South African consumer sentiment towards marketing : A longitudinal analysis

Business performance variables such as productivity, customer satisfaction and marketing effectiveness have traditionally been measured and managed at the micro-level. And for good reason. In the ‘old’ economy the emphasis was on mass production and consumption of commodities (Fornell, Johnson, Anderson, Cha & Bryant, 1996:7) and competition was mainly inter-firm competition. In the modern economy the production and consumption of products are increasingly differentiated (Fornell et al., 1996: 7) and competition, which is increasingly of a global nature, is characterised by slower growth in mature markets (Fornell, 1992:7).


Introduction
Business performance variables such as productivity, customer satisfaction and marketing effectiveness have traditionally been measured and managed at the micro-level.And for good reason.In the 'old' economy the emphasis was on mass production and consumption of commodities (Fornell, Johnson, Anderson, Cha & Bryant, 1996:7) and competition was mainly inter-firm competition.In the modern economy the production and consumption of products are increasingly differentiated (Fornell et al., 1996: 7) and competition, which is increasingly of a global nature, is characterised by slower growth in mature markets (Fornell, 1992:7).
Against this background attempts have recently been made to measure marketing perceptions (the Marketing Index), productivity (the OECD Productivity Index) and customer satisfaction (the Swedish and later the American National Customer Satisfaction Index) at macro rather than at micro level.The objective of the American Customer Satisfaction Index, for instance, is to provide a customer-based measuring system for evaluating and enhancing the performance of firms, industries, economic sectors, and eventually national economies (Fornell et al., 1996:7).Measuring and quantifying consumer perceptions and attitudes at a macro level is thus useful, as they are a precursor of future consumer behaviour and particularly buying behaviour.
Marketing is a discipline that is particularly concerned about its image (Gaski & Etzel 1986) because it often faces severe criticism.Much of this criticism has been focused on the role of intermediaries (retailers and wholesalers) in marketing, and on the role of advertising.Critics argue that intermediaries simply add costs to a final product without adding any value to the process of getting products to consumers.Advertising is criticised for making consumers buy products they do not really need, and for often being misleading and untruthful.Others have argued that packaging and labelling are often deceptive and that packaging pollutes the environment (Lamb, Hair, McDaniel, Boshoff & Terblanché, 2000).These rather negative sentiments are determined by consumer perceptions and attitudes and eventually influence their buying behaviour.

Consumer sentiment
Consumer attitude or sentiment is a 'learned' experience that is formed when the consumer experiences something (concrete or abstract), evaluates it on the basis of the available information, and according to his or her own perception of it, forms a specific predisposition.Because an attitude is a predisposition it can be assumed to influence consumers in their behaviour, which could be a response that is either favourable or unfavourable (Du Plessis & Rousseau, 1999:218).This response in the buying situation involves the consumer making a decision on whether or not to purchase.
The outcome of the buying decision will depend on many influencing variables both individual and environmental.Environmental variables of specific importance to marketers include economic demand factors, and business and marketing influences.Economic demand factors refer to potential purchasing power based on availability of money, whereas business and marketing influences refer to direct contact that the consumer has with firms through, for example, advertising and experience of the product (Du Plessis & Rousseau, 1999:81) such as a sample or a test drive.More specifically marketing influences or stimuli can be described as elements of the marketing mix or the four Ps: product, price, place and promotion and relate to how consumers respond to changes in these elements of the marketing mix (Kotler, Armstrong, Saunders & Wong, 1996:270).
Marketing planning and the execution of a marketing strategy are at the best of times demanding assignments.Doing so in a turbulent environment is a challenge, to say the least.

