INTRODUCTION

Barone et al1 define cause-related marketing (CRM) as a strategy that is designed to achieve marketing objectives through supporting social causes. Unlike sponsorship of a cause, CRM involves a commercial transaction from the consumers’ side. With CRM, the company does not give towards the cause unless the consumer engages in a revenue-providing exchange with the company whereas with sponsorship the company gives with the hope that it will induce a positive consumer reaction.2 Brands use CRM to influence consumer choice by capitalising on the halo effect of associating with a cause and in essence borrow from the equity and goodwill of these causes. The extent of this influence has been tied to a number of variables including the fit of the cause to the brand, the level of involvement of the purchase decision and the price premium perception. In terms of the decision-making process, Strahilevitz and Meyers3 highlight the positive influence of bundling charity donations with products and how this offers consumers two distinctive positive outcomes called affect-based complimentarity. Firstly, they get the product they want and secondly, the added utility of satisfaction from giving to the charity. Similarly, Linville and Fischer4 found that consumers prefer to have two positive outcomes together than separately. CRM also assists consumers to distinguish between responsible and non-responsible organisations1 and therefore aids in the consumer's decision-making process.

It has been shown that the use of CRM messages can be valuable in making advertising more effective. For example, the inclusion of a cause claim in a message significantly increases unaided recall of an advertisement.5, 6 Nan and Heo7 recently found that consumers have a more positive response to a brand when the advertisement includes a CRM message regardless of how good the fit between the cause and the brand. Nan and Heo7 also found that when consumers have high brand consciousness, the brand/cause fit becomes more relevant. Different results have been found in other studies indicating that consumers’ attitudes to brands are unaffected by the CRM message. This suggests that the presence of a CRM message does not necessarily influence the consumers’ evaluation of a product or brand, but that it influences conation or purchase intention.8

CONSUMER INVOLVEMENT WITH THE CAUSE AND MESSAGE FRAMING

Two of the key factors in CRM are the level of involvement a consumer has with the cause and how the message is communicated. 9 According to Laaksonen,10 consumer involvement accounts for the differences in the degree of mental and physical effort a consumer is willing to devote to consumption-related activities. An important aspect of CRM is the level of involvement a consumer has with the cause which is attached to the brand.11 The more important a cause is to a consumer, the more positive their response is to CRM.6 Broderick et al11 argue that the level of involvement a consumer has with the cause influences the intensity with which a consumer will process a charity marketing message from a brand which is associated with the cause. A high level of involvement with the cause translates into a more intense interpretation of the message and higher likelihood to participate. In advertising and CRM campaigns, Zaichowsky12 defines involvement as ‘a person’s perceived relevance of the advertisement based on inherent needs, values, and interests’. Based on the literature therefore, it is proposed that the degree of familiarity or involvement with a cause or charity is a key influencer of CRM.

Grau and Folse9 found that message framing is also an important factor in CRM campaigns. Message framing relates to whether a message is focused on positive or negative factors. While some researchers argue that positive framing is more effective,13 others support negative framing to improve outcomes.14 Consumers with low cause involvement are more inclined to respond to messages which are framed positively and communicate gains as opposed to highly involved consumers who respond more to risk avoidance type messages (that is, negatively framed messages).9 Ellen et al15 found that the immediacy of the cause also influenced a consumer's response. A message framed to address a problem such as a natural disaster, for example, was more positively received than one supporting an ongoing cause.

Relatively limited research has been done in the realm of CRM within the South African context and in particular regarding the criteria for a successful campaign. In their study, Human and Terblanche16 found that race had some influence on how consumers perceived messages in the South African context. They found that some whites hold residual guilt from Apartheid while successful black females feel especially guilty about their success when compared with many of their less successful counterparts. These feelings of guilt meant that respondents were more responsive to messages which in some way allowed them to alleviate some of this guilt by doing something positive for those less fortunate. It is possible that these societal idiosyncrasies may in fact lead to different results compared with what obtains in other countries. This study therefore investigates how both cause involvement and message framing interact in the South African context. Based on prior research findings, the expected result is that positive message framing is more effective for low-involvement consumers, but given the inherent guilt present, positive framing may also be preferred by high-involvement consumers.

