CONSUMER PATRONAGE BEHAVIOUR: AN ANALYSIS OF AN URBAN GROCERY SYSTEM by MICHAEL PETER LEW B.A., University of B.C., 1968 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF BUSINESS ADMINISTRATION i n the Faculty of Graduate Studies We accept t h i s thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA SEPTEMBER, 1970 In presenting t h i s thesis i n p a r t i a l f u l f i l l m e n t of the requirement for an advanced degree at the U n i v e r s i t y of B r i t i s h Columbia, I agree that the Library s h a l l make i t f r e e l y a v a i l a b l e for reference and study. I further agree that permission for extensive copying of t h i s thesis for scholarly purposes may be granted by the Head of my Department or by his representatives. It i s understood that copying or p u b l i c a t i o n of t h i s thesis for f i n a n c i a l gain s h a l l not be allowed without my written permission. Department of Commerce and Business Administration The University of B r i t i s h Columbia Vancouver 8, Canada Date: September, 1970 ABSTRACT Most marketers and r e t a i l e r s agree that the l o c a t i o n of a p a r t i c u l a r r e t a i l outlet may be instrumental i n that outlet*s sales and p r o f i t s . One of the proposed " solutions to t h i s problem of r e t a i l store l o c a t i o n i s Huff's gravity model. Using Huff's gravity model as a s t r u c t u r a l guide, the major purpose of t h i s thesis i s to analyse consumer patronage behaviour within an urban grocery system so as to a s s i s t the r e t a i l grocery store l o c a t i o n analyst. The analysis centers around those consumer patronage behaviour variables suggested by the author as being important i n the consumer's decision to patronize a r e t a i l grocery o u t l e t . The analysis w i l l also serve as a guide to r e t a i l grocery l o c a t i o n analysis i n s p e c i f i c and to the possible effects of p a r t i c u l a r consumer patronage behaviour variables on r e t a i l o u tlets i n general. Relationships are tested using simple and multiple regression analyses and t - t e s t analysis on data c o l l e c t e d on a l l grocery stores i n the Lower Mainland Area of Vancouver, B r i t i s h Columbia* Some of the more important findings are: 1) neither the number of customer checkouts nor the number of parking spaces provided can be employed to predict sales per square foot of supermarket s e l l i n g area. 2) a discount p r i c e p o l i c y w i l l not y i e l d greater mean sales per square foot of supermarket s e l l i n g area than a r e t a i l p r ice p o l i c y . 3) i n general a combination of good i n t e r n a l and external appearances y i e l d s greater mean sales per square foot of supermarket s e l l i n g area than a combination of low in t e r n a l and external appearances. 4) i n general the addition of i n t e r n a l and/or external f a c i l i t i e s y i e l d s greater mean sales per square foot of supermarket s e l l i n g area than those encountered before the addition or additions of such f a c i l i t i e s . The study also indicates that i n some s p e c i f i c cases supermarket s i z e should be accounted for i n the analysis. i i TABLE OF CONTENTS CHAPTER PAGE I INTRODUCTION The Importance of R e t a i l Store Location 1 The Problem of S i t e Selection 2 Purpose of the Study 3 Study Organization 4 II THE DEVELOPMENT OF AN INTRA-REGIONAL LOCATION MODEL: THE GRAVITY MODEL Introduction 6 William J . R e i l l y : The Laws of R e t a i l Gravitation 7 Digression: Frank A. Fetter: The Law of Market Areas 8 The Significance of R e i l l y 1 s Laws 9 David L. Huff: The Gravity Model 11 Expected Consumer U t i l i t y vs. The Prob a b i l i t y of Consumer Patronage 13 Store Size and Distance to be Travelled 14 Chapter Summary and Conclusions •• 25 i i i CHAPTER PAGE III THE DISTANCE EXPONENT AND CONSUMER PATRONAGE BEHAVIOUR Introduction 26 The Significance of C 27 The Components of f : Consumer Patronage Behaviour 30 The Randomness of Selection 31 Degree of S u b s t i t u t a b i l i t y 31 Expected Absolute Price D i f f e r e n t i a l s .... 34 The Absolute Price i n Relation to Consumer Income ...... 37 "Psychic Income" and S o c i a l Class 40 Advertising 43 Chapter Summary and Conclusions .......... 46 IV SPECIFIC STORE VARIABLES - THE SUPERMARKET Introduct ion .... 48 Number of Parking Spaces 49 Number of Checkouts 53 Product Quality Image 56 Store Image • • 58 Store Cleanliness Image 59 Friendliness of Store Personnel Image .... 62 Additional F a c i l i t i e s 65 Chapter Summary and Conclusions 68 i v CHAPTER PAGE V THE ESTABLISHMENT OF WORKING HYPOTHESES Introduction 69 The Available Data 70 The Setting of the Hypotheses 71 Degree of Product S u b s t i t u t a b i l i t y 71 Expected Absolute Price D i f f e r e n t i a l s .... 72 The Absolute Price i n Relation to Consumer Income ........... 73 "Psychic Income" and S o c i a l Class 73 Advertising 74 Number of Parking Spaces 74 Number of Checkouts 75 Product Quality Image 76 Store Cleanliness Image • 76 Friendliness of Store Personnel Image .... 77 Additional F a c i l i t i e s 77 Store Sales and the Pr o b a b i l i t y of Consumer Patronage 78 Chapter Summary and Conclusions .......... 79 VI HYPOTHESIS TESTING Introduction 81 The Data 82 t- t e s t Analysis 86 V CHAPTER PAGE The Hypotheses Tested 86 The Analysis 87 Price Policy 88 Overall Internal-External Appearance Image 91 Additional F a c i l i t i e s (Internal) 93 Bakery or Bake-off 93 Delicatessen 96 Additional F a c i l i t i e s (External) 98 Conclusions: t - t e s t Analyses 100 The Correlation Matrix 101 Conclusions: The Correlation Matrix ..... 104 Corr e l a t i o n - Regression Analysis ........ 108 The Hypotheses Tested 108 I n i t i a l Analysis 109 The Regression Analysis • I l l Conclusions: Regression Analyses 115 Digression: Sales per Square Foot 115 Chapter Summary and Conclusions 118 VII STUDY SUMMARY AND CONCLUSIONS Purpose of the Study • 120 Huff's Gravity Model 121 v i CHAPTER PAGE Additional Consumer Behaviour Variables ... 122 The Data 123 The Hypotheses 124 S t a t i s t i c a l Analysis 125 Conclusions 127 BIBLIOGRAPHY 128 APPENDIX A THE TRIP (TRIANGULAR REGRESSION PACKAGE) COMPUTER PROGRAM 134 APPENDIX B THE COMPUTER PROGRAM SUBROUTINE UTILIZED IN THE SIMPLE AND MULTIPLE REGRESSION ANALYSES 136 APPENDIX C THE SIMCORT COMPUTER PROGRAM 138 APPENDIX D THE COMPUTER PROGRAM SUBROUTINE UTILIZED IN THE t-TESTS INVOLVING OVERALL MEAN SALES PER SQUARE FOOT 140 APPENDIX E THE COMPUTER PROGRAM SUBROUTINE UTILIZED IN THE t-TESTS INVOLVING MEAN SALES PER SQUARE FOOT WITHIN A GIVEN SIZE RANGE 143 v i i LIST OF TABLES TABLE PAGE I Overall Mean Sales Per Square Foot t-Test Results 87 II Price P o l i c y t-Test Results (Within a Given Size Range 90 III Overall Internal-External Appearance Image t-Test Results (Within a Given Size Range) 92 IV Bakery or Bake-off t-Test Results (Within a Given Size Range) 95 V Delicatessen t-Test Results (Within a Given Size Range) .. 97 VI Complementary Area t-Test Results (Within a Given Size Range) 99 VII The Corr e l a t i o n Matrix 102 VIII Interdependent Correlation Results 107 IX Results: Simple Regression Analysis: Sales Correlated With -Sales Area Number of Parking Spaces Parking Spaces per Square Foot .... 112 X Results: Multiple Regression Analysis: Sales Correlated With -Sales Area Number of Parking Spaces Parking Spaces per Square Foot .... 113 v i i i LIST OF FIGURES FIGURE PAGE 1. Huff's Expected Consumer U t i l i t y - Store Size Relationship (A) 15 2. Huff's Expected Consumer U t i l i t y - Store Size Relationship (B) 16 3. Huff's P r o b a b i l i t y of Consumer Patronage -Store Size Relationship .... 17 4. Huff's Expected Consumer U t i l i t y - Distance Relationship 18 5. Huff's P r o b a b i l i t y of Consumer Patronage -Distance Relationship ...... 19 6. Consumer Base - Distance I l l u s t r a t e d 20 7. Size - Distance E f f e c t Upon _£ I l l u s t r a t e d 28 8. The Pr o b a b i l i t y of Consumer Patronage -Product S u b s t i t u t a b i l i t y Relationship 33 9. The P r o b a b i l i t y of Consumer Patronage -Expected Absolute Price D i f f e r e n t i a l s Relationship 36 10. The Pr o b a b i l i t y of Consumer Patronage -Consumer Income and Product Price Relationship 39 11. The Pr o b a b i l i t y of Consumer Patronage -"Psychic Income" Relationship ........ 42 12. The Pr o b a b i l i t y of Consumer Patronage - Amount of Advertising Relationship 45 13. The Expected Consumer U t i l i t y - Number of Parking Spaces Provided Relationship (A) ... 51 14. The Expected Consumer U t i l i t y - Number of Parking Spaces Provided Relationship (B) ... 52 i x FIGURE PAGE 15. The P r o b a b i l i t y of Consumer Patronage - Number of Parking Spaces Provided Relationship .... 52 16. The Expected Consumer U t i l i t y - Number of Checkouts Provided Relationship 54 17. The Pr o b a b i l i t y of Consumer Patronage - Number of Checkouts Provided Relationship 55 18. The Expected Consumer U t i l i t y - Product Quality Relationship 56 19. The P r o b a b i l i t y of Consumer Patronage -Product Quality Relationship 57 20. The Expected Consumer U t i l i t y - Store Cleanliness Image Relationship 59 21. The P r o b a b i l i t y of Consumer Patronage - Store Cleanliness Image Relationship 61 22. The Expected Consumer U t i l i t y - Personnel Friendliness Image Relationship 62 23. The P r o b a b i l i t y of Consumer Patronage -Personnel Friendliness Image Relationship •• 63 24. The Expected Consumer U t i l i t y - Additional F a c i l i t i e s Provided Relationship 65 25. The Pr o b a b i l i t y of Consumer Patronage -Additional F a c i l i t i e s Provided Relationship 67 26. Zero Correlation - Strong Quadratic Relationship I l l u s t r a t e d 103 ACKNOWLEDGEMENT The author i s indebted to a number of people who gave invaluable assistance i n the preparation of t h i s study. S p e c i f i c thanks are due to Mr. C. Campbell of K e l l y Douglas and Company Limited for w i l l i n g l y supplying the data on which t h i s study i s based; and to Mrs. S. Lee and Mrs. D. Loo for t h e i r accurate and rapid preparation of the manuscript. To my thesis advisor, Dr. J.D. Forbes, I wish to express my sincere thanks for his many hours of assistance and constructive c r i t i c i s m . M. P. LEW CHAPTER I INTRODUCTION The Importance of R e t a i l Store Location Poor se l e c t i o n of a r e t a i l store s i t e may seriously hamper the most s k i l l f u l of merchandisers, whereas a good loc a t i o n may greatly o f f s e t the de f i c i e n c i e s of mismanagement. Good lo c a t i o n i s essential to success i n r e t a i l i n g . The lo c a t i o n of a r e t a i l store plays a v i t a l part i n i t s success; t h i s i s because the lo c a t i o n determines to a large degree the sales made and the p r o f i t s r e a l i z e d . Some r e t a i l e r s , such as those s e l l i n g v a r i e t y store merchandise and women's apparel, consider location such an important factor that they prefer to pay a larger-than-usual rental to obtain desirable s i t e s , even i f t h i s means they must r e s t r i c t other expenses. 1 Thus s e l e c t i o n of a r e t a i l store s i t e may be instrumental i n the outcome of a p a r t i c u l a r business' operations. D.J. Duncan and C.F. P h i l l i p s , R e t a i l i n g P r i n c i p l e s and Methods. Homewood, I l l i n o i s , Richard D. Irwin, Inc., 1963, page 93. 2 The Problem of S i t e Selection The importance of r e t a i l s i t e analysis has been recognized for many years. However, recognition of the problem does not automatically lead to i t s solu t i o n . Marketers have yet to find the optimal solution to the problem of r e t a i l store location and consumer patronage behaviour. Through experience they have found that consumer patronage behaviour i s , at best, d i f f i c u l t to measure. As a re s u l t they have found that one of the prime areas of d i f f i c u l t y i n loc a t i o n analysis i s the j u s t i f i c a t i o n and measurement of relevant consumer behaviour varia b l e s . In many cases the solution proposed by the academic i s rejected by the r e t a i l e r . Retailers argue that these proposals are impractical. They argue that such solutions are find i n theory. However, they are too s i m p l i s t i c i n t h e i r basic assumptions to be adapted to the " r e a l world". They argue that the relevant variables have not been accounted for or properly measured. Thus, the problem of loc a t i o n analysis and consumer patronage behaviour i s one of providing for the r e t a i l e r p r a c t i c a l tools for the purpose of analysing consumer patronage behaviour. 