Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

An economic analysis of sales flyer advertising frequency by competing grocery manufactures and supermarkets Lung, Cary 2005

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
831-ubc_2005-0258.pdf [ 4.72MB ]
Metadata
JSON: 831-1.0091965.json
JSON-LD: 831-1.0091965-ld.json
RDF/XML (Pretty): 831-1.0091965-rdf.xml
RDF/JSON: 831-1.0091965-rdf.json
Turtle: 831-1.0091965-turtle.txt
N-Triples: 831-1.0091965-rdf-ntriples.txt
Original Record: 831-1.0091965-source.json
Full Text
831-1.0091965-fulltext.txt
Citation
831-1.0091965.ris

Full Text

An Economic Analysis of Sales Flyer Advertising Frequency by Competing Grocery Manufacturers and Supermarkets B y C a r y L u n g B . S c . ( A g r . ) Un ive r s i t y o f B r i t i s h C o l u m b i a 2000 A T H E S I S S U B M I T T E D I N P A R T I A L F U L F I L L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F M A S T E R O F S C I E N C E i n T H E F A C U L T Y O F G R A D U A T E S T U D I E S A g r i c u l t u r a l E c o n o m i c s T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A 2005 © C a r y L u n g , 2005 Abstract The advertising strategies of retailers competing for processed food and beverage product markets are studied by constructing a theoretical model and then testing the model's predictions using statistical analysis of primary data. The advertising decisions are jointly determined by the food manufacturers and food retailers. Categorical data consisting of yes-no advertising decisions for 22 processed food products was collected from the sales flyers of four major supermarket chains and was recorded in 2 X 2 contingency tables. The chi-square test for a 2 X 2 contingency table was then used to test the hypotheses concerning yes-no advertising decisions for a particular product in three pairs of competing supermarkets, and for different brands of a common product by two competing manufacturers within a particular supermarket. In the second category four scenarios were considered: (1) a chosen brand versus the private label, (2) a chosen brand versus any national brand other than the chosen brand, (3) any national brand versus the private label, and (4) the private label of one supermarket versus the private label of another supermarket. The results show overwhelming support for the hypothesis that food manufacturers and retailers are randomly choosing which products to promote and when to promote them. For the within-store analyses, the advertising strategies for national brands and private label products are independent in most cases. Results show that there is a negative correlation for those few cases that are statistically dependent, apparently that the food companies are choosing to advertise different brands in a given period to avoid intensive competition. However, in the supermarket selling a large volume of products with private labels, there is a positive correlation for those cases that are dependent. The retailer is evidently advertising both the national brand and the private label to promote the private label products as substitutes for the national brand products. ii Table of Contents Abstract ii List of Tables vii 1 Introduction 1 1.1 Background 1 1.2 Research Question 3 1.3 Research Procedure 5 1.4 Outline 6 2 Literature Review 7 2.1 Academic Research. 7 2.2 Textbook Perspective 12 2.3 Industry Perspective 13 3 Market Environment of Food Retail Industry 15 3.1 Food Retail Market Structure 15 3.1.1 Overview of Canadian Retail Food Industry 15 3.1.2 Food Retail Market in British Columbia 16 3.2 History and Market Position of Supermarket Chains in this Study 17 3.2.1 Market Place IGA and IGA : 17 3.2.2 Safeway Canada 17 3.2.3 Save-on Foods & Drugs 18 3.2.4 The Real Canadian Superstore 19 4 Theoretical Model 21 4.1 Model Assumption 22 4.2 Consumers' Decisions 23 4.3 Stores' Decisions 23 4.4 Pricing and Advertising Equilibrium of the Model 23 iii 5 Data and Methodology 29 5.1 D a t a D e s c r i p t i o n 29 5.1.1 Se lec t ion o f R e t a i l Supermarket C h a i n s 30 5.1.2 C h o i c e o f F o o d and Beverage Ca tegor ies . . . . 31 5.1.3 Speci f ica t ion o f N a m e Brands 31 5.2 M e t h o d o l o g y 32 5.2.1 Theore t ica l Cons ide ra t ion 32 5.2.2 The Chi -Square Test for the 2 X 2 Con t ingency Table 33 5.3 Hypo theses Tes t ing 36 5.3.1 S ta t i c A n a l y s i s 36 5.3.2 D y n a m i c A n a l y s i s . . . . . . . 37 6 Results 39 6.1 S t a t i c A n a l y s i s 4 0 6.1.1 Resul ts o n Across-Store Adver t i sement 40 6.1.2 Resul ts o n A c r o s s - B r a n d Adver t i sement 44 6.1.2.1 The C h o s e n Brands vs Pr ivate L a b e l at E a c h Supermarket 44 6.1.2.2 A n y N a t i o n a l B r a n d vs Pr ivate L a b e l at E a c h Supermarket 46 6.1.2.3 The C h o s e n B r a n d vs Pr ivate L a b e l at E a c h Supermarket 48 6.2 D y n a m i c A n a l y s i s 50 6.2.1 L e a d e r - F o l l o w e r R e l a t i o n s h i p . .50 7 Summary and Conclusion 54 7.1 S u m m a r y 54 7.2 L i m i t a t i o n 56 7.3 C o n c l u s i o n 57 7.4 Future Recommenda t ion 59 References 60 Appendix A 64 Appendix B 65 i v Appendix C Contingency Tables B a c o n 67 Breakfast Cerea l 68 C a n n e d F i s h 69 C a n n e d Soup 70 D r y Pasta 71 F r o z e n P i z z a 72 F r o z e n P u n c h Beverage 73 F r o z e n Vegetables 74 Fru i t J a m 75 G r o u n d Coffee 76 Instant Coffee 77 Juice from Concentrate 78 Juice N o t from Concentrate 79 K e t c h u p 80 M a r g a r i n e 81 M a y o n n a i s e 82 Pasta Sauce 83 Peanut But ter 84 Potato C h i p s 85 Processed Cheese Sl ices 86 Tea B a g s 87 Waf f l e 88 v List of Tables Table 5.1 Selected Products Categories and Name Brands 30 Table 5.2 2 X 2 Contingency Table for a Product in Cross-Store Analysis 34 Table 6.1 Advertising Frequency for the 22 Products at Each Supermarket 41 Table 6.2 Chi-Square Statistics on Between-Store Advertisement Safeway Canada vs Competing Supermarket Chains (Chosen Brand) 43 Table 6.3 Chi-Square Statistics on Between-Store Advertisement Safeway Canada vs Competing Supermarket Chains (Private Label) 44 Table 6.4 Chi-Square Statistics on Within-Store Advertisement The Chosen Brand vs Private Label at Each Supermarket 46 Table 6.5 Chi-Square Statistics on Within-Store Advertisement Any National Brand vs Private Label at Each Supermarket 48 Table 6.6 Chi-Square Statistics on Within-Store Advertisement The Chosen Brand vs Any Other National Brand 49 Table 6.7 Leader-Follower Relationship for the Breakfast Cereal Product Category. 51 Table 6.8 Leader-Follower Relationship for the Canned Fish Product Category. 52 Table 6.9 Leader-Follower Relationship for the Canned Soup Product Category. 52 Table 6.10 Leader-Follower Relationship for the Peanut butter Product Category. 53 vi Chapter 1 Introduction 1.1 Background A n advert is ing sales f lyer is a unique form o f print advertisement that is w i d e l y used by retail grocery stores. The sales f lyer is a s imple and informative w a y to identify the prices corresponding to a w i d e array o f grocery products. T h i s thesis examines sales f lyer advert is ing frequencies for national brands and private label food products. F lye r s are prepared and printed by retailers; however, there is a ver t ica l interaction between the food manufacturers and grocery retailer i n m a k i n g the advert is ing dec is ion . The thesis, w h i c h is wri t ten f rom an industr ial organizat ion perspective, attempts to capture the hor izonta l compet i t ion o f supermarkets se l l ing a c o m m o n product (e.g. Kra f t peanut butter), the hor izonta l compet i t ion o f food manufacturers se l l ing s imi la r products (e.g. store-brand versus Kraf t peanut butter) and to a certain extend the ver t ica l interaction between food manufacturers and grocery retailers. The printed advertisement is a powerfu l too l for communica t ing pr ice informat ion to consumers o f retail supermarket products because o f the eff ic iency o f this fo rm o f advertisement - re la t ively l o w i n cost and easy access ib i l i ty to mass ive number o f potential consumer. F lye rs are t yp i ca l ly del ivered w e e k l y to household doorsteps as an insert to free communi ty newspaper. B u r t o n et a l . (1999) have indicated that there is a strong posi t ive correlat ion between advertisement coverage i n the sales f lyer and the sales v o l u m e o f the advertised products. Indeed, the sales f lyer a l lows the consumer to get the pr ice informat ion contained i n the sales f lyer wi thout incurr ing search costs. A d v e r t i s i n g is an important component o f a store's market ing program because it influences the purchas ing behavior o f consumers regarding store choice and brand choice , w h i c h i n turn has a s ignif icant impact o n the store's performance, especial ly i n the short run (Ni j i s et a l . , 2001 , and Te l l i s and Zuf ryden , 1995). W h i l e the major f l o w o f research examines the impact o f advertisement and pr ice p romot ion o n the store's performance i n terms o f sales, store traffic, and profits , this thesis v i e w s m a x i m i z i n g the store's performance as the p r imary objective o f supermarket managers and m a x i m i z i n g ind iv idua l product performance as the p r imary objective o f food manufacturers. T h i s thesis is examin ing the jo in t advert is ing strategy o f manufacturers and retailers f rom an industr ial organizat ion perspective. 1 Chapter 1. Introduction 2 There is a close ver t ical relat ionship and many potential interactions between manufacturers and supermarket chains i n the processed food industry because there is almost no intermediaries i n N o r t h A m e r i c a compared to the A s i a n market . Re t a i l pr ice is theoretically determined by manufacturers and retailers together, as it is the sum o f the manufacturer 's wholesale pr ice and the retailers ' markup margin . Thus , there is a certain interaction between the manufacturer and the retailer w h e n they are m a k i n g decis ions concerning pr ice p romot ion and advertisement i n the sales flyer. The pr ice p romot ion and advert is ing dec i s ion is a c r i t i ca l interaction because the agreement is based o n the ind iv idua l f i rm's objective, market pos i t ion , and target segment; furthermore the result has a direct impact o n the f i rm 's performance, competi t iveness, and market share. M c G e e (1988) states that previous empi r i ca l studies demonstrate a pos i t ive relat ionship between industry profits and advert is ing intensity. The profit rate o f industries that have a higher advert is ing expenditure is greater than those industries w h i c h advertise less. Thus , pr ice communica t ion plays an important role i n attracting consumers, spec i f ica l ly the informed and rat ional consumers w h o search for the lowest pr ice among the retailers. M c G e e (1988) also indicates that i n general "adver t is ing lowers the cost o f b r ing ing valuable informat ion to buyers, eases the entry o f new products and n e w f i rms, increases compet i t ion , and lowers p r i ce" (p.372). T h i s is supported by substantial empi r i ca l evidence. Manufacturers and retailers have to agree o n certain condi t ions i n order to come to an agreement i n a jo in t advert is ing strategy. F o r example , manufacturers m a y need to subsidize the retailer 's advert is ing cost, make payments, or provide the retailers w i t h a special de l ivery schedule. O n the other hand, retailers m a y need to defer the advertisement o f other brands and feature a manufacturer 's name brand exc lus ive ly w i t h i n the sales flyer. K i m and Stael in (1999) point out that "retailers seem to be extracting numerous concessions from manufacturers, such as slott ing a l lowances , l oca l advert is ing support, deep discounts for p romot iona l activit ies, and special de l ivery schedules, but retailers cannot transfer these concessions into account ing prof i t" (p.60). L a i and Mess inge r (1996) suggest that retailers have a lower profit marg in o n advertised products than o n unadvertised products, w h i c h impl ies that pr ice promot ions are not rea l ly benef ic ia l to the retailers. K i m and Stael in (1999) suggest that frequent pr ice p romot ion increases consumers ' pr ice sensi t ivi ty across stores. Consumers pay attention to advert is ing and compare pr ice w h e n searching for the lowest pr ice avai lable , thus increasing the leve l o f store swi tching . A s a result, compet i t ion between retailers becomes more intense and retailers earn lower profits. In this case, manufacturers must g ive larger payments to Chapter 1. Introduction 3 retailers for merchandis ing activit ies. However , retailers w i l l not pocket a l l the payment because "merchandiz ing act ivi ty affects the degree o f cross-store shopping w h i c h drives them to a lways pass o n a por t ion o f the side payment to the consumer" (p.72). Howeve r , manufacturers ' profits decrease as consumers ' store choice becomes more sensitive to the cross-store pr ice difference i n aggregate p romot iona l act ivi ty. "S ince the p romot iona l activit ies also affect brand shares w i t h i n a store, each manufacturer feels compe l l ed to 'enter into a p romot iona l w a r ' and give the retailers larger a l lowances" (p.73), i m p l y i n g that profits w i l l decrease for both food retailers and manufacturers. K i m and Stael in (1999) suggest that manufacturers should "assist retailers i n b u i l d i n g store loyal ty i n ways other than promot iona l ac t iv i ty" (p.73). A l t h o u g h Manufacturers and Retai lers have to w o r k together i n setting the jo in t advert is ing strategy, the t w o players do have different incentives and expectations from the advertisement. E a c h player wants to increase its sales vo lume , market share and profits i n the respective industry. Manufacturers want to increase sales v o l u m e and market share o f the products under their name brands; w h i l e retailers expect more f rom pr ice advert is ing, higher customer traffic, and increase overa l l sales v o l u m e o f products carr ied i n i n d i v i d u a l store, regardless o f brands and product categories. T h i s thesis w i l l investigate the advert is ing patterns o f products f rom both nat ional brands and private labels. These advert is ing patterns w i l l affect sales o f a product across two compet ing stores as w e l l as the sales o f compet ing brands w i t h i n a part icular store. W e w i l l first examine yes-no advert is ing decis ions for a c o m m o n product so ld by compet ing supermarkets and f ind out i f the jo in t p romot iona l activit ies by manufacturers and retailers are independent across stores. Independent act ivi t ies do not encourage cross-store shopping. T h e n w e w i l l shift the analysis to across-brand advert is ing patterns w i t h i n a supermarket, to determine the extent that yes-no advert is ing decis ions are independent over t ime. The results w i l l reveal the extent that j o in t advert is ing strategies are independent across i nd iv idua l supermarkets and i f there is dependence, the extent that f i rms are either purposely avo id ing or purposely engaging i n pr ice compet i t ion at the retai l l eve l . The analysis w i l l also show h o w the i nd iv idua l retailer chooses h o w to s imul taneously advertise its private label products at the same t ime as it enters into a j o in t advert is ing strategy w i t h food manufacturers. Chapter 1. Introduction 4 1.2 Research Question T h i s thesis seeks to answer the f o l l o w i n g question: A r e compet ing supermarkets l i k e l y to promote a part icular product (e.g., Kra f t peanut butter) at the same t ime or at different t imes? Moreover , are compet ing name brands o f the same generic product (e.g., Kra f t versus S k i p p y peanut butter) l i k e l y to be promoted at the same t ime or at different t imes? A s imple spatial theoretical m o d e l , w h i c h el iminates m u c h o f the complex i ty o f the r ea l -wor ld interaction, is constructed and used to generate specif ic hypotheses concern ing the above questions. The hypotheses are then tested us ing w e e k l y sales f lyer pr ice p romot ion data. The theoretical m o d e l predicts that compet ing f i rms w i l l choose to promote their products o n a random basis. T h i s outcome impl i e s that the pr ice p romot ion data should show statistical independence, both across retailers and across brands. A static m o d e l is employed to test the probabi l i ty o f two products be ing advertised together i n a g i v e n week. In the across-store section, w e test the probabi l i ty o f the same product be ing advertised i n a pair o f compet ing supermarket i n the same week; the n u l l hypothesis is that the advert is ing behavior among the pair o f retailers tested is independent. In the across-brand section, w e test the probabi l i ty o f two brands i n the same category be ing advert ised o n the f lyer o f each supermarket i n the same week. It is hypothesized that the advert is ing strategies o f t w o compet ing brands w i t h i n a product category i n a supermarket, are independent. Three combinat ions o f two compet ing name brands are tested: (1) the chosen brand versus the private label at each supermarket, (2) the chosen brand versus other nat ional brands, and (3) any national brand versus the private label at each supermarket. In reality, advert is ing decisions by food manufacturers and retailers are made jo in t ly , but the research i n this thesis does not exp l i c i t l y address this j o in t interaction. B o t h manufacturers and retailers use advertisement i n p r o v i d i n g informat ion to consumers i n terms o f existence o f sellers and pr ice informat ion for their products. A c c o r d i n g to L a c h (2002), retailers practice random p r i c ing strategy so that consumers cannot learn w h i c h stores have consistently l o w prices. L i k e w i s e , r andom advert is ing strategy w o u l d create an imperfect ion informat ion environment for the consumers. T h i s research attempts to observe the advert is ing behavior i n the o l igopo ly environment between the major supermarket chains i n the Met ropo l i t an o f Vancouver . I f there is a strong evidence that the j o in t advertisement between the manufacturers and the four retail supermarkets i n this analysis have independent advert is ing strategies o n a part icular food product. F o o d retailers are apparently r andomiz ing their strategies to prevent intense compet i t ion , w h i c h w o u l d Chapter 1. Introduction 5 decrease their overa l l profit. 1.3 Research Procedures A s impl i f i ed ve rs ion o f the spatial m o d e l developed b y R a o and S y a m ( 2 0 0 1 ) is presented i n Chapter Three. It demonstrates that i retailers w i l l co l l ec t ive ly earn higher profits i f they choose to advertise different-goods rather than advert is ing the same good. Howeve r , there is a strong incentive for a retailer to deviate and chose a same-goods strategy i f its competi tor is choos ing a different goods strategy. T h i s outcome impl i e s that there is no pure strategy equ i l i b r ium, w h i c h necessari ly impl i e s that w e are left w i t h a m i x e d advert is ing strategy equ i l i b r ium. Speci f ica l ly , retail supermarkets w i l l r andomly choose what products they are go ing to advertise and when , i n any g i v e n week. The story is s imi la r for food manufacturers. The categorical data used to test the theoretical hypothesis was obtained from a co l l ec t ion o f sales flyers f rom four major supermarket chains w i t h store locat ions i n the Vancouver metropol i tan area dur ing a 5 2 week per iod ( M a y 6, 2 0 0 1 to M a y 5 , 2 0 0 2 ) . T h e observations were put into 2 X 2 cont ingency tables. A chi-square test was emp loyed to test the hypotheses concerning the statistical independence o f the yes-no advert is ing status o f a chosen product at two compet ing supermarkets, for both chosen brands and private label , and different brands o f a c o m m o n product o f compet ing manufacturers w i t h i n a part icular supermarket. However , there are some l imi ta t ions to the static m o d e l employed i n this thesis, because the m o d e l does not encounter the dynamic aspects o f the advert is ing environment . There is no considerat ion o f the dynamic aspect over t ime. A l t h o u g h the m o d e l assumes a Ber t rand compet i t ion between the food retailers, and each week is a n e w game for advert is ing decis ion , yet it is un l ike ly to be the case i n reality. St igler ( 1 9 6 1 ) , and Hayes and M o r r i s ( 1 9 9 1 ) suggest that the se l l e r s 'would repeat product advertisement over t ime to remind the forgetful consumers the existence o f advertised i tems and the loca t ion o f sellers. A n industry contact expla ins that some advert is ing decis ions between manufacturers and retailers are made up to three months pr ior to the effective date o f the sales flyer, but there are some last minute amendments and corrections before the pr in t ing o f the flyer. L i t t l e is k n o w n about the t ime lag between w h e n the dec i s ion is made and w h e n the f lyer is distributed. Chapter 1. Introduction 6 Therefore, the same chi-square testing procedure w i t h cont ingency tables is constructed to f ind out i f there is a lag relat ionship between pairs compet ing supermarkets w i t h different brand combinat ions i n the 52 weeks sample per iod . The yes-no advert is ing status o f a brand i n Store A i n week t is compared w i t h the yes-no advert is ing status o f the same brand i n Store B i n week t-k where k={ 1,2,3,4} for both the chosen nat ional brands and the private labels. F o r example , the advert is ing status o f Kra f t peanut butter i n Safeway i n week 1 is compared w i t h the advert is ing status o f Kra f t peanut butter i n M a r k e t P lace I G A i n week 2. L a g relat ionship, between a pair o f compet ing supermarkets w i t h s imi l a r market pos i t ion and target segment, does emerge, i m p l y i n g that there is a possible leader-follower relat ionship among the retailers, but the results show no part icular pattern o f lag relat ionship across-stores and across-brands. The general result o f this t ime- lag analysis w i l l be discussed i n Chapter 7. Because manufacturers place advertisements alternately at different supermarkets so that the product is exposed to consumers cont inuously over t ime o n different supermarkets ' f lyers, the product get exposed to consumers i n different popula t ion segments. G i v e n that the manufacturers w i s h to place a certain number o f advertisements w i t h i n a 52 weeks per iod , both the manufacturers and retailers w o u l d l i ke to alternate the advert is ing status between the compet ing supermarkets i n order to reduce compet i t ion at the retai l l eve l , and accord ing to our spatial m o d e l , retailers w o u l d advertise opposite-goods i n order to obtain a higher profit . 1.4 Outline of the study The second chapter provides a r ev iew o f the economic and market ing literature o n pr ice advert is ing f rom academic perspective, and general pr ice advert is ing practice i n the food retail industry. A b r i e f ove rv i ew o f the Canad ian food retai l market structure and the market ing environment is presented i n Chapter Three. The market pos i t ion o f each o f the four retail supermarkets studied i n this thesis is also discussed i n this chapter. The theoretical mode l , a s impl i f i ed vers ion o f the mode l by R a o and S y a m (2001), for this study is constructed i n Chapter Four. The data and statistical methodology are described i n Chapter F i v e . Chapter S i x contains the empi r i ca l results and the d iscuss ion. A summary and conc lus ion are presented i n Chapter Seven. Chapter 2 Literature Review 2 . 1 Academic Research Reta i l supermarkets often use sales flyer to advertise their w e e k l y pr ice informat ion and special discounts o n selected products from their merchandise. Supermarket managers select different i tems i n var ious product categories for advertisement i n the sales f lyer to improve the performance o f their organizations. Sales f lyer is a power fu l market ing too l because consumers often make their decis ions about w h i c h store to v i s i t and w h i c h product to buy, based o n the informat ion i n the flyer. M a n y researchers have investigated food retailers ' pr ice promot ions f rom the perspective o f consumer behavior and o f industr ia l Organiza t ion . Prev ious research, such as L a i and Matutes (1994), G u p t a (1988), B e l l at a l (1999), N i j i s , D e k i m p e , Steenkamp and Hanssens (2001), and K o n i s h i and Sandfort (2002), have concentrated o n the f o l l o w i n g three effects o f pr ice p romot ion o n store performance: 1. Sales effect - Does pr ice p romot ion for selected i tems increase the overa l l sales v o l u m e o f the food retailer? 