brand associations constructs: PV, BP and OA (i.e. we compared variances of four stimuli groups in each construct) to investigate whether different brand cultural oriented narratives significantly influence the variances observed within groups. Tests for homogeneity of variances of stimuli groups are presented in Table 5.

The p-values reported in the last column of Table 5 give no evidence to reject the null hypothesis about the equality of variances, and on these bases we concluded that the variances are equal. We sustain that changing the brand narrative does not influence the shape of the distribution of brand association constructs, which is expected by the analysis.

We analyzed the different possibilities to aggregate the scale. We considered the factorial analysis versus simple average. Correlation coefficients between construct items are high and almost equal, that is, for the BP construct correlations between items equals on average 0.75, for PV 0.66 and for OA 0.85 (these coefficients are reported in Table 6).

Factorial analysis yields one factor for each constructs (based on the Kaiser's criterion of eigenvalue magnitude). The correlation between factors and simple average of items is equal for all constructs for the same value of −0.99. Having received these results, we decided to apply in further analysis simple average for the following reasons. First, as the correlations coefficients between items are almost equal, the weights exploited by factorial analyses are almost equal. Second, we prefer to analyze the result within the original 7-point scale, than within transferred scale yields by factorial analysis. Third, as the correlation coefficients between factors and simple average of items are close to 1 (in absolute values) the results of the ANOVA/MANOVA cannot change, no matter which variable is used. To test the effect of the age and brand narrative the items for each brand associations were averaged, and three variables, namely average perceived value (avPV), average brand personality (avBP), and average organizational associations (avOA) where further analyzed.

The brand association evaluations were analyzed within 4 × 3 = 12 experimental groups using two-way MANOVA (i.e. within groups gained from crossing four narratives: culture-neutral, clinic, cruise, road trip, and three responders groups distinguished by age: 30–39, 40–49, 50–60). The results of MANOVA are displayed in Table 7, while the means and standard deviations of the constructs are presented in Table 8.

In this experiment, average assessments of the brand associations constructs do not differ within and between the experimental groups. The average assessment in each experimental groups oscillate around middle of the 7-point scale, and standard deviation is very low and almost equal, and therefore all statistics result in no significant effect on the level of significance equals 0.05. A MANOVA yielded no significant effect on brand narrative, the age group, and interactions between factors as well.

Based on the results of the experiment tested by MANOVA we found no evidence in the sample to confirm H₁ (p-value = 0.365 for both Wilks’ Lambda and Hotelling's T² tests). As a result we found that brand associations as jointly measured by avPV, avBP and avOA do not differ by the type of the brand narrative. Therefore, as the level of significance equals 0.05, we cannot sustain that the cultural oriented brand narrative of retirement has an impact on the brand associations. Next, based on our results, we found no evidence to confirm the second research hypothesis, H₂ (p-value = 0.202 for relevant statistics). Brand associations as jointly measured by avPV, avBP, and avOA do not differ by the age of the experiment participants: 30–39 years, 40–49 years, and 50–60 years old. Similarly, as the level of significance equals 0.05, we cannot confirm that the age of the respondents moderates the impact of the particular brand narrative of retirement on the evaluation of brand associations. Furthermore, no significant interaction between the two factors: brand narrative and age was found on the level of 0.05 (p-value = 0.11 for respective statistics), which supports the conclusion that in our experiment, age and brand narrative do not interact in the evaluation of brand associations.

We believe our results may serve as some inspiration for managers considering applying brand narratives in their brand strategies.

First, our results show that the only component of brand associations affected by the application of particular brand narratives based on various cultural codes is BP. Therefore managers who aim at creating distinctive BP could consider using brand narratives in their brand communication strategies.

Second, our results show that evaluation of BP based on brand narrative differs according to respondents’ age, therefore managers trying to target various demographic segments could consider differentiating the narratives, respectively, age segment of 30–39 evaluated BP under the “road trip” narrative statistically lower, whereas people aged 40–49 years evaluated BP under the same narrative statistically higher. This suggests that age and possibly the stage of family life cycle make people more or less prone to “adventurous” narratives. Our results also indicate that people aged 50–60 years old, who are soon to be retired, evaluate BP under the “cruise” narrative statistically higher compared with the remaining narratives.

We encourage managers to use these suggestions when developing and testing various narratives for particular market segments.

Our study has some limitations. It is based on an experiment with one set of narrative stimuli (verbal only). Using image in the stimuli might extend the findings. The study may also be limited by the use of a single product (voluntary pension saving account), therefore different types of products within the high involvement product category could be further investigated. Other avenues for research remain open in the field of product category context and the relations between the brand narrative and brand associations. Further investigation could explain whether such relations exist in the case of low involvement product categories. Expanding the scope of research would allow to understand better the impact of retirement brand narratives on brand-related effects.

Another future research direction is connected with the constructs measured. Our research revealed that in the case of voluntary pension insurance, the retirement narratives used in experiment did not impact brand associations. Future study could be devoted to other dimensions of brand equity [Aaker, 1991], as brand associations are one of its constructs. Further research could revolve around brand performance indicators examined by De Chernatony et al. [2004] for financial services industry (brand loyalty, consumer satisfaction and brand reputation).

Finally, taking into account multiple research calls for a more socially oriented perspective in branding studies [Golob et al., 2020; Swaminathan et al., 2020], future research could shift away from addressing brand-related metrics into exploring and explaining how using various brand narratives affects consumers’ self-esteem [Machin et al., 2019], empowerment [Hajdas and Kłeczek, 2021] or create a social change [Naidoo and Abratt, 2018].