3 Purpose of the Study The purpose of t h i s thesis i s to analyze consumer patronage behaviour within an urban grocery system so as to a s s i s t the r e t a i l grocery store location analyst. The study i s s t r u c t u r a l l y centered around one of the proposed solutions to r e t a i l store l o c a t i o n analysis, Huff's gravity model. The study suggests c e r t a i n refinements which should be made to Huff's distance decay factor 7± for the purpose of r e t a i l grocery store l o c a t i o n analysis. It w i l l also serve as a guide to the analyst i n v e s t i g a t i n g consumer patronage behaviour i n general and i n the f i e l d of r e t a i l grocery stores. Prediction of consumer patronage behaviour i n a metropolitan system i s , i n the le a s t , d i f f i c u l t . The available data did not allow a complete analysis of the gravity model per se as there are a n c i l l i a r y problems which could not be f u l l y tested. Hence the v a l i d i t y of the gravity model per se was neither proved nor disproved. Rather, using the available data, s p e c i f i c consumer behaviour variables were hypothesized and tested and r e s u l t i n g proposals were made towards the improvement of the gravity model, and the analyst's knowledge of urban consumer behaviour i n general. 4 Study Organization The study i s primarily organized into two sections, d e s c r i p t i v e and a n a l y t i c a l . It begins with a b r i e f d e s c r i p t i o n o f the deterministic models which preceded Huff's p r o b a b i l i s t i c g r a v i t y model* Huff's gra v i t y model i s then discussed, i n d e t a i l , i n regards to i t s d i r e c t i n c l u s i o n of the variables store s i z e and consumer distance to be t r a v e l l e d . Chapter II then ends with a discussion as to how one might apply the gravity model to the " r e a l world". Chapter III opens with a de t a i l e d discussion of Huff's distance exponent X • An example i l l u s t r a t i n g the sign i f i c a n c e of *X i s presented, followed by a discussion of the consumer behaviour components of "X • The study then turns to those consumer patronage variables which, i n addition to those outlined by Huff, may have a s i g n i f i c a n t e f f e c t upon a model designed for r e t a i l outlet l o c a t i o n analysis and may be c h a r a c t e r i s t i c of consumer behaviour. Having discussed the variables involved i n r e t a i l store l o c a t i o n i n general the study turns to the s p e c i f i c problem of supermarket consumer behaviour patterns. 7 5 Chapter IV outlines and describes those s p e c i f i c store variables suggested by the author, which must be taken into account when dealing with the problem of supermarket patron a n a l y s i s . Chapter V begins the a n a l y t i c a l section of the study. It i s concerned with the se t t i n g of hypotheses regarding the e f f e c t of c e r t a i n consumer patronage behaviour variables upon supermarket sal e s . Chapter VI concerns i t s e l f with the s t a t i s t i c a l analysis of those hypothesized variables stated i n Chapter V. The concluding chapter, Chapter VII summarizes the study giving indications as to those additional variables which should be included for consideration i n Huff's gravity model i n general and for the s p e c i f i c problem of supermarket patron a n a l y s i s . In addition i t serves as a summary of the additional information provided by t h i s study which w i l l add to the marketer's knowledge of urban consumer behaviour. CHAPTER II THE DEVELOPMENT OF AN INTRA-REGIONAL LOCATION MODEL THE GRAVITY MODEL Introduction With the passing of time and the growth of c i t i e s the problem of int r a - r e g i o n a l , as opposed to int e r - r e g i o n a l , l o c a t i o n analysis has developed. The importance of new s i t e evaluation and the analysis of e x i s t i n g and future d i s t r i b u t i o n of consumer based a c t i v i t i e s Is r e ceiving growing attention and app l i c a t i o n . 2 Re t a i l e r s , marketers and urban land developers, among others, are asking the following: "Is the present s i t e optional or can an increase i n revenue be brought about through a change i n location?"; "What factors must be considered i n evaluating r e t a i l store location?"; "How important i s the lo c a t i o n of z J.D. Forbes and A.G. Fowler, "Simmulation of a Gravity Model", The Annals of Regional Science. June, 1969, The Western Regional Science Association, page 86. 7 competitive r e t a i l e r s ? " ; and "Is l o c a t i o n a factor to be reckoned with at a l l ? " . R e t a i l store businesses are v i r t u a l l y a l l market-oriented ... The r e t a i l e r must be accessible to people, and i t i s t h i s fact that governs s i t e s e l e c t i o n . 3 At present, one of the p r i n c i p a l tools under consideration concerning r e t a i l s i t e analysis i s the gravity model. Through the use of the gravity model, marketers are attempting to i s o l a t e those variables which they f e e l are of key importance i n the consumer's decision to patronize a s p e c i f i c r e t a i l o u t l e t . William J . R e i l l y : The Laws of R e t a i l Gravitation In an attempt to measure the market areas of p a r t i c u l a r r e t a i l o u t l e ts, William J . R e i l l y developed the "laws of r e t a i l g r a v i t a t i o n " . Through the use of these "laws", R e i l l y was able to determine market areas, knowing only the population and central functions of p a r t i c u l a r centres and the distances between them. R.L. Nelson, The Selection of R e t a i l Store Locations. New York, F.W. Dodge Corporation, 1958, page 44. 8 R e i l l y developed the "breaking-point" equation which stated that the trading boundary between two areas, A and B, i n miles from B i s equal to Miles Between A and B [Size of A* 1 + J Size of B In most applications, R e i l l y used population as an index of s i z e , but number of central functions can be used, since i t i s the fundamental measure of the attractiveness of a center. 4 Digression: Frank A. Fetter: The Laws of Market Areas Much l i k e R e i l l y ' s "breaking-point" equation, Fetter stated that by s u b s t i t u t i n g "attractiveness" or center s i z e for market pr i c e s , and distances for f r e i g h t rates, i t i s possible to determine the boundary l i n e between two geographically competing markets for s i m i l a r goods. The boundary l i n e i s the r e s u l t of the continuum of points where the freight charges plus prices of goods of the two centres i s equal. Hence, a reduction i n p r i c e and/or freight charges increases the market area of a 4 Brian J.L. Berry, Geography of Market Centers and R e t a i l D i s t r i b u t i o n . Englewood C l i f f s , Prentice-Hall Inc. 1967, page 40. 9 given centre. The Significance of R e i l l y ' s Laws R e i l l y ' s equation loses i t s s i g n i f i c a n c e when applied to heavily populated areas. In r u r a l areas, distance plays a major r o l e i n the time and cost a l l o c a t i o n made by the consumer i n h i s s e l e c t i o n of a market center. However, due to the proximity of various r e t a i l outlets i n heavily populated, urban areas and t h e i r respective varying degrees of attractiveness, the break-point becomes "fuzzy". One cannot say that on one side of a s p e c i f i c boundary consumers w i l l patronize outlet A and on the other side they w i l l patronize outlet B. A s p e c i f i c boundary, the break-point, does not e x i s t . It i s "fuzzy" due to the d i f f e r e n t degrees of attractiveness of centers f o r consumers. There are many centers, not just one or two, within the maximum distance the consumer i s w i l l i n g to t r a v e l . Within metropolitan regions there i s no such thing as an absolute breaking-point. The break-point formula simply gives the point at which the proportion of consumers located around the breaking-point s p l i t s equally between two competing alternatives of d i f f e r i n g attractiveness. 5 5 Berry, op. c i t . . page 41 10 It i s i n these densely populated urban areas that R e i l l y ' s o r i g i n a l law of g r a v i t a t i o n becomes applicable. Two centers a t t r a c t trade from intermediate places approximately i n d i r e c t proportion to the sizes of the centers and i n inverse proportion to the square of the distances from these two centers to the intermediate place* froportions of trade rom the intermediate place attracted, by centers A and B sizes of centers A and iJ distances of A and B from the intermediate place. 6 R e i l l y * s o r i g i n a l law of g r a v i t a t i o n i s deterministic i n nature. It merely gives the analyst a means of c a l c u l a t i n g the l i n e or breaking-point between trading areas. It i s simply a measurement, generally i n miles, of the distance between t h i s l i n e and the p a r t i c u l a r center i n question. There i s no consideration made for the number or types of consumers which Berry, op. c i t . . page 41. 11 Inhabit these areas. However, densely populated urban areas are characterized by many business centers, having d i f f e r e n t degrees of attractiveness, within the consumer's desired perimeter of t r a v e l . Thus, i n d i v i d u a l consumer behaviour within such densely populated areas i s p r o b a b i l i s t i c rather than deterministic i n nature. Hence, R e i l l y ' s model i s inadequate for the study of consumer behaviour i n such areas. This led to David L. Huff's development of "the gravity model". David L. Huff: The Gravity Model In general, when present day marketers speak of "the gravity model" they are r e f e r r i n g to Huff's gravity model. As stated e a r l i e r , intra-regional marketers and location analysts are concerned with p r o b a b i l i t y models of urban areas rather than deterministic models such as those proposed by R e i l l y and Fet t e r . As a r e s u l t , Huff's gravity model has been widely discussed as a proposed answer to the problem of location analysis. Huff states that the p r o b a b i l i t y of consumer patronization i s related to the expected u t i l i t y which the consumer feels he w i l l gain through patronization of a p a r t i c u l a r o u t l e t . That i s , 12 j=i -where and P. = the p r o b a b i l i t y of consumer patronage of r e t a i l outlet j U. 3 8 the u t i l i t y gained by the consumer through patronage of outlet j n. YLV5 - 1 with 0 the percentage of income spent on product class k by residents of area i , i s inserted into the model. B i k where A = expected average expenditures made by consumers located at consumer base i at r e t a i l outlet j for a p a r t i c u l a r good k. Using Huff's gravity model the analyst i s now able co determine the optimal l o c a t i o n tor a p a r t i c u l a r r e t a i l o u t l e t . The procedures i s as follows; 8 10 divide the area surrounding the actual or proposed outlet location into s t a t i s t i c a l l y s i g n i f i c a n t units determine the si z e of the s e l l i n g area of the outlet i n square feet determine the distance from unit i to location j empirically estimate a value for the parameter "X apply the gra v i t y model and determine the respective patronage p r o b a b i l i t i e s from each consumer unit calculate the number of consumers i n each unit c a l c u l a t e the percentage of consumer income spent on that p a r t i c u l a r product class by consumers r e s i d i n g i n base i calc u l a t e the r e s u l t i n g expenditures made by consumers at that p a r t i c u l a r l o c a t i o n apply the above procedure to alt e r n a t i v e r e t a i l outlet locations choose that l o c a t i o n which maximizes the r e t a i l outlet owner's p r o f i t . 