UNDERSTANDING CHARITABLE GIVING

In trying to understand how CRM campaigns work, it is also important to understand charitable giving and the factors that influence how charitable an individual may be. Charitable giving is described as a process of exchange involving both economic and social values that is performed for both selfish and altruistic motives.17 Factors that influence charitable giving include income levels. Higher levels of income and wealth lead to higher absolute levels of charitable donations.18, 19, 20 Research findings on the effect of income on the propensity to donate or the relative proportion of income donated are conflicting.21 For example, some findings indicate that the highest and lowest income earners tend to give the greatest proportions of their incomes to charity.22, 23, 24 Other studies suggest that there is a negative relationship between income level and charitable giving.25, 26, 27 However, some researchers, including Douglas,28 argue that charitable giving does not depend on one's class, education, language, race or colour, but is linked to the cultural environment and may be subject to temporal and geographic variation. Therefore, there are many variables that need to be considered when evaluating the decision process and propensity to donate, including but not limited to, how charities ask for support, what media is used, how strong the brand of the charity is and the demographics of the target market.29

RESEARCH HYPOTHESES

According to the Elaboration Likelihood Model,30 the importance of the cause to the consumer, or their level of involvement with the cause, may alter how they respond to a CRM campaign. 11 Based on the literature reviewed, we hypothesise that:

Hypothesis 1:

  • High cause involvement influences purchase intention positively.

Research shows that the way in which a CRM message is framed9, 16 influences the consumer's response to it. Chaiken31 contends that individuals with high levels of involvement are more likely to process a message in detail than those that are less involved. Also, negatively framed messages are more persuasive for highly involved individuals.9, 14 Further, consumers with low cause involvement respond better to positively framed messages.9 An objective of this research was to determine the influence of a cause message framed to induce feelings of guilt (negatively framed message) and a cause message framed to induce positive feelings or associations (positively framed message) and how this interacts with cause involvement and most importantly how this influences the action taken, in this case the propensity to purchase the product advertised. According to Chang and Lee,32 messages that are framed negatively in the context of charitable giving arouse self-relevance and consciousness and induce feelings of guilt and the need to reduce the loss that will occur if no action is taken. Different factors influence an individual's desire or need to give to charity including perceptual reaction,33 that is, how an individual perceives a cause or charity. Perceptual reaction is influenced by self-perception and how supporting the charity/cause fits with this perception. Other influencing factors identified by Sargeant33 are self-esteem, guilt, social justice, empathy, past experiences and judgmental criteria. It is apparent that message framing is a relatively complex exercise and that many factors need to be taken into consideration if a desired response is to be induced. Based on the foregoing, the second and third hypotheses for this study are therefore:

Hypothesis 2:

  • Consumers with low cause involvement respond more positively to CRM messages framed positively.

Hypothesis 3:

  • Consumers with high cause involvement respond more positively to CRM messages framed negatively.

METHODOLOGY

Factorial ANOVA between-subjects with two independent variables (namely cause involvement and message framing) and one dependent variable (purchase intention) was used in the analysis. Each of these independent variables had two levels, hence the 2 (number of independent variables) × 2 (levels of each variable) design. The brand used for the purpose of this study is Irvin and Johnson (I&J), a brand within the South African fast moving consumer goods market. I&J is a frozen foods company that has been in existence for more than one hundred years and has enjoyed significant market share for the duration of its existence. I&J has been chosen as the brand for investigation because it has a strong heritage in the South African market and is well known to consumers. This avoids variations occurring due to lack of brand knowledge or experience with the brand. I&J is also a brand that provides fast moving consumer goods which are low-involvement purchases. By focussing on low involvement products it is possible to remove the influence of decision processes which may differ between high and low involvement purchases. Brand and cause fit are important when designing CRM campaigns. For this reason, a food-related cause was used in the form of a feeding scheme which provides food to hungry children. For every pack of I&J Fish Fingers purchased, a donation would be made towards a fictitious children's feeding scheme. The research cause has purposefully been kept fictitious in order to control for any bias which may be introduced through using well-known and recognised causes.7

SAMPLING

The research focussed on women as AMPS (All Media and Products Survey) data show that the primary target market for I&J is women and because women are typically the shoppers within a household. Quota sampling was used as the probability of one of these women being chosen is not known. Relevant shopping centres (strata) that cater for the brand target market within the Cape Town area were identified and individuals were selected using a simple random sampling technique. Starting from the time the interviewer was at the mall entrance, every fifth female was approached. Each interviewer was given either only the negatively framed stimulus (A print advertisement featuring an image of a box of I&J Fish Fingers with the message ‘Every night, thousands of children go to sleep hungry. Buy I&J Fish Fingers and we will donate one fish finger for every pack you buy to “Feed the Children”. Every pack you buy means a child does not go hungry.’) or the positively framed stimulus (The same advertisement with the message ‘A hungry child can go to bed full because of you. Buy I&J Fish Fingers and we will donate one fish finger for every pack you buy to “Feed the Children”. Every pack you buy helps feed hungry children in South Africa.’) in order to ensure simplicity. The sample size which was initially selected consisted of n=20 for each cell of the 2 × 2 factorial ANOVA matrix. However, the sample size realised was 10 cases per cell distributed as in Table 1.