The South African marketing environment
South Africa has faced tremendous political changes over the past ten years and its business sector and marketers in particular have not been insulated from the impact of these changes.Deteriorating exchange rates, alarmingly high interest rates, a deluge of new government legislation, a growing informal sector, high unemployment, the threat of Aids and increasing crime are but a few examples of the changes and threats facing the South African public and business sector.Changes in external forces translate into changes in consumer demand for consumer and industrial products, as well as services.Not only do these forces result in changes in consumer demand, but also contribute to changes in consumer perceptions and attitudes.
Variables such as unemployment, inflation, interest rates, exchange rates and the economic growth rate (GNP) can be seen as indicators of the state of affairs in the South African economic environment.Willingness to spend or save depends to a large extent on levels of consumer confidence (or lack thereof) in an economy (Du Plessis, Rousseau & Blem, 1995).These environmental factors will thus exert an influence on consumer behaviour, attitude and perception.For example economic factors such as high interest rates can lead to a decrease in consumer spending (Marx, Van Rooyen, Bosch & Reynders, 1998:50).It is generally accepted that one of the most important determinants of buyer behaviour for the economy as a whole is the income of the spenders (Du Plessis & Rousseau, 1999:345).Inflation, interest rates and prices are thus important factors that determine just what that income is worth in terms of buying power.
Due to the constant interaction that occurs between the firm, the consumers it targets and the extraneous variables in the environment, it is important to consider that changes in the environment continuously exert an influence on consumer sentiment and thus behaviour, in terms of how consumers react and perceive the marketing of products and services.

Objectives of the study
In a longitudinal study respondents are questioned on the same topic at different time periods with the purpose to '… examine continuity of response and to observe changes that occur over time' (Zikmund, 2000:178).The primary objective of this longitudinal study is to measure South African consumer sentiment towards marketing and to compare the results with previous studies done in 1990 and 1994.Longitudinal studies of this nature, whereby researchers survey several different samples at different times are sometimes called cohort studies, because similar groups of people (in this case South Africans) are expected to be in each sample (Zikmund, 2000:178).Secondary objectives include an investigation to establish to what extent this sentiment was influenced by demographic variables such as age and gender to facilitate comparisons with similar analyses done in the previous studies.The comparison will reveal whether consumer sentiment has evolved over time and an attempt will be made to identify environmental occurrences that could be responsible for the changing sentiment.

Measuring South African consumer sentiment towards marketing
During 1990 and 1994 empirical studies were conducted to measure the South African consumer sentiment towards marketing (referred to as 'the 1990 study ' and 'the 1994 study' henceforth).The results of these studies were published during 1992 (Boshoff & Du Plessis, 1992) and 1995 (Boshoff & Du Plessis, 1995) respectively.The overriding conclusion of the first study was that South African consumers' attitude towards marketing is slightly negative and '….ought to be a source of concern to the entire marketing fraternity' (Boshoff & Du Plessis, 1992).Older consumers in particular voiced their disapproval.Pricing was singled out as the most important bone of contention.The pricing and pricing practices used by marketers were seen as often unfair and unreasonable.Advertising was also viewed rather sceptically.The only marketing mix element that escaped reasonably unscathed was distribution or retailing.Retailing, defined in the study as the role and quality of service provided by retailers to buyers, was the only marketing mix element to return a positive index.
The second study was conducted in 1994 (the results were published during 1995) and assessed whether sentiments towards marketing had changed during the interim period (Boshoff & Du Plessis, 1995).The South African Marketing Index as well as three of the four individual indices had decreased considerably during the period.The advertising index had changed from negative to positive and the retailing index from positive to negative.The deteriorating attitude towards pricing was once again a major source of concern.The changes in the various indices were attributed to some extent to changing economic and political circumstances occurring in South Africa during that time.