25 Chapter Summary and Conclusions In t h i s chapter we have been introduced to the o r i g i n and use of Huff's gravity model. A short hi s t o r y of the development of Huff's model was presented along with the basic underlying assumptions behind i t . A detailed description of the model was then followed by a procedural ou t l i n e for i t s p r a c t i c a l a p p l i c a t i o n . Chapter I I , i s then, an attempt to f a m i l i a r i z e the reader with the gravity model and i t s use as a t o o l for lo c a t i o n analysis. In the chapters which follow, the consumer behaviour variables which make up ^ w i l l be discussed, followed by the formulation and t e s t i n g of hypotheses drawn on the various e f f e c t s of these v a r i a b l e s . CHAPTER III THE DISTANCE EXPONENT £ AND CONSUMER PATRONAGE BEHAVIOUR Introduction We have seen i n Chapter II that a r e l a t i v e l y simple model, the gravity model, has been developed and discussed as being a t o o l for lo c a t i o n analysis. The model s p e c i f i c a l l y takes into account two var i a b l e s , the distance to be t r a v e l l e d by the consumer and the size of the store. In addition, the distance exponent £ i s inserted into the model i n order to account for the non-linearity of the distance e f f e c t . Chapter III i l l u s t r a t e s the si g n i f i c a n c e of the distance exponent £ and i t s i n c l u s i o n of s p e c i f i c consumer behaviour components as outlined by Huff. In addition, those variables outlined by Forbes as being c r i t i c a l i n the analysis of consumer behaviour w i l l be discussed. As mentioned i n the Thesis Introduction, Chapter III and IV are centered around a discussion of the distance exponent £ and consumer patronage behaviour. Chapter III discusses those variables which are considered to be of a more general nature, whereas Chapter IV 27 makes more s p e c i f i c reference to those variables concerning supermarket patron analysis. Following the discussion of these va r i a b l e s , Chapter V w i l l e s t a b l i s h various hypotheses regarding t h e i r relationships with sales, thus se t t i n g the stage for the analysis of actual data. The Significance of £ The symbol £ i s then a " c a t c h - a l l " exponent designed to take into account those consumer patronage variables, r e s u l t i n g i n the e f f e c t of non-linearity, which are d i f f i c u l t to measure or not measurable at a l l . However, the exponent has great s i g n i f i c a n c e upon the outcome of gravity model c a l c u l a t i o n s . Witness the following example. Suppose we are looking at three locations each having a store of equal s e l l i n g space, 100,000 square feet. The distances from consumer base are as follows; 28 FIGURE 7 SIZE-DISTANCE EFFECT UPON % ILLUSTRATED D=l 3 (100,000 sq.ft.) 1 (100,000 sq.ft.) 2 (100,000 sq.ft.) I n i t i a l l y we set £ = 1 then D i j i j j=l 'j D c i j PA1 100.000 1 100t000 + 100.000 + 100.000 ( l ) x I 2 ) x ( 3 r = .5454 29 Changing k = 1 to £ = 2 and holding size of outlet constant we have 100.000 ( I ) 2 r A l 100.000 + 100.000 + 100.000 (1) (2) (3) - .7346 Notice that by changing the exponent £ from 1 to 2 the p r o b a b i l i t y of consumers at base A, t r a v e l l i n g to l o c a t i o n 1 r i s e s from .5454 to .7346. This i s a 35 percent increase i n p r o b a b i l i t y ! Because such changes i n the value of £ a f f e c t the outcome so markedly, i t i s desirable to have £ measurements as accurate as possible. In order to give the analyst a more fe a s i b l e model, then, i t i s f i r s t necessary to determine a l l those variables concerned with the decay of the distance factor and consumer patronage behaviour« 30 The Components of P : Consumer Patronage Behaviour It i s known that consumer patronage behaviour i s dependent, not only upon the si z e of a p a r t i c u l a r r e t a i l outlet and i t s distance from the consumer base, but on a multitude of other factors. Consumers also display differences i n terms of t h e i r willingness to t r a v e l various distances for d i f f e r e n t types of products. This accounts for the distance exponent £ . These variati o n s can be attributed to the value differences that various goods and services possess to the consumer. Such values are based on: 1. the degree of s u b s t i t u t a b i l i t y of various products; 2. the expected absolute p r i c e d i f f e r e n t i a l between d i f f e r e n t products; 3. the absolute p r i c e of a product i n r e l a t i o n to a consumer's income l e v e l ; and, 4. the degree of "psychic income" 7 anticipated from d i f f e r e n t products. In addition to the factors mentioned above, " s o c i a l c l a s s , advertising and store v a r i a b l e s " a l l a f f e c t consumer patronage behaviour. D.L. Huff, A P r o b a b i l i s t i c Analysis of Consumer Sp a t i a l Behaviour. Real Estate Research Program, Graduate School of Business Administration, University of C a l i f o r n i a , Los Angeles, page 19. J.D. Forbes, Consumer Patronage Behaviour. Presentation to the American Marketing Association Meetings, Denver, Colorado, August 29, 1968, page 1. 31 The Randomness of Selection In introducing and describing the consumer patronage behaviour variables, which are pertinent to the analysis, the term "randomness of s e l e c t i o n " w i l l be used quite frequently. By t h i s i t i s meant that i n some cases consumer patronage i s random i n nature, i e . someone new to the neighbourhood or merely " t r a v e l l i n g through", randomly s e l e c t i n g a r e t a i l outlet to patronize. These consumers, i n i t i a l l y , are not aware of consumer "grapevines" with regards to "good and bad" r e t a i l o u t l e t s . Hence, r e t a i l outlet s e l e c t i o n for t h i s portion of consumers may i n i t i a l l y be, random i n nature. Degree of S u b s t i t u t a b i l i t y The nature of the product i t s e l f i s a determinant of consumer patronage behaviour. In the eyes of the consumer, there may be a c e r t a i n degree of product s u b s t i t u t a b i l i t y . That i s , ice-cream may be substituted for fresh strawberries as the dessert item for an evening's meal or packaged cakes for "bon-ton". Hence, the consumer may patronize that grocery outlet o f f e r i n g ice-cream, which i s located nearest to his base, rather than t r a v e l a greater distance i n order to patronize an outlet having a fresh produce section or a bakery shop, " a l l other things being equal". The consumer's willingness to 32 t r a v e l to an outlet o f f e r i n g a s p e c i f i c item for purchase i s affected by his willingness to substitute other items, offered at more proximate locations. On the other hand, i t may not be p r a c t i c a l to substitute one good for another. Here the cost of substitution outweighs the benefits derived from a reduction i n t r a v e l l i n g costs and time. A Chevrolet d i s t r i b u t o r cannot be substituted with a Ford d i s t r i b u t o r , nor a Volkswagen rotor for a Chrysler rotor. The engine being repaired w i l l not function. Hence, the savings benefits derived from patronizing a more proximate Ford or Volkswagen dealer are far outweighed by the ultimate costs. Thus the consumer must patronize his nearest Chevrolet or Chrysler parts dealer, not his nearest auto parts dealer, and he i s w i l l i n g to do so. Brand preferences also enter into the determination of consumer behaviour. In shopping tor bread, some housewives w i l l choose any brand. They w i l l patronize any supermarket or grocery store which handles bread i n general. Others, i n shopping for bread, are looking for a p a r t i c u l a r brand, e.g. Skylark. They have found through experience that they prefer Skylark and hence, are not w i l l i n g to substitute other brands. 33 They are w i l l i n g to t r a v e l longer distances i n the search for Skylark rather than just any bread. As a r e s u l t , they w i l l patronize t h e i r nearest Safeway store, not just t h e i r nearest grocery store. Thus, depending upon the degree of both product type and brand s u b s t i t u t a b i l i t y which the consumer i s w i l l i n g to t o l e r a t e , the e f f e c t of distance to a p a r t i c u l a r r e t a i l outlet from the consumer base may be large or small. The p r o b a b i l i t y curve with respect to consumer patronage and the degree of s u b s t i t u t a b i l i t y of the products offered by a p a r t i c u l a r outlet would then be; FIGURE 8 Degree of S u b s t i t u t a b i l i t y Note: i t i s assumed that a l l variables except the degree of s u b s t i t u t a b i l i t y of the derived product are constant. 34 Here we see that for consumers i n search of a product which cannot be substituted at a l l , a s p e c i a l t y good, the p r o b a b i l i t y of consumer patronage of the only outlet which off e r s t h i s good i s one. That i s , anyone d e s i r i n g to purchase e l e c t r i c i t y i n Vancouver and not wishing to generate his own by i n s t a l l i n g a private power plant w i l l d e f i n i t e l y become a customer of B.C. Hydro. This would be the case i n a l l product and brand name monopolies. As the degree of product s u b s t i t u t a b i l i t y increases, the p r o b a b i l i t y of consumer patronage, then, decreases, approaching zero. Zero p r o b a b i l i t y i s never reached, for even with a p e r f e c t l y substitutable good there i s s t i l l a random chance of consumer patronage. Expected Absolute Price D i f f e r e n t i a l s ••Woodward's $1.49 Day", "Eaton's Trans-Canada Sale", "Bay Day". What draws the consumer to these respective sales? She expects to save money by purchasing sale goods. Some consumers w i l l t r a v e l "half-way across town" i n order to attend these sales. Absolute price d i f f e r e n t i a l s are important to them. 35 The " r a t i o n a l " consumer weighs t r a v e l costs against the absolute price of the item or items which she expects to purchase. Monetary savings must outweigh the additional t r a v e l costs suffered by the consumer i n patronizing a less proximate r e t a i l l o c a t i o n . Here we are not only concerned with the monetary increases i n cost brought about through an increase In the distance to be t r a v e l l e d but also time and psychological expenses. There i s an opportunity cost Involved. The consumer i s foregoing time which may be put to other uses, i n order to t r a v e l t h i s longer distance. At the same time she may be subjected to greater frustrations brought about by longer periods of t r a f f i c congestion, longer queuing times at the outlet which i s holding the sale and the p o s s i b i l i t y that a l l the sale goods may have been sold upon her a r r i v a l . Thus expected absolute price d i f f e r e n t i a l s are of prime consideration to the r a t i o n a l consumer. P l o t t i n g the p r o b a b i l i t y of consumer patronage with respect to expected absolute price d i f f e r e n t i a l s we have the following; 36 FIGURE 9 THE PROBABILITY OF CONSUMER PATRONAGE -EXPECTED ABSOLUTE PRICE DIFFERENTIALS RELATIONSHIP +1 Pr o b a b i l i t y of Consumer Patronage Expected Absolute Price D i f f e r e n t i a l s Note: i t i s assumed that a l l variables except expected absolute p r i c e d i f f e r e n t i a l s are constant. At zero expected absolute price d i f f e r e n t i a l s we do not have a zero p r o b a b i l i t y of consumer patronage. Here the consumer i s i n d i f f e r e n t as to the l o c a t i o n which he w i l l patronize, as a l l prices are equal. His choice of outlet patronization i s then affected by some other variable or i f they are a l l assumed to be constant, i t i s random and each store w i l l have an equal p r o b a b i l i t y of being v i s i t e d . 37 As the expected savings i n price d i f f e r e n t i a l s r i s e the difference between monetary savings and t r a v e l costs can be expected to increase. The more po s i t i v e t h i s difference, the greater the p r o b a b i l i t y of r a t i o n a l consumer patronage. Hence, the p r o b a b i l i t y curve r i s e s . Holding a l l other variables constant, i t should reach one for the " r a t i o n a l " consumer residing within "reasonable" distance of t h i s p a r t i c u l a r r e t a i l o u t l e t . Notice that i t has been assumed that the consumer has, at his disposal, a l l of the information regarding product prices at p a r t i c u l a r o u t l e t s . He i s i n a state of perfect market knowledge. The Absolute Price of an A r t i c l e and Consumer Income No consumer has an i n f i n i t e income. Whether he i s poor or r i c h , there i s some a r t i c l e , which the consumer may wish i to purchase, whose price i s such that i t w i l l r e s u l t i n a considerable drain on his income. It may be an average wage earner wishing to purchase a new automobile or Howard Hughes looking for a new hotel i n which to invest. Although the degree of the u t i l i t y of money d i f f e r s from i n d i v i d u a l to i n d i v i d u a l , monetary u t i l i t y i s common to a l l . 