2. Profi t effect - Does the pr ice reduct ion or loss leader p r i c i n g has a negative effect o n the store's profi t? 3. Traffic effect - Does pr ice p romot ion for selected items influence the f l o w o f consumer shopping i n the store? E m p i r i c a l research b y Walters and M a c K e n z i e (1988) tests the direct and indirect effects o f pr ice p romot ion w i t h loss leading p r i c i n g o n grocery sales, store traffic and profit , advertised and unadvertised in-store specials, and double coupons promot ions . T h e y conclude that these pr ice promot ions can have a significant impact o n the store's performance i n the short run, but the effect o f loss leader p r i c i n g is essential ly zero i n the long run. Moreove r , the pr ice p romot ion does not have a direct impact o n sales and profit; the reason that the supermarket has a prompt increase i n sales and profit i n the p romot ion per iod is a result o f an increase i n store traffic w i t h a store swi tch ing effect. The pr ice p romot ion at retail supermarket attracts consumers f rom compet ing stores w h i l e retaining its patrons. Because there is a transportation cost incurred by consumers for each shopping tr ip, consumers rarely purchase o n l y the promoted products w h e n they make a shopping 7 Chapter 2. Literature Review 8 tr ip to the grocery store. P romot ion , therefore, stimulates the sale o f overa l l merchandise. The corresponding increase i n store traffic leads to an indirect effect o f higher sales and profit performance. N i j s , D e k i m p e , Steenkamp and Hanssens (2001) y ie lds a s imi l a r conc lus ion from the perspective o f demand effects: pr ice p romot ion results i n category demand that is re la t ively stable i n the long run. A l t h o u g h the long-term impact is essentially zero, the net short-term effect o f pr ice p romot ion o n product demand is substantial. However , the demand effect is on ly related to pr ice advertisement, showing that the consumers are ve ry pr ice sensitive. Non-p r i ce advertisement (i.e. image b u i l d i n g , qual i ty signal) reduces the impact o f pr ice p romot ion o n product demand. A s the compet i t ion i n the product category increases, the short-run pr ice p romot ion effect decreases. Las t ly , N i j s et a l . conclude that r iva ls general ly do not react to the product advertisement and pr ice p romot ion . In other words , the advert is ing strategies between compet ing f i rms are independent. The research o f Walters and M a c K e n z i e (1988) also finds that unadvert ised in-store specials appear to have no effect o n store profit , sales, or traffic. M o s t consumers do not have the pr ice knowledge for each supermarket, and large search costs prevent consumers from shopping around compet ing supermarkets to locate the lowest pr ice . Therefore, consumers are l i k e l y to make their store choice base o n their overa l l reservation pr ice . Supermarkets w i t h unadvertised in-store specials do not general ly earn higher profits , because there is incomplete informat ion i n the market. Un les s the consumer rout inely patronizes a part icular supermarket, then he or she m a y have perfect pr ice knowledge for that supermarket on ly and s t i l l be uncertain about prices i n other retail supermarkets. A price p romot ion is more effective w h e n the temporary discount is advertised. T h i s a l lows the reduced pr ice informat ion to reach large numbers o f consumers, so that the food retailer can achieve the goal o f ga in ing store traffic b y stealing consumers from its rivals. Bur ton , Lichtens te in , and Netemeyer (1999) studied the associat ion between consumers ' exposure to sale flyers and the sales v o l u m e i n retail supermarkets. The i r results indicate that there is a significant relat ionship between the purchasing behavior o f an ind iv idua l consumer w h o is exposed to the advertisement, the number o f advert ised products purchased, and the amount spent o n these products. The results also indicate that there is a posi t ive correlat ion between the advertisement, the pr ice sensitivity, and the age o f a consumer. C o m p a r e d to unadvertised products, the sales v o l u m e doubles w h e n Chapter 2. Literature Review 9 products are advertised. T h i s impl ies that a supermarket manager can influence the consumers ' product choice and purchasing behavior i n order to achieve their goa l o f profit m a x i m i z i n g . A c c o r d i n g to the study by Urbany , D i c k s o n , and Sawyer (2000), "supermarket managers often misjudge consumers ' pr ice search behavior and bel ieve that consumers are very pr ice sensitive w i t h a h igh wi l l ingness- to-pay for pr ice informat ion i n the market". However , their f indings suggest that consumers general ly devote a higher effort to within-store pr ice search than across-stores pr ice search, i m p l y i n g that the increased sales v o l u m e i n the short-run is a result o f l o y a l customers w h o wai t and stock up w h e n a pr ice p romot ion is advertised rather than n e w customers w h o v i s i t the store w h e n they are exposed to the advertisement. Because there is a transportation cost incurred o n each shopping tr ip, consumers l ook more carefully for in-store specials once they already shopping at a store, instead o f reading through the flyers from i n d i v i d u a l retail supermarkets and searching for the lowest prices pr ior to their shopping trip. In another words , an increase i n store traffic is more l i k e l y to be a result o f customer retention rather than customer acquis i t ion from compet ing supermarkets. It is imposs ib le for retail supermarkets to advertise a l l the products w i t h pr ice p romot ion o n the sales flyer; consumers o n l y k n o w about the advertised prices, so there is s t i l l imperfect informat ion i n the market. The analysis o f Raj iv , Dut ta and Dha r (2002) examines h o w the advertisements o f pr ice p romot ion for stores i n different market posi t ions compete for store traffic w i t h the appl ica t ion o f a game theory. T h e y conclude that "promot iona l advert is ing is mot ivated by both traffic b u i l d i n g and customer retention considerations; and the relative importance o f these considerations is related to the store's service posi t ioning"(p93) . A s s u m i n g that the supermarkets have s imi l a r cost structure, the high-service stores w o u l d have a higher marg in ; therefore, they can afford to advertise at a higher frequency than those low-service stores w i t h a lower margin . The analysis indicates that high-service stores advertise a sha l low discount at a higher frequency and p lay a more prominent role i n offensive advert is ing, w h i l e the low-serv ice stores offer deeper cuts at a lower frequency for customer retention purposes. In addi t ion to examin ing the advert is ing strategies o f compet ing supermarkets, researchers have also studied the advert is ing strategies o f i nd iv idua l supermarkets. F o o d manufacturers pay retailers for promot iona l or merchand iz ing act ivi t ies , i nc lud ing Chapter 2. Literature Review 10 advertisement i n sales flyers, in-store p romot ion , and even the d isp lay pos i t ion o f the products, i m p l y i n g that the manufacturers influence the select ion o f the brands o n the sales flyer. A l oca l i zed study o f R e e d and Robb ins (1981) compares the informat ion o f advertised specials to the manufacturers ' advert is ing a l lowances and the b u y - i n pr ice o f the products from the corresponding manufacturers for retai l chains. T h e y f ind that "adver t i s ing a l lowance by manufacturers p lays an important role i n ind ica t ing what products should be adver t i sed"(pl9) . Howeve r , each retailer has their o w n degree o f influence through their spatial , product, and service differentiations. T h i s develops a specif ic market pos i t ion or so-cal led " f i r m image" i n terms o f storewide merchandis ing activit ies and p r i c i n g strategy i n both regular and temporary pr ice p romot ion . Te l l i s and Zuf ryden (1995) indicate that there are two significant characteristics o f retailers ' promot ions . Firs t , un l ike manufacturers, retai l supermarkets are not interested i n the effect o f pr ice p romot ion o n ind iv idua l product performance; they are more concerned w i t h h o w market ing and product m i x affect overa l l store performance i n category sales and profits. Somet imes the marg in o f a promoted product is higher than the marg in o f an unadvertised product; otherwise, advertisement that leads to brand swi tch ing by i t se l f is not profitable. Secondly, manufacturers affect retailers ' margins that fluctuate over a p lann ing cyc le based o n deals; thus, the length o f pr ice reduct ion and the adver t is ing frequency is direct ly related to the manufacturer 's deal. G u p t a (1988) examines the effect o f pr ice p romot ion b y focusing o n the increase i n sales v o l u m e dur ing the p romot ion per iod; the study indicates that brand swi tch ing causes the sales v o l u m e to rise dur ing the pr ice promot ion . B o t h Gup ta (1988) and Tel l i s and Zuf ryden (1995) conclude that the discounts have o n l y a short-run sales effect; the accelerated purchase appears at the t ime o f the discount per iod, but w h e n prices return to their regular l eve l , sales subsequently decl ine due to the s tockpi l ing effect. A l t h o u g h nat ional brand manufacturers exert ver t ica l influence o n the supermarket manager 's dec i s ion regarding the select ion o f product advertisement, manufacturers s t i l l face hor izonta l compet i t ion from private labels. Mess inge r and N a r a s i m h a n (1995), Ros toks (2002), and W a r d et a l . (2002) discuss the rapid invas ion o f private labels into food industries. T h e y show that the market share o f private labels is expanding s ignif icant ly relative to the growth o f nat ional brands. The evidence c o m p i l e d by W a r d et a l . suggests that an increase i n the share o f private- label goods is correlated w i t h a rise i n the pr ice o f name-brand goods. The i r conc lus ion is consistent w i t h the f indings o f Jafri et a l . (1993). A Chapter 2. Literature Review 11 private label is usual ly subcontracted to a manufacturer b y the supermarkets and thus the cost is re la t ively l o w compared w i t h the costs o f most nat ional brands. Therefore, the pr ice o f a private label product is at a consistently l o w price regardless o f its share i n the market. Because a nat ional brand devotes more resources to research and development for product improvement , and makes a greater effort i n product market ing, the nat ional-brand manufacturers offer fewer discounts and non-price promot ions . The nat ional brand also achieves a h i g h l eve l o f product differentiation as a react ion to the invas ion o f pr ivate labels into the market. Sex ton and L a v o i e (2001) noted that, un l ike most A s i a n countries, the ver t ica l relat ionship between food manufacturers and retailers is very close, w i t h few interventions by intermediaries. F o o d manufacturers trade di rect ly w i t h supermarket chains; the manufacturers and the retailers are the m a i n parties w h o j o i n t l y decide the retai l pr ice . Re ta i l pr ice varies w i t h i n a large range depending o n the profit marg in o f the retailers, and there is a temporal price dispers ion w i t h price va ry ing over t ime such as the w e e k l y pr ice p romot ion offered b y different supermarket chains. The theoretical m o d e l o f Var i an (1980) indicates that the retai l market is characterized by a larger degree o f pr ice d ispers ion and the price dispers ion can persist i n markets where at least some consumers behave i n a rat ional manner. The empi r i ca l study b y L a c h (2002) indicates "pr ice d ispers ion across stores is prevalent and differs across products i n reasonable w a y s " ( p l ) . B o t h Var i an and L a c h ' s f indings support our result o f independent advert is ing decis ions o n the part o f t w o compet ing retailers w h o randomize prices i n an attempt to reduce the difference i n shopping decisions between informed and un informed consumers. Las t ly , R a o and S y a m (2001), a study that is c lose ly related to m y research, constructed a m o d e l w i t h i n w h i c h compet ing f i rms chose to advertise one o f two goods i n each per iod . It is more profitable for compet ing f i rms to have pure opposite goods advert is ing strategies than pure same goods advert is ing strategies. Howeve r , there is a strong incentive for a f i r m to deviate from the pure equ i l i b r i um for opposite goods advert is ing and offer same goods advert is ing, w h i c h y ie lds a higher profit , w h i l e its competi tor is offering oppose-good advert is ing. In Chapter three, w e w i l l discuss R a o and Syam' s study i n more detai l and w i l l develop a s imp l i f i ed ve rs ion o f their m o d e l to show that compet ing f i rms choose to randomize their advert is ing strategies, advert is ing the same product some o f the t ime and different goods at the other t imes. Chapter 2. Literature Review 12 2.2 Textbook Perspective The Stmcture-Conduct-Performance (SCP) paradigm is an essential study in most Industrial Organization textbooks, and it provides a framework for investigating market behavior. The structure of an industry determines the conduct of firms, and the behavior of firms determines how well the industry performs. The relationships of Advertising with Market Concentration, Prices and Profits are discussed from an Industrial Organization perspective in assessing the behavior of firms in different market environments. There are two categories of advertising: pure informative and persuasive. The informative advertising provides consumer with information such as existence and location of sellers and products, and prices; while persuasive advertising attempts to increase consumer's preference for the firms' product. McGee (1998) and Hayes and Morris (1991) explain that information is important to the functioning of the market, but persuasive advertising is indeed less desirable in welfare terms. Informative advertisement serves to identify the existence of sellers and the brand names that they carry in their merchandise. Supermarket managers choose the sales flyer as their marketing communication of choice because it is efficient in delivering price information to large numbers of consumers at a relatively low cost comparing Wi th television and radio commercials. The flyer is then combined with other aspects of marketing programs, such as product, pricing, arid distribution strategies from the marketing mix of the firm. The profit margin of individual products is differs across the overall merchandise for sale in retail supermarkets. The profit margin on national brand items is less than the private label products because manufacturers of national brand products spend enormous resources on research and development to continuously improve the quality of products, service information, and promotional activities such as recipes and contests. Also non-price advertisement sends signals of high quality to consumers. On the other hand, all private labels are subcontracted by the manufacturer and produce copies of national name brand products at a much lower cost, seeking to gain a market share in the product category by targeting price sensitive consumers. Because the cost of private label products is low, they usually sell at a consistently low price and if there is a price promotion on private labels, the discount would not be as deep as for the national brands. Chapter 2. Literature Review 13 A l t h o u g h private label and nat ional brands compete for market share w i t h i n the supermarket, and accord ing to Gupta ' s f indings (1998) brand swi t ch ing has a significant effect o n profit as a result o f price p romot ion , brand swi tch ing does not influence the store performance i n general. I f bo th nat ional and private labels are advertised i n the same week, the advertisement w o u l d become inefficient for the nat ional brand manufacturers because this w o u l d reduce the effect o f pr ice p romot ion o n i nd iv idua l product performance. Hence , supermarket managers a v o i d advert is ing the nat ional and private label together i n the same week and a l l o w the consumer to focus o n the discounted nat ional brand i n the sales flyer. In the across-brand advertisement analysis for compet ing nat ional brands, the associat ion o f advert is ing frequency is expected to be h i g h i n most product categories. T h i s si tuation is different f rom the compet i t ion between national brands and private brands as the nat ional brands i n the same product category compete intensively for market share. There are exceptions, however, w h e n there is a name brand domina t ion i n the product category; i n this case the associat ion o f advert is ing frequency is expected to be low. 2.3 Industry Perspective A c c o r d i n g to an industry insider, D e b b y M c K i n n o n (personal communica t ion , N o v e m b e r 29 , 2002), a retail supermarket has a very large number o f items that are promoted each week i n var ious product categories; there are usual ly over 7,000 i tems i n the grocery section o f the supermarket o n sale o n average each week. A p p r o x i m a t e l y 1,000 items ( inc lud ing different f lavours and sizes) are advertised i n the f lyer weekly , but not a l l products advertised are o n p romot ion ~ some o f the products are advertised at regular prices. The select ion o f advertised products i n the f lyer is t yp i ca l ly p lanned three months i n advance, especial ly for products w i t h seasonality; for instance, ice-cream and barbeque sauce i n the summer, and turkey and f lour near Chr i s tmas t ime. Genera l ly , there are 10 to 15 loss leader i tems every week, and it is possible to have 3 to 4 o f the loss leader products advertised i n the flyer. There is some last minute adjustment for pr ice changes and corrections pr ior to the pr in t ing o f the advertisement; however , the t ime lag between the adjustment and the pr in t ing is u n k n o w n . M c K i n n o n indicates that it is usual ly the manufacturers w h o initiate the temporary pr ice p romot ion and w h o pay retailers to advertise their products i n the supermarket 's w e e k l y sales flyer. It is c o m m o n for manufacturers to exercise bargaining power and negotiate the terms o f the p romot ion , such as w h i c h product is to be advertised and Chapter 2. Literature Review 14 discounted, the level of discount, and the length of promotion. Manufacturers also have to pay for merchandising activities, such as advertisement in the sales flyer and shelving position in the store. Consumers choose at which store to shop and which brand to buy according to their personal preferences and the price information. Roughly 7-12% of customers only shop for the promotional products. Various supermarket card programs also offer special discounts to customers; they do not only build store loyalty, but also obtain valuable customer information for marketing purposes. Chapter 3 Market Environment of the Food Retail Industry This chapter provides some background information on the market environment and the market structure of the Canadian food retail industry. Section 4.1 describes the Canadian food retail market; section 4.1.1 provides an overall view of the Canadian food retail market environment based on information from a report by Drake (2001); and section 4.1.2 discusses the food retail market environment of the Vancouver metropolitan area studied in this research in terms of market concentration, social demography, and recent trends in the food retail industry. Section 4.2 describes the background and the market positions of the four supermarkets under study in this thesis. 3.1 Food Retail Market Structure 3.1.1 Overview of Canadian Food Retail Industry According to the Canadian grocery sales report by Drake (2001), the Canadian grocery business has been growing steadily across the country. Grocery sales in Canada increased 3.9% in 2000 and reached a total of $56.63 billion. The food retail sector is a competitive industry involving domestic food distributors and retail chains as well as American ones. Domestic food distributors and food retail chains are expanding in various regions in order to gain a greater market share in the industry, and the report by Woodcock (1999) indicates that U.S. based companies are also seeking opportunities to enter the growing market in Canada. However, the market structure varies across provinces because the grocery business is a localized industry; most promotional activities occur only at the selling point. This is obvious from the perspective of consumers; they are not willing to travel a long distance to purchase groceries unless they live in remote areas. Life styles and food trends may also vary across regions due to demographic and geographic differences. Drake (2001) also indicates that in the year 2,000, the food retail industry had a market concentration in a range of 61.7 - 67.1% in the Atlantic provinces, Ontario, and the Prairies; Alberta had the highest market concentration at 73.2%; and Quebec had the lowest market concentration at 32.9%. The market concentration is 65.5% in British Columbia. Although large corporations are continuously expanding across Canada, the trend is to move towards 15 Chapter 3. Market Environment of Food Retail Industry 16 smaller communi ty grocery retail stores, w h i c h sel l fresh food i tems as w e l l as del icacies , to appeal to the i m p r o v i n g l i fe style o f consumers w h o increas ingly demand good service and h i g h qual i ty food. 3.1.2 Food Retail Market in British Columbia The retail food sector i n B r i t i s h C o l u m b i a is a compet i t ive industry w i t h a h i g h market concentration dominated by a sma l l number o f large supermarket chains. In 2,000, w i t h a g r o w i n g popula t ion hav ing an increas ingly large disposable income, the food retail sector i n B r i t i s h C o l u m b i a had 14 .5% o f the market share i n Canada and an increase o f 3 .6% i n p rov inc i a l sales from 1999 accord ing to D r a k e (2001). In B r i t i s h C o l u m b i a , the market compet i t ion i n the retail food industry is intense w i t h var ious food retail formats and sizes. E x a m p l e s are: independent corner stores, convenience store chains, convent ional supermarkets, and c lub warehouses. T h e largest five food retail companies are Safeway, Overwai tea , L o b l a w , H Y L o u i e , and Thr i f ty ' s . Together they control 9 0 % o f the total food dis t r ibut ion (Woodcock , 1999). M o s t supermarket chains were expanding to larger stores to achieve eff ic iency o f scale and a greater select ion o f merchandise i n order to attract price sensitive consumers at a lower marg in and thereby achieve a higher sales vo lume . T h i s includes those t ime constrained consumers w h o want one-stop shopping w i t h selections other than the convent ional food and grocery products. T h i s expansion enables them to compete w i t h the invas ion o f warehouse c lubs and survive the intense compet i t ion o f r iva ls , thus main ta in ing their market share. In the 1980's, large firm size and m a x i m u m market share were the ph i losophy o f most prosperous f i rms; therefore, supermarket corporations invested huge amounts o f money i n expanding, either though acquis i t ion , b u i l d i n g new stores, or renovat ion o f ex is t ing stores. The socia l demographic is cont inuously changing i n the L o w e r M a i n l a n d as people emigrate from a l l over the w o r l d or relocate from other cit ies i n N o r t h A m e r i c a . The statistics from B r i t i s h C o l u m b i a M u n i c i p a l and R e g i o n a l Dis t r ic t Popu la t ion Est imates ( B C Statistics, 2001) indicate that there is an increasing d ivers i f ica t ion i n the cul tura l background o f the popula t ion. A t the same t ime, the economy is g rowing , so the consumer 's wi l l ingness to pay for h igh qual i ty food is increasing. F o o d retailers p l an their market ing strategies, intended to m a x i m i z e market share and profits, i n accordance w i t h this changing market environment . Chapter 3. Market Environment of Food Retail Industry 17 3.2 History and Market Position of Supermarket Chains in this Study 3.2.1 IGA and Market Place IGA In B r i t i s h C o l u m b i a there are eight corporate offices and 48 franchised supermarkets contro l led b y the H . Y . L o u i e C o m p a n y L i m i t e d , under the I G A banner. T h i s is a Vancouver-based company w i t h headquarters i n the C i t y o f Burnaby . A c c o r d i n g to Lazarus (2001), the H Y L o u i e Company , the wholesaler for I G A i n B r i t i s h C o l u m b i a , faced a c r i t i ca l s i tuation i n the 1980's w h e n large format competi tors (i.e. The R e a l Canad ian Superstore) entered the market. W h i l e most supermarkets were expanding i n 1980's and the early 90s, hop ing to ga in a greater share o f the market, L o u i e decided instead to reposi t ion. In 1999 the company asked I G A retailers to reinvest i n ex is t ing stores by renovat ing arid reducing the stores' size to the 25-30,000 square foot range. N o w most I G A stores i n the Greater Vancouver area have been transformed into the new market place format - M a r k e t P lace I G A . The n e w M a r k e t Place I G A introduced the " food bout ique" concept, a more inv i t i ng and homey environment", to differentiate themselves f rom their competi tors (Lazarus , 2001) . Lazarus indicates that the n e w M a r k e t P lace I G A has shifted its pos i t ion i n the market and reinforced its n e w image as a ne ighbourhood supermarket w i t h strong customer relationships and strong presence i n the communi ty" . The n e w I G A stores make the store layout different f rom other supermarkets w i t h brighter l igh t ing , and ha rdwood f loor ing under the displays i n the produce section. It also provides demonstrat ion ki tchens i n some o f the new stores w h i c h feature the I G A house labels Our Compliments and Smart Choice. T h e y launched a " c o m m u n i t y ca rd" ini t ia t ive i n September 2000 , w h i c h does not track i nd iv idua l purchases but raises money to buy computers for l o c a l schools . M a r k e t P lace I G A focuses o n the l o w - v o l u m e but higher marg in customer w i t h a more expensive l ifestyle. The transformation was a success w i t h an average 17% increase i n customers and a 2 2 % increase i n sales. 