Sample
Questionnaires were mailed to a random sample of 2000 respondents.An address list provided by ListSearch was used as the sampling frame.The sample frame consisted of individuals throughout South Africa, between the ages 18 to 60, earning in excess of R32 000 per annum.Even numbers of male and female, as well as English and Afrikaans speaking persons were randomly selected from the list.In total 305 useable questionnaires were returned and a response rate of 15.25% was realised.The demographic composition of the respondents in this study as well as the studies of 1990 and 1994 are depicted in Table 1.
A major difference regarding the respondents of this study, in comparison with the previous studies, is the age of respondents.The majority of respondents who participated in this study were in the age group 55 -64.A possible explanation for this is that older persons have more time to complete the questionnaires and are probably more willing to provide information.Although every effort was made to ensure a representative sample comparable to those used in the earlier surveys, there can be no doubt that the realised sample is skewed in terms of age and does not accurately approximate the SA population.(See limitations section).Also a slightly larger number of respondents were employers or self-employed compared with the previous studies.This could be attributed to the different population (address lists) used compared with the previous studies, which both used the SABC license holder lists as sampling frame.In comparison with the previous studies, a more even response between males and females was obtained, but a larger number of the respondents lived in or near an urban area/city.The level of education in the current study is slightly higher than in previous studies and more English speaking persons participated in this study than in previous years.

Measuring instrument
The questionnaire, which served as measuring instrument in this study, as well as the previous two studies, was originally developed by Gaski and Etzel (1986).It measures consumer sentiment towards marketing via attitudes towards the four major elements of marketing practice.These elements are product, price, promotion and personal selling/retailing.Statements were generated with the intention to measure the respondent's perceptions of each of the four elements, linked to a 5-point Likert-type scale.By marking a 5, respondents indicated strong agreement with the statement while a 1 signalled strong disagreement.
Besides ensuring comprehensibility through extensive pretesting, the instrument was subjected, by Gaski and Etzel (1986), to thorough reliability assessments during the original development phase.Cronbach's alpha was utilised for this purpose.The alpha values suggested a high degree of reliability (Gaski & Etzel, 1986).A principal axis common factor analysis with an oblique rotation confirmed the discriminant validity of the four scale categories of the questionnaire.The convergent validity of the instrument was also confirmed by the original developers.

Data analysis
Although the validity and reliability of the Marketing Index instrument was established by Gaski and Etzel (1986), two of their measures to assess the psychometric properties of the instrument were repeated for this study, namely the calculating of Cronbach alpha coefficients and an exploratory factor analysis.The data analysis occurred in three phases.Firstly, to evaluate the internal reliability of the Marketing Index instrument, Cronbach alpha coefficients were calculated.Secondly, the discriminant validity of the instrument was assessed by means of an exploratory factor analysis procedure.Thirdly, the various indices were calculated using the formula suggested by Gaski and Etzel (1986:73) and used as dependent variables in a series of five regression analyses (the marketing index and the 4 Ps).The following demographic variables served as independent variables: gender, age, marital status, level of education, home language, position in labour market and area of residence (urban versus rural).

Internal reliability
The internal reliability of the measuring instrument was assessed by calculating the Cronbach alpha coefficients using the computer programme SAS (SAS Institute, 1990).Both the final alpha values and the item-to-total correlations of individual items were compared (Tables 2 and 3) with the findings of the 1990 and 1994 South African studies and Gaski and Etzel's American study.The Cronbach alpha values of the various categories (product, price, advertising and retailing) are consistent with previous studies (Table 3) and all are above the 0.7 cut-off point suggested by Peterson (1994).
For the sake of brevity Gaski and Etzel (1986:74) removed the two (of the seven) items with the lowest item-to-total correlations in all four instances (product, price, advertising and retailing) during the purification process of developing the Marketing Index measuring instrument.By contrast all 28 items were retained in the three studies conducted in South Africa.
It is shown in Table 2 that two advertising items (Advert 1 and Advert 6) and two retail items (Retail 4 and Retail 5) returned item-to-total correlations lower than 0.4.This situation may have necessitated their exclusion from further statistical analysis.Three reasons mitigated against this course of action: (1) direct comparisons with the findings of the 1990 and 1994 studies would not have been possible; (2) only one item had a particularly low item-to-total correlation of less than 0.3 (Retail 5); and (3) all the alpha coefficients are above 0.7, confirming the reliability of the Marketing Index instrument.
For these reasons, and particularly the first, all seven items were retained for subsequent analyses.
Table 3 shows that all four Ps returned acceptable Cronbach alpha coefficients in this study, comparing favourably with those in the previous studies.