38 In each case, the pot e n t i a l purchaser becomes more cautious i n hi s shopping patterns. He r e a l i z e s that the purchase of such a good means the spending of a good proportion of his income. He i s w i l l i n g to "shop around", making many shopping t r i p s without purchasing the desired item. He may return to a s p e c i f i c outlet many times, comparing i t s product qu a l i t y , p r i c e and/or "psychic income" with the products of other o u t l e t s . He i s w i l l i n g to t r a v e l many miles i n search of the "best" product. One would i n t u i t i v e l y guess that the greater the proportion of the " r a t i o n a l " consumer's income to be spent on a p a r t i c u l a r a r t i c l e , the greater the p r o b a b i l i t y that he w i l l v i s i t a l l of the outlets i n the area which o f f e r t h i s a r t i c l e . The following i n t u i t i v e p r o b a b i l i t y curve then r e s u l t s ; 39 FIGURE 10 THE PROBABILITY OF CONSUMER PATRONAGE -CONSUMER INCOME AND PRODUCT PRICE RELATIONSHIP + 1-P r o b a b i l i t y of Consumer Patronage Proportion of Price to Consumer Income Note: i t i s assumed that a l l variables except the proportion of absolute price to consumer income are constant. The preceding curve i s much the same as that derived for the p r o b a b i l i t y of consumer patronage and expected absolute price d i f f e r e n t i a l s , Figure 9. However, at zero proportion of pri c e to consumer income the good i s free. Hence, given the above assumption, the p r o b a b i l i t y of patronage i s one and the curve Is discontinuous at t h i s point. 40 As the proportion of income to be spent r i s e s , the p r o b a b i l i t y of consumer patronage r i s e s . Notice that, as stated e a r l i e r , consumer patronage refers to the consumer 1s willingness to patronize a p a r t i c u l a r o u t l e t . However, he does not necessarily have to buy anything. The p r o b a b i l i t y curve then r i s e s , approaching a p r o b a b i l i t y of one. Here, where the " r a t i o n a l " consumer i s spending a majority of his income on a p a r t i c u l a r good, he w i l l v i s i t as many outlets as he can j u s t i f y using a cost-benefit type analysis. "Psychic Income" and S o c i a l Class For some, a shopping t r i p i s more than a duty. It i s an afternoon spent with "the g i r l s " . It i s a chance to "get out of the house". It i s an opportunity to meet new people or old friends. It i s a change i n atmosphere. "Psychic income", i n some ways, explains why the older Chinese i n the Oakridge Area would rather shop i n Vancouver's Chinatown Area for the same goods which may be purchased i n Woodward's Food Department i n Oakridge. Shopping i n Chinatown gives these older people an opportunity to converse i n t h e i r native language. They speak to both friends and clerks i n Chinese. They meet old friends who l i v e 41 tn the Chinatown Area or who are shopping there, as they are. They may arrange a "mah jong" game for the following week or t a l k about the "old country". They are more "comfortable" i n t h i s Chinese, rather than Western, atmosphere. Educational l e v e l s , family and ethnic background, income, education and other demographic c h a r a c t e r i s t i c s have been used i n various ways by s o c i o l o g i s t s to define an amorphic term c a l l e d s o c i a l c l a s s . 9 There i s psychological value to be gained by the consumer through conforming. He i s considered to be a member of a c e r t a i n s o c i a l class and he does what they do, shopping where they shop. "Psychic income" i s gained through shopping where "everyone" does. In t h i s way, the consumer reinforces his f e e l i n g of belonging and at the same time avoids r i d i c u l e or o s t r a c i z a t i o n for not j o i n i n g the group. The s o c i a l climber may, rather than shoppingsat an outlet compatible to h i s s o c i a l l e v e l , patronize an outlet which i s considered to be a class or classes above his own. He feels that he i s better than the class to which he belongs and reinforces t h i s through patronization of "upper c l a s s " stores. He gains psychological value through such pr a c t i c e s . Forbes, op. c i t . . page 5. 42 The "psychic income" and s o c i a l class components of £ , although d i f f i c u l t to measure, may then account for the " i r r a t i o n a l " shopping behaviour of many consumers. Assuming that one i s able to measure the degree of "psychic income" gained by the consumer during a shopping t r i p , i t i s i n t u i t i v e l y f e l t that the p r o b a b i l i t y curve of consumer patronage with regards to the degree of "psychic income" gained i s as follows; FIGURE 11 THE PROBABILITY OF CONSUMER PATRONAGE -"PSYCHIC INCOME" RELATIONSHIP +1 Pro b a b i l i t y of Consumer Patronage Degree of "Psychic Income" Gained Note: i t i s assumed that a l l variables except the degree of "psychic income" gained are constant. 43 Much l i k e the p r o b a b i l i t y d i s t r i b u t i o n derived for consumer patronage and expected absolute p r i c e d i f f e r e n t i a l s (Figure 9) we see that the above curve does not originate from the o r i g i n . At zero "psychic income", there i s s t i l l a random chance of patronization. As the degree of "psychic income" r i s e s , so r i s e s the p r o b a b i l i t y of consumer patronage. The curve continues to r i s e u n t i l an optimal degree of "psychic income" i s reached. At t h i s point i t le v e l s o f f , a l l other things equal. Advertising R e t a i l stores advertise for one or both of the following reasons: to pass on information to the consumer concerning products and product prices and/or to b u i l d or create a store image. In eithe r case the r e t a i l e r i s attempting to make the consumer aware of some advantage he can gain through shopping at a p a r t i c u l a r o u t l e t . The consumer may be informed as to the date and nature of an upcoming sa l e . He may be made aware of the " f a c t " that "The Bay" deals only i n " q u a l i t y " goods. Such information may a t t r a c t the consumer into t r a v e l l i n g longer distances than he normally does on a shopping t r i p . He does so for he feels that the advantages to be gained through t r a v e l l i n g longer distances outweigh the costs. 44 One should r e a l i z e that i t i s not just the amount of advertising that we are concerned with, but also the subject matter of advertisements. The r e t a i l e r must be informed as to what "advantages" the majority of his po t e n t i a l customers desire. Is i t absolute price differences, the qu a l i t y of goods, the s o c i a l significance of patronizing a p a r t i c u l a r outlet, brand or product preference or a combination of a l l of these "advantages"? I n t u i t i v e l y , then, one must say that advertising does a f f e c t consumer patronage behaviour. But l i k e the "psychic income" and s o c i a l class components of £ i t i s also d i f f i c u l t to measure the consumer e f f e c t s of advertising. However, assuming that the r e t a i l e r i s advertising the "advantages" to patronizing his p a r t i c u l a r outlet which are consistent with those desired by a majority of probable consumers, one may assume the following p r o b a b i l i t y curve; 45 FIGURE 12 THE PROBABILITY OF CONSUMER PATRONAGE -AMOUNT OF ADVERTISING RELATIONSHIP +1 Pro b a b i l i t y of Consumer Patronage Amount of Advertising At zero advertising, we do not assume a zero p r o b a b i l i t y of consumer patronage as i t i s possible to learn of "advantages" to be gained through patronizing a p a r t i c u l a r outlet through word of mouth. We must also allow for randomness i n the consumer patronage decision. With zero advertising and a l l other things equal, there i s a random chance of se l e c t i o n for each o u t l e t . Assuming that advertisement content i s that which i s desired by the consumer, that i s , p r i c e information, q u a l i t y information, etc., the p r o b a b i l i t y curve for the amount spent on advertising and the p r o b a b i l i t y of consumer patronage w i l l r i s e as more i s spent on advertising. That i s , with proper media s e l e c t i o n and message content, a greater expense on advertising i s assumed to bring about a greater spread of 46 information to consumers, hence, a r i s e i n the p r o b a b i l i t y of consumer patronage, other things being equal• Taking these assumptions into account, one may conclude, then, that as advertising expenditures r i s e , the p r o b a b i l i t y of consumer patronage r i s e s to some point and then " l e v e l s o f f " . This " l e v e l i n g o f f " occurs because, f i r s t of a l l , at a given point i n time, there tends to be an upper l i m i t to the t o t a l p o t e n t i a l demand for any p a r t i c u l a r product. The easier sales prospects are sold f i r s t ; the more r e c a l c i t r a n t sales prospects remain. As the upper l i m i t i s approached i t becomes increasingly expensive to stimulate sales. In the second place, as a company steps up i t s marketing e f f o r t , i t s competitors are l i k e l y to do the same, with the net r e s u l t that each company experiences increasing sales resistance. 10 Chapter Summary and Conclusions In Chapter III we have seen, through example, the arithmetic s i g n i f i c a n c e of the distance exponent £ , and its a f f e c t upon the outcome of the gravity model. Holding the siz e and distance variables constant, the value of £* was changed from one to two. Thus, any change i n p r o b a b i l i t y was attributed to a change i n the value ftf £ • The significance of these changes was then noted. 10 P h i l i p Kotler, Marketing Management Analysis. Planning and Control. Prentlce-Hall Inc., Englewood C l i f f s , 1967, page 271. 47 Those consumer patronage variables which Huff described as being the components of £ were also outlined and explained. In addition, the s o c i a l class and advertising factors mentioned by Forbes were outlined as being important to the consumer's decision to t r a v e l various distances for various products. Thus, the more general variables concerning consumer patronage behaviour and r e t a i l store location were covered. Upon completing Chapter I I I , one begins to entertain the thought that the distance exponent £ plays an important r o l e i n gravity model c a l c u l a t i o n s . Perhaps minute changes i n £ may r e s u l t i n marked changes i n the outcome of the gravity model. I f so, i t i s important to attempt to i s o l a t e and quantify the relationships of the outlined components of £ wherever possible. In addition to the variables mentioned i n Chapter III Forbes mentions the factor of store v a r i a b l e s . Because of t h e i r s p e c i f i c nature, store variables w i l l be outlined i n the chapter which follows. CHAPTER IV SPECIFIC STORE VARIABLES - THE SUPERMARKET Introduction Store variables d i f f e r i n importance for various store types. Because of these differences i t was decided that from t h i s point on, reference w i l l be made s o l e l y to the problem of r e t a i l supermarket patrons. Such s p e c i f i c references, i n addition to a s s i s t i n g supermarket owners, may also serve as a procedural example for those analyzing consumer patronage behaviour for other store types. Store variables include such things as the siz e of the store, the number of checkouts, the a v a i l a b i l i t y of parking, f r i e n d l i n e s s of personnel, etc. They d i f f e r i n importance for various store types. The number of customer checkout i n s t a l l a t i o n s i s an important supermarket v a r i a b l e . Whereas i n a men's clothing store, the number of checkouts provided i s probably less important. 