3.2.2 Canada Safeway Canada Safeway is a d i v i s i o n o f Safeway U . S . . There are 215 Canada Safeway stores located across western Canada, a l l o f them corporately o w n e d and operated w i t h headquarters i n Calgary , A lbe r t a . M e r c h a n d i z i n g and procurement functions are central ly control led i n Calgary , w i t h key warehouses and dis t r ibut ion points located i n Vancouver , Chapter 3. Market Environment of Food Retail Industry 18 Calgary , E d m o n t o n and W i n n i p e g . Canada Safeway dominated the food retai l industry i n western Canada before the entrance o f L o b l a w ' s R e a l Canad ian Superstore f rom eastern Canada, and before the expans ion o f the J i m Patterson G r o u p ' s Save -On-Foods i n the west. Howeve r , the company s t i l l has an advantage i n the market, w i t h its strength l y i n g i n the large number o f stores and their strategic locat ions i n in tens ively populated urban areas (Marke t ing M a g a z i n e p g 1 0 , 1 1 , 1 9 9 6 ) . Safeway stores are 30-35,000 square feet i n size, and l a i d out i n a cohesive format offering fu l l services w i t h a bakery, meat counter, de l i , fresh produce, a f lora l sect ion, houseware sections, and a pharmacy. O p e n and spacious aisles are w e l l s tocked and w e l l organized w i t h some cut-case display. F a m i l y s ize products are also avai lable . Safeway maintains its share i n the western food retail industry b y hav ing a strong merchandis ing program - Safeway C l u b C a r d - w h i c h tracks i n d i v i d u a l purchases. It has also shifted to the "neighbourhood store" concept w i t h its focus o n value added customer service. Because o f the payments from manufacturers for p l ac ing advertisements i n their sales flyer, Safeway is p l a y i n g a leading role v i s a v i s its competi tors i n flyer advert is ing. A c c o r d i n g to D e b b y M c K i n n o n (personal communica t ion , N o v e m b e r 29 , 2002) , Safeway advertises over 1,000 i tems i n its flyer each week, i n c l u d i n g sizes and flavours. Hence , there is h i g h advert is ing frequency i n most o f the product categories, and Safeway 's advert is ing strategy is used as a parameter i n the cross-store analyses. 3.2.3 Save-On Foods Save-On-Foods is one o f the k e y retail banners operating w i t h i n the Overwa i t ea F o o d G r o u p , as a d i v i s i o n o f the J i m Pat t ison G r o u p . The Overwai t ea G r o u p has 60 Save-On-Foods locat ions throughout B r i t i s h C o l u m b i a and A l b e r t a , w i t h the head office located i n Langley , B r i t i s h C o l u m b i a . The Save-On-Foods banner has been the g rowth banner w i t h i n the Overwa i t ea F o o d G r o u p , w i t h f ive new locations i n 1999. A l l these stores have adopted the new trend to smal ler "ne ighbourhood markets" w i t h sizes ranging from 25,000 to 30,000 square feet, instead o f the tradit ional large format o f 70 to 80,000 square feet (Woodcock , 1999). A l l B C stores have long hours, from 8 a.m. to 12 midnight , 7 days a week, convenient for customers w i t h l im i t ed shopping t ime. M o s t o f the stores provide fu l l services w i t h a bakery, meat counter, seafood, a floral boutique, photo f in ish ing , cosmet ic department, and pharmacy. T h e y try to Chapter 3. Market Environment of Food Retail Industry 19 appeal to people w h o have a more envirormiental-fr iendly and healthier l i fe style. F o r example , they introduced a n e w brand ca l led Br igh t l i f e Na tu ra l Foods , and set up a r ecyc l ing centre ca l l ed Changes. T h e y also try to appeal to price-sensi t ive consumers, w i t h a l o w cost, h i g h v o l u m e bu lk food sect ion - an example o f l o w marg in market ing . 3.2.4 Real Canadian Superstore The R e a l Canad ian Superstore is an independent store under L o b l a w ' s C o m p a n y L i m i t e d , the most important subsidiary o f George Wes ton L i m i t e d , founded i n 1882 as a food processing company spec ia l i z ing i n baked goods. L o b l a w ' s , based i n eastern Canada , is the largest food distr ibutor i n Canada. A c c o r d i n g to the report b y W o o d c o c k (1999), L o b l a w ' s has annual sales o f $2,596 m i l l i o n . It has expanded into the west and n o w has a lmost 150 supermarkets under three different banners i n western Canada: R e a l Canad ian Superstore ( R C S S ) , Super V a l u and E x t r a Foods . The three different banners, w i t h var ious store sizes and product selections, are designed to target different market segments. The majori ty o f its business, approximate ly 7 0 % , is generated through the R e a l Canad ian Superstore, w i t h a regional head office i n Calgary , A lbe r t a . A l l R C S S outlets are corporately owned . The market ing pos i t ion for the R e a l Canad ian Superstore is pr ice-oriented and offers an " E v e r y d a y L o w P r i c e " to its customers; services and informat ion are a l ower pr ior i ty and therefore not as complete as its r ivals . The concept o f Superstore is to be b i g i n size, inventory, and selection, i n order to achieve eff ic iency i n scale w h i c h makes l o w prices feasible at m i n i m u m costs. R C S S ' s store format, select ion, and she lv ing are s imi la r to the warehouse style w i t h displays that are huge and plent i ful , thereby enabl ing it to compete effectively w i t h c lub warehouses l i ke Cos tco . The R C S S stores are t yp i ca l l y 65-70,000 square feet i n size and feature large grocery, dairy, meat, and bakery departments, and var ious special ty departments such as photo f in i sh ing and electronics, c lo th ing , automotive, hardware sections, and even furniture. In-store merchandis ing is systematic and consistent w i t h the company ' s strategy o f support ing their o w n private label brands, President 's C h o i c e and N o N a m e . The spacious aisles are we l l - s tocked w i t h m a n y cut-case d isplays , where private label brands dominate the shelves. W i t h their rock bot tom prices, the R C S S is able to attract consumers f rom a distance, approximate ly w i t h i n a 1 0 k m radius (Marke t i ng M a g a z i n e p l O - 1 6 , 1996). Because the store locat ions are large i n size, most o f the R C S S stores are located i n industr ial areas or major commerc ia l areas, and the number o f stores is re la t ively smal ler Chapter 3. Market Environment of Food Retail Industry 20 than its competitors. T h i s helps to main ta in a l o w operating cost. The R C S S has other ways to l i m i t its expenditures such as l o w advert is ing costs: since the R C S S exercises the " E v e r y d a y L o w P r i c e " opt ion, it o n l y offers l i m i t e d selections for pr ice p romot ion i n w e e k l y advert is ing flyers. Information and services avai lable f rom the R C S S are l i m i t e d ; the company provides l i t t le pr ice informat ion and few value added services. Chapter 4 Theoretical Model Similar to other profit-driven firms, food retailers would like to maximize their overall store performance by increasing store profit, sales volume, and market share. Each year supermarket chains spend millions of dollars on price advertising and temporary price discounts for food products. By offering discounts and providing price information supermarkets attempt to increase store traffic and thus attract potential customers as well as retain existing customers. Consumers are unlikely to buy the discounted items if those items will be their only purchase at a particular store, due to the transportation cost incurred for each shopping trip. Therefore, supermarkets aim for strategies other than just discounts; these strategies are intended to maximize store traffic and thus increase sales volume, store profit, and the store's market share in the industry. The subsequent research studies the price advertising pattern of major supermarket chains and investigates the price advertising decisions of the food retailers in two categories: (1) same goods advertisement in competing stores in a given period. (2) competing goods within a store in a given period. The model developed in this section is a simple version of the one developed by Rao and Syam (2001). Before explaining the specifics of this model, it is useful to first discuss the general results. Firms can pursue one of three strategies when selecting which products to advertise: (1) promote the product at the same time as their competition; (2) promote the product at a time different from that of their competition; (3) make a random choice regarding which product will be promoted and when. This choice is the same for two competing supermarkets choosing when to advertise a particular product, or competing food manufacturers choosing when to advertise their respective products within one retail outlet. To simplify the discussion, the model is constructed from the perspective of two competing supermarkets, although the general result obtained does seem to apply to both situations. Rao and Syam (2001) demonstrate that the equilibrium outcome of their price promotion game is the third strategy: randomize which product will be promoted and when. They argue that randomized advertising reduces competition and thus maximizes profit for 21 Chapter 4. Theoretical Model 22 food retailers. Because consumers make their store choice based on advertised prices, if the retailers know in advance the goods that will be advertised by a competitor, there is a strong incentive for retailers to undercut each other's price in seeking a greater market share. Hence, competition becomes intense and yields a lower profit for the retailers. With randomized advertising, retailers advertise the same goods sometimes and different goods other times; thus, head-to-head competition is reduced and profit is greater for retailers advertising the same goods, at the same time. Rao and Syam (2001) extend and alternate the work of Lai and Matutues (1994), which studies the pricing strategy of supermarkets and the effect of price advertising on increasing store traffic from the perspective of loss-leader pricing. The main difference is that Lai and Matutes focus on retail pricing and advertising in a multistage game framework, treating the decisions of both stores and consumers as endogenous. They find that loss-leader pricing can be the equilibrium outcome in a multiproduct situation. There is currently no price data available for unadvertised promoted products, so a simple model is developed and utilized to show why randomization is arranged as an equilibrium. The variable p in the general model of Rao and Syam is very important as it is chosen endogenously in the third stage of the game; however, p does not play a useful role in the modified model of this thesis. 4.1 Model Assumptions There are two identical retailers, A and B, located at the two end points of a straight line with unit length: A on the left side and B on the right. Each store carries two identical goods, 1 and 2. These two goods are neither complements nor substitutes, the marginal cost of these procuring two goods are constant and identical for both stores. The advertising cost is zero. Consumers whom are uniformly distributed along the line, have identical preferences. Each consumer will purchase one of each of both goods provided that the price of the good is less than the consumer's reservation price, R. A consumer located a distance m, 0 < m < 1, from Store A is located at a distance (1 - m) from Store B. The unit transportation cost for each consumer is c/2; thus, the consumer in question would incur a cost of cm for a round-trip visit to Store A, and a cost of c(l - m) for a round-trip visit to Store B. Chapter 4. Theoretical Model 23 4.2 Consumer's Decision Consumers are assumed to have full knowledge of both prices, advertised and unadvertised, at both stores. Consumers can choose to visit no store, visit only one store (A or B), or visit both stores, depending on product prices relative to R, and transportation costs. The price of the advertised goods is R - DT where DT is the advertised discount of price i, i=A, B. Let the price of the nonpromoted goods be R - u. In the model, u is specified exogenously rather than chosen by the retailers; moreover, u is assumed to be equal for both retailers. 4.3 Stores'Decision Retailers can advertise the same good, Store A Store B Product 1 R-DA R - D B Product 2 R-u R-u or opposite goods, Store A . Store B Product 1 R-DA R-u Product 2 R-u R-DB In the first table, product 1 is jointly advertised, and in the second table A advertises Good 1 and Store B advertises Good 2. There is no need to consider the complementary case because the results are symmetric. 4.4 Pricing and Advertising Equilibrium of the model Rao & Syam present the price communication model as a three-stage game. In this simple three-stage game, retailers make the advertising decision in the first stage by selecting which goods to advertise and how much to discount. The price is revealed in the second stage, and consumers make their store choice based on the advertised prices. Following Rao & Syam, the subgame perfect equilibrium to this multistage game is derived for the cases of same-goods advertising, opposite-goods advertising, and mixed strategy. Same-good Case Suppose both stores devise a pure strategy to advertise the same goods with probability one in the first stage. In this case, each consumer will purchase both goods from just one of the Chapter 4. Theoretical Model 24 retailers. Let mo be the location of the marginal consumer who has no preference for either Store A or Store B. The equation that implicitly defines mo can be written as (4.1) cm0+R-DA+R-u = c(l-m0) + 2R-DB-u c + D. — DR This equation can be solved to obtain m0 = The profit function of Store A is 2c (4.2) maxx = (2R-DA u c + DA-DB^ J DA ' " \ 2c The first-order condition for maximizing profits is dn (4.3) 'c + DA-DB dDA V 2c + 2R-DA -u 2c Solving the first-order condition given by equation (4.3) for the choice variable, DA , results in 2R-u-c-D„ (4.4) DA = By symmetry, DA = DB = D * . Setting DA = DB = D* in equation (4.4) and solving implies that (4.5) D* =2R-u-c This equation can be used to show that mo* = 0.5. Now solve for the profit for Store A by substituting the expression in equation (4.5) into equation (4.2), and then reducing c 2 (4.6) *A=-Again by symmetry, n*A =n*B = — . The discount for the advertised good is the same at both stores and both stores share the market equally and earn the same level of profits. Opposite-good Case Now suppose that both stores devise a pure strategy to advertise opposite goods with probability one, with Store A advertising Goods 1 and Store B advertising Goods 2. Chapter 4. Theoretical Model 25 Consumers near the end points will generally prefer to purchase both goods at either Store A or Store B, whereas consumers near the middle will generally find it optimal to travel to both stores and purchase the advertised goods at each location. Let mi, 0 < mx < 1 , be the location of the marginal consumer who has no preference for buying either Goods 1 or Goods 2 at Store A or buying the discounted good at either store. The implicit equation for m\ is shown by (4.7) cml + 2R - D A -u = cml + R - DA + c(l -mx)+ R - DB Solving this expression gives * x c + u-D„ (4.8) mx = B-Let rri2, 1 2 2 , be the location of the marginal consumer who has no preference for either Goods 1 or Goods 2 at Store B or for buying the discounted good at either store. The associated equation is (4.9) c(l-m2)+2R-DB -u = cm2 + R-DB + c(\-m2)+R-DA and the solution to this equation is (4.10) m 2 = ^ ^ -Now substitute m\ and mi into Store A's profit function (4.11) m a x 7 r = (R-DA DA-u + {R-u) c + u-D, The appropriate first-order equation is (4,2) f ^ . i ^ W l W ) dDA V 2 J \2) Solving the first-order condition given by equation (4.12) gives Chapter 4. Theoretical Model 26 (4.13) D'A = *±H-By Symmetry, DA = D*B =D*. Now substitute the expression forD*A and D'B into the expression for mx and m2 given by equation (4.8) and (4.10) to obtain 2c (4.14) m\i=l- and = 2 1 2c Substitute the expression for£>*, m\ and m2 given by equation (4.13) and (4.14) into profit function given by equation (4.11) to obtain (4.15) 7rA=7rB=R-u-y 4c As before, symmetry ensures %\* — KB* Comparison of Pure Strategy Outcomes Now we would like to compare the firms' profit from Case 1 to Case 2 to see which advertising strategy is preferred. Keep in mind that from equation (4.6) and (4.15) nl =- and n1 =R-u-s ' 2 4c In Appendix A it is established that if the parameters, R, u, and c, satisfy the feasible restriction R >u> V8c , then n2 > nx is true and firms collectively prefer opposite goods advertising rather than same goods advertising. Advertising opposite goods with this set of parameter restrictions results in a relatively less elastic implicit demand curve facing each firm and thus each firm is able to exert relatively more market power. Non-existence of Pure Strategy Equilibrium Even though the two firms jointly prefer to advertise opposite goods rather than the same goods, it is important to check to ensure that the opposite-goods outcome is indeed an equilibrium. It is shown below that when firm B sets price according to the opposite-goods R -I- ix advertising strategy,^ =—-— (see equation 4.13); firm A can earn higher profits by advertising the same goods as B and charging the appropriate optimal price. This incentive to deviate implies that a pure strategy equilibrium does not exist. Chapter 4. Theoretical Model 27 Now let's formally establish these results. If Store B continues to price at D*B = —-—, the profit function for A, given that A deviates by advertising the same goods as B, can be written as (4.16) nA = {2R - D A - ^ j-c The price discount that maximizes this profit function is given by (4.17) D-TJ^y If this expression is substituted back into equation (4.16), an expression for optimized profit emerges. ( 4 , 8 ) + 16c In order to show that Store A has an incentive to deviate, the parameters R, u, and c, need to fit the restriction R - u < ^ (see Appendix B for the proof). If this restriction holds, then JT — 7T "> 0 deviate no-deviate ' Mixed Strategy Advertising Equilibrium It has now been established that firms collectively prefer opposite-goods advertising to same goods advertising. Moreover, a pure strategy equilibrium for opposite-good advertising does not exist since there is a strong incentive for each retailer to deviate from the pure strategy. These two results imply that the equilibrium must involve a mixed strategy. A mixed strategy for Store A consists of advertising Goods 1 with probability P\> 0, and Goods 2 with probability of 1 - PA- Similarly, a mixed strategy, for Store B consists of advertising Goods 1 with probability PB> 0, and Goods 2 with probability 1 - P B . The implication of this result is that, from a data generating perspective, stores are randomly choosing when and which product to advertise. The mixed strategy equilibrium includes both outcomes of competing stores advertising opposite goods or same goods, and it is consistent with our hypothesis that the advertising strategies of competing firms are independent. Chapter 4. Theoretical Model 28 Rao and Syam find that supermarkets use unadvertised specials to retain consumers who patronize the store because they are familiar with the everyday unadvertised prices of the store's merchandise. In-store discount can reduce the probability for consumers to shop around and purchase only advertised items from each supermarket. Supermarkets also use unadvertised specials to attract new customers and to increase store traffic. Consumers are likely to buy other unadvertised items as well as the advertised product because they have already paid the transportation cost for the shopping trip, and their savings may decrease if they make a shopping trip to another supermarket. From this model, we know that all consumers purchase both goods at the same store when competing retailers advertise the same goods, given that consumers have no price knowledge of the unadvertised products. However, when competing retailers advertise different goods, some consumers shop around, and some purchase both goods at one store depending on the distance they have to travel and their transportation costs. The model yields a positive expected profit for Store A and Store B, keeping in mind that transportation costs play an important role in drawing store traffic. Therefore, supermarkets implement a mixed strategy in advertising, and randomize the price advertising for their selection. Supermarkets would choose to advertise opposite goods most of the time to avoid intense competition and to capture higher profit, and have same-goods advertising occasionally. Chapter 5 Data and Methodology The data set is described in Section 5.1; Section 5.2 describes the theoretical consideration and the procedures for testing the hypotheses concerning the advertising strategies of two competing supermarkets. 5.1 Data Description The data set for this research was obtained by collecting advertising sales flyers from four major retail supermarket chains in the Vancouver metropolitan area during the period from May 6, 2001 to May 5, 2002. The data therefore gives us 52 weeks of price promotion information. The sales flyer advertising frequency of different brands in 22 processed food and beverage categories was recorded for four supermarket chains. The data only records whether a product is advertised or not advertised in the sales flyer in a given week, not the level of price discount for the product. If a brand is advertised in the sales flyer, regardless of size and flavour, it is counted as "advertised" for the brand. Advertisement frequency is recorded in 2 X 2 contingency tables in the Appendix. From the collection of sales flyers, we see that discounts are often quite similar both across stores and across time. The 22 product categories and the corresponding name brands are listed in Table 5.1 and an explanation of how these products and name brands were selected is provided below. 