Table 1 2X2 factorial ANOVA matrix
Full size table

To establish cause involvement, a 20-item seven-point semantic differential scale was used based on previous work done by Maheswaran and Meyers-Levy34 and Grau and Folse9 with adaptations from the Personal Involvement Inventory (PII) scale.12 This scale asked respondents to indicate whether the cause in the campaign is: Unimportant or Important, Means nothing or Means a lot, Is personally relevant or Is not personally relevant, Matters a great deal or Does not matter a great deal, Is of concern or Is of no concern, and so forth. The positive and negative statements varied in position from left to right to avoid a tendency to only read the adjective placed on the left. The PII scale was tested for validity and reliability. The Cronbach's α of 0.93 realized is comparable with the reported average of 0.95 to 0.97 of the 20-item PII scale used by Zaichowsky,35 indicating that the scale is reliable. Validity was checked using the ‘corrected item-total correlation’ and exceeded the 0.3 minimum suggested by Pallant36 for all items used in the scale.

Cause involvement was only determined after being measured because it was not possible to assign respondents to cells (high or low involvement) a priori. As per Zaichowsky,35 a middle point was assigned based on the data gathered in order to separate respondents into these categories. The data provided sufficient information to provide 10 high involvement and 10 low involvement respondents using this technique. Respondents with total involvement scores of greater than 32 were assigned to lower involvement and those less than 32 were assigned to higher involvement.

Respondents were randomly assigned to positively and negatively framed campaigns in order to avoid the influence of having to choose between the positive and negative campaign as this would influence results.9 The negatively framed message was designed to induce feelings of guilt while the positively framed message was designed to focus on the positive contribution the respondent can make. The advertisements were pre-tested with 10 randomly selected subjects not participating in the study to ensure that the campaign messages communicated as desired.

RESULTS AND DISCUSSION

Before testing for the main effects, the existence of interaction (that is, the effect of one independent variable depending on the effect of the other independent variable) between the independent variables was assessed. As reflected in Hypotheses 2 and 3 interaction was expected based on previous research.9 An examination of the profile plot of the results shown in Figure 1 suggested the existence of interaction between level of involvement and message framing due to the non-parallel lines. However, the results in Table 2 show that there is in fact no interaction between level of involvement and message framing with regard to influence on purchase propensity [F(1, 67.8)=0.212, P=0.648]. Hypotheses 2 and 3 are therefore rejected. Further, the R2 value (0.113) (Table 2) is relatively small indicating that some other factors account for the majority of the variance observed. This is contrary to Maheswaran and Meyers-Levy's34 findings that higher involved respondents responded better to negatively framed messages and lower involved respondents responded better to positively framed messages.

Table 2 Two-way ANOVA results
Full size table
Figure 1
figure 1

Profile plot.

Full size image

Although there is no interaction between the independent variables, there is a main effect caused by level of involvement [F(1, 67.8)=0.4.301, P=0.045]. This means that the level of involvement with a cause influences purchase intention. Hypothesis 1 is therefore not rejected (P=0.045). However, the effect size (that is, partial η2=0.107) caused by involvement is moderate according to Cohen's (1988) criteria.37 There is no main effect for message framing [F(1, 67.8)=0.053, P=0.819], meaning that message framing does not influence purchase intention in this study.

CONCLUSION

Considering the growing importance of CRM in emerging economies that face many social issues, it is imperative to understand what factors influence the success of CRM campaigns and how these are perceived. This study sought to provide a better understanding of how South African middle-income consumer perceives CRM; specifically how their level of involvement with a cause and how the CRM message is framed influenced their intention to purchase. Contrary to our expectations, we found that consumers with low cause involvement did not respond more positively to CRM messages framed positively; and that consumers with high cause involvement did not respond positively to messages framed negatively. However, we found that only the level of involvement influenced intention to purchase.

The findings of this study reemphasise that the context for the CRM campaign is an important factor. Being able to tailor a CRM campaign towards a defined audience can therefore be used to differentiate a brand and to build brand equity while giving back to society. Based on this study, it appears that cultivating consumer involvement with a cause is more likely to yield positive results (that is, intention to purchase) rather than just focusing on the message. This suggests that marketing effort should be directed at igniting and growing consumer involvement with social causes in order to maximise impact and campaign success.

Although a weakness of this study is that it was based on a small sample and limited to one city, it presents an opportunity for replication and improvement on a more spatially diversified scale.