49 Chapter IV, then, w i l l be concerned with those p a r t i c u l a r store values which are f e l t to be of prime concern i n supermarket patron analysis. Like Chapter III i t i s s t r i c t l y d e s c r i p t i v e i n nature. It i s not u n t i l Chapters V and VI that hypotheses are formulated and tested using actual data. Number of Parking Spaces It i s i n t u i t i v e l y f e l t that by adding parking f a c i -l i t i e s to a supermarket that consumers w i l l look upon t h i s as an increase i n the conveniences provided by that p a r t i c u l a r o u t l e t . The consumer no longer has to drive around looking for a place to park h i s automobile. He i s no longer forced to " f i g h t " t r a f f i c i n order to f i n d a parking space. However, at c e r t a i n times of the day, week, month and/or year these parking f a c i l i t i e s may become crowded. Congestion mounts and customers find themselves parking on the perimeter of the parking l o t . The larger the parking area, the further the walk for those who park on the perimeter of the l o t . I n i t i a l l y , then, during c e r t a i n shopping times, there i s a point of diminishing consumer u t i l i t y for larger parking areas, x, Figure 13. 50 FIGURE 13 THE EXPECTED CONSUMER UTILITY -NUMBER OF PARKING SPACES PROVIDED RELATIONSHIP (A) Expected Consumer U t i l i t y Number of Parking Spaces Provided The curve is discrete in nature, as the addition of parking spaces is in whole, rather than in fractioned spaces. Notice that the curve does not begin at the origin. A l l other things held equal, a supermarket with zero parking f a c i l i t i e s is s t i l l patronized and therefore has consumer u t i l i t y . 51 Every individual has the a b i l i t y to learn. Hence, the i n i t i a l l y frustrated customer finds that by altering her shopping times, she is able to avoid incurring the d i s u t i l i t i e s referred to above. However, in doing so, the consumer may be sub-optimizing aggregate shopping satisfaction. Holding a l l other factors constant, the consumer may now be shopping at a time which is less convenient for her. Thus, she is satisficing, rather than optimizing one of her consumer patronage goals in favour of available parking, Figure 14. Satisficing may also occur in another form. Should the consumer find that i t is inconvenient to patronize her most optimal choice of supermarkets at a different time, she may patronize a supermarket having a less congested parking area. In doing so, she has chosen a supermarket which is less than optimal in one or a l l of the remaining patronage factors, with the exception of parking. Here, she outweighs these less than optimal factors with the fact that there is less congestion and/or a shorter walking distance in the less attractive supermarket's parking area, Figure 14. Therefore, the curve of consumer u t i l i t y related to the number of parking spaces provided, is as follows; 52 FIGURE 14 THE EXPECTED CONSUMER UTILITY -NUMBER OF PARKING SPACES PROVIDED RELATIONSHIP (B) Expected Consumer U t i l i t y Number of Parking Spaces Provided The r e s u l t i n g consumer patronage p r o b a b i l i t y curve i s then; FIGURE 15 THE PROBABILITY OF CONSUMER PATRONAGE -NUMBER OF PARKING SPACES PROVIDED RELATIONSHIP +1 Pro b a b i l i t y of Consumer Patronage Number of Parking Spaces Provided Note: i t i s assumed that a l l variables except the number of parking spaces provided are constant. 53: As stated e a r l i e r , there i s no point of zero p r o b a b i l i t y due to the randomness of s e l e c t i o n , a l l other variables equal. Thus, the p r o b a b i l i t y curve originates from a point greater than zero on the v e r t i c a l a x i s . The curve r i s e s to some point, x, where i t then begins to l e v e l o f f . This l e v e l l i n g o f f i s the r e s u l t of the attainment of maximum consumer u t i l i t y for parking upon the completion of the e a r l i e r mentioned learning process. It i s discontinuous i n nature due to the reasons given e a r l i e r . Number of Checkouts Like parking spaces, additional checkouts provide added convenience for the consumer. Waiting time i s reduced and therefore, the d i s u t i l i t y caused by such i s reduced. Unlike parking f a c i l i t i e s , however, the consumer receives no d i s u t i l i t y from a greater than optimal number of checkouts. (One could argue that too many checkouts reduce potential s e l l i n g space.) However, due to the cost of providing a greater than optimal number of checkouts, equipment costs, personnel costs, etc., i t i s f e l t that an over-supply of checkouts w i l l not be provided. 54 FIGURE 16 THE EXPECTED CONSUMER UTILITY -NUMBER OF CHECKOUTS PROVIDED RELATIONSHIP Expected Consumer U t i l i t y Number of Checkouts Provided The preceding curve i s again discontinuous, as the number of checkouts can be either 1, 2, 3 etc., not f r a c t i o n a l u n i t s . Notice that the curve begins at the o r i g i n as a supermarket with zero checkouts w i l l not function. The peak and corresponding downturn are merely i l l u s t r a t i v e of the point made i n the previous paragraph. The r e s u l t i n g p r o b a b i l i t y curve would then be as follows; 55 FIGURE 17 THE PROBABILITY OF CONSUMER PATRONAGE -NUMBER OF CHECKOUTS PROVIDED RELATIONSHIP +1 Prob a b i l i t y of Consumer Patronage Number of Checkouts Provided Note: i t i s assumed that a l l other variables except the number of checkouts are constant* Again, assuming that we are looking only at the upward sloping portion of the consumer u t i l i t y curve, we have the p r o b a b i l i t y curve r i s i n g and approaching one* The curve i s discontinuous and begins at the o r i g i n for the reasons given above. 56 Product Quality Image Product q u a l i t y rather than price may be an important variable i n the determination of consumer patronage behaviour. It i s through t h i s variable that consumer brand preferences may be recognizede In pre f e r r i n g one brand of a product to another, the consumer feels that she i s getting the highest q u a l i t y good within a p a r t i c u l a r grocery product's p r i c e range. Hence, for that portion of a supermarket's customers which i s characterized by brand preferences, consumer u t i l i t y r i s e s as product q u a l i t y r i s e s within a given price range. It Is i n t u i t i v e l y f e l t that the relevant consumer u t i l i t y curve for a supermarket's product q u a l i t y image within a given range of price i s as follows; FIGURE 18 THE EXPECTED CONSUMER UTILITY-PRODUCT QUALITY IMAGE RELATIONSHIP Expected Consumer U t i l i t y Overall Product Quality Image Within a Given Price Range 57 At a zero q u a l i t y image, the consumer experiences zero utility„ As the image r i s e s , consumer u t i l i t y r i s e s , but at a slow rate. When an acceptable l e v e l of q u a l i t y i s attained, consumer u t i l i t y r i s e s at an increasing rate to the point of optimal u t i l i t y where i t l e v e l s o f f . > From the above curve, i t i s possible to i n t u i t i v e l y determine the following p r o b a b i l i t y of consumer patronage curve; FIGURE 19 THE PROBABILITY OF CONSUMER PATRONAGE -PRODUCT QUALITY IMAGE RELATIONSHIP +1 Pro b a b i l i t y of Consumer Patronage Overall Product Quality Image Within a Given Price Range Note: i t i s assumed that a l l variables except the o v e r a l l product q u a l i t y image within a given p r i c e range are constant. f 58 The above i l l u s t r a t i v e p r o b a b i l i t y curve i s much the same as the preceding u t i l i t y curve. However, at a zero product q u a l i t y image, we have a minute p o s i t i v e p r o b a b i l i t y of patronage due to the randomness of s e l e c t i o n . As the product q u a l i t y image r i s e s , so r i s e s the p r o b a b i l i t y of patronage but at a far lower rate. Upon attaining an acceptable l e v e l of qualit y , the p r o b a b i l i t y curve begins to r i s e at an increasing rate, l e v e l i n g o f f at an optimal point which i s at some l e v e l below a patronage p r o b a b i l i t y of 1.0. Store Image In discussing store image, we are concerned with those store variables which are attributed to the store i t s e l f . We are e s p e c i a l l y concerned with those variables which appeal to the majority of supermarket patrons. We have seen two possibly e f f e c t i v e store image variables; product price and product q u a l i t y . We have also seen the expected e f f e c t of advertising these " a t t r i b u t e s " to the consumer. In addition to these general image variables of product price and product qu a l i t y , one may also hypothesize about the importance of cleanliness and f r i e n d l i n e s s of personnel images of a supermarket. 59 Store Cleanliness Image For a majority of supermarket patrons, store cleanliness i s an important c r i t e r i o n for store s e l e c t i o n . The Western World stresses cleanliness i n cooking and eating habits. This ultimately begins at the place of food purchase, the supermarket. Whether consciously or subconsciously, a majority of Western World supermarket patrons would probably rather shop at supermarkets which have an image of c l e a n l i n e s s . In t h i s vein, i t i s possible to i n t u i t i v e l y derive the following consumer u t i l i t y curve for store cleanliness; FIGURE 20 THE EXPECTED CONSUMER UTILITY -STORE CLEANLINESS IMAGE RELATIONSHIP Expected Consumer U t i l i t y Degree of Store Cleanliness Projected 60 The consumer group i l l u s t r a t e here i s that which puts primary in t e r e s t on store cleanliness, a l l other things equal. In the Western World, i t i s assumed that t h i s group would also form a majority. At zero cleanliness, consumer u t i l i t y i s zero. The curve then begins to r i s e but at a very slow rate, u n t i l some minimum l e v e l of cleanliness i s reached. As the degree of cleanliness then continues to r i s e , i t i s f e l t that consumer u t i l i t y w i l l also r i s e but at an exponential r a t e . After reaching an "acceptable" degree of cleanliness the curve continues to r i s e but at a very slow rate. From t h i s curve, i t i s possible to derive a p r o b a b i l i t y of consumer patronage curve; 61 FIGURE 21 THE PROBABILITY OF CONSUMER PATRONAGE -STORE CLEANLINESS IMAGE RELATIONSHIP Note: I t i s assumed that a l l variables except the degree of store cleanliness, projected through advertising and "word-of-mouth", are constant. Taking the above assumption into consideration, we find that, once again, due to the randomness of s e l e c t i o n at zero cleanliness, the p r o b a b i l i t y curve does not begin at the o r i g i n . From i t s point of o r i g i n , the p r o b a b i l i t y curve r i s e s much the same as the u t i l i t y curve, meeting a suitable l e v e l of cleanliness before r i s i n g . The curve then begins to " l e v e l ouf" a f t e r reaching an "acceptable" degree of c l e a n l i n e s s . 62 Friendliness of Store's Personnel Image Turning now to the f r i e n d l i n e s s of personnel employed at a p a r t i c u l a r o u t l e t , we find that the consumer u t i l i t y curve can take on a negative, as well as a p o s i t i v e aspect. Because man i s a s o c i a l l y oriented animal, we fi n d f r i e n d l i n e s s to be important to h i s s o c i a l well being. "No man can l i v e alone". Although the degree of s o c i a l o r i e n t a t i o n d i f f e r s among indi v i d u a l s , i t does e x i s t , no matter how minutely, i n a l l men. For that f r a c t i o n of the population which puts prime concern on fr i e n d l i n e s s of personnel, we have the following consumer u t i l i t y curve; FIGURE 22 THE EXPECTED CONSUMER UTILITY -PERSONNEL FRIENDLINESS IMAGE RELATIONSHIP Expected Consumer ^ — -U t i l i t y f Degree of Personnel Friendliness 63 Notice that i t i s quite possible to have negative consumer u t i l i t y . This would be the case for surliness of personnel. As the degree of surliness declines, so declines the degree of negative consumer u t i l i t y . At zero f r i e n d l i n e s s and surliness of personnel, the consumer gathers zero u t i l i t y , a l l other factors held constant. As the degree of f r i e n d l i n e s s r i s e s , so does consumer u t i l i t y . However, t h i s u t i l i t y curve "peaks out" as store patrons become suspicious or annoyed with clerks who are too f r i e n d l y . The resultant p r o b a b i l i t y curve i s ; FIGURE 23 THE PROBABILITY OF CONSUMER PATRONAGE -PERSONNEL FRIENDLINESS IMAGE RELATIONSHIP Pr o b a b i l i t y of Consumer Patronage Note: Degree of Personnel Friendliness i t i s assumed that a l l variables with the exception of the degree of personnel f r i e n d l i n e s s are constant. 