29 Chapter 5. Data and Methodology 30 Table 5.1. Selected Products Categories and Name Brands Product Category Specified Brand Bacon Olympics Breakfast Cereal Kellogg's Canned Fish Clover Leaf Canned Soup Campbell Dry Pasta Catelli Frozen Pizza Kraft Frozen Punch McCain Frozen Vegetables Green Giant Fruit Jam Kraft Ground Coffee Maxwell Instant Coffee Maxwell Juice from Concentrate Sun Rype Juice Not from Concentrate Tropicana Ketchup Heinz Margarine Canola Harvest Mayonnaise Kraft Pasta Sauce Prego Peanut Butter Kraft Potato Chip Lay's Processed Cheese Slices Kraft Tea Bag Tetley Waffle Kellogg's 5.1.1 Selection of Retail Supermarket Chains Most food retailers place their sales flyers in the free community paper, including independent food retailers with one location, small local supermarket franchises, and national supermarket chains. However, not all food retailers can afford to have weekly advertisement throughout the year with plentiful coverage of their merchandise. Four major supermarket chains with numerous store locations in the Vancouver metropolitan area were selected in order to obtain consecutive observations within the 52 week sample period. Chapter 5. Data and Methodology 31 5.1.2 Choice of Food and Beverage Categories There are usually more than 10,000 items in a medium size supermarket. For the purpose of control, twenty-two processed food and beverage categories were selected for this research given their non-perishable nature and standard quality. Fresh produce such as fruits and vegetables were avoided because of their seasonal nature and variation in quality: data collected in this study applies only to products which do not vary in nature over a 52 week period and are advertised in a more or less consistent manner. To minimize the seasonality effect, all of the product categories selected are processed food and beverage. The data set shows that most of the products have been consistently advertised throughout the year, however, there is an exception for Frozen Vegetable. Frozen Vegetable tends to be advertised more often during the winter months when the supply of Fresh Vegetable is not as sufficient as in the summer months. 5.1.3 Specification of Name Brands The market competition for each food and beverage category is different according to the nature of the products, yet supermarkets carry numerous brands and labels for each grocery product in different packaging and sizes. It is typical for the large supermarket chains to have over 500 items on the sales flyer each week, but the advertisement frequency of most individual items is very low within the 52 week period. The different flavours, packaging, and sizes of the branded product are disregarded for simplicity. Mature name brands are selected for each category to prevent the "interest saturation" which sometimes occurs in the introductory period of a new product and because they show up relatively frequently in the flyer. It is interesting to see that Breakfast Cereals and Canned Soup have the highest advertising frequency among the 22 product categories; and Kellogg's, Campbell's and Kraft's are the three food manufacturers who have the highest advertising frequency. All of them are mature established-product manufacturers. Our data shows similarity to LeBlanc's finding in 1998 that a mature established-product industry often use "cheap" information advertising and direct mainly to local markets. Although informative price advertising often neither creates new nor expand existing markets, it is an important feature of competition in many established markets because firms are aiming to capture the largest possible share of a fixed market. Chapter 5. Data and Methodology 32 5.2 Methodology As explained above, the empirical component of this research investigates the statistical correlation of the advertising strategies of competing retail supermarkets and competing brands within a supermarket. The yes-no advertising status for each chosen product was collected during the sample period and analyzed with a simple test of independence. Specifically a chi-square test for 2 X 2 contingency table was used to test the various hypotheses. The between-store analysis compares the yes-no advertising status of specified national brands within various product categories between a pair of competing supermarkets. The study then turns to within-store analysis by comparing the advertising status of two competing brands within individual supermarkets. Section 5.2.1 presents the definition of the contingency table and explains the theoretical consideration of employing the chi-square test. Section 5.2.2 describes the techniques and test procedure as explained by Samuels (1989). Then the hypotheses testing for the between-store and within-store analyses are described in sections 5.3. 5.2.1 Theoretical Consideration The theoretical analysis suggests that the joint decision of food manufacturers and retailers is to randomly choose when and what products to promote in the weekly sales flyer. This suggests that the data should reveal statistical independence. The advertising frequency of two competing supermarkets is constructed in a two-way contingency table. Rosner (2000) defines a 2 X 2 contingency table as a table composed of two rows cross-classified by two columns. It is appropriate to display data that can be classified by two different variables, each of which has only two possible outcomes. One variable is arbitrarily assigned to the rows and the other to the columns". The association between the advertising strategies of store A and Store B (or product A and product B) is studied by testing the hypothesis of independence using the chi-square test. The chi-square test used in this study closely follows the test procedure as described in Samuels (1989). The chi-square test is a method of analyzing categorical variables rather than quantitative variables. A variable is categorical when each observation can be classified into one of two or more categories rather than taking on a numerical value. A categorical variable is said to be dichotomous if there are only two possible categories. The advertising status for each product category is a dichotomous variable in nature, as a product is either advertised or not advertised. Therefore, the chi-square test is employed to test the independence of advertising strategies within and between competing supermarkets. Chapter 5. Data and Methodology 33 5.2.2 The chi-square Test for the 2X2 Contingency Table The observed categorical data is summarized into a two-way contingency table as shown in Table 5.2, which shows a general 2 X 2 contingency table used in our between-store analysis. There is a single sample of advertising flyer data with size n = 52 weeks observed with respect to two dichotomous variables - advertising status (brand is advertised or not) and store (i.e. A or B). A 2 X 2 contingency table, as described by the name, consists of two rows and two columns. Each category in the contingency table is called a cell; hence, there are four cells in a 2 X 2 contingency table. The observed advertising frequencies are recorded in the four cells. There is an additional column and row to show the marginal frequencies which are the sum of the observed advertising frequencies across each row or each column, and the grand total of all the cell frequencies which should equal to the same size, n. Let i denote the advertising status for Store A, where /=1 implies the brand is advertised in the sales flyer of Store A and i=2 implies the brand is not advertised on the flyer of Store A. Similarly, let j denote the advertising status for Store B, where j=\ represents the brand advertised in the sales flyer of Store B and j=2 implies the brand is not advertised at Store B. Each ny represents the observed count of positive advertising outcomes within the 52-week period, so we haveO<«, y <52 = 1,2) subject to the constraint n = «n + nn + « 2 i + «22 = 52. For example, i f Kraft peanut butter is observed to be advertised at both Store A and Store B 14 times during the sample period, then «n = 14. Let xt (z=l,2) denote the marginal total frequency, the sum of advertising frequency observed at Store A, and let yt (/=1,2) denote the marginal total frequency the sum of advertising frequency at Store B. Therefore, we have the following notations: rtn = the observed frequency that the brand is advertised at both Store A and B in the same week «n = the observed frequency that the brand is not advertised at Store A but advertised at Store B «2i = the observed frequency that the brand is advertised at Store A but not advertised at Store B H22 = the observed frequency that the brand is not advertised at both Store A and Store B p = the expected frequency that the brand is advertised at both Store A and B in the same week p = the expected frequency that the brand is not advertised at Store A but advertised at Store B ft = the expected frequency that the brand is advertised at Store A but not advertised at Store B ^ = the expected frequency that the brand is not advertised at both Store A and Store B Chapter 5. Data and Methodology 34 = the marginal frequency that the brand is advertised at Store A x2 = the marginal frequency that the brand is not advertised at Store A yi = the marginal frequency that the brand is advertised at Store B y2 - the marginal frequency that the brand is not advertised at Store B n = the sample size o f the observation Table 5.2 2X2 Contingency Table for a product in Cross-Store Analysis Store A Advertised Not Advertised Total Store B Advertised Not Advertised A A "21 ( / * 2 1 ) « 1 2 ( / / 1 2 ) A " 2 2 0 2 2 ) n\\+n\2=y\ N2i+n22=y2 Total « 1 1 + " 2 1 = * 1 «12+"22 = *2 n The chi-square statistic measures discrepancy between the observed frequencies and the expected frequencies. It is calculated to test the independence of weekly advertisement between two competing supermarkets. Independence means that the probability of the product being advertised is the same for both stores; in other words, there is no direct association with the other store. For the case of competing supermarkets, the following hypotheses are to be tested: Ho: Advertising strategies for the chosen brand between two competing supermarkets are independent HA : Advertising strategies for the chosen brand between two competing supermarkets are dependent The null, represented by Ho, states that the advertising strategies between two competing supermarkets for a specified brand are independent; and the alternative, represented by HA, states that the advertising strategies between two competing supermarkets for a specified brand are dependent. Similar hypotheses can be written for the case of two food manufacturers promoting their products within a store. Chapter 5. Data and Methodology 35 (5.1) The chi-square test statistic for testing Ho is calculated as follows: A where the sum is taken over all four cells in the 2 X 2 contingency table. When the null hypothesis is true, i.e., the advertising strategies for a specified brand between two competing supermarkets are independent, the test statistic^2 follows an approximate chi-square distribution with one degree of freedom. The degree of freedom is calculated as the product of the number of rows in the table less one and the number of columns in the table less one. When the null hypothesis is true, the expected frequencies in contingency table can be calculated using the marginal total frequencies as follows: (5.2) My n The value of p.tj in equation (5.2) is the expected value of «y when the null hypothesis is true. A 95% confidence interval for the proportion of the time when a product is advertised in both stores can be constructed as follows: A V (5.3) f A Pi-Pi C.I.= rp>n i -n P - P j For x1 •> df = 1 at the 5% significant level; if x1 > 3-84, Ho is rejected, where 3.84 is the critical value such that there is a 5% probability that a test statistic exceeding the critical value is observed when the null is true. Chapter 5. Data and Methodology 36 There are two conditions that must hold true for the chi-square test to be valid: 1. It must be reasonable to regard the data as a random sample of categorical observations from a large population. Observations must be independent of each other over time. 2. The sample size must be large enough. Both the confidence interval and test are approximate, and the approximation is best for large samples. It is important to note that the chi-square test can only test the association between the joint advertising strategies of any two competing retailers and the corresponding food manufacturers; it does riot provide information on the existence of a positive or negative correlation i f the advertising strategies are indeed dependent. Yet this information can still be obtained from the contingency tables (see Appendix B). If the frequencies nu and n22 are relatively high then there is a positive correlation between the advertising strategies of Store A and Store B (product A or product B). On the other hand, i f the frequencies n\2 and n 2 1 are high then there is a negative correlation between the advertising strategies of Store A and Store B (product^ or product B). 5.3 Hypotheses Tests 5.3.1 Static Analysis The chi-square test described in section 5.2 is applied to the case of two competing retailers promoting the chosen brand in three different pairs of supermarkets: (1) Safeway and Market Place IGA, (2) Safeway and Save-on Foods & Drugs, and (3) Safeway and The Real Canadian Superstore. In the between-store analysis, the advertising frequency of Safeway Canada is used as a benchmark in pairing with other supermarket chains. Safeway was chosen because it is the advertising leader in the number of products advertised each week, and in the advertising frequency of the 22 product categories selected. Similarly, the chi-square test is applied to the case of two competing food manufacturers promoting their products within a store in three scenarios: (1) the chosen brand and the private label at each supermarket, (2) any national brand versus the private brand at each supermarket, and (3) the chosen brand versus any other national brand. A problem occurs when marginal total frequency takes on the value zero, i.e. when the product was not advertised during the 52 weeks sample period. Referring to equation (5.2), Chapter 5. Data and Methodology 37 when one of the marginal total frequencies (x, or yi) is zero, then the expected frequency of the corresponding cell becomes zero. When one of the four expected frequencies in the 2 X 2 contingency table is zero, then the chi-square test statistic becomes undefined according to equation (5.1). A n undefined test statistic does not provide information for our hypotheses. In order to eliminate this problem, a constant number, 0.5, is added to all observations in the problematic contingency table. This would increase each marginal total frequency, x, and yh by 1 with the n added up to 54 instead of 52. This procedure only applies to the contingency tables with x, = 0 or y( = 0. The chi-square test statistics are not sensitive to this adjustment because the calculation of the test statistic, in equation (5.1), only considers the conditional probability of advertisement rather than the quantitative frequencies. It is important to note that the chi-square test can only test the association between the advertising strategies of any two competing food manufacturers and retailers; it does not provide information on the existence of a positive or negative correlation i f the advertising strategies are indeed dependent. Yet this information can still be obtained from the contingency table. If the frequencies n\\ and «22 are relatively high then there is a positive correlation between the advertising strategies of Store A and Store B (product A or product B). On the other hand, i f the frequencies nn and «2i are high then there is a negative correlation between the advertising strategies of Store A and Store B (product A or product B). 5.3.2 Dynamic Analysis The hypotheses tests in Section 5.3.1 can only analyze the static relationship among the advertising strategies between two competing supermarkets in the same week. In order to capture the dynamic aspects of the advertising environment over time, another set of chi-square testing procedure, as described in 5.2.2, is constructed for the time-lag analysis between pairs of competing supermarkets and with different brands combination in the 52 weeks sample period. The yes-no advertising status of a product in Store A in week t is compared with the yes-no advertising status of the same product in Store B in week t-k, where k={ 1,2,3,4}. The dynamic analysis tests the potential advertising leader and follower relationship among the supermarkets in a 4 weeks cycle. For instance, Store A is assumed to be a leader in advertising Brand A in week 4 and it is tested against the advertising status of Brand B at Chapter 5. Data and Methodology 38 Store B , as a fol lower , i n the weeks 1,2, 3, and week 4. E a c h supermarket is assumed to be a pr ice leader; then it is pa i red w i t h each o f the three compet ing supermarkets as a pr ice fo l lower and tested w i t h four different pairs o f brands combina t ion : 1) Speci f ied B r a n d at L e a d i n g Supermarket and Spec i f i ed B r a n d at F o l l o w i n g Supermarket 2) Private L a b e l at L e a d i n g Supermarket and Pr ivate L a b e l at F o l l o w i n g Supermarket 3) P r i ce Leader ' s Spec i f ied B r a n d and F o l l o w e r ' s Pr ivate L a b e l 4) P r i ce Leader ' s Private L a b e l and F o l l o w e r ' s Spec i f ied B r a n d Therefore, w e have a different set o f notations for the cont ingency tables as s h o w n i n section 5.2.2, for example: «n = the observed frequency that the brand is advertised at Store A in week t and at B in week t-k K12 = the observed frequency that the brand is not advertised at Store A in week t but advertised at Store B in week t-k «2i = the observed frequency that the brand is advertised at Store A in week t but not advertised at Store B in week t-k «22 = the observed frequency that the brand is not advertised at Store A in week t and Store B in week t-k Thus , different hypotheses w i t h the pr ice leader-fol lower i n the t ime- lag analysis : Ho: A d v e r t i s i n g strategies for the chosen brand between Store A i n week t and Store B i n week t-k are independent HA : A d v e r t i s i n g strategies for the chosen brand between Store A i n week t and Store B i n week t-k are dependent The n u l l , represented by Ho, states that the advert is ing strategies between Store A i n week t and Store B i n week t-k for a specif ied brand are independent; and the alternative, represented b y HA, states that the advert is ing strategies between Store A i n week t and Store B i n week t-k for a specif ied brand are dependent. Chapter 6 Results The data concerning the advert is ing frequency o f the products for the static analysis are recorded i n the cont ingency tables i n A p p e n d i x C . I f one fails to reject the n u l l hypothesis o f independence, then the advert is ing strategies o f two compet ing retailers are said to be statistically independent. I f the n u l l hypothesis o f independence is rejected, then w e can see i f there is a posi t ive or negative correlat ion between the advert is ing strategies o f the two compet ing retailers f rom the cel ls i n the corresponding cont ingency table i n A p p e n d i x C . Table 6.1 reports the advert is ing frequency ( x f s and y i ' s i n the cont ingency tables) o f different brands i n each supermarket for each product category. F o r a l l discussions o f correlat ion i n this chapter, please refer to the cont ingency tables i n A p p e n d i x C , and for discussions about the selected nat ional brands benchmark, refer to Table 5.1. There are 22 pages i n A p p e n d i x C , w i t h one page for each product category. E a c h page has four sets o f cont ingency tables: (1) the table for the chosen brand versus the private label at each supermarket i n the upper left hand corner, (2) the table for the chosen brand versus any other nat ional brand at each supermarket i n the upper right hand corner, (3) the table for any nat ional brand versus the private label at each supermarket i n the l ower left hand corner, and (4) the table for the advert is ing status o f the chosen brand so ld by three pairs o f compet ing supermarkets i n the lower right hand corner o f the page. The chi-square statistics o f each hypothesis testing for each product category are reported i n the result tables. E a c h chi-square statistic measures the discrepancy between the observed data and the expected values under the n u l l hypothesis o f independence. The test statistic is very sma l l i n most cases, because the large value o f test statistic x2 indicates evidence against the n u l l hypothesis o f independence, thus the sma l l values o f test statistics i n most o f our results indicate that the data provides sufficient evidence that the j o in t advert is ing strategies o f two compet ing retail stores and the corresponding manufacturers, are statistically independent o n the selected product categories. Gene ra l ly there is a strong evidence showing that compet ing f i rms randomly choose what product to advertise and when , coherent w i t h the f indings o f previous researches by Varian ' (1980) and L a c h (2002). 