Validity of results
All 28 items were subjected to an exploratory factor analysis procedure to assess the discriminant validity of the measuring instrument.The computer progamme BMDP4M (Frane, Jennrich & Sampson, 1990) was used, specifying maximum likelihood as the method of initial factor extraction and a Direct Quartimin oblique rotation of the original factor matrix (Jennrich & Sampson, 1996).As the theory suggests (Gaski & Etzel 1986) that sentiment towards marketing will be determined by attitudes towards the four so-called marketing mix elements (product, price, advertising and retailing) the extraction of four factors was specified.Table 4 shows that all four factors returned eigen values above the customary cut-off point of one.It also shows all the items except PROD3, PROD6, PRICE4, ADVERT1 and RETAIL5 loaded to a significant extent (p > 0.3) on a different factor as expected.These results provide at least some evidence of the factorial discriminant validity of the measuring instrument.Individual item questions can be seen in Annexure A

An index of consumer sentiment towards marketing
To calculate the different indices, responses to all questions were scored from +2 to -2, with the most favourable answer receiving the highest score.Scores for each item in each category were importance-weighted, scored on a 5-point scale of extremely important (5) to not important at all (1).These responses were then fitted to the following formula, suggested by Gaski and Etzel (1986) to quantify consumer sentiment towards marketing practices.
The measure of overall consumer sentiment towards marketing is then:

w
where: Xij = scale item response i in category j, Wj = importance weight for the marketing mix category j, m = number of items in category (7), and n = number of categories (4).
This represents an attitude score for an individual respondent.The mean score across all respondents can then be calculated and converted to an index number as a measure of general sentiment.
Consumer attitudes or sentiment towards each of the four Ps and marketing in general are shown in Table 5 and graphically depicted in Figure 1.The SA Marketing Index has changed from -16.262 in 1990 and -34.125 in 1994 to +25.658 in 1999.
From Table 5 and Figure 1 it can be seen that the improvement in the overall marketing index is primarily due to considerable increases in the price and retail indices.The price index rose from -24.368 in 1994 to +13.764 in 1999 and the retailing index from -0.423 in 1994 to +22.258 in 1999.The product and the advertising indices have remained relatively stable over the period and the low values of these indices could indicate that a certain amount of scepticism still exists with regard to the credibility of advertising, and products and product quality.The influence of demographic variables on the marketing index To evaluate the influence that the demographic variables (gender, age, marital status, level of education, home language, position in the labour market and area of residence) have on the dependent variable (the marketing index and the four marketing instruments), a series of five regression analyses were conducted, using the PROC GLM procedure available on the computer package SAS (SAS Institute, 1990).It was, however, established that none of the demographic variables has a significant influence on any of the five dependent variables (the overall marketing index and the product, price, advertising and retail indices).In other words, perceptions of marketing and marketing activities are not influenced by respondents' gender, age, marital status, level of education, home language, position in the labour market or area of residence This is in contrast to the previous studies where it was revealed, in the 1990 study for instance, that older people reported a particularly poor attitude towards marketing (Boshoff & Du Plessis, 1992: 37).In the 1994 study, language of respondents, their marital status, and the level of education exerted a significant impact on the marketing index.Afrikaansspeaking consumers were particularly dissatisfied as were respondents who were married or in stable personal relationships as opposed to those in other marital status groups (never married).The academically better-qualified respondents who participated in the 1994 study were also more dissatisfied with marketing activities than those who were less well qualified (Boshoff & Du Plessis, 1995:54).
Although the amount of variation accounted for by the demographical variables in both earlier studies was less than 10% in most cases, it did provide some insight into differing market perceptions. -