64 We are assuming here that the consumer i s concerned with the personnel image put forth by the outlet i n general. It i s i n t u i t i v e l y f e l t that although the f r i e n d l i n e s s of personnel i s considered i n the patronage decision, i t i s not as important as, for example, the a v a i l a b i l i t y of parking. Hence, the consumer, i n t h i s case, does not wish to s a t i s f i c e a l l other variables i n order to optimize the degree of personnel f r i e n d l i n e s s . Because of the randomness of s e l e c t i o n , the curve does not begin at the o r i g i n . Similar to the store cleanliness curve, i t r i s e s exponentially to a point where i t begins to " l e v e l o f f " . However, due to the assumption that personnel can be "too" f r i e n d l y , the curve forms a peak, in d i c a t i n g a decrease i n consumer u t i l i t y . As stated e a r l i e r , i t i s d i f f i c u l t to imagine the f r i e n d l i n e s s of personnel as a prime consideration of r e t a i l store s e l e c t i o n . It i s f e l t , however, that i t plays, at the l e a s t , a secondary p o s i t i o n i n the consumer patronage decision. For t h i s reason, the above u t i l i t y and patronage p r o b a b i l i t y curves were i n t u i t i v e l y derived. 65 Additional F a c i l i t i e s In addition to s e l l i n g groceries, supermarket owners are now opening bakery shops, butcher shops and/or delicatessens within t h e i r stores. In doing so, supermarket owners are attempting to increase the attractiveness of t h e i r p a r t i c u l a r stores. The consumer i s now able to purchase not only groceries, but baked goods and/or fresh meat i n one store. Consumer u t i l i t y then increases as the number of f a c i l i t i e s , provided at a single location, r i s e s . This i s one of the theories behind the shopping center where there are many f a c i l i t i e s i n the immediate area. The consumer u t i l i t y curve for a dditional store f a c i l i t i e s i s then; FIGURE 24 THE EXPECTED CONSUMER UTILITY -ADDITIONAL FACILITIES PROVIDED RELATIONSHIP Expected Consumer . • * * U t i l i t y A dditional F a c i l i t i e s Provided 6b Like the consumer u t i l i t y curve for the number of checkouts provided, t h i s curve i s discontinuous i n nature. The supermarket can either add a bakery shop or not. Differences i n bakery s i z e , product p r i c e , product q u a l i t y , etc, appear as increases or decreases on the v e r t i c a l consumer u t i l i t y a x is. The curve originates at a point above the beginning of a curve which i s i n i t i a l l y characterized by randomness. This i s because the consumer gains some u t i l i t y from a supermarket handling groceries alone, be i t product price d i f f e r e n t i a l s , store cleanliness, product q u a l i t y , etc. The u t i l i t y curve r i s e s as f a c i l i t i e s are added, " l e v e l i n g o f f " at some optimal number of additional f a c i l i t i e s . Such a " l e v e l i n g o f f " occurs as there i s only so much a consumer can do i n one shopping t r i p , a f t e r which any other f a c i l i t i e s cannot be u t i l i z e d due to a lack of time. The curve of the p r o b a b i l i t y of consumer patronage with respect to additional store values can be i l l u s t r a t e d as follows; 67 FIGURE 25 THE PROBABILITY OF CONSUMER PATRONAGE -ADDITIONAL FACILITIES PROVIDED RELATIONSHIP +1 Prob a b i l i t y of Consumer Patronage Additional F a c i l i t i e s Provided Note: i t i s assumed that a l l other variables with the exception of additional f a c i l i t i e s are constant. The p r o b a b i l i t y curve, with the assumption of a l l other things remaining equal, originates at the point of random sel e c t i o n as i l l u s t r a t e d e a r l i e r . It i s discontinuous, as outlined through the i l l u s t r a t i o n of consumer u t i l i t y , and i t s v e r t i c a l p o s i t i o n i s determined by the same reasoning. The " l e v e l i n g o f f " of the p r o b a b i l i t y curve i s a r e s u l t of the lack of u t i l i t y to be gained from those f a c i l i t i e s which are not u t i l i z e d . 68 Chapter Summary and Conclusions This chapter s p e c i f i c a l l y outlined those store variables, described by Forbes, which are applicable to supermarket patron analysis and distance decay. Thus, i n supermarket consumer behaviour analysis, the analyst must take into account the relevant general variables described i n Chapter I I I and the s p e c i f i c variables described i n Chapter W. The previous chapters have presented the background and t h e o r e t i c a l portion of the t h e s i s . The gravity model has been outlined and additional variables have been postulated as having a marked e f f e c t upon the model's outcome and needing further consideration when studying supermarket customer behaviour. The chapters which follow are more p r a c t i c a l i n nature. They involve the s e t t i n g of various hypotheses concerning the e f f e c t or e f f e c t s of the previously described variables upon sales, and hypotheses t e s t i n g . CHAPTER V THE ESTABLISHMENT OF WORKING HYPOTHESES Introduction Chapter V concerns i t s e l f with the formulation of hypotheses regarding the possible e f f e c t , or effects of various consumer behaviour variables upon supermarket sales* In addition to Huff's suggested product p r i c e variable, hypotheses regarding s p e c i f i c consumer behaviour variables and supermarket buying habits, other than those allowed for i n /C , w i l l be made. Chapter V w i l l be concerned with the establishment of testable hypotheses* That i s , hypotheses which can be tested using currently available data. As a r e s u l t , some of the variables discussed i n the preceding t h e o r e t i c a l portion of the thesis w i l l be omitted. In each case the c r i t e r i a for omission w i l l be discussed. It i s hoped that as data c o l l e c t i o n techniques improve, one w i l l be able to f u l l y test a l l consumer behaviour variables and t h e i r e f f e c t s upon r e t a i l store patronage i n general. 70 In the chapters which follow, hypotheses w i l l be established and tested using currently available data. Testing w i l l include simple and multiple regression techniques and T-test analyses. Upon completion of the various t e s t s , thesis conclusions w i l l be drawn. The Available Data The t e s t i n g of hypotheses i s constrained by the a v a i l a b i l i t y of appropriate data. Hence, Chapter IV opens with a d e s c r i p t i o n of the available data. The data was c o l l e c t e d under the following sub-sections; 1. Annual Sales 2. S e l l i n g Area (Square Feet) 3. Parking Spaces - Car Spaces 4. Complementary Area - The evaluation of r e t a i l stores and services i n the immediate area. 5. Bakery A. F u l l B. Bake-off, Service Bakery 6. Delicatessen - yes or no 7. Discount Prices - yes or no 8. External Appearance 71 9• Internal Appearance 10. Number of Customer Checkouts 11. Store Opening Year (Limited Data) The establishment of testable hypotheses i s , then, constrained by the above form of data. The Setting of the Hypotheses As stated e a r l i e r , hypotheses w i l l be set, so as to be tested i n the following chapter using the currently available data. In each case, a p a r t i c u l a r r e l a t i o n s h i p between the chosen variable and store sales w i l l be hypothesized. In addition, the n u l l hypothesis w i l l be stated so as to est a b l i s h a basis for hypothesis t e s t i n g . Degree of Product S u b s t i t u t a b i l i t y The consumer's willingness to t r a v e l to a supermarket o f f e r i n g s p e c i f i c food items or products for sale i s affected by his willingness to substitute other products, offered at more proximate locations. Within the c l a s s i f i c a t i o n of groceries, with the exception of speci a l t y goods, we find that many products can be substituted for one another at l i t t l e expense to the consumer. Packaged r o l l s may be substituted 72 for fresh r o l l s . Canned, frozen and fresh vegetables are interchangeable as meal supplements. Products themselves may be interchanged, peas for corn, potatoes for r i c e , and oranges for apples. Therefore, product s u b s t i t u t a b i l i t y , as such, was not allowed for i n the formulating of hypotheses. Expected Absolute Price D i f f e r e n t i a l s The i n c l u s i o n of pri c e p o l i c y information i n the coll e c t e d data allows the formulation of the following hypothesis; hypothesis: the mean sales per square foot of s e l l i n g area of a supermarket employing a discount p r i c e p o l i c y w i l l be s i g n i f i c a n t l y d i f f e r e n t from those of a supermarket employing a r e t a i l price p o l i c y n u l l hypothesis: the mean sales per square foot of s e l l i n g area of a supermarket employing a discount p r i c e p o l i c y w i l l not be s i g n i f i c a n t l y d i f f e r e n t from those of a supermarket employing a r e t a i l p r i c e p o l i c y 73 The Absolute Price In Relation to Consumer Income The absolute price of groceries i n r e l a t i o n to consumer income may play an important r o l e i n the consumer's decision to patronize a p a r t i c u l a r supermarket. However, i t w i l l not be d i r e c t l y allowed for, and tested i n the analysis. Should the consumer's perceived subsistance l e v e l of food i n r e l a t i o n to other goods "required" by her, require excess consumer expenditures, the consumer may become aware of absolute prices i n r e l a t i o n to her income. I f t h i s i s the case, It i s f e l t that the majority of such consumers w i l l begin to "search out" lower price image supermarkets. Thus, i f the absolute price i n r e l a t i o n to consumer Income i s prevalent i n the consumer patronage decision, i t should be accounted for by the variable expected product p r i c e d i f f e r e n t i a l s . "Psychic Income" and S o c i a l Class As stated i n Chapter I I I , both the "psychic income" and s o c i a l class variables of consumer patronage behaviour are, i n the l e a s t , d i f f i c u l t . Although i t i s i n t u i t i v e l y f e l t that both variables play some r o l e i n the patronage of supermarkets, measurement by marketers of the degree and e f f e c t of each has 74 not r e a l l y been successful. Hence, due to the lack of a t o o l of analysis, both the "psychic income" and the s o c i a l class variables w i l l be omitted from variable t e s t s . Advertising The advertising variable w i l l not be included i n variable tests for two reasons; 1. marketers have not been very successful i n quantifying nor determining the e f f e c t s of advertising i n general 2. the amount of advertising done i n the supermarket trade for the area to be studied and simulated, Vancouver, B r i t i s h Columbia, i s roughly equal among the major outlets i n the area. Number of Parking Spaces It was postulated i n Chapter IV that an increase i n the number of parking spaces provided at a p a r t i c u l a r supermarket would r e s u l t i n an increase i n supermarket sales. Formulating a hypothesis, then, re s u l t s i n the following; 75 hypothesis: an increase i n the portion of a parking space per square foot of supermarket s e l l i n g space w i l l r e s u l t i n an increase i n sales per square foot of supermarket s e l l i n g space* n u l l hypothesis: an increase i n the portion of a parking space per square foot of supermarket s e l l i n g space w i l l not r e s u l t i n an increase i n sales per square foot of supermarket s e l l i n g space* Number of Checkouts As postulated i n Chapter IV, the hypothesis regarding the number of customer checkouts i s ; hypothesis: an increase i n the portion of a customer checkout per square foot of supermarket s e l l i n g space w i l l r e s u l t i n an increase i n sales per square foot of supermarket s e l l i n g space* n u l l hypothesis: an increase i n the portion of a customer checkout per square foot of supermarket s e l l i n g space w i l l not r e s u l t i n an increase i n sales per square foot of supermarket s e l l i n g space. 