39 Chapter 6. Results 40 The result of the dynamic analysis for the leader-follower relationship is discussed in section 6.2. Each supermarket is assumed to be a price leader; then it is paired with each of the three competing supermarkets as a price follower and tested with four different pairs of brands combination. Because there are twelve hypothesis tests for each product, and there are 22 products in total, hence, the results are not formally presented here. 6.1 Static Analysis 6.1.1 Results on Across-Store Advertisement The chi-square statistics of the across-store chi-square tests for each product category, chosen brand at Store A vs. chosen brand at Store B, are reported in Table 6.2. There is strong evidence that two competing firms randomly choose what product to advertise and when. The small values of test statistics in most of our across-store results indicate that the data provides sufficient evidence that the advertising decisions of competing stores are statistically independent. In addition to the across-store analysis for the chosen brands as reported in Table 6.2, another across-store analysis for the private labels, private label of Store A vs. private label of Store B, has been carried out. Since the manufacturers have a certain degree of vertical intervention to the retailers' advertising consideration for the national name brands, the private label chi-square tests provide statistical information on the advertising strategies for the private labels, and see i f the result would be similar to the findings in the across-store chosen brands analysis. The set of chi-square statistics for the across-store chi-square tests, private label of Store A vs. private label of Store B, for each product category, are reported in Table 6.3. Although the null hypothesis of independent advertising is not rejected in majority of the cases, there are some exceptions. The chi-square test rejects the null hypothesis of independence in the comparison between Safeway and Market Place IGA in three categories: canned fish, frozen vegetables, and peanut butter. Data shows a positive correlation in these three cases, implying that there is intense competition between Safeway and Market Place IGA for those products because they are advertised together in the same week more than the expected frequency during the 52 weeks of the sample period. Another hypothesis of independence is rejected in the Safeway and The Real Canadian Superstore comparison for the canned fish category with a positive correlation in the advertising A l t h o u g h the n u l l hypothesis o f independent advert is ing is not rejected i n majority o f the cases, there are some exceptions. The y Product Category Market Place IGA Safeway Canada Save-on Foods The Real Canadian Superstore A B C A B C A B C A B C Bacon 17 31 0 8 31 4 0 40 2 0 0 42 Breakfast Cereal 28 27 12 48 47 1 40 37 19 11 14 8 Canned Fish 34 21 11 25 28 5 9 28 21 8 0 5 Canned Soup 39 13 14 31 12 0 29 12 12 5 2 7 Dry Pasta 15 17 18 19 15 0 9 20 20 0 4 7 Frozen Pizza 20 30 0 13 26 12 13 33 10 3 6 19 Frozen Punch Beverage 7 33 14 0 24 21 6 25 7 3 6 13 Frozen Vegetable 13 0 16 13 1 22 17 1 11 4 2 2 Fruit Jam 10 19 12 6 7 0 3 29 20 2 10 13 Ground Coffee 23 26 14 4 22 0 15 28 10 6 2 26 Instant Coffee 11 28 0 6 5 0 6 15 4 5 4 6 Juice from Concentrate 36 19 3 17 9 18 30 8 0 4 6 8 Juice Not from Concentrate 23 27 7 23 10 4 21 9 16 4 4 13 Ketchup 11 0 16 10 0 0 11 2 11 9 0 9 Margarine 1 20 22 10 44 2 2 32 16 2 8 14 Mayonnaise 40 6 11 26 12 0 12 17 0 11 7 1 Pasta Sauce 6 32 19 6 33 1 2 34 22 0 7 13 Peanut Butter 24 9 19 24 11 4 10 10 24 2 6 18 Potato Chips 12 34 12 19 16 7 14 19 15 1 2 19 Processed Cheese Slices 18 19 10 36 6 6 12 10 8 1 0 16 Tea Bag 14 20 19 7 14 18 20 31 4 4 9 9 Waffle 14 0 10 19 2 16 17 3 21 5 0 4 * A - the advertising frequency of the chosen brand * B - the advertising frequency of any national brand other than the chosen brand * C - the advertising frequency of the private label at the corresponding supermarket Chapter 6. Results 42 frequencies. The competing supermarkets have the same advertising status in 31 weeks of the 52 weeks sample period. Referring to Table 5.1 for the national brands benchmark, note that all rejected cases are positively correlated. This implies that both food manufacturers and the retailers launched an aggregate merchandizing activity at the retail level to enhance the performance of both parties, with an increase of overall profit for the supermarkets and with an increase of sales for the individual products. The positive correlation in the advertising frequencies of Safeway and Market Place IGA, implies that Safeway and Market Place IGA have a similar target population segment of high profit margin, high service retailers. This is why the food manufacturers are advertising their products at both stores: they wish to increase sales in that particular market segment. Similar to the across-store chosen brands results, there is strong evidence that two competing firms randomly choose what product to advertise and when in the across-store private labels analysis. The small values of test statistics in most of our across-store results indicate that the data provides sufficient evidence that the advertising strategies between competing stores are statistically independent. While the advertising decision for national brand is made jointly by manufacturers and retailers, the across-store private label advertising interaction does not demonstrate significant intervention by the manufacturers. However, in the five scenarios of rejected hypothesis in four product categories (Instant Coffee, Mayonnaise, Processed Cheese, and Waffle), the private labels advertising strategies between stores are negatively correlated compared to the positive correlation in the name brand analysis. With the national brands, when the across-store advertisement is not statistically independent, the product has the same advertising status; but when the across-store advertisement is not statistically independent for the private labels, the product has opposite advertising status instead. It is interesting to see that the chosen brands tend to be advertised together in a pair of competing retailers in a given week, where as the private labels of the supermarkets, tested pair-wise, have an opposite advertising schedule. Since the manufacturers have a certain influence on the retailers' advertising strategies on the national name brands, the same-good advertising across-stores would have been the manufacturers' marketing strategy in price advertising to increase product exposure and consumer awareness. With the assumption of Bertrand competition, and retailers do not have information of their competitors' action or strategy, it is doubtful that how the retailers achieve in avoiding head-on competition of their private label products. But the numbers on the contingency tables show that the negative correlation of the private labels in the scenarios of rejected hypothesis is often a result of low advertising frequency for both retailers, meaning that both Chapter 6. Results 43 supermarkets do not advertise the product in most of the sample period. Thus, the supermarket might not be intentionally having same good advertising but the good is not advertised due to other reasons. Across-store analyses for both national brands and private labels show that the advertising strategy of a supermarket is statistically independent from another supermarket. Supermarkets randomly advertise promoted items, thus randomly setting the promoted price of the advertised products. Although the advertising decision is somewhat influenced by the manufacturers, the jointly determined advertising behavior, by the manufacturers and the retailers, is to maximize individual firm's profit and optimize its performance. This result is consistent with the findings of Varian (1980) and Lach (2001), with informed and uninformed consumers making decisions on the basis of price. Random price advertising also reduces the intensity of competition among supermarket chains. Table 6.2 Chi-Square Statistics on Across-Store Advertisement: Safeway Canada vs. a Competing Supermarket Chain (National Name Brand) Product Category IGA Save-On Superstore V x1 x2 Bacon 0.134 0.815 0.815 Breakfast Cereal 0.026 1.769 2.162 Canned Fish 5.385* 0.133 11.355* Canned Soup 0.027 0.949 0.955 Dry Pasta 0.109 0.293 0.073 Frozen Pizza 0 1.675 0.118 Frozen Punch Beverage 0.999 1.239 2.695 Frozen Vegetable 4.137* 3.525 1.444 Fruit Jam 1.615 0.415 0.271 Ground Coffee 1.663 0.031 0.769 Instant Coffee 0.603 0.175 0.722 Juice from Concentrate 0.243 1.721 0.590 Juice Not from Concentrate 0.009 2.381 0.780 Ketchup 0.243 0.495 1.268 Margarine 0.243 0.495 1.268 Mayonnaise 0 0 0.115 Pasta Sauce 3.156 3.014 1.239 Peanut Butter 5.056* 0.189 0.012 Potato Chips 1.219 3.508 1.771 Processed Cheese Slices 0.085 0.048 0.453 Tea Bags 1.044 0.334 0.674 Waffle 3.508 0.234 0.029 * Indicates the chi-square statistic that is larger than 3.84, null hypothesis is rejected at 5% significance level Chapter 6. Results 44 Table 6.3 Chi-Square Statistics on Between-Store Advertisement: Safeway Canada vs. a Competing Supermarket Chain (Private Label) Product Category IGA Save-On Superstore t t t Bacon 2.013 0.173 0.093 Breakfast Cereal 0.306 0.587 0.185 Canned Fish 1.484 0.000 0.589 Canned Soup 0.251 0.375 0.999 Dry Pasta 0.098 0.053 0.999 Frozen Pizza 0.375 0.066 0.177 Frozen Punch Beverage 2.859 0.469 0.266 Frozen Vegetables 0.020 1.292 0.050 Fruit Jam 0.375 0.053 0.307 Ground Coffee 0.251 0.551 0.000 Instant Coffee 12.995* 2.013 1.239 Juice from Concentrate 6.014* 0.098 0.035 Juice not from Concentrate 0.674 1.926 1.444 Ketchup 0.162 0.455 0.669 Margarine 0.050 0.361 0.563 Mayonnaise 0.455 12.995* 6.123* Pasta Sauce 1.771 0.748 0.340 Peanut Butter 0.014 0.539 0.002 Potato Chips 0.138 0.774 0.139 Processed Cheese Slices 0.868 1.233 4.103* Tea Bags 2.748 0.308 0.391 Waffle 0.495 4.493* 0.068 * Indicates the chi-square statistic that is larger than 3.84 and thus the null hypothesis is rejected at 5% significance level 6.1.2 Results on Across-Brand Advertisement 6.1.2.1 The Chosen Brand vs. Private Label at Each Supermarket The test statistics fail to reject the null hypothesis of independence between the advertising strategies of the chosen brand and the private label within a particular supermarket in the majority of cases. Table 6.4 reports the test statistics for the across-brand scenario between the chosen brand and the private label at each supermarket. The null hypothesis is rejected at IGA in the ketchup category, at Safeway for frozen vegetables with a negative correlation between opposite-goods advertising and processed cheese slices, but with a positive correlation with same-good advertising. A test statistic, %2 > 3.84, rejected the null hypothesis of independence for instant coffee at Save-on Foods, and for juice from concentrate in Superstore; both have a strong positive correlation implying same-good Chapter 6. Results 45 advertising between the chosen brand and the private labels in those retailers. From the five tests which the hypotheses are rejected, three of them have a positive correlation indicating a same-good advertising; however, the numbers in the contingency tables show that both the chosen brand and the private label were not advertised in more than 35 weeks in the sample period. The five product categories do not have high advertising frequency in the supermarket corresponding to the rejected cases. There is a strong evidence to show that the advertising strategies, for the chosen brand and the private label, at each supermarket are independent in most of the tests. This indicates that the manufacturer of the chosen brand and the retailer are advertising on a random basis in general, so the two brands are being advertised together some of the time and alternately at other times. It is possible that the retailers have received payments or promotional allowances, so the private labels are not advertised at the same time as the national brands to avoid head-on competition for sales volume within the supermarket. In the rejected scenarios, the negative correlation means that the two brands avoid matching advertisement in a given week to reduce competition; the positive correlation implying the two brands are likely to be advertised in a given week. However, all of the three rejected hypotheses have positive correlation because there is no advertisement for both the chosen brand and private label for more than 35 week in the sample period. Since a designated budget is allocated for advertising, it is inefficient for the supermarket to advertise both the national brand and private label in the same week, while price promotion for one brand of a particular product is adequate to attract consumer to visit the store. Chapter 6. Results 46 Table 6.4 Chi-Square Statistics on Within-Store Advertisement: The Chosen Brand vs. Private Label at Each Supermarket Product Category IGA Safeway Save-On Superstore x2 x2 r2 x2 Bacon 0.127 0.788 3.835 0.551 Breakfast Cereal 0.126 0.085 0.896 0.084 Canned Fish 0.724 0.145 1.491 0.091 Canned Soup 1.173 0.036 0.042 3.344 Dry Pasta 1.990 0.073 0.121 0.999 Frozen Pizza 0.053 0.578 0.578 1.246 Frozen Punch Beverage 0.011 0.023 1.055 2.948 Frozen Vegetable 1.926 8.509* 0.085 0.173 Fruit Jam 0.066 1.239 0.035 0.693 Ground Coffee 0.259 2.142 2.142 0.754 Instant Coffee 0.455 1.239 33.641* 0.722 Juice from Concentrate 0.010 1.371 0.023 3.989* Juice Not from Concentrate 0.547 0.058 4.493 1.444 Ketchup 6.933* 0.551 1.217 0.292 Margarine 0.748 0.495 0.924 5.646 Mayonnaise 1.537 0 0.375 0.274 Pasta Sauce 0.530 0.133 0.050 0.307 Peanut Butter 1.045 3.714 3.408 0.218 Potato Chips 1.910 1.727 1.979 0.587 Processed Cheese Slices 0.117 8.798* 2.836 0.453 Tea Bags 1.886 0.008 2.708 0.907 Waffle 1.802 3.154 2.980 1.180 * Indicates the chi-square statistic that is larger than 3.84 and thus the null hypothesis is rejected at 5% significance level 6.1.2.2 Any National Brand vs. Private Label at Each Supermarket Table 6.5 presents the result of the tests among national brands and private label at each supermarket. The number of rejected null hypotheses of independence is relatively higher than the last two analyses in section 6.1.1 and 6.1.2.1. The results indicate that the advertising strategies of any national brand and private brands in a supermarket are dependent and are mostly negatively correlated. Ten cases are rejected at Market Place IGA, four cases at Safeway, seven cases at Save-on Foods, and surprisingly only one case is rejected at Superstore. All rejected cases in Market Place IGA, Safeway and Save-on Foods show a negative correlation between the national brands and the private labels. In these scenarios, both manufacturers and retailers would make the same decision to advertise opposite goods and Chapter 6. Results 47 avoid head-on competition. From the perspective of manufacturers, it is inefficient to advertise their brand along with the private labels from the same category because this would reduce the effect of price promotion on the sales volume of the product. From the perspective of retailers, advertising two national brands in same category would increase their advertising costs. This increases traffic volume and sales increase as consumers are unlikely to purchase only the promoted items once they are at the store. Advertising products in the same category does not yield a sales increase effect as each consumer would only purchase one brand within the same product category; there is no brand switching effect. Consumers exhibit their usual purchasing behavior by stocking up when the price is reduced, and therefore retailers lose. Therefore, the advertising strategy between the national brands and private brands within a supermarket is either independent to conduct a random advertisement or has opposite-goods advertisement. Unlike the three other retail chains, the only case rejected in Superstore is juice from concentrate with a positive correlation, a clear deviation from The Real Canadian Superstore and the other three supermarkets. In the cases where the null hypothesis is rejected, Superstore is the only retailer that shows a positive correlation between the national brands and its private label. This implies that Superstore has a strong private label program and it advertises the private label products at the same time as national brands. Since the retailer has a higher profit margin on private labels, Superstore attempts to increase its profit by featuring its private label products as substitutes for the national brand products, at a lower price. The differentiation in the results show that Superstore has a different market position and marketing strategy compared to the other three supermarket chains studied in this thesis. Chapter 6. Results 48 Table 6.5 Chi-Square Statistics on Within-Store Advertisement: Any National Brand vs. Private Label at Each Supermarket Product Category IGA Safeway Save-On Superstore x2 x2 x2 x2 Bacon 1.900 0.752 6.933* 0.551 Breakfast Cereal 0.112 6.144 0.014 0.174 Canned Fish 6.281* 5.005* 0.156 0.091 Canned Soup 7.139* 0.307 1.997 2.284 Dry Pasta 5.967* 0.036 0.624 0.674 Frozen Pizza 0.203 0.069 0.334 0.004 Frozen Punch Beverage 2.472 7.077* 2.810 1.376 Frozen Vegetable 1.926 2.255 0.019 3.014 Fruit Jam 3.328 0.375 8.799* 2.311 Ground Coffee 4.535* 0.001 8.945* 1.486 Instant Coffee 0.001 0.862 2.935 1.420 Juice from Concentrate 0.408 0.340 0.073 4.690* Juice Not from Concentrate 0.008 0.244 11.455* 0.495 Ketchup 6.933* 0.042 1.883 0.292 Margarine 4.938* 2.384 1.806 0.302 Mayonnaise 0.658 0.203 0.006 0.495 Pasta Sauce 5.409* 0.540 0.234 0.495 Peanut Butter 4.348* 8.922* 17.093* 1.477 Potato Chips 6.370* 0.729 2.703 0.014 Processed Cheese Slices 0.001 6.280* 6.405* 0.453 Tea Bags 15.117* 1.000 8.677* 0.878 Waffle 1.802 3.794 1.817 1.180 * Indicates the chi-square statistic that is larger than 3.84 and thus the null hypothesis is rejected at 5% significance level 6.1.2.3 The Chosen Brand vs. Any Other National Brand at Each Supermarket The chi-square statistics are reported in Table 6.6. Again, more tests with x1 statistics rejecting null hypotheses compared to the analysis in Section 6.1.1 and 6.1.2.1. Eight cases are rejected at IGA, eight cases rejected at Safeway, but only two cases are rejected at Save-on Foods, and four at Superstore. Similar to the analysis of national brands and private labels, Market Place IGA, Safeway, and Save-on Foods have a negative correlation for the rejected cases. The products in the same category are advertised alternatively during the 52 weeks. It is obvious that manufacturers have some level of influence on these advertising decisions. They do not want to create an intensive competition with other national brands because this would reduce the overall profit in the industry. The incentive for collusion is high if the market concentration Chapter 6. Results 49 is low in the manufacturing level. The differentiation between Superstore and the other three supermarket chains occurs in the result in this section as well. All four scenarios of rejection have positive correlation, with over 45 weeks of none advertising between the chosen brand and any other national brand. Adequate information demonstrate that Superstore have a higher advertising frequency for its private labels compared to national brands, it is obvious that Superstore has a different marketing strategy focusing on its private label rather than the national brands. Table 6.6 Chi-Square Statistics on Within-Store Advertisement: The Chosen Brand vs. Any Other National Brand at Each Supermarket Product Category IGA Safeway Save-On Superstore x2 x2 x2 x2 Bacon 40.411* 8.717* 0.375 12.995* Breakfast Cereal 3.714 0.265 0.112 0.542 Canned Fish 4.912* 0.662 0.387 0.815 Canned Soup 4.137* 0.599 1.258 3.903* Dry Pasta 10.239* 2.487 1.213 2.013 Frozen Pizza 0.711 5.026* 2.239 1.482 Frozen Punch Beverage 4.247* 0.006 2.681 9.479* Frozen Vegetable 0.307 0.340 0.495 0.173 Fruit Jam 0.117 0.060 4.014* 0.495 Ground Coffee 9.433* 0.532 0.437 3.014 Instant Coffee 1.665 12.728* 2.750 0.461 Juice from Concentrate 0.009 5.286 0.090 17.098* Juice Not from Concentrate 11.028* 5.881* 1.491 1.828 Ketchup 0.375 0.559 0.558 0.669 Margarine 0.637 22.599* 0.117 0.378 Mayonnaise 2.770 10.833* 4.207* 0.229 Pasta Sauce 0.382 11.780* 0.218 0.999 Peanut Butter 0.013 11.958* 2.948 3.014 Potato Chips 1.631 0.279 1.886 0.041 Processed Cheese Slices 0.122 1.178 1.193 6.123 Tea Bags 11.974* 3.755 2.884 0.179 Waffle 0.251 0.163 1.670 1.560 * Indicates that chi-square statistic that is larger than 3.84 and thus the null hypothesis is rejected at 5% significance level Chapter 6. Results 50 6.2 Dynamic Analysis 6.2.1 Leader - Follower Relationship The dynamic analysis is performed to f ind out i f there is any potential leader-fol lower relationship i n the advert is ing status among the compet ing supermarkets over the sampl ing per iod. A s stated i n Sec t ion 5.3.2, there are 16 tests for each product category, since w e are on ly interested i n cases, w i t h n u l l hypotheses are rejected at 5% significant l eve l , ind ica t ing a possible leader-fol lower relat ionship; hence, the tables i n this section o n l y present results when the chi-square statistic is larger than 3.84. There are no significant patterns o f leader-fol lower relat ionship emerged i n these hypothesis tests. Howeve r , products categories, and supermarkets w i t h a higher advert is ing frequency tend to have a higher poss ib i l i ty o f advert is ing leader-fol lower relat ionship between two supermarkets over t ime. In this study, I G A and Safeway are the supermarkets that have higher advert is ing rates than Save-on Foods , w h i l e Superstore advertises least among the four supermarkets for the 22 selected products. The results o f the f ive products, Breakfast Cerea l , Canned F i s h , Canned Soup, Peanut Butter and Potato C h i p s , are as shown i n Table 6.7 to Table 6.11. Tests w i t h n u l l hypotheses rejected at 5 % signif icant l eve l for the Breakfast Cerea l product category is shown i n Table 6.7. There is no pattern for a leading store and a particular f o l l o w i n g store. It is obvious that there are more n u l l hypotheses be ing rejected w h e n k = 1 and k = 3. There are quite a few cases w i t h Private Labe l s as advert is ing fol lowers , and this is coherent w i t h the w o r k by Sex ton and L a v o i e (2001). Reta i ler is advert is ing the Private L a b e l f o l l o w i n g the advertisement by another retailer. N o t e that four o f the five cases w i t h Private L a b e l as a fo l lower appear w h e n k = 3 and k = 4, this shows that retailer want to m a x i m i z e the effect o f the advertisement by responding to the competi tor 's act ion i n advert is ing it 's private label three weeks or four weeks after the competi tor 's advertisement. Because consumers w o u l d respond to the first wave o f advertisement and make purchase at the advert is ing supermarket, i f another retailer advertises immedia te ly i n the week after, w h e n k = 1; the cost o f that part icular advertisement w o u l d not be very effective i n increasing sale v o l u m e o f the product. Breakfast Cerea l is a non-perishable product, and food is an infer ior good; consumers m a y have stocked up w i t h the sale advertisement or have not yet f inished consuming the product, g iven the general packaging size o f Breakfast Cerea l . Therefore, consumer are u n l i k e l y to purchase the same product i n the per iod w h e n k = 1. Howeve r , the advertisement o f the Speci f ied Brands are not as f lexib le , because manufacturers have scheduled the advert is ing Chapter 6. Results 51 patterns w i t h supermarkets months i n advance, and it is u n l i k e l y for them to make last minute changes compared to the Private Labe l s . Table 6.7 Leader-Follower Relationships for the Breakfast Cereal Product Category Lag Leader(Brand) Follower (Brand) x2 k=l IGA (specified) Safeway (specified) 4.833 IGA (private) Save-on (private) 31.19 Superstore (private) Safeway (specified) 3.84 k=2 Save-on (specified) IGA (specified) 4.788 k=3 IGA (private) Save-on (private) 9.228 Safeway (private) Superstore (specified) 3.981 Save-on (private) Superstore (private) 4.472 Superstore (specified) IGA (private) 5.478 k=4 Safeway (specified) Save-on (specified) 6.701 Safeway (specified) Save-on (private) 7.273 In table 6.8, there are 3 cases w i t h rejected n u l l hypotheses w h e n k = 2 i n the Canned F i s h product category. O n l y one n u l l hypothesis is rejected w h e n k - 3 and w h e n k = 4. Compared to the difference results i n Table 6.7, it is l i k e l y to be associated w i t h the package size and the number o f cans is consumed i n preparing meals , resul t ing i n a different purchasing pattern. It looks l i ke that Save-on Foods is p l ay ing a fo l lower role to its competitor, especia l ly Safeway, but it shows that Private L a b e l is be ing advertised i n three o f the s ix cases w h e n Save-on Foods is the fol lower . W e k n o w that retailer has the complete control o f the advert is ing status for the Private L a b e l , w h i l e the advert is ing pattern o f the Speci f ied B r a n d is p lanned by the manufacturer. In this situation, we do not have fu l l knowledge i f Save-on Foods is targeting Safeway and is act ing as a fo l lower i n responding to Safeway 's advert is ing schedule i n the Canned F i s h product category; or manufacturer selects the advert is ing sequence among the supermarkets. The data o f this study shows the jo in t advert is ing dec i s ion between the manufacturers and the retailers, and cannot d is t inguish the ind iv idua l p layer ' s ini t iat ive decis ion. Further Research is required to differentiate the advert is ing dec i s ion between retailer and manufacturer. Chapter 6. Results 52 Table 6.8 Leader-Follower Relationships for the Canned Fish Product Category Lag Leader (Brand) Follower (Brand) x2 k=l Safeway (private) Save-on (private) 3.868 Superstore (specified) Save-on (private) 6.122 k=2 IGA (specified) Save-on (specified) 4.480 Safeway (specified) Save-on (specified) 3.899 Safeway (specified) Save-on (private) 5.773 k=3 Safeway (private) IGA (specified) 6.639 k=4 IGA (private) Save-on (specified) 4.769 Unexpectedly , there are very few n u l l hypotheses rejected i n the Canned Soup product category. Campbe l l ' s , as the Speci f ied B r a n d , is hav ing a very h i g h market share i n the Canned Soup industry, thus most o f the advertisement i n the flyers are p laced by the same manufacturer, this explains w h y the advert is ing rate is one o f the highest among the product categories but there are s t i l l on ly 5 cases w i t h rejected n u l l hypotheses. The result shows that the manufacturer i n this category w o u l d schedule the advertisement prec ise ly among the supermarkets i n order to m i n i m i z e advert is ing costs and achieve the goal o f hav ing sufficient exposure to reminder consumers o f the existence o f the brand. A g a i n , Save-on Foods is act ing as a fo l lower by advert is ing its private label but to no specific leader. It is be l i eved that Save-on Foods is responding to the competi tor ' s advertisement o f both Speci f ied B r a n d and Private Labe l s by adjusting the advert is ing schedule o f its Private L a b e l . Table 6.9 Leader-Follower Relationships for the Canned Soup Product Category Lag Leader(Brand) Follower (Brand) x2 k=l Safeway (specified) Superstore (specified) 3.868 Superstore (specified) Save-on (private) 4.554 k=2 Safeway (private) Save-on (private) 4.633 Superstore (private) Save-on (private) 4.160 k=4 IGA (private) Save-on (specified) 4.769 Peanut Butter has the highest number o f rejected n u l l hypotheses o f the 22 product categories. It has a re la t ive ly h i g h advert is ing frequency and the industry is rather compet i t ive when compared to other product categories. R a n d o m leader-fol lower relat ionship occurs over t ime, w i t h no specif ic leader o r fo l lower store-wise. Coherent w i t h the previous discuss ion, more n u l l hypotheses are b r ing rejected w h e n k > 1; the product is Chapter 6. Results 53 not advertised immedia te ly i n the first week after the product is be ing advertised b y a competitor. T h e randomness i n this sect ion c o u l d be a result o f var ious package sizes for Peanut Butter, because the data records the advert is ing pattern for Peanut Butter regardless o f package size. Yet , it is diff icul t to compare this w i t h the result i n the Breakfast Cerea l category, since the consumpt ion rate o f the two products are different and consumers ' purchasing behavior w o u l d be different as w e l l . Further research is required to d iscover the detai l relat ionship o f the advert is ing dec i s ion w i t h regard to different packag ing sizes o f a product. Table 6.10 Leader-Follower Relationships for the Peanut Butter Product Category Lag Leader(Brand) Follower (Brand) x2 k = l Safeway (private) IGA (specified) 3.858 Save-on (specified) Superstore (specified) 8.534 k=2 IGA (specified) Save-on (specified) 5.128 Superstore (specified) IGA (specified) 6.254 Superstore (private) IGA (private) 4.740 k=3 IGA (specified) Safeway (specified) 16.593 IGA (private) Superstore (specified) 3.925 Superstore (private) IGA (private) 4.013 k=4 Save-on (specified) IGA (private) 7.257 Save-on (specified) Superstore (specified) 6.074 Superstore (private) Save-on (specified) 4.042 In general, there are r andom leader-fol lower relat ionships between different pairs o f supermarkets, but there is no specific ind ica t ion for a strong leader-fol lower relat ionship between supermarkets, or a definite suggestion for a part icular store as the leader or as the fol lower. A l t h o u g h Save-on Foods is act ing as a fo l lower i n the Can n ed F i s h and Canned Soup product categories, w e cannot c lear ly d is t inguish i f that is a result o f the manufacturer 's advert is ing schedule or pure dec i s ion by the retailer as a direct response to compet i tor ' s act ion. The random occurrence o f advert is ing leader-fol lower relat ionships differs among product categories; i t is possible that the advert is ing pattern is schedule depending o n the consuming habits, package sizes, and the nature o f the products, and further research is required i n f ind ing out the answer to that. Chapter 7 Summary and Conclusion In this thesis we have looked at the association between the advertising patterns in sales flyers of competing supermarkets, and the advertising strategy of each supermarket that is jointly determined by competing manufacturers and the corresponding retailers. Our results show that there is strong evidence to support the null hypothesis that competing supermarkets have mixed strategies in product advertising. The results are consistent with our hypothesis of independent advertising strategies on the part of two competing firms. 