Implication for marketing
The remarkable increase in the price index may, in part, be attributed to the continual decrease in the inflation rate (cf. Figure 2) over the past few years and its currently being at its lowest level since the 1980s.Although the response may not have been immediate the persistent decline in the inflation rate (Figure 2) and the resultant stability in consumer prices coincided with a substantial increase in the price index.It is also true that the price index was at its worst level during periods of high inflation [just below 15% at the time (1990) when the first survey was conducted].
Although one can obviously not deduce causality it is a relationship that may justify future attention.Another factor that may have contributed to the improved price index is increased competition in many South African markets.Globalisation and the resulting increase in competition, increased local competition from chains such as Mr Price and Spar and even informal traders such as street hawkers and spaza shops added to deregulation in industries such as agriculture may all have contributed towards the lower price index.The positive price index could also be an outcome of the price positioning efforts of larger retailers such as Shoprite-Checkers, Mr Price and Pick and Pay.Shoprite-Checkers, for instance, positions itself as 'always cheaper' while Pick and Pay positions itself as 'fighting for the consumer' and has recently again called for the deregulation of the petrol price.
An increase in the number of garage 'shops' and an increase in franchise stores such as Spar, 7-eleven, Rite-value, amongst others, could have contributed to the improvement in the retail index.Most of these stores provide a wide variety of convenience goods and are open longer hours than normal, thus better satisfying the retail needs of many consumers.
The advertising index has moved from marginally positive to marginally negative and a possible explanation for this could be the attention that advertising has drawn due to government interference with regard to the advertising of tobacco products.It may also be that the publicity generated by controversial advertising campaigns such Nando's blind women and Charlize Theron's anti-rape campaign, may have had a negative impact on the advertising industry as a whole.That the advertising industry has not improved its image among consumers is clear.For instance, 44.3% of respondents agreed or strongly agreed with the statement that most advertising is annoying and 41.0% agreed or strongly agreed that most advertising is intended to deceive rather than to inform consumers.Only 35.4% said that they enjoy most advertising.On the positive side, 40% acknowledged that advertising provides consumers with essential information as well.
It must be pointed out, however, that the advertising index is the one index that has probably been most affected by the skewness, in terms of age, of the realised sample (Table 1).Older people have consistently reported more critical perceptions of advertising in both local (Boshoff & Du Plessis, 1992:38) and international research (Pollay & Mittal, 1993).In other words, a more representative sample in terms of age may have yielded a more positive advertising index.Also, considering that the range of the index scores is -70 to +70, an index of -0.714 may not be that alarming.Still, it is sad that a South African industry that is truly world class is not recognised as such by its own people.
Despite the efforts of the Freedom of Commercial Speech Trust, the Advertising Standards Authority and others we again call for a better effort to market the advertising industry to South African consumers.
The product index has remained at a consistent level in all three surveys and returned indices of around minus nine (-9.459 and -9.939) in both of the last two surveys.In this study less than a third of respondents (31.5%) agreed or strongly agreed with the statement that the quality of most products they buy today is as good as can be expected and only 23.3% are of the opinion that product quality has improved over the last few years.