76 Product Quality Image Much l i k e the advertising v a r i a b l e , the store product q u a l i t y image w i l l not be tested, as; 1. i t i s very d i f f i c u l t to meastire how the consumer perceives product q u a l i t y 2. within a given price range, product qual i t y i s roughly equal for goods offered at supermarkets i n the area, Store Cleanliness Image Using the i n t e r n a l and external ratings for store appearance as a means of judging the store cleanliness image, the hypothesis i s as follows; hypothesis: the mean sales per square foot of s e l l i n g area of a supermarket with an o v e r a l l , high, internal-external appearance r a t i n g w i l l be s i g n i f i c a n t l y d i f f e r e n t from those of a supermarket with an o v e r a l l , low, i n t e r n a l -external appearance r a t i n g . n u l l hypothesis: the mean sales per square foot of s e l l i n g area of a supermarket with an o v e r a l l , high, internal-external appearance r a t i n g w i l l not be s i g n i f i c a n t l y d i f f e r e n t from those of a supermarket with an o v e r a l l , low, i n t e r n a l -external appearance r a t i n g . 77 Friendliness of Store Personnel Image As discussed i n Chapter IV, the f r i e n d l i n e s s of personnel image of a p a r t i c u l a r supermarket may be a factor considered by the consumer i n her decision to patronize that p a r t i c u l a r o u t l e t . However, due to the d i f f i c u l t y of measuring, and hence, the lack of s u f f i c i e n t data, hypotheses w i l l not be formulated i n regards to t h i s p a r t i c u l a r consumer patronage v a r i a b l e . Additional F a c i l i t i e s Using the data provided, i t i s possible to include additional f a c i l i t i e s within a p a r t i c u l a r supermarket, i e . bakery, meat counter, etc., and those i n the immediate area, i e . l i q u o r store, clothing store, hardware store, etc. i n t h i s v a r i a b l e . The hypothesis i s then; hypothesis: the mean sales per square foot of s e l l i n g area of a supermarket with additional f a c i l i t i e s within, or i n the immediate area of the premises, w i l l be s i g n i f i c a n t l y d i f f e r e n t from those of a supermarket without such f a c i l i t i e s within, or i n the immediate area of the premises. 78 n u l l hypothesis: the mean sales per square foot of s e l l i n g area of a supermarket with additional f a c i l i t i e s within, or i n the immediate area of the premises, w i l l not be s i g n i f i c a n t l y d i f f e r e n t from those of a supermarket without such f a c i l i t i e s within, or i n the immediate area of the premises. Store Sales and the P r o b a b i l i t y of Consumer Patronage One might theorize that an increase i n a p a r t i c u l a r store's sales i s a r e s u l t of a r i s e i n the p r o b a b i l i t y of consumer patronage. Such a theory may a r i s e from the following; 1. a greater number of customers patronizing the outlet 2. the same number of customers spending a greater share of t h e i r income at the p a r t i c u l a r o u t l e t . In either case, one could assume that because store sales have r i s e n , the p r o b a b i l i t y of consumer patronage has also r i s e n . The above may be a v a l i d theory. However, a lack of s u f f i c i e n t data makes i t impossible to test at t h i s time. 79 Chapter Summary and Conclusions In t h i s chapter, working hypotheses have been formulated i n regards to the relationships between many of the previously discussed consumer behaviour variables and supermarket sales. However, a l l previously discussed variables were outlined and i f needed, j u s t i f i c a t i o n s of omissions from hypotheses formulation were presented. The c r i t e r i o n for omission, a lack of testable data, was also discussed and an outline of the currently available data was presented. In addition, a digression regarding store sales and the p r o b a b i l i t y of consumer patronage was presented. In the chapter which follows, the hypotheses formulated i n Chapter V w i l l be tested, and either the hypothesis or the n u l l hypothesis accepted. One may conclude from Chapter V that the p r a c t i c a l analyst w i l l be hampered by; 1. a lack of currently suitable data 2. a lack of currently available tools of analysis. 80 The hypotheses to be tested by the data i n the next chapter are: hypothesis: hypothesis: hypothesis: hypothesis: hypothesis: an increase i n the portion of a parking space per square foot of supermarket s e l l i n g space w i l l ( w i l l not) re s u l t i n an increase i n sales per square foot of supermarket s e l l i n g space an increase i n the portion of a customer checkout per square foot w i l l ( w i l l not) r e s u l t i n an increase i n sales per square foot of supermarket s e l l i n g space the mean sales per square foot of s e l l i n g area of a supermarket employing a discount price p o l i c y w i l l ( w i l l not) be s i g n i f i c a n t l y d i f f e r e n t from those of a supermarket employing a r e t a i l p r ice p o l i c y the mean sales per square foot of s e l l i n g area with an o v e r a l l , high, internal-external appearance r a t i n g w i l l ( w i l l not) be s i g n i f i c a n t l y d i f f e r e n t from those of a supermarket with an o v e r a l l , low, internal-external appearance r a t i n g the mean sales per square foot of s e l l i n g area of a supermarket with additional f a c i l i t i e s within, or i n the immediate area of the premises, w i l l ( w i l l not) be s i g n i f i c a n t l y d i f f e r e n t from those of a supermarket without such f a c i l i t i e s within, or i n the immediate area of the premises* CHAPTER VI HYPOTHESES TESTING Introduction Chapter VI i s concerned with the s t a t i s t i c a l analysis of the hypotheses stated i n Chapter V. The analysis was carried out i n order to test the v a l i d i t y of the hypotheses. The analysis also served to enlighten the analyst's knowledge of consumer shopping habits i n a metropolitan area. Testing the hypotheses involved t - t e s t s and simple and multiple regression analysis, using the University of B r i t i s h Columbia SIMCORT and TRIP (Triangular Regression Package) computer programs. The SIMCORT program was u t i l i z e d i n order to perform t-tests on the data provided. SIMCORT was chosen as i t allows the analyst to carry out a t- t e s t upon two variable means for which the number of observations per v a r i a b l d i f f e r s . TRIP was u t i l i z e d to calculate means, standard deviations, c o r r e l a t i o n c o e f f i c i e n t s and, to perform both simpl and multiple regression analyses. 82 Chapter VI begins with a de s c r i p t i o n of the data which was e a r l i e r outlined. Following t h i s , the analyses and t h e i r respective r e s u l t s are presented. The chapter analysis i s divided into two sections; the f i r s t , involving those hypotheses tested using a t - t e s t to test the difference between two means and the second involving those hypotheses tested using simple and multiple regression techniques. In each case, the hypotheses to be tested w i l l be stated, the r e s u l t s presented and the conclusions drawn. Several computer programs u t i l i z e d i n data transform-ation i n preparation for the main analysis appear i n Appendices. The Data The data outlined i n Chapter V was accumulated for supermarkets and grocery stores ranging i n estimated annual sales of less than $300,000 to greater than $9,000,000. Hence, the data includes a l l grocery stores located i n the Lower Mainland Area of Vancouver, B r i t i s h Columbia. The complete census consisted of 1283 grocery stores and included the following, e a r l i e r described data; 83 !• Annual Sales 2. Sales Area (Square Feet) 3. Parking Spaces - Car Spaces 4. Complementary Area - the evaluation of r e t a i l stores and services i n the immediate area. Excellent 8 Very Good 6 Good 5 F a i r 4 Poor 3 Bad 2 None 0 5. Bakery F u l l 1 Bake-off, Service Bakery 2 None 0 6. Delicatessen - yes 1 - no 0 or blank 7. Discount Prices - yes 1 - no G or blank 8. External Appearance - same scale as Complementary Area 9. Internal Appearance - same scale as Complementary Area 10. Number of Customer Checkouts 11. Store Opening Year (Limited Data) 84 The subjective evaluations of appearance and complementary area were kept as consistent as possible through the minimization of the number of observers and evaluators. Cross-checks of evaluations showed very consistent r e s u l t s * It was f e l t that the evaluation of i n t e r n a l and external appearance should be tested as s i n g l e , rather than separate parameters. In t h i s way, the consumer's o v e r a l l opinion of a p a r t i c u l a r appearance would be simulated. The consumer, i n looking at the appearance and cleanliness of a supermarket, i s i n i t i a l l y drawn or repelled by the external projection of each. Should t h i s external projection meet the consumer's standards, she w i l l enter the supermarket. I f not, assuming that t h i s variable plays a major ro l e i n her decision to patronize a p a r t i c u l a r supermarket, she w i l l choose to patronize another o u t l e t . Should she enter and f i n d that the i n t e r n a l appearance and cleanliness projections are unsatisfactory, she w i l l again choose to patronize another establishment. Hence, i t i s a combination of both the i n t e r n a l and external appearance and cleanliness projections which attr a c t s t h i s p a r t i c u l a r form of consumer. This single parameter was determined through the averaging of the i n t e r n a l and external appearance evaluations. 85 In each case, the evaluations involving appearance and complementary area were grouped on the following basis; 1) "good or better" (an average evaluation greater than or equal to 5,0) 2) " f a i r or worse" (an average less than 5.0) Those supermarkets which were c l a s s i f i e d as having a bakery were defined as; supermarkets i n which the "raw" bakery dough was prepared, formed and baked into the f i n a l product within the premises. Those supermarkets which were c l a s s i f i e d as having a bake-off were defined as; supermarkets i n which the "raw" bakery dough was prepared and formed outside the premises and baked into the f i n a l product within the premises. The ef f e c t of having a bakery or bake-off within the store was taken to be equivalent, as In each case, the goods are fr e s h l y baked on the premises. Hence, there was no d i s t i n c t i o n made between these two and the analysis was c a r r i e d out on a bakery or bake-off and a no bakery or bake-off basis. 86 t-Test Analysis The Hypotheses Tested U t i l i z i n g t - t e s t techniques three hypotheses were tested; 1. the mean sales per square foot of s e l l i n g area of a supermarket employing a discount p r i c e p o l i c y w i l l ( w i l l not) be s i g n i f i c a n t l y d i f f e r e n t from those of a supermarket employing a r e t a i l price p o l i c y 2. the mean sales per square foot of s e l l i n g area of a supermarket with an o v e r a l l , high, internal-external appearance r a t i n g w i l l ( w i l l not) be s i g n i f i c a n t l y d i f f e r e n t from those of a supermarket with an o v e r a l l , low, internal-external appearance r a t i n g 3. the mean sales per square foot of s e l l i n g area of a supermarket with additional f a c i l i t i e s within, or i n the immediate area of the premises, w i l l ( w i l l not) be s i g n i f i c a n t l y d i f f e r e n t from those of a supermarket without such f a c i l i t i e s within, or i n the immediate area of the premises. 87 TABLE I OVERALL MEAN SALES PER SQUARE FOOT t-TEST RESULTS MEAN VARIABLE n# SALES/SQ.FT. t-PROB.