Embed Size (px)
Citation preview
Ram Bezawada & Koen Pauwels
What Is Special About MarketingOrganic Products? How Organic
Assortment, Price, and PromotionsDrive Retailer Performance
Higher sales and margins are key goals for retailers promoting emerging products, such as organics, but little isknown about their marketing effectiveness and their cross-effects on conventional product sales. Extant researchreports conflicting results about price and promotional sensitivity for organic products and does not address theimpact of organic assortment. This article calculates long-term own- and cross-elasticities of organic andconventional product sales in response to changes in assortment, price, and promotions. Using a rich data set of56 categories, the authors test hypotheses on how different costs and benefits of organic products affect theseelasticities. They find that enduring actions, such as assortment and regular price changes, have a higher elasticityfor organics than for conventional products. In contrast with common wisdom, even "core" organic consumers aresensitive to these actions. Increasing organic assortment and promotion breadth yields higher profits for the totalcategory, as do more frequent promotions on conventional products. The category comparison yields specificadvice with regard to where larger assortment and lower prices versus more and deeper promotions are mosteffective.
Keywords: organic products, food mari<eting, empirical generalizations, cross-category, mari<eting mix, vectorautoregressive models
Faced with intense competition and razor-thin marginson mature products, retailers are constantly searchingfor the "next big thing"—that is, groups of products
that attract customers to the store and also generate highermargins (Ceslfo 2011). Such "emerging" product groupsinclude deli, ready-to-serve entrées, health and Wellnessproducts (e.g., food supplements, weight loss bars), organicand natural foods (e.g., organic milk, natural yogurt), andprivate labels (Beverage Industry 2010; Drug Store News2008; Food Marketing Institute 2009). Often, these productgroups are first bought by a small group of devoted cus-tomers and then spread to the general shopper population.This creates challenges for the retailer because emergingand mature products are often substitutes in the same cate-
Ram Bezawada is Assistant Professor of Marketing, School of Manage-ment, State University of New York, Buffalo (e-mail: bezawada ©buffalo,edu). Koen Pauwels is Professor of Marketing, Graduate School of Busi-ness, Özyegin University (e-mail: [email protected]). Theauthors thank the managers at two large northeastern United States retailchains for providing valuable insights into the issues addressed in thisarticle. They also thank the Center for Relationship Marketing at theSchool of Management and the Baldy Center for Law and Social Policy atthe State University of New York, Buffalo, for the data and support. Theythank the anonymous JM reviewers and seminar participants at the Uni-versity of Groningen and the 2008 BBCRST and INFORMS 2009 Market-ing Science conferences for constructive comments. Bob Leone servedas area editor of this article.
gory. Apart from assessing the effectiveness of emergingcategories for their marketing programs, retailers need tounderstand the intracategory cross-effects of promotionactivities and their impact on overall category and store per-formance (Progressive Grocer 2008).
In today's marketplace, many retailers perceive a keyopportunity in organic products, whose U.S. sales havegrown I7%-21% each year, compared with 2%-\% growthin nonorganic (hereinafter "conventional") product sales(Progressive Grocer 2009). The Great Recession has notdampened this growth (Brandweek 2009), which may befurthered by the appointment of an organics expert to theU.S. Agriculture Department's number-two post with a bud-get allocation of $50 million specifically to fund neworganic initiatives. The majority of U.S. consumers eatorganic products at least occasionally, and organic productsare now available in more than 70% of traditional super-markets, such as Kroger and Safeway (The Hartman Group2008).
Retailers hope that promoting organic products willincrease total category margins and store revenues, in addi-tion to enhancing stores' long-term image, equity, and dif-ferentiated positioning (Chain Store Age 2009). Our inter-views with retail managers of two large northeastern UnitedStates supermarkets revealed their belief that organic prod-ucts will become more established, thus generating tangiblebenefits for retailers that are willing to invest in them. How-ever, key questions remain as to where such investments
© 2013, American Marketing AssociationISSN: 0022-2429 (print), 1547-7185 (electronic) 31
Journal oi MarketingVolume 77 (January 2013), 31-51
pay off most and how they affect conventional product salesand retailers' category and store performance.
