Regionality and drivers of consumer liking: the case of Australian Shiraz in the context of the Australian domestic wine market. Trent E. Johnson A thesis submitted for the degree of Doctor of Philosophy University of Adelaide Faculty of Sciences School of Agriculture, Food and Wine Wine Science and Business Group May 2013
Thesis Summary Understanding the needs of consumers is a fundamental principle of marketing and Shiraz is arguably the most important grape variety produced in Australia, as it is the variety most widely associated, both domestically and globally, with Australia s wine industry. This three part project examined consumers in the Australian domestic wine market in respect of their liking of Australian Shiraz and provided up to date market intelligence on that market. The first stage of the project consisted of a study that segmented the Australian domestic market using a newly developed Fine Wine Instrument (FWI) that consisted of three variables, as the base. This instrument identified three segments in the market which were denoted: Connoisseurs ; Aspirants ; and No Frills wine consumers. The Connoisseur segment consumed more wine, spent more money on wine and was more knowledgeable about wine than the other segments identified in the market. The results demonstrated that this segment of consumers was important to the Australian wine industry, as they offered large potential lifetime earnings to the industry. This project also identified a number of stable segments within that market and provided updated information on the market. The next stage introduced the Shiraz variety into the project and was motivated by Wine Australia s Regional Heroes brand segmentation. The aim of the project was to identify those sensory attributes that might differentiate Shiraz wines from a number of delimited Australian Shiraz producing regions. The study employed a number of wine experts to undertake a sorting task, where wines that were perceived as similar by the experts were sorted together in groups. If wines from a single region were perceived as similar, then ii
they were sorted together. The data from this task identified three dimensions that separated the wines and these dimensions generally represented sensory attributes associated with Australian Shiraz. The wines also underwent sensory descriptive analysis which confirmed that the wines occupied diverse sensory spaces. However, identifying specific sensory attributes that differentiated wines from different regions was problematic and we concluded that any future studies of this type should concentrate on a single wine region, with a large cross section of wines from that region, rather than examining a number of wines from many regions. The last study combined elements of the previous two, where a diverse sub set of twelve of those Shiraz wines was tasted by a cohort of Australian Shiraz wine consumers and the consumers rated their acceptability, or liking, of each wine. Those consumers also completed a questionnaire so that they might be segmented using the FWI developed in the first study. The sensory data for each of the wines was married with the consumers acceptability data and the sensory attributes that drove the liking (and disliking) of the wines were identified. A similar exercise was undertaken with a cohort of wine experts and the results compared. The results demonstrated that as consumers wine knowledge and wine involvement increased, their wine preferences mimicked those of the wine experts and they preferred more elegant and complex wines. By comparison, the consumers with lower levels of wine knowledge and involvement tended to prefer wines that demonstrated more one dimensional fruit and oak characters. This technique of marrying sensory with consumer data can be transferred to any wine style and identified consumer segment. iii
The project provides the wine industry with tools that might enable producers to better identify and meet the needs of their consumers. This, in turn, might improve their profitability and increase consumer satisfaction, both admirable goals. iv