* R e t a i l Price P o l i c y Discount Price Policy High Overall Appearance Image Low Overall Appearance Image No Bakery or Bake-off Bakery or Bake-off No Delicatessen Delicatessen Good Complementary Area Poor Complementary Area * a t - p r o b a b i l i t y of .05 or less indicates a 57. or less chance of the difference between the tested means being a r e s u l t of randomness # the t o t a l number of v a l i d observations equals 1281 1264 17 124.10 142.30 0.4741 561 720 1239 42 1257 24 78 1203 144.70 108.50 123.00 164.30 123.80 154.90 183.50 120.50 0.0000 0.0041 0.0059 0.0000 88 Price Policy The analysis thus I l l u s t r a t e d that the o v e r a l l mean sales per square foot of s e l l i n g area of a supermarket with a r e t a i l price p o l i c y ($124,10) was not s i g n i f i c a n t l y d i f f e r e n t from the mean sales per square foot of s e l l i n g area of a supermarket with a discount p r i c e p o l i c y ($144.70) (t-Prob. = 0,4741) (Table I ) . Further t e s t i n g yielded the re s u l t s displayed In Table I I . A discount price p o l i c y did not r e s u l t i n s i g n i f i c a n t l y d i f f e r e n t mean sales/sq. f t . for a l l size ranges tested. However these differences were i n the ri g h t d i r e c t i o n for the 0-12,000 sq. f t . and 12,000 - 15,000 sq. f t . ranges. Such res u l t s may indicate one, or a combination of the following; 89 a) as far as expected absolute p r i c e d i f f e r e n t i a l s are concerned, a majority of supermarket customers may be r a t i o n a l i n t h e i r decision to patronize a s p e c i f i c o u t l e t . That i s , the majority of supermarket customers may weigh t r a v e l costs against the absolute p r i c e and hence, savings for the item or items which they expect to purchase. b) as a r e s u l t of inadequate advertising procedures, the price conscious sector of supermarket patrons may not be kept well informed as to price d i f f e r e n t i a l s . c) there may be a large sector of s o c i a l l y conscious supermarket customers who s t r i v e for "psychic income" rather than d o l l a r savings. d) absolute price d i f f e r e n t i a l s may not be sought by a majority of supermarket customers. e) the less able supermarket operators may resort to a discount price p o l i c y which i s apparently not an e f f e c t i v e competitive t o o l i n the market s i t u a t i o n . f) at the time the study was performed the number of discount supermarkets was l i m i t e d as t h i s concept of marketing had only recently been implemented i n the Vancouver area. Thus, the data for the tests was l i m i t e d to the f i r s t discounters i n the area. Since the data was c o l l e c t e d the trend to discounting has been strong. Data c o l l e c t e d now may produce d i f f e r e n t r e s u l t s . TABLE II PRICE POLICY t-TEST RESULTS (WITHIN A GIVEN SIZE RANGE) PRICE POLICY STORE SIZE (SQ.FT.) n# MEAN SALES/SQ.FT. t-PROB.* R e t a i l Discount 0-12,000 1229 7 123.00 151.30 0.4232 Re t a i l Discount 12,000-15,000 8 6 130.00 163.90 0.5945 R e t a i l Discount 15,000-18,000 15 1 136.50 51.95 -0.0000 R e t a i l Discount 18,000- above 12 3 197.90 108.30 0.1471 * a t- p r o b a b i l i t y of .05 or less indicates a 5% or less chance of the difference between the tested means being a resu l t of randomness # the t o t a l number of v a l i d observations equals 1281 91 Overall Internal-External Appearance Image The o v e r a l l mean sales per square foot of s e l l i n g area of a supermarket with a high o v e r a l l appearance average ($144.70) was s i g n i f i c a n t l y d i f f e r e n t from that of a supermarket with a low o v e r a l l appearance average ($108.50) (t-Prob. - 0.0000) (Table I ) . Testing the hypothesis on a given s i z e basis produced i n t e r e s t i n g r e s u l t s . Only i n the 0-12,000 sq. f t . of store s e l l i n g space was a s i g n i f i c a n t difference i n mean sales/sq. f t . recorded. In the 15,000 - 18,000 sq. f t . range the difference was i n s i g n i f i c a n t due to the number of observations recorded. In the remaining two s i z e ranges; 12,000 - 15,000 sq. f t . and 18,000 - above sq. f t . the differences were i n s i g n i f i c a n t and i n the wrong d i r e c t i o n . Such may have come about as a r e s u l t of one, or a combination of the following; a) a majority of supermarket customers may be highly conscious of both the i n t e r n a l and external appearance images projected by small supermarkets b) such stores may be those sought out by customers seeking "psychic income" or s o c i a l image and t h i s sector of the market may be i n the majority c) when a store's sales are not s a t i s f a c t o r y and p r o f i t s are low, i t s appearance may not be maintained. Therefore, appearance may be a r e s u l t of, rather than a cause of poor sales. TABLE III OVERALL INTERNAL-EXTERNAL APPEARANCE IMAGE t-TEST RESULTS (WITHIN A GIVEN SIZE RANGE) OVERALL INTERNAL-EXTERNAL APPEARANCE IMAGE STORE SIZE (SQ.FT.) MEAN SALES/SQ.FT, t-PROB.* High Appearance Image Low Appearance Image High Appearance Image Low Appearance Image High Appearance Image Low Appearance Image High Appearance Image Low Appearance Image 0-12,000 12,000-15,000 15,000-18,000 18,000- above 524 144.20 712 108.00 10 140.70 4 154.20 15 136.50 1 51.95 12 179.60 3 181.40 -0.0000 0.7971 -0.0000 .9267 * a t - p r o b a b i l i t y of .05 or less indicates a 57. or less chance of the difference between the tested means being a result of randomness # the t o t a l number of v a l i d observations equals 1281 93 Additional F a c i l i t i e s (Internal) Bakery or Bake-off The o v e r a l l mean sales per square foot of s e l l i n g area of a supermarket with a bakery or bake-off ($164.30) was s i g n i f i c a n t l y d i f f e r e n t from that of a supermarket without a bakery or bake-off ($122.00) (t-Prob. - 0.0041) (Table I ) . The re s u l t s of t e s t i n g for differences between mean sales of supermarkets with or without bakeries or bake-offs appear i n Table IV. With the exception of the 12,000 - 15,000 sq. f t . range, s i g n i f i c a n t l y higher mean sales/sq. f t . were recorded for supermarkets with bakeries or bake-offs. However, the difference i n mean sales/sq. f t . within t h i s size range was i n the r i g h t d i r e c t i o n . These r e s u l t s may indicate one, or a combination of the following; a) the supermarket customer, when shopping for groceries, may also wish to purchase bakery goods, and may patronize that supermarket which o f f e r s such goods for the sake of t r a v e l l i n g convenience b) the supermarket customer may prefer freshly baked goods to pre-packaged ones and i n deciding to purchase such goods, they may choose a convenient outlet which offers both fr e s h l y baked goods and groceries on the premises 94 c) the a b i l i t y to form fresh "raw" dough on the premises allows the baker to make specia l t y goods which cannot be purchased i n the pre-packaged form, thus adding to the attractiveness of the supermarket d) there may be "psychic income" or s o c i a l image gained through patronizing a supermarket with a bakery or bake-off. TABLE IV BAKERY OR BAKE-OFF t-TEST RESULTS (WITHIN A GIVEN SIZE RANGE) BAKERY OR BAKE-OFF STORE SIZE (SQ.FT.) MEAN n# SALES/SQ.FT, t-PROB.* No Bakery or Bake-off Bakery or Bake-off No Bakery or Bake-off Bakery or Bake-off No Bakery or Bake-off Bakery or Bake-off No Bakery or Bake-off Bakery or Bake-off 0-12,000 12,000-15,000 15,000-18,000 18,000- above 1224 123.00 12 157.50 8 137.00 b 154.70 4 106.10 12 139.60 3 97.17 12 200.70 .0505 .7445 .0150 .0114 * a t-p r o b a b i l i t y of .05 or less indicates a 5% chance or less of the difference between the tested means being a resu l t of randomness # the t o t a l number of v a l i d observations equals 1281 96 Delicatessen The o v e r a l l mean sales per square foot of s e l l i n g area of a supermarket with a delicatessen ($159.90) was s i g n i f i c a n t l y d i f f e r e n t from that of a supermarket without a delicatessen ($123.80) (t-Prob. = .0059) (Table I ) . Testing on a more s p e c i f i c basis proved i n t e r e s t i n g . A f a i r l y s i g n i f i c a n t difference i n mean sales/sq. f t . was recorded i n the 0-12,000 sq. f t . range and i n the 15,000 -18,000 sq. f t . range i s shown i n Table V. The difference i n the 12,000 - 15,000 sq. f t . range was i n the ri g h t d i r e c t i o n but not s i g n i f i c a n t . Notice the 18,000 - above sq. f t . range, the difference i s not s i g n i f i c a n t and i n the wrong d i r e c t i o n . These re s u l t s may indicate one, or a combination of the following; a) there may be "psychic income" or s o c i a l image derived from patronizing a supermarket with a delicatessen b) those customers who patronize a supermarket for i t s delicatessen may fi n d d i s u t i l i t y i n large s i z e stores. TABLE V DELICATESSEN t-TEST RESULTS (WITHIN A GIVEN SIZE RANGE) STORE SIZE MEAN DELICATESSEN (SQ.FT.) n# SALES/SQ.FT. t-PROB.* No Delicatessen Delicatessen 0-12,000 1227 9 123.10 157.00 .0837 No Delicatessen Delicatessen 12,000-15,000 13 1 144.30 147.50 .2494 No Delicatessen Delicatessen 15,000-18,000 8 8 117.10 145.40 .0443 No Delicatessen Delicatessen 18,000- above 9 6 189.50 165.70 .4693 * a t - p r o b a b i l i t y of .05 or less indicates a 57. chance or less of the difference between the tested means being a re s u l t of randomness # the t o t a l number of v a l i d observations equals 1281 98 Additional F a c i l i t i e s (External) The mean sales per square foot of store s e l l i n g area of a supermarket with a high complementary area r a t i n g ($183.50) was s i g n i f i c a n t l y d i f f e r e n t from that of a supermarket with a low complementary area r a t i n g ($120.50) (t-Prob. - 0.0000). That i s , supermarkets with other complementary stores nearby had higher sales per square foot than stores without such additional shopping opportunities nearby (Table I ) . The r e s u l t s of more s p e c i f i c t e s t i n g appear i n Table VI. S i g n i f i c a n t l y higher mean sales/sq. f t . were recorded i n a l l si z e ranges except the 18,000 - above range. However the difference i n t h i s range was i n the r i g h t d i r e c t i o n . These r e s u l t s may indicate that; a) the consumer may s t r i v e to maximize the u t i l i t y of a shopping t r i p through patronizing as many r e t a i l outlets as possible i n one t r i p b) the consumer may find the concept of a "one stop shopping center" convenient, as i t maximizes her exposure to a l l the types of goods she desires and minimizes t r a v e l l i n g time c) the complementary area appearance may be a r e f l e c t i o n of the general economic conditions i n the area and t h i s may be r e f l e c t e d i n the sales of supermarkets located i n these areas. TABLE VI COMPLEMENTARY AREA t-TEST RESULTS (WITHIN A GIVEN SIZE RANGE) COMPLEMENTARY AREA STORE SIZE (SQ.FT.) n# MEAN SALES/SQ.FT. t-PROB.* Good Complementary Area Poor Complementary Area 0-12,000 41 1195 165.70 120.60 -0.0000 Good Complementary Area Poor Complementary Area 12,000-15,000 9 5 175.90 88.10 .0043 Good Complementary Area Poor Complementary Area 15,000-18,000 15 1 131.50 127.40 .0005 Good Complementary Area Poor Complementary Area 18,000- above 13 2 189.40 118.60 1.0000 * a t- p r o b a b i l i t y of .05 or less Indicates a 5% chance or less of the difference between the tested means being a r e s u l t of randomness # the t o t a l number of v a l i d observations equals 1281 VO SO 100 Conclusions: T-test Analyses From the t-t e s t s which were performed, one must conclude that i n general; 1. the mean sales per square foot of s e l l i n g area of a supermarket employing a discount p o l i c y are not s i g n i f i c a n t l y d i f f e r e n t from those of a supermarket employing a r e t a i l price p o l i c y 2. the mean sales per square foot of s e l l i n g area of a supermarket with an o v e r a l l , high internal-external appearance r a t i n g are s i g n i f i c a n t l y d i f f e r e n t from those of a supermarket with an o v e r a l l , low, internal-external appearance r a t i n g 3. the mean sales per square foot of s e l l i n g area of a supermarket with additional f a c i l i t i e s within, or i n the immediate area of the premises are s i g n i f i c a n t l y d i f f e r e n t from those of a supermarket without such f a c i l i t i e s within, or i n the immediate area of the premises* However on a more s p e c i f i c basis store s i z e must be considered i n the analysis of supermarket appearance and additional f a c i l i t i e s within and i n the immediate area of a p a r t i c u l a r o u t l e t . 101 The Correlation Matrix P r i o r to the regression analyses, a c o r r e l a t i o n matrix was produced from the data and i s presented i n Table VII. The matrix y i e l d s a c o r r e l a t i o n c o e f f i c i e n t , r , which i s a measure of the l i n e a r association between two var i a b l e s . The value of r ranges from -1 to +1; -1 indicates a p e r f e c t l y negative l i n e a r r e l a t i o n s h i p between two variables and +1 indicates a p e r f e c t l y p o s i t i v e l i n e a r r e l a t i o n s h i p between the va r i a b l e s . The clo s e r r i s to zero, the weaker the l i n e a r r e l a t i o n s h i p between the va r i a b l e s . TABLE VII THE CORRELATION MATRIX 4J -u .p ^ 4J C r-l 5 r-4 CU *-> C r- l 3 , CO § «} fc co c «o cu >-i co S co cr*w ri 3 3 to w <3 co to o co < w