7.1 Summary The sales flyer is a commonly used form of advertisement among retailers, especially in the food retailing industry. It is an effective tool for supermarkets to communicate price information to large numbers of consumers, given the many products discounted every week and their corresponding price changes. This research focuses on the advertising frequency of 22 selected product categories from four supermarket chains with sales flyers distributed in the Vancouver lower mainland, and investigates the association of the explicit advertising strategies of competing retailers with the advertising patterns in the sales flyers over a sample period of 52 weeks. The results of our across-store hypothesis testing show strong evidence to support the null hypothesis that the advertising strategies of two competing retailers are independent. These results are consistent with the conclusion of Varian (1980) and Lach (2002) although their approach is other than the spatial model used in this thesis. Most of the advertisement in the flyers is planned approximately three months in advance of the effective dates of the sales flyers with some last minute changes and corrections prior to the printing of flyers. Thus, it is believed that there is a lag relationship between two competing supermarkets with similar market position and target population segment during the sample period of 52 weeks, as both manufacturers and retailers attempt to maximize their profit by reducing their competition in the market. From the result in the time lag analysis, lagged relationships emerge in a random pattern across-brand and across-store, but the occurrences of lagged relationship for most of the cases concentrate in lag period with k = {1,3,}. 54 Chapter 7. Summary and Conclusion 55 A retailer does not have information about its competitors' pricing and marketing strategies in Bertrand competition, thus each advertisement in the sales flyers is treated as a independent promotional event. Because the competing supermarkets are selling products from the same major food processors, the wholesale price of the same product to these same sized supermarkets would be very similar. The price variation among the retailers is determined by their mark-up margin on the products, and retailers with similar target segment and market position are likely to set products in similar price range, and the competition between these supermarkets would be more rigorous compared to supermarkets in different market segment. Although the price of private label products is lower than the national branded products, the retailer's profit margin on its private label product is higher than those of the national branded products because the products are contracted to food processors in attempting to mimic the products of the popular brand names. Therefore, an increase in the sales of private label would yield a higher profit compared to an increase in the sales of national brands proportionally. On the other hand, food retailers receive promotional allowances and payments from national brand manufacturers who want to form strategic alliances with individual retailers, with the goal of reducing brand switching within individual stores. Some of the test statistics reject the null hypothesis of independence between the advertising strategies of all national brands available within a store, and those of any national brand and the private label at each supermarket. This implies that the competition across brands is reduced if the manufacturers give payments or advertising allowances to the retailers. However, this is not the case for The Real Canadian Superstore even though there is a positive correlation rather than a negative correlation in most other cases. As indicated in the work of Rostoks (2002), Superstore has the strongest private label program in Canada, and the profit margin of the private label is higher than that of the national brand. To maximize profit, Superstore lowers its advertising cost by featuring the promotion of national brand only but encourage brand switching, once the consumer has visited the store, by positioning the private label product next to the advertised brand and highlight the price difference as indicated by Morton and Zettelmeyer (2001) and Sayman et al. (2002). In fact, consumers are unlikely to purchase the promoted item only but also other unadvertised goods as well; therefore, Superstore "uses store brands to exploit the marginal-average cost gap of national brands"(p23). In general, manufacturers and each retailer determine their joint advertising strategies from two different perspectives but they would want to maximize the overall industry profit by reducing the competition at the retail level in randomizing advertisement Chapter 7. Summary and Conclusion 56 in the supermarket sales flyers. 7.2 Limitation There are several limitations to this thesis because of the data available and the static model applied as described in Chapter 4. Sexton and Lavoie (2001) point out that in perfect competition, firms recognize no active rivalries, but in imperfection competition, rival firms would response or react to one another; thus, an oligopoly would have a Cournot competition and an leader-follower relationship among the sellers is likely to emerge overtime. Since the model does not encounter the dynamic aspects of the advertising environment, additional tests are complied to analyze the lag relationship between the advertising strategies between the retailers. Same chi-square testing procedure as in Chapter 5 is constructed to test the existence of lag relationship between pairs of competing supermarket chains and with different brand combination of leader-follower in a Cournot setting. The yes-no advertising status of a brand in Store A in week t is compared with the yes-no advertising status of the same brand in Store B in week t - 1, where k = {1,2,3,4}. There are four leader-follower brand combinations: (1) chosen brand - chosen brand (2) chosen brand - private label (3) private label - chosen brand (4) private label - private label. The result of the time lag analyses varies with product categories and shows no particular pattern of leader-follower relationship across-store and across-brand. The lag relationships emerge in a random basis between stores in different brand combination. For example, Save-on Foods tends to be a follower in the Canned Fish and Canned Soup categories with different brand & store combination, and no significant lag relationships emerge between Save-on Foods and other retailers in other product categories. The null hypothesis of independence between the advertising strategies is rejected in many cases in the time lag analyses, and there are two interesting points. Firstly, most of the tests with rejected null hypotheses are in the product categories with higher advertising frequency. Secondly, most of the lag relationships emerge when k = {1,3,}. It is obvious that if the advertising frequency is high in the 52 week period, the probability of having a lag relationship overtime is higher. The existence of lag relationship across-store is noticeably higher with k = {1,3,}. If Store A is advertising a product in week t, Store B is likely to advertise a product in the same category in week t - 1 and t - 3. This is coherent with the Sexton and Lavoie (2001) that oligopoly sellers response or react to the action of one Chapter 7. Summary and Conclusion 57 another with a Cournot competition. However, the manufacturers also play an important role in advertising, since there is no particular pattern indicating a specific retailer as a leader or follower, the lag relationship across-store may be a schedule for manufacturer to advertise its product at each supermarket in different time frame because it is too expensive for manufacturer to have price advertising every week with each retailer. The advertisement is more efficient this way, in reminding the forgetful consumers who patronized with different supermarket, and maximize the manufacturer's sales and profit. Therefore, the assumption of Bertrand competition and independence of advertisement overtime may be unrealistic. Another question arises in the course of this study, there are private label products being advertised in the supermarket flyers without indicating the price. Although this situation is very rare, this occurred twice in Safeway only with the Private Label product during the 52 weeks sample period. However, it is often seen in the Superstore flyer that Superstore advertises its private label product by comparing the price of national brand, indicating the amount saved in large font while the actual price of the product printed in a much smaller font. The main feature of advertising flyer is providing price information to mass consumer, even though the image of the product also plays an important role of reniinding the forgetful consumer the existence of the product in a visual form, it is very unlike for a product to be advertised without the price information. In some issues of Safeway's flyer, the pictures of a set of Private Label products are group together and the prices are layout separately underneath the pictures. There may be problem with this form of advertising layout as the consumers cannot immediately associate the price information with the image of the particular product, and this would reduce the effectiveness of the advertisement. 7.3 Conclusion The purpose of advertising, given the specification of the product in all objective respects and given prices, is to increase the number of consumer who will prefer that product to its competitors. There are two ways in achieving the purpose: purely informative, provide consumers with information to exercise their choice and persuasive, increase the preference for the product. This thesis focuses on the information advertising because information is essential to the functioning of markets. Information affects the conduct of all firms in the market, and the simultaneously vertical and horizontal interactive behavior between firms. Chapter 7. Summary and Conclusion 58 With assumption of Bertrand competition between the supermarket chains, the hypotheses testing provide strong evidence that the across-store advertising strategies, which are jointly determined by the manufacturers and the corresponding supermarket, are statistically independent of each other in a static model. In the across-brand analyses within each supermarket, sufficient evidence shows that the jointly determined advertising strategies of chosen brand and the private label at each supermarket are statistically independent, but the null hypothesis of independence is rejected in a considerable number of cases when the advertising status of any national brand is tested against the chosen brand and when it is tested against the private label at each supermarket. Certain product categories in IGA, Safeway, and Save-on Foods show negative correlation when the hypotheses are rejected; this implies an opposite advertising situation between the brands. The behavior is understood because it is inefficient to advertise more than one brand of the same product category from the perspectives of both manufacturers and retailers. Manufacturer would want to have its product advertised solely in a given week, because when competing brands of the same category is advertised, the objective of identifying the specific product to consumer cannot be achieved as the effect of advertising is reduced. In general, the advertising strategies between two competing brands are statistically independent in Superstore. But it shows a positive correlation when the hypotheses of independence are rejected in the across-brand analyses. A positive correlation should imply same-good advertising between two competing brands, and this result contradicts with our prediction, but the contingency tables in Appendix C indicate that the result of same-good advertising emerges because the two competing brands have a no-no advertising status in most of the sample period. Superstore's behavior has a clear distinction from its horizontal competitors, low advertising frequency keeps advertising expenditure low and this is the operation objective in setting the market position of Superstore as discussed in section 3.2.4. However, the static model employed in this study does not consider the dynamic aspect of the advertising environment for a possible leader-follower relationship across-store or across-brand over time, and repetitive advertisement desired by both manufacturers and retailers to remind forgetful consumers of the identification of product and the existence of sellers. Price cut can be carried out in a very short time, but an advertising campaign takes time to mount; therefore further research is required to attend the dynamic environment of advertising and to distinguish the role of vertical and horizontal players in advertising strategy if possible. Chapter 7. Summary and Conclusion 59 7.4 Scope for Future Research This study has examined the independence of the across-store and across-brand advertising strategies, which are jointly determined by manufacturer and retailer. The result follows the theory of previous researches showing that competing firms want to randomize their price advertising. Although the purpose of advertisement is to provide information for consumer to make their choice in which brand and where to purchase, a carefully designed advertising campaign by the manufacturers and retailers determines a selection of products to be advertised on the sales flyer in particular time frame. This would actually create an imperfect information environment then rational consumers cannot learn which store has the lowest price for specific brands. According to Lach (2002), this allows price dispersion to persist, thus reduce the competition in the food retailing industry and the suppliers, in this case both manufacturer and retailer, can obtain a higher profit than the profit at perfect competition equilibrium. The study only considered the data in form of yes-no advertising status patterns in the sales flyer, with no pricing information on advertised and unadvertised products in the supermarket. If both the price-advertising data and price information are available, it would be interesting to find out if there is any correlation between the pricing and the advertising strategy of a product over time. A more complete dynamic model can be constructed in future research to interpret if a product is advertised only when there is a price promotion, with a price reduction to boost up sales volume and gain market share in short run, or a product is simply advertised repetitively to increase consumer's awareness in the long run. References BC Statistics. 2001. British Columbia Municipal and Regional District Population Estimates, www.bcstats.gov.bc.ca Bell, David R., Jeongwen Chiang, and V. Padmanabhan. 1999. The Decomposition of Promotional Response: An Empirical Generalization. Marketing Science. 18(4): 504-526. Burton, Scot, Donald R. Lichtenstein, and Richard G. Netemeyer. 1999. Exposure to Sales Flyers and Increased Purchases in Retail Supermarkets. Journal of Advertising Research. 39(5): 7-14. Drake, Julia. 2001. Did sales sizzle or fizzle? [National Market Survey 2001]. Canadian Grocer. 116(1): 16-25. Gupta, Sunil. 1988. Impact pf Sales Promotions on When, What, and How Much to Buy. Journal of Marketing Research. 3 625(4): 342-355. Hayes, Donald A., and Dereck J. Morris. 1991. Industrial Economics and Organization: Theory and Evidence 2n d ed. New York: Oxford University Press. Jafri, S. Hussain Ali, Trey Rodgers, and D. I. Padberg. 1993. Price Competition between National Brand and Private Label Foods Products: 1966-1992. Journal of Food Distribution Research. 24(2): 49-55. Kim, Sang Yong, and Richard Staelin. 1999. Manufacturer Allowances and Retailer Pass-Through Rates in a Competitive Environment. Marketing Science. 18(1): 59-76. Konishi, Hideo, and Michael T. Sandfort. 2002. Expanding Demand Through Price Advertisement. International Journal of Industrial Organization. 20(2): 965-994. Lach, Saul. 2002. Existence and Persistence of Price Dispersion: An Empirical Analysis. Working Paper 8737. National Bureau of Economic Research. 60 Lai, Rajiv, and Carman Matutes. 1994. Retail Pricing and Advertising Strategies. The Journal of Business. 67(3): 345-370. Lai, Rajiv, and Chakravarthi Messinger. 1996. The Inverse Relationship between Manufacturer and Retailer Margins: A Theory. Marketing Science. 15(2): 132-151. Lazarus, Eve. 2001. Worth a look: IGA Market Place. Canadian Grocer. 115(3): 36-37. LeBlanc, Greg. 1998. Informative Advertising Competition. The Journal of Industrial Economics. 46(1): 63-77. Marketing Magazine. 1996. Food Fight: Safeway's Old Dominance is just a Fading Memory in the West's Grocery Rivalry. 101 (7): 10,11. Marketing Magazine. 1996. Report on Western Canada. 101(7): 10-16 McGee, John S. 1998. Industrial Organization. Englewood Cliffs, N.J. : Prentice Hall. Messinger, Paul R. and Chakravarthi Narasimhan. Has Power shifted in the Grocery Channel? Marketing Science. 14(2): 189-223. Morton, Fiona S. and Florian Zettelmeyer. 2001. The Strategic Positioning of Store Brands in Retailer - Manufacturer Negotiations. Unpublished dissertation, University of California Berkeley, Berkeley. Nijs, Vincent R., Marnik G Dekimpe, Jan-Benedict E. M. Steenkamp, and Dominique M. Hanssens. 2001. The Category-Demand Effects of Price Promotions. Marketing Science. 20(1): 1-22. Rajiv, Surendra, Shantanu Dutta, and Sanjay K. Dhar. 2002. Asymmetric Store Positioning and Promotional Advertising Strategies: Theory and Evidence. Marketing Science. 21(1): 74-96. Reed, Michael R., and Lynn W. Robbins. 1981. An Analysis of Supermarket Advertising 61 Patterns. Journal of Food Distribution Research. 12(3): 19-32. Rao, Ram C , and Niladri Syam. 2001. Equilibrium Price Communication and Unadvertised Specials by Competing Supermarkets. Marketing Science. 20(1): 61-81. Rostocks, Lisa. 2002. Mastering Private Label. Canadian Grocer. 116(1): 26-27,29, 31. Samuels, Myra L. 1989. Statistics for the life Sciences. San Fransico: Dellen Pub. Co. Sayman, Serder, Stephen J. Hoch, and Jagmohan S. Raju. 2002. Positioning of Store Brands. Marketing Science. 21(4): 378-397. Sexton, Richard J., and Natalie Lavoie. 2001. Food Processing and Distribution. In B. Garner and G Rausser, (Eds.) Handbook of Agricultural Economics Vol. IB Marketing, Distribution and Consumers. (pp863-932). Amsterdam; New York : Elsevier. Stigler, George J. 1961. The Economics of Information. The Journal of Political Economy. 69(3): 213-225. Tellis, Gerard J., and Fred S. Zufryden. 1995. Tackling the Retailer Decision Maze Which Brands to Discount, How Much, When and Why? Marketing Science. 14(3): 271-299. Urbany, Joel E., Peter R. Dickson, and Alan G. Sawyer. 2000. Insights into Cross- and Within-Store Price Search: Retailer Estimates vs. Consumer Self-Reports. Journal of Retailing. 76(2): 243-258. Varian, Hal R. 1980. A Model of Sales. The American Economic Review. 70(4): 651-659. Walters, Rochkney G and Scott B. MacKenzie. 1988. A Structural Equations Analysis of the Impact of Price Promotions on Store Performance. Journal of Marketing Research. 25(1): 51-63. Ward, Michael B.Jay P. Shimshack, Jeffrey M. Perloffand J. Michael Harris. 2002. Effects of the Private-Label Invasion in Food Industries. American Journal of Agricultural Economics. 84(4):961-973 62 Woodcock, Brian. 1999. Canada Retail Food Sector Western Canada Food Sector Report. USDA Foreign Agricultural Services. 63 Appendix A Compare the firms' profit from Case 1 (same-good advertising) and Case 2 (opposite-good advertising) and find out which strategy is preferred. Recall that: nx c 2 and 4c Generally speaking, it is believed that firms would chose opposite good advertising to avoid intensive competition. n2 > nx is required to show that this is true. If R-u Let A = ( R - u ) > — is true, then firms would choose opposite-goods advertising 4c 2 R-u- (R-u)2 4c A>0 Find out what value of c makes A = 0 and solve for c* using quadratic formula c .= • If we assume that (R - u) = 0, then c'hj h > 0 and c*aw>0 Consumers make their store choice depend on the transportation cost, c, therefore, we would like to find out how the change in c affects the profit difference, A, between Case 1 and Case 2. Take the first order condition for A with respect to c dA _ R-u-y/Sc dc 4c If the FOC is positive then firms would prefer Case 2 over Case 1. Recall that c* = —(R -U)>0, then the restriction for the parameters R, u is R>u>y/Sc for firms to choose opposite-goods advertising strategy. 64 Appendix B We can see if Store A has an intention to deviate from the pure strategy of opposite-good advertising by comparing the profits if deviate or not. This is similar to the procedure in Appendix A. Let A = ndeviate - nno_devjale > 0 Set A = 0 and solve fore* using quadratic formula. _(3R-3u + 2cf 16c R-u (R-uf Ac = 0 . _-(R + u)±[(R + u)-yfu(R-u) 2 c = If we assume that (R - u) = 0, then c L A > 0 and c*low > 0 Now we would like to see how the change in c affects the profit difference, A, with deviation and without deviation. Take the derivative of A with respect to c dA _ 3u-3R + l ~dc~~ 4? Recall that c > 0, so the restriction for the parameters R, u for the firm to deviate is 1 1 R-u>- for A>0. 