Summary and future perspectives
Although the marketing index decreased during the period 1990 to 1994, it improved dramatically during the period 1994 to 1999.The advertisement index on the other hand declined slightly and the product index established in 1999 remained relatively stable, whilst the retailing and price indices improved considerably.In contrast to the previous studies in 1990 and 1994 it was established that demographic factors had no significant influence on the sentiment towards marketing.
Changes in external forces translate into changes in consumer demand for products and services.Not only do these forces result in changes in consumer demand, but also contribute to changes in consumer perceptions and attitudes.The buying patterns and behaviour of consumers are affected by variables in the environment.
The deterioration in the marketing index from 1990 to 1994 was attributed to changing economic circumstances and the political developments in the country, both of which contributed towards the feelings of insecurity and anxiety and dampened the national mood.The dramatic increase during the period 1994 to 1999 could once again be attributed to the same factors.A certain degree of political stability has returned to the country but developments in the economic environment have been both stable in parts (the government's sound macroeconomic policies, lower interest rates, stability in the value of the rand, etc.) and unstable in other areas such as 'the Asian crisis', reams of new legislation, calls for 'transformation' and affirmative procurement, to name but a few factors.
The future prospects of the South African economy look promising and will benefit perceptions and attitudes towards marketing.The South African economy entered its sixth year of positive economic growth in 1998, albeit at a considerably slower pace than earlier in the upturn.In the early stages of the recovery, growth in real gross domestic product accelerated robustly to 3.5% in 1995 and 3% in 1996, but tapered off to about 1.5% in 1997.Growth diminished even further during 1998, and an annualised economic growth rate of 0,5 % was realised over this period (Domestic Economic Developments: 1999).The economic growth rate (GDP) reflects the strength of the economy and its future prospects.Consumer perceptions and attitudes are influenced by their expectation of the future, which could be influenced by how consumers perceive the future growth prospects of the economy (as reflected by the GDP).
Interest rates reached a record high in mid-1998, but declined steadily during 1999 to 15.5%, only slightly higher than that in 1990 (14.3%).This low level of interest rates coupled with inflation at its lowest levels in 10 years both augur well for future prospects of South African marketers.
Considering the importance of price in many consumers' buying behaviour, South African marketers will also need to take cognisance of the impact that Internet shopping will have on prices as well as on retailing.Internet shopping will cut out intermediaries in a variety of distribution channels and thus bring about cost savings for consumers in the form of reduced prices on the one hand, and increased variety, convenience and timesaving, on the other.
Marketers also need to be aware that attitudes and sentiments can change over time and proactive steps need to be taken to ensure that these changes are positive regardless of the changes in the external environment.If economic indicators suggest changes in the environment, then marketers should take steps to adapt the 4 Ps to suit the prevailing circumstances.

Limitations
Although every effort was made to ensure a representative sample comparable to those used in the earlier surveys, the realised sample is skewed in terms of age.However, the results of the multiple regression tests showed that there is no relationship between age and the marketing index or any of the other indices.From a statistical point of view it thus appears as if the potential bias of sampling error may not have had a significant impact on the results.We therefore believe that the results of this survey are comparable to the previous studies although future studies need to ensure a more even distribution in terms of all demographic variables.

Annexure A: Individual question items
The quality of most products I buy today is as good as can be expected Product 1 I am satisfied with most products I buy Product 2 Most products I buy wear out too quickly (-) Product 3 Products are not made as well as they used to be (-) Product 4 Too many products I buy are defective in some way (-) Product 5 The companies that make products don't care enough about how well the product performs (-) Product 6 The quality of products has consistently improved over the years Product 7 Most products I buy are overpriced Price 1 Business could charge lower prices and still be profitable (-) Price 2 Most prices are reasonable considering the high cost of doing business Price 3 Competition between businesses keeps prices reasonable Price 4 Businesses are unjustified in charging the prices they charge (-) Price 5 Most prices are fair Price 6 In general, I am satisfied with the prices I pay Price 7 Most advertising provides consumers with essential information Advert 1 Most advertising is very annoying (-) Advert 2 Most advertising makes false claims (-) Advert 3 If most advertising was eliminated, consumers would be better off (-) Advert 4 I enjoy most advertisements Advert 5 Advertising should be more closely regulated (-) Advert 6 Most advertising is intended to deceive rather than to inform consumers (-) Advert 7 Most shops serve their customers well Retail 1 Because of the way I'm treated in shops, most of my shopping is unpleasant (-) Retail 2 I find most salespeople to be very helpful Retail 3 Most shops provide an adequate selection of goods Retail 4 In general, most retailers make excessive profits (-) Retail 5 When I need assistance in a shop, I am usually not able to get it (-) Retail 6 Most shops provide adequate service Retail 7 (-) Negative items were reverse scored