Declaration I declare that this thesis is a record of original work and contains no material which has been accepted for the award of any other degree or diploma in any university or other tertiary institution. To the best of my knowledge and belief, this thesis contains no material previously published or written by another person, except where due reference has been made in the text. The manuscripts included in this thesis have not been previously submitted for the award of any degree at the University of Adelaide or other University. I give consent to this copy of my thesis when deposited in the University library, being made available for loan and photocopying, subject to the provisions of the Copyright Act 1968. The author acknowledges that copyright of work contained within this thesis that may be published resides with the copyright holders of those works. I also give permission for the digital version of my thesis to be made available on the web, via the University s digital research repository, the Library catalogue, the Australasian Digital Thesis Program (ADTP) and also through web search engines, unless permission has been granted by the University to restrict access for a period of time. Trent Johnson May 2013 v
Panel of Supervisors Dr Sue Bastian, Principal Supervisor School of Agriculture, Food and Wine The University of Adelaide Mr Brian Croser, AO, External Supervisor Managing Director, Tapanappa Wines Former Deputy Chancellor, The University of Adelaide vi
Acknowledgements This thesis is dedicated to my late father, John E Johnson. I hope I have made him proud. This has been quite a journey and I have met a whole lot of people who have helped, mentored and motivated me throughout it. If I have omitted anyone, it is not deliberate and I apologise in advance. Sandra, Anna, Amy, Crystal, Damo, Renata and Chris have all played a part along the way and your support is much appreciated. Many thanks to Brian Croser for his early input and impetus for the project and his ongoing help during its duration and to Peter Dry for his input during the submission stage. Thanks to the University of Adelaide for their support via a scholarship and also to Orlando for their support. Thank you to Anne Hasted for her invaluable statistical advice. To Karel and Barb who were always supportive and let me know that there was life outside of study - like running, latterly riding, travelling and above all, friendship, fun, food and wine! vii
Through my training I met Cyd and Gary from Modesto in the US. Our friendship with them has blossomed and we have now spent time in each others' homes. Now that the journey has finished, I am confident that the friendship will not. To Mum and Dad in the early stages and latterly Mum - thanks for all of your support and love. Last but certainly not least, I would like to thank two amazing women, without whom, none of this would have happened. The first is my principal supervisor, Sue Bastian, but she is far more than that. She is the person who first sowed the seed of a PhD in my mind and then took on the responsibility of guiding me through the process. She has a wealth of knowledge, is forever positive and kept me positive throughout the journey and I am proud to call her a friend. Finally I come to my wife, Brenda. I said a few years ago that she had put up with a lot when I decided to take on post grad study at a time when she might have contemplated slowing down herself. Ten years on and not much has changed - I studied and she was the bread winner. Uncomplaining about working, always supporting, encouraging (and occasionally cajoling), she has been my strength, my rock and my love. I couldn't have done it without her. viii
Table of Contents Thesis Summary...ii Declaration... v Panel of Supervisors... vi Acknowledgements... vii Table of Contents... ix Table of Tables... xii Table of Figures... xiv 1 Introduction... 1 1.1 Background to the research... 1 1.2 Research problem... 3 1.3 Justification of the research... 6 1.4 Methodology... 7 1.5 Outline of this report... 9 1.6 Delimitations of scope and key assumptions... 11 1.7 Conclusion... 11 2 Literature Review... 12 2.1 The Market Segmentation Concept... 12 2.1.1 Wine Market Segmentation Studies of the Australian Domestic wine market... 20 2.1.2 Other wine market segmentation studies... 25 2.2 Fine Wine Consumers... 29 2.2.1 How to measure fine wine consumer behaviour... 32 2.3 Knowledge as a construct... 34 2.3.1 Objective Wine Knowledge... 35 2.3.2 Subjective Wine Knowledge... 37 2.3.3 Australian Wine Knowledge Studies... 40 2.4 Involvement as a Construct... 41 2.4.1 Levels of Involvement... 42 2.4.2 Measuring Involvement... 43 2.4.3 Involvement and Wine... 43 2.5 Sorting Tasks and Multidimensional scaling (MDS)... 47 2.6 The use of wine experts in wine related research... 49 2.7 Descriptive Analysis... 51 2.7.1 Descriptive Analysis A Brief History... 51 2.7.2 Quantitative Descriptive Analysis (QDA)... 54 2.7.3 Descriptive Analysis and Wine... 56 2.8 Wine consumers and wine research... 59 2.8.1 Preference Mapping... 59 2.8.2 Preference Mapping and wine... 62 2.9 The project's objectives... 64 2.10 Literature Review Summary... 66 Chapter 3. Identification of Australian fine wine consumers and comparisons of their wine related behaviour with other segments in the domestic wine market... 69 3.1 Introduction... 73 3.2 Materials and Methods... 75 3.2.1 The data collection questionnaire... 77 3.2.2 Statistical analyses... 79 ix