3 65 Appendix C Contingency Tables 66 Contingency Tables for Bacon Olympic vs Private Label within Store 0\ Olympic IGA Advertise Not Advertise Total Private Label Advertise. IGA Not Advertise 0 (0.00) 0 (0.00) 17 (17.00) 35 (35.00) 0 52 Total 17 35 52 Olympic Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 0 (0.62) 4 (3.38) 8 (7.38) 40 (40.62) 4 48 Total 8 44 52 Olympic Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 0 (0.00) 2 (2.00) 0 (0.00) 50 (50.00) 2 50 Total 0 52 52 Olympic Superstore. Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 0 (0.00) 42 (42.00) 0 (0.00) 10 (10.00) 42 10 Total 0 52 52 Private Label vs Other National Brands within Store National Brands IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 0 (0.00) 0 (0.00) 48 (48.00) 4 (4.00) 0 52 Total 48 4 52 National Brands Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 2 (2.77) 2 (1.23) 34 (33.23) 14 (14.77) 4 48 Total 36 16 52 National Brands Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 0 (1.54) 2 (0.46) 40 (38.46) 10 (11.54) 2 50 Total 40 12 52 National Brands Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 0 (0.00) 42 (42.00) 0 (0.00) 10 (10.00) 42 10 Total 0 52 52 Olympic vs Other National Brands within Store Olympic IGA Advertise Not Advertise Total Other . Advertise IGA Not Advertise 0 (10.46) 32 (21.54) 17 (6.54) 3 (13.46) 32 20 Total 17 35 52 Olympic Safeway Advertise Not Advertise Total Other Advertise Safeway Not Advertise 1 (4.77) 30 (26.23) 7 (3.23) 14 (17.77) 31 21 Total 8 44 52 Olympic Save-on Advertise Not Advertise Total Other Advertise Save-on Not Advertise 0 (0.02) 40 (40.00) 0 (0.00) 12 (12.02) 40 12 Total 0 52 52 Olympic' Superstore Advertise Not Advertise Total Other Advertise Superstore Not Advertise 0 (0.00) 0 (0.00) 0 (0.00) 52 (52.00) 0 52 Total 0 52 52 Olympic vs Olympic across Store Olympic Safeway Advertise Not Advertise Total Olympic Advertise IGA Not Advertise 3 (2.55) 14 (14.38) 5 (5.45) 30 (29.62) 17 35 Total 8 44 52 Olympic Safeway Advertise Not Advertise Total Olympic Advertise Save-on Not Advertise 0 (0.00) 0 (0.00) 8 (8.00) 44 (44.00) 0 52 Total 8 44 52 Olympic Safeway Advertise Not Advertise Total Olympic Advertise Superstore Not Advertise 0 (0.00) 0 (0.00) 8 (8.00) 44 (44.00) 0 52 Total 8 44 . 52 Contingency Tables Kellogg's vs Private Label within Store Kellogg's IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 7 (6.47) 5 (5.54) 21 (21.54) 19 (18.46) 12 40 Total 28 24 52 Kellogg's Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 1 (0.92) 0 (0.08) 47 (47.08) 4 (3.92) 1 51 Total 48 4 52 Kellogg's Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 16 (14.63) 3 (4.39) 24 (25.39) 9 (7.62) 19 33 Total 40 12 52 Kellogg's Superstore Total Advertise Not Advertise ON Private Label Advertise 2 (1.69) 6 6.31 8 OO Superstore Not Advertise 9 (9.31) 35 34.69 44 Total 11 41 52 for Breakfast Cereal Kellogg's vs Other National Brands within Store Kellogg's IGA Advertise Not Advertise Total Other Advertise IGA Not Advertise 18 (14.54) 9 (12.46) 10 (13.46) 15 (11.54) 27 25 Total 28 24 52 Kellogg's Safeway Advertise Not Advertise Total Other Advertise Safeway Not Advertise 44 4 4 0 48 4 Total 48 4 52 Kellogg's Save-on Advertise Not Advertise Total Other Advertise Save-on Not Advertise 28 (28.46) 9 (8.54) 12 (11.54) 3 (3.46) 37 15 Total 40 12 52 Kellogg's Superstore Advertise Not Advertise Total Other Advertise Superstore Not Advertise 2 (2.96) 12 (11.04) 9 (8.04) 29 (29.96) 14 38 Total 11 41 52 Private Label vs Other National Brands within Store National Brands IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 9 (8.54) 3 (3.46) 28 (28.46) 12 (11.54) 12 40 Total 37 15 52 National Brands Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 1 (1.00) 0 (0.00) 51 (51.00) 0 (0.00) 1 51 Total 52 0 52 National Brands Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 18 (17.90) 1 (1.10) 31 (31.10) 2 (1.90) 19 33 Total 49 3 52 National Brands Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 3 (3.54) 5 (4.46) 20 (19.46) 24 (24.54) 8 44 Total 23 29 52 Kellogg's vs Kellogg's across Store Kellogg's Safeway Advertise Not Advertise Total Kellogg's Advertise IGA Not Advertise 26 (25.85) 2 (2.15) 22 (22.15) ' 2 (1.85) 28 24 Total 48 4 52 Kellogg's Safeway Advertise Not Advertise Total Kellogg's Advertise Save-on Not Advertise 38 (36.92) 2 (3.08) 10 (11.08) 2 (0.92) 35 9 Total 48 4 52 Kellogg's Safeway Advertise Not Advertise Total Kellogg's Advertise Superstore Not Advertise 9 (10.15) 2 (0.85) 39 (37.85) 2 (3.15) 41 Total 48 4 52 Contingency Tables for Canned Fish Clover Leaf vs Private Label within Store 0\ Clover Leaf IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 6 (7.19) 5 (3.81) 28 (26.81) 13 (14.19) 11 41 Total 34 18 52 Clover Leaf Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 2 (2.40) 3 (2.60) 23 (22.60) 24 (24.40) 5 47 Total 25 27 52 Clover Leaf Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 2 (3.64) 19 (17.37) 7 (5.37) 24 (25.64) 21 31 Total 9 43 52 Clover Leaf Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 1 (0.77) 4 (4.23) 7 (7.23) 40 (39.77) 5 47 Total 8 44 52 Private Label vs Other National Brands within Store National Brands IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 7 (9.52) 4 (1.48) 38 (35.48) 3 (5.52) 11 41 Total 45 7 52 National Brands Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 2 (3.94) 3 (4.06) 39 (37.06) 8 (9.94) 5 47 Total 41 11 52 National Brands Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 14 (13.33) 7 (7.67) 19 (19.67) 12 (11.33) 21 31 Total 33 19 52 National Brands Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 1 (0.77) 4 (4.23) 7 7.23 40 (39.77) 5 47 Total 8 44 52 Clover Leaf vs Other National Brands within Store Clover Leaf IGA Advertise Not Advertise Total Other Advertise IGA Not Advertise 10 (13.73) 11 (7.27) 24 (20.27) 7 (10.73) 21 31 Total 34 18 52 Clover Leaf Safeway Advertise Not Advertise Total Other Advertise Safeway Not Advertise 12 (13.46) 16 (14.54) 13 (11.54) 11 (12.46) 28 24 Total 48 27 52 Clover Leaf Save-on Advertise Not Advertise Total Other Advertise Save-on Not Advertise 4 (4.85) 24 (23.15) 5 (4.15) 19 (19.85) 28 24 Total 9 43 52 Clover Leaf Superstore Advertise Not Advertise Total Other Advertise Superstore Not Advertise 0 (0.00) 0 (0.00) 8 (8.00) 44 (44.00) 0 52 Total 8 44 52 Clover Leaf vs Clover Leaf across Store Clover Leaf Safeway Advertise Not Advertise Total Clover Leaf Advertise IGA Not Advertise 19 (13.89) 15 (17.65) 6 (11.11) 12 (9.35) 34 18 Total 25 27 52 Clover Leaf Safeway Advertise Not Advertise Total Clover Leaf Advertise Save-on Not Advertise 4 (4.63) 5 (17.65) 21 (20.37) 22 (22.33) (14) (1) Total 25 27 52 Clover Leaf Safeway Advertise Not Advertise Total Clover Leaf Advertise Superstore Not Advertise 6 (1.85) 2 (4.67) 19 (23.15) 25 (22.85) 8 44 Total 25 27 52 Contingency Campbell vs Private Label within Store - J O Campbell IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 12 (10.50) 2 (3.50) 27 (28.50) 11 (9.50) 14 38 Total 39 13 52 Campbell Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 0 (0.00) 0 (0.00) 31 (31.00) 21 (21.00) 0 52 Total 31 21 52 Campbell Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 7 (6.69) 5 (5.31) 22 (22.31) 18 (17.69) 12 40 Total 29 23 52 Campbell Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 2 (0.67) 5 (6.33) 3 (4.33) 42 (40.67) 7 45 Total 5 47 52 for Canned Soup Campbell vs Other National Brands within Store Campbell IGA Advertise Not Advertise Total Other Advertise IGA Not Advertise' 7 (9.75) 6 (3.25) 32 (28.25) 7 (9.75) 13 39 Total 39 13 52 Campbell Safeway Advertise Not Advertise Total Other Advertise Safeway Not Advertise 6 (7.15) 6 (4.85) 25 (28.85) 15 (16.15) 12 40 Total 48 21 52 Campbell Save-on Advertise Not Advertise Total Other Advertise Save-on Not Advertise 5 (6.69) 7 (5.31) 24 (22.31) 16 (17.69) 12 40 Total 29 23 52 Campbell Superstore Advertise Not Advertise Total Other Advertise Superstore Not Advertise 1 (0.19) 1 (1.81) 4 (4.81) 46 (45.19) 2 50 Total 5 47 52 Private Label vs Other National Brands within Store National Brands IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 13 (8.89) 1 (5.12) 20 (24.12) 18 (13.89) 14 38 Total 33 19 52 National Brands Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 0 (0.00) 0 (0.00) 13 (13.00) 39 (39.00) 0 52 Total 13 39 52 National Brands Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 8 (9.69) 4 (2.31) 34 (32.31) 6 (7.69) 12 40 Total 42 10 52 National Brands Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 3 (1.48) 4 (5.52) 8 (9.52) 37 (35.48) 7 45 Total 11 41 52 Campbell vs Campbell across Store Campbell Safeway Advertise Not Advertise Total Campbell Advertise IGA Not Advertise 23 (23.25) 16 (15.75) 8 (7.75) 5 (5.25) 39 13 Total 31 21 52 Campbell Safeway Advertise Not Advertise Total Campbell Advertise Save-on Not Advertise 19 (17.29) 10 (11.71) 21 (13.71) 11 (9.29) 7 12 Total 40 21 61 - Campbell Safeway Advertise Not Advertise Total Campbell Advertise Superstore Not Advertise 4 (2.98) 1 (2.02) 27 (28.11) 20 (18.98) 5 47 Total 31 21 52 Contingency Tables for Canned Dry Pasta Catelli vs Private Label within Store Catelli IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 3 (5.19) 15 (12.81) 12 (9.81) 22 (24.19) 18 34 Total 15 37 52 Catelli Safeway . Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 0 (0.00) 0 (0.00) 19 (19.00) 33 (33.00) 0 52 Total 19 33 52 Catelli Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 3 (3.46) 17 (16.54) 6 (5.54) ' 26 (26.46) 20 32 Total 9 43 52 Catelli Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 0 (0.00) 7 (7.00) 0 (0.00) 45 (45.00) 7 45 Total 0 52 52 Private Label vs Other National Brands within Store National Brands IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 7 (11.08) 11 (6.92) 25 (20.92) 9 (13.08) 18 34 Total 32 20 52 National Brands Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 0 (0.00) 0 (0.00) 31 (31.00) 21 (21.00) 0 52 Total 31 21 52 National Brands Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 9 (10.39) 11 (9.62) 18 (16.62) 14 (15.39) 20 32 Total 27 25 52 National Brands Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 0 (0.54) 7 (6.46) 4 (3.46) 41 (41.54) 7 45 Total 4 48 52 Catelli vs Other National Brands within Store Catelli IGA Advertise Not Advertise Total Other Advertise IGA Not Advertise 0 (4.90) 17 (12.10) 15 (10.10) 20 (24.90) 17 35 Total 15 37 52 Catelli Safeway Advertise Not Advertise Total Other Advertise Safeway Not Advertise 3 (5.48) 12 (9.52) 16 (13.52) 21 (23.48) 15 37 Total 48 33 52 Catelli Save-on Advertise Not Advertise Total Other Advertise Save-on Not Advertise .2 (3.46) 18 (16.54) 7 (5.54) 25 (26.46) 20 32 Total 9 43 52 Catelli Superstore Advertise Not Advertise Total Other Advertise Superstore Not Advertise 0 (0.00) 4 (4.00) 0 (0.00) 48 (48.00) 4 48 Total 0 52 52 Catelli vs Catelli across Store Catelli Safeway Advertise Not Advertise Total Catelli Advertise IGA Not Advertise 6 (5.48) 9 (9.52) 13 (13.52) 24 (23.48) 15 37 Total 19 33 52 Catelli Safeway Advertise Not Advertise Total Catelli Advertise Save-on Not Advertise 4 (3.29) 5 (5.71) 15 (15.71) 28 (27.29) (2) (12) Total 19 33 52 Catelli Safeway Advertise Not Advertise Total Catelli Advertise Superstore Not Advertise 0 (0.00) 0 (0.00) 19 (19.00) 33 (33.00) 0 52 Total 19 33 52 Contingency Tables for Frozen Pizza Kraft vs Private Label within Store Kraft IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 0 (0.00) 0 (0.00) 20 (20.00) 32 (32.00) 0 52 Total 20 32 52 Kraft Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 2 (3.00) 10 (9.00) 11 (10.00) 29 (30.00) 12 40 Total 13 39 52 Kraft Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 4 (3.00) 8 (9.00) 9 (10.00) 31 (30.00) 12 40 Total 13 39 52 Kraft Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 2 (1.10) 17 (17.90) 1 (1.90) 32 (31.10) 19 33 Total 3 49 52 Kraft vs Other National Brands within Store Kraft IGA Advertise Not Advertise Total Other Advertise IGA Not Advertise 13 (11.54) 17 (18.46) 7 (8.46) 15 (13.54) 30 22 Total 20 32 52 Kraft Safeway Advertise Not Advertise Total Other Advertise Safeway Not Advertise 3 (6.50) 23 (19.50) 10 (6.50) 16 (19.50) 26 26 Total 48 39 52 Kraft Save-on Advertise Not Advertise Total Other Advertise Save-on Not Advertise 6 (8.25) 27 (24.75) 7 (4.75) 12 (14.25) 33 19 Total 13 39 52 Kraft Superstore Advertise Not Advertise Total Other Advertise Superstore Not Advertise 1 (0.35) 5 (5.65) 2 (2.65) 44 (43.35) 6 46 Total 3 49 52 Private Label vs Other National Brands within Store National Brands IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 0 (0.00) 0 (0.00) 37 (37.00) 15 (15.00) 0 52 Total 37 15 52 National Brands Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 8 (7.62) 4 (4.39) 25 (25.39) 15 (14.62) 12 40 Total 33 19 52 National Brands Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 7 (7.69) 3 (2.31) 33 (32.31) 9 (9.69) 10 42 Total 40 12 52 National Brands Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 3 (2.92) 16 (16.08) 5 (5.08) 28 (27.92) 19 33 Total 8 44 52 Kraft vs Kraft across Store Kraft Safeway Advertise Not Advertise Total Kraft Advertise IGA Not Advertise 5 (5.00) 15 (15.00) 8 (8.00) 24 (24.00) 20 32 Total 13 39 52 Kraft Safeway Advertise Not Advertise Total Kraft Advertise Save-on Not Advertise 5 (3.25) 8 (9.75) 8 (9.75) 31 (29.25) (5) (21) Total 13 39 52 Kraft Safeway Advertise Not Advertise Total Kraft Advertise Superstore Not Advertise 1 (0.75) 2 (2.25) 12 (12.25) 37 (36.75) 3 49 Total 13 39 52 Contingency Tables for Frozen Punch Beverage McCain vs Private Label within Store McCain IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 2 (1.89) 12 (12.12) 5 (5.12) 33 (32.89) 14 38 Total 7 45 52 McCain Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 0 (0.00) 22 (22.00) 0 (0.00) 30 (30.00) 22 30 Total 0 52 52 McCain Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 0 (0.81) 7 (12.25) 6 (5.19) 39 (39.81) 7 45 Total 6 46 52 McCain Superstore Total Advertise Not Advertise Private Label Advertise 2 (0.75) 11 (12.25) 13 Superstore Not Advertise 1 (2.25) 38 (36.75) 39 Total 3 49 52 Private Label vs Other National Brands within Store National Brands IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 8 (10.23) 6 (3.77) 30 (27.77) 8 (10.23) 14 38 Total 38 14 52 National Brands Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 5 (9.69) 16 (11.31) 19 (14.31) 12 (16.69) 21 31 Total 24 28 52 -National Brands Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 2 (4.04) 5 (2.96) 28 (25.96) 17 (19.04) 7 45 Total 30 22 52 National Brands Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 3 (1.75) 10 (11.25) 4 (5.25) 35 (33.75) 13 39 Total 7 45 52 McCain vs Other National Brands within Store McCain IGA Advertise Not Advertise Total Other Advertise IGA Not Advertise 2 (4.44) 31 (28.56) 5 (2.56) 14 (16.44) 33 19 Total 7 45 52 McCain Safeway Advertise Not Advertise Total Other Advertise Safeway Not Advertise 0 (0.00) 24 (24.00) 0 (0.00) 28 (28.00) 24 28 Total 48 52 52 McCain Save-on Advertise Not Advertise Total Other Advertise Save-on Not Advertise 1 (2.89) 24 (22.12) 5 (3.12) 22 (23.89) 25 27 Total 6 46 52 McCain Superstore Advertise Not Advertise Total Other Advertise Superstore Not Advertise 2 (0.35) 4 (5.65) 1 (2.65) 45 (43.35) 6 46 Total 3 49 52 McCain vs McCain across Store McCain Safeway Advertise Not Advertise Total McCain Advertise IGA Not Advertise 0 (0.00) 7 (7.00) 0 (0.00) 45 (45.00) 7 45 Total 0 52 52 McCain Safeway Advertise Not Advertise Total McCain Advertise Save-on Not Advertise 0 (0.00) 6 (6.00) 0 (0.00) 46 (46.00) (6) (46) Total 0 52 52 McCain Safeway Advertise Not Advertise Total McCain Advertise Superstore Not Advertise 0 (0.00) 3 (3.00) 0 (0.00) 49 (49.00) 3 49 Total 0 52 52 Contingency Tables for Frozen Vegetable Green Giant vs Private Label within Store Green Giant IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 2 (4.00) 14 (12.00) 11 (9.00) 25 (27.00) 16 36 Total 13 39 52 Green Giant Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 1 (5.50) . 21 (16.50) 12 (7.50) 18 (22.50) 22 30 Total 13 39 52 Green Giant Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 4 (3.60) 7 (7.40) 13 (13.40) 28 (27.60) 11 41 Total 17 35 52 Green Giant Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 0 (0.15) 2 (1.85) 4 (3.85) 46 (46.15) 2 50 Total 4 48 52 Green Giant vs Other National Brands within Store Green Giant IGA Advertise Not Advertise Total Other Advertise IGA Not Advertise 0 (0.00) 0 (0.00) 13 (13.00) 39 (39.00) 0 52 Total 13 39 52 Green Giant Safeway Advertise Not Advertise Total Other Advertise Safeway Not Advertise 0 (0.25) 1 (0.75) 13 (12.75) 38 (38.25) 1 51 Total 48 39 52 Green Giant Save-on Advertise Not Advertise Total Other Advertise Save-on Not Advertise 0 (0.33) 1 (0.67) 17 (16.67) 34 (34.33) 1 51 Total 17 35 52 Green Giant Superstore Advertise Not Advertise Total Other Advertise Superstore Not Advertise 0 (0.15) 2 (1.85) 4 (3.85) 46 (46.15) 2 50 Total 4 48 52 Private Label vs Other National Brands within Store National Brands IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 2 (4.00) 14 (12.00) 11 (9.00) 25 (27.00) 16 36 Total 13 39 52 National Brands Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 1 (2.96) 10 (8.04) 13 (11.04) 28 (29.96) 11 41 Total 14 38 52 National Brands Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 4 (3.81) 7 (7.19) 14 (14.19) 27 (26.81) 11 41 Total 18 34 52 National Brands Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 1 (0.32) 1 (1.77) 5 (5.77) 45 (44.23) 2 50 Total 6 46 52 Green Giant vs Green Giant across Store Green Giant Safeway Advertise Not Advertise Total Green Giant Advertise IGA Not Advertise 6 (3.25) 7 (9.75) 7 (9.75) 32 (29.25) 13 39 Total 13 39 52 Green Giant Safeway Advertise Not Advertise Total Green Giant Advertise Save-on Not Advertise 7 (4.25) 10 (12.75) 6 (8.75) 29 (26.25) 17 35 Total 13 39 52 Green Giant Safeway Advertise Not Advertise Total Green Giant Advertise Superstore Not Advertise 0 (1.00) 4 (3.00) 13 (12.00) 35 (36.00) 4 48 Total 13 39 52 Contingency Tables for Fruit Jam Kraft vs Private Label within Store Kraft IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 2 (2.31) 10 (9.69) 8 (7.69) 32 (32.31) 12 40 Total 10 42 52 - Kraft Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 0 (0.00) 0 (0.00) 6 (6.00) 46 (46.00) 0 52 Total 6 46 52 Kraft Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 1 (1.15) 19 (18.85) 2 (1.85) 30 (30.15) 20 32 Total 3 49 52 Kraft Superstore Total Advertise Not Advertise --4 Private Label Advertise 0 (0.50) 13 (12.50) 13 Superstore Not Advertise 2 (1.50) 37 (37.50) 39 Total 2 50 52 Private Label vs Other National Brands within Store National Brands IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 3 5.77 9 (6.23) 22 (19.23) 18 (20.77) 12 40 Total 25 27 ' 52 National Brands Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 0 0.00 0 (0.00) 12 (12.00) 40 (40.00) 0 52 Total 12 40 52 National Brands Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 9 (4.62) 11 (15.39) 3 (7.39) 29 (24.62) 20 32 Total 12 40 52 National Brands Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 1 (3.00) 12 (10.00) 11 (9.00) 28 (30.00) 13 39 Total 12 40 52 Kraft vs Other National Brands within Store Kraft IGA Advertise Not Advertise Total Other Advertise IGA Not Advertise 3 (3.46) 15 (14.54) 7 (6.54) 27 (27.46) 18 34 Total 10 42 52 Kraft Safeway Advertise Not Advertise Total Other Advertise Safeway Not Advertise 1 (0.81) 6 (6.19) 5 (5.19) 40 (39.81) 7 45 Total 48 46 52 Kraft Save-on Advertise Not Advertise Total Other Advertise Save-on Not Advertise 0 (1.67) 29 (27.33) 3 (1.33) 20 (21.67) 29 23 Total . 3 49 52 Kraft Superstore Advertise Not Advertise Total Other Advertise Superstore Not Advertise 0 (0.39) 10 (9.62) 2 (1.62) 40 (40.39) 10 42 Total 2 50 52 Kraft vs Kraft across Store Kraft Safeway Advertise Not Advertise Total Kraft Advertise IGA Not Advertise 0 (1.15) 10 (8.85) 6 (4.85) 36 (37.15) 10 42 Total 6 46 52 Kraft Safeway Advertise Not Advertise Total Kraft Advertise Save-on Not Advertise 0 (0.35) 3 (2.65) 6 (5.56) 43 (43.35) (3) (37) Total 6 46 52 Kraft Safeway Advertise Not Advertise Total Kraft Advertise Superstore Not Advertise 0 (0.23) 2 (1.77) 6 (5.77) 44 (44.23) 2 50 Total 6 46 52 Contingency Tables for Ground Coffee Maxwell vs Private Label within Store Maxwell IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 7 (6.19) 7 (7.81) 16 (16.81) 22 (21.19) 14 38 Total 23 29 52 Maxwell Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 0 (0.00) 0 (0.00) 4 (4.00) 48 (48.00) 0 52 Total 4 48 52 Maxwell. Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 1 (2.89) 9 (7.12) 14 (12.12) ' 28 (29.89) 10 42 Total 15 37 52 Maxwell Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 2 (3.00) 24 (23.00) 4 (3.00) 22 (23.00) 26 26 Total 6 46 52 Maxwell vs Other National Brands within Store Maxwell IGA Advertise Not Advertise Total Other Advertise IGA Not Advertise 6 (11.50) 20 (14.50) 17 (11.50) 9 (14.50) 26 26 Total 23 29 52 Maxwell Safeway Advertise Not Advertise Total Other Advertise Safeway Not Advertise 1 (1.69) 21 (20.31) 3 (2.31) 27 (27.69) 22 30 Total 48 48 52 Maxwell Save-on Advertise Not Advertise Total Other Advertise Save-on Not Advertise 7 (8.08) 21 (19.92) 8 (6.92) 16 (17.08) 28 24 Total 15 37 52 Maxwell Superstore Advertise Not Advertise Total Other Advertise Superstore Not Advertise 1 (0.23) 1 (1.77) 5 (5.77) 45 (44.23) 2 50 Total 6 46 52 Private Label vs Other National Brands within Store National Brands IGA Advertise . Not Advertise Total Private Label Advertise IGA Not Advertise 9 (11.58) 5 (2.42) 34 (31.42) 4 (6.58) 14 38 Total 43 9 52 National Brands Safeway Advertise - Not Advertise Total Private Label Advertise Safeway Not Advertise 0 (0.00) 0 (0.00) 25 (25.00) 27 (27.00) 0 52 Total 25 27 52 -National Brands Save-on Advertise . Not Advertise Total Private Label Advertise Save-on Not Advertise 3 (6.92) 7 (3.08) 33 (29.08) (9) (12.92) 10 24 Total 36 (2) 34 National Brands Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 2 (3.50) 24 (22.50) 5 (3.50) 21 (22.50) 26 26 Total 7 45 52 Maxwell vs Maxwell across Store Maxwell Safeway Advertise Not Advertise Total Maxwell Advertise IGA Not Advertise 3 (1.77) 20 (21.23) 1 (2.23) 28 (26;77) 23 29 Total 4 48 52 Maxwell Safeway Advertise Not Advertise Total Maxwell Advertise Save-on Not Advertise 1 (1.15) 14 (13.85) 3 (2.85) 34 (34.15) (13) (31) Total 4 48 52 Maxwell Safeway Advertise Not Advertise Total Maxwell Advertise Superstore Not Advertise 1 (0.46) 5 (5.54) 3 (3.54) 43 (42.46) 6 46 Total 4 48 52 Contingency Tables for Instant Coffee Maxwell vs Private Label within Store <1 Maxwell IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 0 (0.00) 0 (0.00) . 11 (11.00) 41 (41.00) 0 52 Total 11 41 52 Maxwell Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 0 (0.00) 0 (0.00) 6 (6.00) 46 (46.00) 0 52 Total 6 46 52 Maxwell Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 0 (0.00) 0 (0.00) 2 (6.00) 50 (50.00) 0 52 Total 2 50 52 Maxwell Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 0 (0.58) 6 (5.42) 5 (4.42) 41 (41.58) 6 46 Total 5 47 52 Maxwell vs Other National Brands within Store Maxwell IGA Advertise Not Advertise Total Other Advertise IGA Not Advertise 2 (3.81) 16 (14.19) 9 (7.19) 25 (26.81) 18 34 Total 11 41 52 Maxwell Safeway Advertise Not Advertise Total Other Advertise Safeway Not Advertise 3 (0.58) 2 (4.42) 3 (5.42) 44 (41.58) 5 47 Total 48 46 52 Maxwell Save-on Advertise NotAdvertise Total Other Advertise Save-on Not Advertise 0 (1.73) 15 (13.27) 6 (4.27) 31 (32.73) 15 37 Total 6 46 52 Maxwell Superstore Advertise Not Advertise Total Other Advertise Superstore Not Advertise 0 (0.39) 4 (3.62) 5 (4.62) 43 (43.39) 4 48 Total 5 47 52 Private Label vs Other National Brands within Store National Brands IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 0 (0.00) 0 (0.00) 27 (27.00) 25 (25.00) 0 52 Total 27 . 25 52 National Brands Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 0 0.00 0 (0.00) 8 (8.00) 44 (44.00) 0 52 Total 8 44 52 National Brands Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 0 (1.62) 4 (2.39) 21 (19.39) 27 (28.62) 4 48 Total 21 31 52 National Brands Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 0 (1.04) 6 (4.96) 9 (7.96) 37 (38.04) 6 46 Total 9 43 52 Maxwell vs Maxwell across Store Maxwell Safeway Advertise Not Advertise Total Maxwell Advertise IGA Not Advertise 2 (0.69) 5 (5.31) 4 (4.73) 41 (40.69) 7 45 Total 6 46 52 Maxwell Safeway Advertise Not Advertise Total Maxwell Advertise Save-on Not Advertise 1 (0.69) 5 (5.31) 5 (5.31) 41 40.69 (4) 46 Total 6 46 52 Maxwell Safeway Advertise . NotAdvertise Total Maxwell Advertise Superstore Not Advertise 0 (0.58) 5 (4.42) 6 (5.42) 41 (41.58) 5 47 Total 6 46 52 Contingency Tables for Sun Rype vs Private Label within Store Sun Rype IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 2 (2.08) 1 (0.92) 34 (33.92) 15 (15.08) 3 49 Total 36 16 52 Sun Rype Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 4 (5.89) 14 (12.12) 13 (1.12) 21 (22.89) 18 34 Total 17 35 52 Sun Rype Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 0 (0.00) 0 (0.00) 30 (30.00) 22 (22.00) 0 52 Total 30 22 52 Sun Rype Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 2 (0.62) 6 (7.39) 2 (3.39) 42 (40.62) 8 44 Total 4 48 52 Juice from Concentrate Sun Rype vs Other National Brands within Store Sun Rype IGA Advertise Not Advertise Total Other Advertise IGA Not Advertise 13 (13.15) 6 (5.85) 23 (22.85) 10 (10.15) 19 33 Total 36 16 52 Sun Rype Safeway Advertise Not Advertise Total Other Advertise Safeway Not Advertise 0 (2.94) 9 (6.06) 17 (14.06) 26 (28.94) 9 43 Total 48 35 52 Sun Rype Save-on Advertise Not Advertise Total Other Advertise Save-on Not Advertise 5 (4.62) 3 (3.39) 25 (25.39) 19 (18.62) 8 44 Total . 