3.3 Results... 80 3.3.1 Demographic Data... 80 3.3.2 Wine purchase and consumption information... 82 3.3.3 Wine purchase and consumption information by gender... 83 3.3.4 Wine purchase and consumption information by fine wine consumer segment... 85 3.3.5 Fine Wine Consumer Segments... 97 3.3.6 The economic importance of each segment... 99 3.3.7 Correlations between the six wine scale values... 100 3.4 Discussion... 101 3.4.1 Fine Wine Consumers... 101 3.4.2 Purchase channels in the Australian domestic wine market... 105 3.4.3 The structure of the Australian domestic wine market... 108 3.5 Study limitations... 109 3.6 Conclusion... 110 3.7 Acknowledgements... 111 Chapter 4 Multidimensional scaling (MDS), cluster and descriptive analyses provide preliminary insights into Australian Shiraz wine regional characteristics...... 112 4.1 Introduction... 117 4.1.1 Wine regionality and typicality... 118 4.1.2 The use of expert panellists... 119 4.1.3 Sorting Tasks and Multidimensional scaling (MDS)... 120 4.1.4 Study Purpose... 121 4.2 Materials and Methods... 122 4.2.1 Wines... 122 4.2.2 Expert Panellists... 124 4.2.3 Sorting tasks... 125 4.2.4 Descriptive analysis of 29 Shiraz wines... 127 4.2.5 Statistical analyses... 131 4.3 Results... 131 4.3.1 Chemical composition of the wines... 131 4.3.2 Experts Hedonic (Liking) and Technical Quality Ratings... 133 4.3.3 MDS Analysis... 134 4.3.4 Drivers of the experts liking and technical quality scores... 140 4.3.5 Descriptive Analysis (DA) and principal component analysis of the 29... Shiraz wines... 141 4.4 Discussion... 147 4.4.1 The drivers of the experts liking and technical quality scores... 147 4.4.2 Sorting task analysis... 149 4.4.3 DA panel data... 150 4.4.4 Can a true Australian regional Shiraz character be determined?... 151 4.4.5 Study Limitations... 153 4.5 Conclusion... 154 4.6 Acknowledgements... 154 Chapter 5 The sensory drivers of Australian consumers liking of Australian Shiraz... 156 5.1 Introduction... 160 5.2 Materials and Methods... 163 5.2.1 Wines and chemical analysis... 163 x
5.2.2 Wine Experts... 165 5.2.3 Consumers and consumer wine tasting... 165 5.2.4 Descriptive analysis of 12 Shiraz wines... 167 5.2.5 Statistical analyses... 171 5.3 Results... 172 5.3.1 Chemical composition of the wines... 172 5.3.2 Wine consumer demographics and wine behaviour... 174 5.3.3 Expert hedonic (Liking) rating of the 12 wines... 177 5.3.4 Consumer hedonic rating of the 12 wines... 178 5.3.5 Cluster analysis of the consumers hedonic scores... 181 5.3.6 Descriptive Analysis (DA) and principal component analysis (PCA) of the 12 Shiraz wines... 185 5.3.7 Internal preference mapping... 187 5.3.8 Partial least squares (PLS) regression analysis... 188 5.4 Discussion... 193 5.4.1 Was the sample representative of Australian Shiraz wine consumers?... 193 5.4.2 Consumers hedonic scores and the price relationship... 194 5.4.3 The drivers of consumer liking of Australian Shiraz... 195 5.4.4 The drivers of liking of Australian Shiraz of the Connoisseur and No Frills' consumer segments... 198 5.4.5 Implications for the Australian wine industry... 200 5.4.6 The relationship between consumers and experts wine liking scores... 201 5.5 Study limitations... 203 5.6 Conclusion... 203 5.7 Acknowledgements... 204 Chapter 6 Conclusion... 206 6.1 Stage 1: Identification of Australian fine wine consumers and comparisons of their wine related behaviour with other segments in the domestic wine market... 207 6.2 Stage 2: Multidimensional scaling (MDS), cluster and descriptive analyses provide preliminary insights into Australian Shiraz wine regional characteristics.... 212 6.3 Stage 3: The sensory drivers of Australian consumers liking of Australian Shiraz 215 Reference List... 219 Appendix 1 Details of the FWI... 265 Appendix 2 Exploratory Factor Analysis data of the FWI.... 267 Appendix 3 CFA Data for the three FWI variables... 274 Appendix 4 Chapter 4 as published in Food, Quality and Preference... 280 Appendix 5 Statement of the contributions of jointly authored papers... 293 xi