30 22 52 Sun Rype Superstore Advertise Not Advertise Total Other Advertise Superstore Not Advertise 3 (0.46) 3 (5.54) 1 (3.54) 45 (42.46) 6 46 Total 4 48 52 Private Label vs Other National Brands within Store National Brands IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 2 (2.42) 1 (0.58) 40 (39.58) 9 (9.42) 3 49 Total 42 10 52 National Brands Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 8 (9.00) 10 (9.00) 18 (17.00) 16 (17.00) 18 34 Total 26 26 52 National Brands Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 0 (0.00) 0 (0.00) 33 (33.00) 19 (19.00) 0 52 Total 33 19 52 National Brands Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 3 (1.08) 5 (6.92) 4 (5.92) 40 (38.08) 8 44 Total 7 45 52 Sun Rype vs Sun Rype across Store Sun Rype Safeway Advertise Not Advertise Total Sun Rype .Advertise IGA Not Advertise 11 11.77 25 (24.23) 6 (5.23) 10 (10.77) 36 16 Total 17 35 52 Sun Rype Safeway Advertise Not Advertise Total Sun Rype Advertise Save-on Not Advertise 12 (9.81) 18 (20.19) 5 (7.19) 17 (14.81) (8) (10) Total 17 35 52 Sun Rype Safeway Advertise Not Advertise Total Sun Rype Advertise Superstore Not Advertise 2 (1.31) 2 (2.69) 15 (15.69) 33 (32.31) 4 48 Total 17 35 52 Contingency Tables for Tropicana vs Private Label within Store Tropicana IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 4 (3.10) 3 (3.90) 19 (19.90) 26 (25.10) 7 45 Total 23 29 52 Tropicana Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 2 (1.77) 2 (2.23) 21 (21.23) 27 (26.77) 4 48 Total 23 29 52 Tropicana Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 3 (6.46) 13 (9.54) 18 (14.54) 18 (21.46) 16 36 Total 21 31 52 Tropicana Superstore -. Advertise NotAdvertise Total Private Label Advertise Superstore Not Advertise 0 (1.00) 13 (12.00) 4 (3.00) 35 (36.00) 13 39 Total 4 48 52 Private Label vs Other National Brands within Store National Brands IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 6 (5.92) 1 (1.08) 38 (38.08) 7 (6.92) 7 45 Total 44 8 52 National Brands Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 2 (2.46) 2 (1.54) 30 29.54 18 (18.46) 4 48 Total 32 20 52 National Brands Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 3 (8.62) 13 (7.39) 25 (19.39) 11 (16.62) 16 36 Total 28 24 52 National Brands Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 1 (1.75) 12 (11.25) 6 (5.25) 33 (33.75) 13 39 Total 7 45 52 Not from Concentrate Tropicana vs Other National Brands within Store Tropicana IGA Advertise Not Advertise Total Other Advertise IGA Not Advertise 6 (11.94) 21 (15.06) 17 (11.06) 8 (13.94) 27 25 Total 23 29 52 Tropicana Safeway Advertise Not Advertise Total Other Advertise Safeway Not Advertise 1 (4.42) 9 (5.58) 22 (18.58) 20 (23.42) 10 42 Total 48 29 52 Tropicana Save-on Advertise Not Advertise Total Other Advertise Save-on Not Advertise 2 (3.64) 7 (5.37) 19 (17.37) 24 (25.64) 9 43 Total 21 31 52 Tropicana Superstore Advertise Not Advertise Total Other Advertise Superstore Not Advertise 1 (0.31) 3 (3.69) 3 (3.69) 45 (44.31) 4 48 Total 4 48 52 Tropicana vs Tropicana across Store Tropicana Safeway Advertise Not Advertise Total Tropicana Advertise IGA Not Advertise 10 (10.17) 13 (12.83) 13 (12.83) 16 (16.17) 23 29 Total 23 29 52 Tropicana Safeway Advertise Not Advertise Total Tropicana Advertise Save-on Not Advertise 12 (9.29) 9 (11.71) 11 (13.71) 20 (16.17) 0 31 Total 23 29 52 Tropicana Safeway Advertise Not Advertise Total Tropicana Advertise Superstore Not Advertise 1 (1.85) 3 (2.15) 23 (22.15) 25 (25.85) 4 48 Total 24 28 52 Contingency Tables for Ketchup Heinz vs Private Label within Store Heinz IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 0 (3.69) 16 (12.31) 12 (8.31) 24 (27.69) 16 36 Total 12 40 52 Heinz Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 0 (0.00) 0 (0.00) 10 (10.00) 42 (42.00) 0 52 Total 10 42 52 Heinz Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 1 (0.00) 10 (8.67) 10 (10.00) 31 (32.33) 11 41 Total 11 41 52 Heinz Superstore Total Advertise Not Advertise OO Private Label Advertise 1 (1.56) 8 (7.44) 9 o Superstore Not Advertise 8 (7.44) 35 (35.56) 43 Total 9 43 52 Private Label vs Other National Brands within Store National Brands IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 0 (3.69) 16 (7.44) 12 (8.31) 24 (27.69) 16 36 Total 12 40 52 National Brands Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 0 (0.00) 0 (0.00) 10 (10.00) 42 (42.00) 0 52 Total 10 42 52 National Brands Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 1 (2.75) 10 (8.25) 12 (10.25) 29 (30.75) 11 41 Total 13 39 52 National Brands Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 1 (1.56) 8 (7.44) 8 (7.44) 35 (35.56) 9 43 Total 9 43 52 Heinz vs Other National Brands within Store Heinz IGA Advertise Not Advertise Total Other Advertise IGA Not Advertise 0 (0.00) 0 (0.00) 12 (12.00) 40 (40.00) 0 52 Total 12 40 52 Heinz Safeway Advertise Not Advertise Total Other Advertise Safeway Not Advertise 0 (0.00) 0 (0.00) 10 (10.00) 42 (42.00) 0 52 Total 48 42 52 Heinz Save-on Advertise Not Advertise Total Other Advertise Save-on Not Advertise 0 (0.42) 2 (1.58) 11 (10.58) 39 (39.42) 2 50 Total 11 41 52 Heinz Superstore Advertise Not Advertise Total Other Advertise Superstore Not Advertise 0 (0.00) 0 (0.00) 9 (9.00) 43 (43.00) 0 52 Total 9 ' 4 3 52 Heinz vs Heinz across Store Heinz Safeway Advertise Not Advertise Total Heinz Advertise IGA Not Advertise 3 (2.31) 9 (9.69) 7 (7.69) 33 (32.31) 12 40 Total 10 42 52 Heinz Safeway Advertise • Not Advertise Total Heinz Advertise Save-on Not Advertise 4 (2.12) 7 (8.89) 6 (7.89) 35 (33.12) (5) 41 Total 10 42 52 Heinz Safeway Advertise Not Advertise Total Heinz Advertise Superstore Not Advertise 2 (1.73) 7 (7.27) 8 (8.27) 35 (34.73) 9 43 Total 10 42 52 Contingency Tables for Margarine Canola Harvest vs Private Label within Store Canola Harvest IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 0 (0.42) 22 (21.58) 1 (0.58) ' 29 (29.42) 22 30 Total 1 51 52 Canola Harvest Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 0 (0.39) 2 (1.62) 10 (9.62) 40 (40.39) 2 50 Total 10 42 52 Canola Harvest Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 0 (0.62) 16 (15.39) 2 (1.39) 34 (34.62) 16 36 Total 2 50 52 Canola Harvest Superstore Total Advertise Not Advertise OO Private Label Advertise 2 (0.54) 12 (13.46) 14 Superstore Not Advertise 0 (1.46) 38 (36.54) 38 Total 2 50 52 Private Label vs Other National Brands within Store National Brands IGA . Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 5 (8.89) 17 (13.12) 16 (12.12) 14 (17.89) 22 30 Total 21 31 52 National Brands Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 1 (1.73) 1 (0.27) 44 (43.27) 6 (6.73) 2 50 Total 45 7 52 National Brands Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 8 (10.15) 8 (5.85) 25 (22.85) 11 (13.15) 16 36 Total 33 19 52 National Brands Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 2 (2.69) 12 (11.31) 8 (7.31) 30 (30.69) 14 38 Total 10 42 . 52 Canola Harvest vs Other National Brands within Store Canola Harvest IGA Advertise Not Advertise Total Other Advertise IGA Not Advertise 0 (0.39) 20 (19.62) 1 (0.62) 31 (31.39) 20 32 Total 1 51 52 Canola Harvest Safeway Advertise Not Advertise Total Other Advertise Safeway Not Advertise 2 (7.69) 38 (32.31) 8 (2.31) 4 (9.69) 40 12 Total 48 42 52 Canola Harvest Save-on Advertise Not Advertise Total Other Advertise Save-on Not Advertise 1 (1.23) 31 (30.77) 1 (0.77) 19 (19.23) 32 20 Total 2 50 52 Canola Harvest Superstore Advertise Not Advertise Total Other Advertise Superstore Not Advertise 0 (0.31) 8 (7.69) 2 (1.69) 42 (42.31) 8 44 Total 2 50 52 Canola Harvest vs Canola Harvest across Store i Canola Harvest Safeway Advertise Not Advertise Total Canola Harvest Advertise IGA Not Advertise 0 (0.19) 1 (0.81) 10 (9.81) 41 41.19 51 Total 10 42 52 Canola Harvest Safeway Advertise Not Advertise Total Canola Harvest Advertise Save-on Not Advertise 0 (0.39) 2 (1.62) 10 (9.62) 40 (40.39) (2) 50 Total 10 42 52 Canola Harvest Safeway Advertise Not Advertise Total Canola Harvest Advertise Superstore Not Advertise 1 (0.39) 1 (1.62) 9 (9.62) 41 (40.39) 2 50 Total 10 42 52 Contingency Tables for Mayonnaise Kraft vs Private Label within Store Kraft IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 10 (8.46) 1 (2.54) 30 (31.54) 11 (9.46) 11 41 Total 40 12 52 Kraft Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 0 (0.00) 0 (0.00) 26 (26.00) 26 (26.00) 0 52 Total 26 26 52 Kraft Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 0 (0.00) 0 (0.00) 12 (12.00) 40 (40.00) 0 52 Total 12 40 52 Kraft Superstore Total Advertise Not Advertise OO Private Label Advertise 0 (0.21) 1 (0.79) 1 to Superstore Not Advertise 11 (10.79) . 40 (10.21) 51 Total 11 41 52 Private Label vs Other National Brands within Store National Brands IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 10 (9.10) 1 (1.90) 33 (33.90) 8 (7.10) 11 41 Total 43 9 52 National Brands Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 0 (0.00) 0 (0.00) 37 (37.00) 15 (15.00) 0 52 Total 37 15 52 National Brands Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 0 (0.00) 0 (0.00) 28 (28.00) 24 (24.00) 0 52 Total 28 24 52 National Brands Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 0 (0.33) 1 (0.67) 17 (16.67) 34 (34.33) 1 51 Total 17 35 52 Kraft vs Other National Brands within Store Kraft IGA Advertise Not Advertise Total Other Advertise IGA Not Advertise 3 (4.62) 3 (1.39) 37 (35.39) 9 (10.62) 6 46 Total 40 12 52 Kraft Safeway Advertise Not Advertise Total Other Advertise Safeway Not Advertise 1 (6.00) 11 (6.00) 25 (20.00) 15 (20.00) 12 40 Total 48 26 52 Kraft Save-on • Advertise Not Advertise Total Other Advertise Save-on Not Advertise 1 (3.92) 16 (13.08) 11 (8.08) 24 (26.92) 17 35 Total 12 40 52 Kraft Superstore Advertise Not Advertise Total Other Advertise Superstore Not Advertise 1 (1.48) 6 (5.52) 10 (9.52) 35 (35.48) 7 45 Total 11 41 52 Kraft vs Kraft across Store Kraft Safeway Advertise Not Advertise Total Kraft Advertise IGA Not Advertise 20 (20.00) 20 (20.00) 6 (6.00) 6 (6.00) 40 12 Total 26 26 52 Kraft Safeway Advertise Not Advertise Total Kraft Advertise Save-on Not Advertise 6 (6.00) 6 (6.00) 20 (20.00) 20 (20.00) 0 40 Total 26 26 52 -Kraft Safeway Advertise Not Advertise Total Kraft Advertise Superstore Not Advertise 5 (5.50) 6 (5.50) 21 (20.50) 20 (20.50) 11 41 Total 26 26 52 Contingency Tables for Pasta Sauce Prego vs Private Label within Store Prego IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 3 (2.19) 16 (16.81) 3 (3.81) 30 (29.19) 19 33 Total 6 46 52 Prego Safeway Advertise Not Advertise Total Private Label Advertise Safewav Not Advertise 0 (0.12) 1 (0.89) 6 (5.89) 45 (45.12) 1 51 Total 6 46 52 Prego Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 1 (0.85) 21 (21.15) 1 (1.15) 29 (28.85) 22 30 Total 2 50 52 Prego Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 0 (0.00) 13 (13.00) 0 (0.00) 39 (39.00) 13 39 Total 0 ' 52 52 Private Label vs Other National Brands within Store National Brands IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise • 9 (12.79) 10 (6.21) ' 26 (22.21) 7 (10.79) 19 33 Total 35 17 52 National Brands Safeway Advertise Not Advertise Total Private Label Advertise . . Safeway Not Advertise 1 (0.65) 0 (0.35) 33 (33.35) 18 (17.65) 1 51 Total 34 18 52 National Brands Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 14 (14.81) 8 (7.19) 21 (20.19) 9 (9.81) 22 30 Total 35 17 52 National Brands Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 1 (1.75) 12 (11.25) 6 (5.25) 33 (33.75) 13 39 Total 7 45 52 Prego vs Other National Brands within Store Prego IGA Advertise Not Advertise Total Other Advertise IGA Not Advertise 3 (3.69) 29 (28.31) 3 (2.31) 17 (17.69) 32 20 Total 6 46 52 Prego Safeway Advertise Not Advertise Total Other Advertise Safeway Not Advertise 0 (3.81) 33 (29.19) 6 (2.19) 13 (16.81) 33 19 Total 48 46 52 Prego Save-on Advertise Not Advertise Total Other Advertise Save-on Not Advertise 1 (1.31) 33 (32.69) 1 (0.69) 17 (17.31) 34 18 Total 2 50 52 Prego Superstore Advertise Not Advertise Total Other Advertise Superstore Not Advertise 0 (0.00) 7 (7.00) 0 (0.00) 45 (45.00) 7 45 Total 0 52 52 Prego vs Prego across Store Prego Safeway Advertise Not Advertise Total Prego Advertise-IGA Not Advertise 2 (0.69) 4 (5.31) 4 (5.31) 42 (40.69) 6 46 Total 6 46 52 Prego Safeway Advertise Not Advertise Total Prego Advertise Save-on Not Advertise 1 (0.23) 1 (1.77) 5 (5.77) 45 (44.23) 2 50 Total 6 46 52 Prego Safeway Advertise Not Advertise Total Prego Advertise Superstore Not Advertise 0 (0.00) 0 (0.00) 6 (6.00) 46 (46.00) 0 52 Total 6 46 52 Contingency Tables for Peanut Butter Kraft vs Private Label within Store OO Kraft IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 7 (8.77) 12 (10.23) 17 (15.23) 16 (17.77) 19 33 Total 24 28 52 Kraft Safeway Advertise NotAdvertise Total Private Label Advertise Safeway Not Advertise 0 (1.85) 4 (2.15) 24 (22.15) 24 (25.85) 4 48 Total 24 28 52 Kraft Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 2 (4.62) 22 (19.37) 8 (5.39) 20 (22.62) 24 28 Total 10 42 52 Kraft Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 1 (0.69) 17 (17.31) 1 (1.31) 33 (32.69) 18 34 Total 2 50 52 Private Label vs Other National Brands within Store National Brands IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 7 (10.60) 12 (8.40) 22 (18.40) 11 (14.60) 19 33 Total 29 23 52 National Brands Safeway Advertise NotAdvertise Total Private Label Advertise Safeway Not Advertise 0 (2.69) 4 (1.31) 35 (32.31) 13 (15.69) 4 48 Total 35 17 52 National Brands Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 2 (9.23) 22 (14.77) 18 (10.77) 10 (17.23) 24 28 Total 20 32 52 National Brands Superstore Advertise Not Advertise Total Private Label Advertise Superstore ' Not Advertise 1 (2.42) 17 (15.58) 6 (4.58) 28 (29.42) 18 34 Total 7 45 52 Kraft vs Other National Brands within Store Kraft IGA Advertise Not Advertise Total Other Advertise IGA Not Advertise 4 (4.15) 5 (4.85) 20 (19.85) 23 (23.15) 9 43 Total 24 28 52 Kraft Safeway Advertise Not Advertise Total Other Advertise Safeway Not Advertise 0 (5.08) 11 (5.92) 24 (18.92) 17 (22.08) 11 41 Total 48 28 52 Kraft Save-on Advertise Not Advertise Total Other Advertise Save-on Not Advertise 0 (1.92) 10 (8.08) 10 (8.08) 32 (33.92) 10 42 Total 10 42 52 Kraft Superstore Advertise Not Advertise Total Other Advertise Superstore Not Advertise 1 (0.23) 5 (5.77) 1 (1.77) 45 (44.23) 6 46 Total 2 50 52 Kraft vs Kraft across Store Kraft Safeway Advertise Not Advertise Total Kraft Advertise IGA Not Advertise 4 (1.85) 0 (2.15) 20 (22.15) 28 (25.85) 4 48 Total 24 28 52 Kraft Safeway Advertise Not Advertise Total Kraft Advertise Save-on Not Advertise 4 (4.62) 6 (5.39) 20 (19.39) 22 (22.62) 10 42 Total 24 28 52 Kraft Safeway Advertise Not Advertise Total Kraft Advertise Superstore Not Advertise 1 (0.92) 1 (1.08) 23 (23.08) 27 (26.92) 2 50 Total 24 28 52 Contingency Tables for Potato Chips Lay's vs Private Label within Store OO Lay's IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 1 (2.77) 11 (9.23) 11 (9.23) 29 (30.77) 12 40 Total 12 40 52 Lay's Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 1 (2.56) 6 (4.44) 18 (16.44) 27 (28.56) 7 45 Total 19 33 52 Lay's Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 2 (4.04) 13 (10.96) 12 (9.96) 25 (27.04) 15 37 Total 14 38 52 Lay's Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 0 (0.37) 19 (18.64) 1 (0.64) 32 (32.37) 19 33 Total 1 51 52 Private Label vs Other National Brands within Store National Brands IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 6 (9.23) 6 (2.77) 34 (30.77) 6 (9.23) 12 40 Total 40 12 52 National Brands Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 3 (4.31) 4 (2.96) 27 (25.%) 18 (19.04) 7 45 Total 30 22 52 National Brands Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 6 (8.65) 9 (6.35) 24 (21.35) 13 (15.65) 15 37 Total 30 22 52 National Brands Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 1 (1.10) 18 (17.90) 2 (1.90) 31 (31.10) 19 33 Total 3 49 52 Lay's vs Other National Brands within Store Lay's IGA Advertise Not Advertise Total Other Advertise IGA Not Advertise 6 (7.85) 28 (26.15) 6 (4.15) 12 (13.85) 34 18 Total 12 40 52 Lay's Safeway Advertise Not Advertise Total Other Advertise Safeway Not Advertise 5 (5.85) 11 (10.15) 14 (13.15) 22 (22.85) 16 36 Total 48 33 52 Lay's Save-on Advertise Not Advertise Total Other Advertise Save-on Not Advertise 3 (5.12) 16 (13.89) 11 (8.89) 22 (24.12) 19 33 Total 14 38 52 Lay's Superstore Advertise Not Advertise Total Other - • Advertise Superstore- Not Advertise 0 (0.04) 2 (1.96) 1 (0.96) 49 (49.04) 2 50 Total 1 51 52 Lay's vs Lay's across Store Lay's Safeway Advertise Not Advertise Total Lay's Advertise IGA Not Advertise 6 (4.39) - 6 (7.62) 13 (14.62) 27 (25.39) 12 40 Total 19 33 52 Lay's Safeway Advertise Not Advertise Total Lay's Advertise Save-on Not Advertise 8 (5.12) 6 (8.89) 11 (13.89) 27 (24.12) 14 38 Total 19 33 52 Lay's Safeway. Advertise Not Advertise Total Lay's Advertise Superstore Not Advertise 1 (0.37) 0 (0.64) 18 (18.64) 33 (32.37) 51 Total 19 33 52 Contingency Tables for Processed Cheese Slices Kraft vs Private Label within Store OO ON Kraft IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 3 (3.46) 7 (6.54) 15 (14.54) 27 (27.46) 10 42 Total 18 34 52 Kraft Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 1 (4.15) 5 (1.85) 35 (31.85) 11 (14.15) 6 46 Total 36 16 52 Kraft Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 0 (1.85) 8 (6.15) 12 (10.15) 32 (33.85) 8 44 Total 12 40 52 Kraft Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 0 (0.31) 16 (15.69) 1 (0.69) 35 (35.31) 16 36. Total 1 51 52 Private Label vs Other National Brands within Store National Brands IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 6 (5.96) 4 (4.04) 25 (25.04) 17 (16:96) 10 42 Total 31 21 52 National Brands Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 2 (4.50) 4 (1.50) 37 (34.50) 9 (11.50) 6 46 Total 39 13 52 National Brands Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 0 (3.23) 8 (4.77) 21 (17.77) 23 (26.23) 8 44 Total 21 31 52 National Brands Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 0 (0.31) 16 (15.69) 1 (0.69) 35 (35.31) 16 36 Total 1 51 52 Kraft vs Other National Brands within Store Kraft IGA Advertise Not Advertise Total Other Advertise IGA Not Advertise 6 (6.58) 13 (12.42) 12 (11.42) 21 (21.58) 19 33 Total 18 34 52 Kraft Safeway Advertise Not Advertise Total Other Advertise Safeway Not Advertise 3 (4.15) 3 (1.85) 33 (31.85) 13 (15.15) 6 46 Total 48 16 52 Kraft Save-on Advertise Not Advertise Total Other Advertise Save-on Not Advertise 1 (2.31)' 9 (7.69) 11 (9.69) 31 (32.31) 10 42 Total 12 40 52 Kraft Superstore Advertise Not Advertise Total Other Advertise Superstore Not Advertise 0 (0.00) 0 (0.00) 1 (1.00) 51 (51.00) 0 52 Total 1 51 52 Kraft vs Kraft across Store Kraft Safeway Advertise Not Advertise Total Kraft Advertise IGA Not Advertise 12 (12.46) 6 (5.54) 24 (23.54) 10 (10.46) 18 34 Total 36 16 52 Kraft Safeway Advertise Not Advertise Total Kraft Advertise Save-on Not Advertise 8 (8.31) 4 (3.69) 28 (27.69) 12 (12.31) 12 40 Total 36 16 52 Kraft Safeway Advertise Not Advertise Total Kraft Advertise Superstore Not Advertise 1 (0.37) 0 (0.31) 35 (35.31) 16 (15.69) 1 51 Total 36 16 52 Contingency Tables for Tea Bags Tetley vs Private Label within Store Tetley IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 3 (5.12) 16 (13.89) . 11 (8.89) 22 (24.12) 19 33 Total 14 38 52 Tetly Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 1 (1.08) 7 (6.92) 6 (5.92) 38 (38.08) 8 44 Total 7 45 52 Tetly Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 0 (1.54) 4 (2.46) 20 (18.46) 28 (29.54) 4 48 Total 20 32 52 Tetly Superstore Total Advertise Not Advertise OO Private Label Advertise 0 (0.69) 9 (8.31) 9 Superstore Not Advertise 4 (3.31) 39 (39.69) 43 Total 4 48 52 Private Label vs Other National Brands within Store National Brands IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 6 (12.42) 13 (6.58) 28 (21.58) '.. 5 (11.42) 19 33 Total 34 18 52 National Brands Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 1 (2.15) 7 (5.85) 13 (11.85) 31 (32.15) 8 44 Total 14 38 52 National Brands Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 1 (3.23) 3 (0.77) 41 (38.77) 7 (9.23) 4 48 Total 42 10 52 National Brands Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 1 (2.08) 8 (6.92) 11 (9.92) 32 (33.08) 9 43 Total 12 40 52 Tetley vs Other National Brands within Store Tetly IGA Advertise Not Advertise Total Other Advertise IGA Not Advertise 0 (5.39) 20 (14.62) 14 (8.62) 18 (23.39) 20 32 Total 14 38 52 Tetly Safeway Advertise Not Advertise Total Other Advertise Safeway Not Advertise 4 (1.89) 10 (12.12) 3 (5.12) 35 (32.89) 14 38 Total 48 45 52 Tetly Save-on Advertise Not Advertise Total Other Advertise Save-on Not Advertise 9 (11.92) 22 (19.08) 11 (8.08) 10 (12.92) 31 21 Total 20 32 52 Tetly Superstore Advertise Not Advertise Total Other Advertise Superstore Not Advertise . 1 (0.69) 8 (8.31) 3 (3.31) 40 (39.69) 9 43 Total 4 48 52 Tetly vs Tetly across Store Tetly Safeway Advertise Not Advertise Total Tetly Advertise IGA Not Advertise 3 (1.89) 11 (12.12) 4 (5.12) 34 (32.89) 14 38 Total 7 45 52 Tetly Safeway Advertise Not Advertise Total Tetly Advertise Save-on Not Advertise 2 (2.69) 18 (17.31) 5 (4.31) 27 (27.69) 20 32 Total 7 45 52 Teuy Safeway Advertise Not Advertise Total Tetly Advertise Superstore Not Advertise 0 (0.54) 4 (3.46) 7 (6.46) 41 (41.54) 4 48 Total 7 45 52 Contingency Tables for Waffle Kellogg's vs Private Label within Store Kellogg's IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 1 2.69 9 (7.31) 13 (11.31) 29 (30.69) 10 42 Total 14 38 52 Kellogg's Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 3 (5.85) 13 (10.15) 16 (13.15) 20 (22.87) 16 36 Total 19 33 52 Kellogg's Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 4 (6.87) 17 (14.14) 13 (10.14) 18 (20.87) 21 31 Total 17 35 52 Kellogg's Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 1 (0.39) 3 (3.62) 4 (4.62) 44 (43.39) 4 48 Total 5 47 52 Kellogg's vs Other National Brands within Store Kellogg's IGA Advertise Not Advertise Total Other Advertise IGA Not Advertise 0 (0.00) 0 (0.00) 14 (14.00) 38 (38.00) 0 52 Total 14 38 52 Kellogg's Safeway Advertise Not Advertise Total Other Advertise Safeway Not Advertise 1 (0.73) 1 (1.27) 18 (18.27) 32 (31.73) 2 50 Total 48 33 52 Kellogg's Save-on Advertise Not Advertise Total Other Advertise Save-on Not Advertise 2 (0.98) I (2.02) 15 16.02 34 (32.98) 3 49 Total 17 35 52 Kellogg's Superstore Advertise Not Advertise Total Other Advertise Superstore Not Advertise 0 (0.00) 0 (0.00) 5 (5.00) 47 (47.00) 0 52 Total 5 47 52 Private Label vs Other National Brands within Store National Brands IGA Advertise Not Advertise Total Private Label Advertise IGA Not Advertise 1 (2.69) 9 (7.31) 13 (11.31) 29 (30.69) 10 42 Total 14 38 52 National Brands Safeway Advertise Not Advertise Total Private Label Advertise Safeway Not Advertise 3 (6.15) 13 (9.85) 17 (13.85) 19 (22.15) 16 36 Total 20 32 52 National Brands Save-on Advertise Not Advertise Total Private Label Advertise Save-on Not Advertise 5 (7.27) 16 (13.73) 13 (10.73) 18 (20.17) 21 31 Total 18 34 52 National Brands Superstore Advertise Not Advertise Total Private Label Advertise Superstore Not Advertise 1 (0.39) 3 (3.62) 4 (4.62) 44 (43.39) 4 48 Total 5 47 52 Kellogg's vs Kellogg's across Store Kellogg's Safeway Advertise Not Advertise Total Kellogg's . Advertise IGA Not Advertise 8 (5.12) 6 (8.89) 11 (13.89) 27 (24.12) 14 38 Total 19 33 52 Kellogg's Safeway Advertise Not Advertise Total Kellogg's Advertise Save-on Not Advertise 7 (6.21) 10 (10.79) 12 (12.79) 23 (22.21) 17 35 Total 19 33 52 Kellogg's Safeway Advertise Not Advertise Total Kellogg's Advertise Superstore Not Advertise 2 (1.83) 3 (3.17) 17 (17.17) 30 (29.83) 5 47 Total 19 33 52 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0091965/manifest

Comment

Related Items