Table of Tables Table 2.1 Classification of Segmentation Bases 16 Table 2.2 Classification of methods used for Segmentation 18 Table 2.3 McKinna s 1986 Australian wine consumer segments 22 Table 2.4 Comparison of Australian Wine Market Segmentation Studies 24 Table 2.5 QDA Advantages and Disadvantages 54 Table 2.6 A non-exhaustive list of Descriptive Analysis and Wine studies 56 Table 2.7 Comparison of Internal and External Preference Mapping. 61 Table 3.1 Confirmatory Factor Analysis data for the three factors 77 suggested by EFA. Table 3.2 Demographic data of the survey s respondents (n = 1017 80 respondents) Table 3.3 Wine consumption and purchase data of the respondents 82 Table 3.4 Selected wine consumption and purchase data of the respondents 84 sorted by gender Table 3.5 Percentage of wine purchases made in the HORECA category, 85 by gender and age group. Table 3.6 Cluster Centroids following AHC 86 Table 3.7 Demographic data of the three fine wine segments. Data are 87 percentages. Table 3.8 Objective and subjective wine knowledge, wine involvement 89 scores and other relevant data calculated for each fine wine consumer segment (standard deviations are in parentheses). Table 3.9 Alcoholic beverages and wine styles consumed by fine wine 90 consumer segments. Data are percentages. Table 3.10 Purchase driver responses by fine wine segment. 94 Table 3.11 Correlation matrix of the wine related scales/variables 101 Table 4.1 Geographical Indication, vintage and expert score details of the 123 29 wines used in the study. Table 4.2 Colour, aroma and palate vocabulary generated by the DA panel, 129 with agreed definitions and reference standards of the significant attributes. Table 4.3 Cluster analysis results of the 27 wines based on the sorting task 138 and DA panel consensus data. Table 4.4 One way ANOVA of the DA sensory data with the regions as 146 the source of variation Table 5.1 Details of the 12 Shiraz wines used in the consumer tasting. 164 Table 5.2 Colour, aroma and palate attribute vocabulary with agreed 169 definitions and reference standards of the 17 attributes that significantly differentiated the wines, as generated by the DA panel. Table 5.3 Chemical composition data of the 12 Shiraz wines tasted by the 173 consumers. Standard deviations are in parentheses. Table 5. 4 Australian Shiraz Wine Consumer Demographics 174 Table 5.5 Pearson correlation coefficients of the three wine related scales administered to the consumers. 176 xii
Table 5.6 Table 5.7 Table 5.8 Mean Shiraz wine liking scores of the various knowledge, involvement and fine wine segments. Pearson correlation coefficients of the wine experts and the three objective wine knowledge segments and three fine wine segments mean liking scores. Mean liking score for each Shiraz wine by cluster as determined by cluster analysis of the consumer liking scores. Numbers in parentheses indicate the percentage of consumers per cluster. 179 181 183 xiii
Table of Figures Figure 2.1 Example of a Category Scale used in FP 53 Figure 2.2 Project Schematic 65 Figure 3.1 Average price points spent on a bottle of wine, by fine wine 92 consumer segments. Figure 3.2 Wine purchases within the HORECA channel by respondent 93 gender, age and fine wine segment level. Values with different superscripts are significantly different (t Test, p < 0.05). Figure 3.3 PCA plot of variables used to characterise the three segments with the segments' bi plots projected. 96 Figure 3.4 Comparison of each segments' relative size and percentage of 100 wine spend. Figure 4.1 PCA plot of the 29 wine's chemical data with wine bi - plot 133 scores also projected. Figure 4.2 Three dimensional MDS solution of the 29 sorted wines (A = 136 dimensions 1 and 2 Blackberry, plum, pepper and spice; and Herbal, vanilla, cedar and berry jam; B = dimensions 1 and 3 Blackberry, plum, pepper and spice; and Earthy, savoury, dusty and meaty). Clusters identified by AHC (Table 3A) are circled and labelled C 1-C5. Figure 4.3 PCA plot of MDS solution, experts quality and hedonic scores, 139 wine MJT and RRP data with wine bi-plot scores for PC also projected. Chemical data are superimposed as supplementary variables. Figure 4.4 PLS regression coefficients of the 29 wines with the RRP, MJT 141 and MDS solution data as the X variables and the experts liking and quality scores as the Y variables. Figure 4.5 PCA plot of DA and MDS data with wine bi - plot scores 143 projected. Figure 4.6 Bi - plot of the wines' DA data with clusters identified by AHC 145 (Table 3B) circled and labelled DA CX Figure 5.1 Shiraz wine preference segmentation of the 179 consumers. The 184 aggregated consumer results are shown on the right and the results of the 4 clusters are shown on the left. Figure 5.2 Vector plot of the significant attributes identified by the DA 186 panel, with the bi-plot of the 12 wines overlaid. Figure 5.3 Internal preference map of the 12 Shiraz wines. 188 Figure 5.4 PLS regression coefficients of the 12 wines scored by the overall consumer cohort, the four identified clusters and the wine experts (A) and the three fine wine segments (B) (Y variables). The sensory attributes described by the DA panel were the X variables. 190 xiv