Current marketing literature is rich in how consumersmake trade-offs among different conventional products in acategory and how price and promotions affect such trade-offs (e.g., Bijmolt, Van Heerde, and Pieters 2005; Sethura-man and Srinivasan 2002). However, retailers are unsureabout how these general findings apply to organic products,given the mixed evidence on price elasticity (from -9.73 inGlaser and Thompson [2000] to -.001 in Kiesel and Villas-Boas [2007]), the surprising recent findings on promotionalelasticity (negative in Ngobo [2011]), and the absence ofresearch on how organic assortment benefits organic sales,category margin, and store revenues. Conceptually, somestudies predict higher own marketing elasticities for organ-ics because of the high price premium over conventionalproducts (Glaser and Thompson 2000; Verhoef 2005). Incontrast, Ngobo (2011) postulates lower own marketingelasticities (even of an opposite sign to conventional prod-ucts) because consumers associate low prices and promo-tions with low-quality and "popular" products, jeopardizingthe special status of organics. Indeed, research has not evenestablished that cross-elasticities with conventional prod-ucts are asymmetric in favor of (higher-priced) organicproducts (e.g., Blattberg and Wisniewski 1989; Sethuramanand Srinivasan 2002). Violating this general rule, the onlyeconometric analysis on the subject reports asymmetry infavor of conventional products (Glaser and Thompson2000). Finally, survey-based research indicates that con-sumer response to organic product marketing may differ bycategory and consumer segment. What is lacking is a large-scale study of what this means for the effectiveness of mar-keting organic products in driving organic sales and retailerperformance.
Why might organics be "special" compared with otherexpensive products? Consumers state different motivationsfor buying and consuming organics, such as health, envi-ronment, and animal welfare concerns (Bourn and Prescott2002; Fotopoulos and Krystallis 2002; Makatouni 2002;Zanoli and Naspetti 2002). Consumers have also expressedskepticism whether these motivations can be fulfilled inmainstream supermarket chains, and researchers have ques-tioned the use of traditional marketing actions to promoteorganics (Ngobo 2011). In this context, our specificresearch questions are as follows:
1. What is the long-term own-effect of assortment, regularprice, discount breadth and depth, and price specials fororganic products versus conventional products?
2. How does the marketing of organic products stimulate pur-chases across different levels of consumer organic usage(i.e., "core" organic vs. "noncore")?
3. How large are the cross-effects of organic product market-ing activities on conventional product sales, and vice versa?
4. Which types of conventional products (i.e., top-tier andsecond-tier national brands and private labels) are afïectedthe most by marketing actions of organics, and vice versa?
5. What is the effect of marketing organic products on cate-gory and store performance?
On the basis of the perceived benefits and costs oforganics, we propose that enduring retail actions (assort-ment and regular price) generate higher consumer responsefor organics than for conventional products, but temporaryactions do not. Our analysis across 56 categories identifiesand quantifies how consumer response differs, which yieldsconcrete insight for retailers into where and how to devotetheir marketing resources to increase category and store per-formance. In contrast with recent advice that retailers shouldkeep organic prices high and avoid point-of-purchase pro-motions (e.g., Ngobo 2011), we find that organic salesincrease strongly with lower regular prices, even for con-sumers with high intrinsic value for organics (the coreorganic segment). We also find substantial benefits ofincreasing organic assortment to overall category margin,especially in produce categories. In contrast with Glaserand Thompson (2000), we find that (price) promotions ofconventional products do little harm to organic productsales, thus offering specific guidelines to retailers on how tostrike a balance between emerging and mature products.
More generally, this article contributes to the burgeon-ing literature on the marketing and consumer adoption ofsustainable/ethical products (e.g., Henderson and Arora2010). Recently, issues pertaining to sustainability havereceived considerable attention not only from governmentalagencies (e.g., U.S. Environmental Protection Agency greenproduct programs) but also from firms (e.g., Clorox GreenWorks), which are investing considerable resources into thedesign and marketing of products or initiatives that createlong-term societal value (e.g.. Kotier 2011). Thus, implica-tions from our research are germane to the design of pro-grams that influence public policy, resource management,and health behavior.
Research BackgroundOrganic Food Products at Conventional RetailOutletsCurrently, consumers in the United States buy more organicproducts in traditional supermarkets than in other outlets(TABS 2012). At the same time, traditional supermarketsare increasingly promoting organic products through vari-ous in-store marketing programs (e.g., increasing variety,displays). Because organics have higher gross margins—30% to 50% versus 20% to 25% for conventional products(Oberholtzer, Green, and Lopez 2006; Roheim and D'Silva2009) —promoting them should enhance total categoryprofits and store revenues. However, academic literaturehas yet to verify such performance effects of marketingactions for organics because it has focused instead on othersupply-side and demand-side issues (Thompson 1998).
On the supply side, Dimitri and Oberholtzer (2009) findthat organic fanners sometimes struggle to provide suffi-cient supply to keep up with the rapid growth in demand,and Ciu (2008) reports that some farmers have struggled toobtain the necessary certification to market produce asorganic. Finally, Tondel and Woods (2006) find that organicsupply is becoming more competitive and efficient, lower-ing prices throughout the supply chain.
32 / Journal of Marketing, January 2013
On the demand side, previous research falls into threebroad categories: (1) self-report surveys and interviews thatuncover the motivations for consumers to buy organic prod-ucts, and the category factors that favor organic adoption,(2) studies on product health claims and labeling, and (3)econometric analyses of how individual household charac-teristics and retail prices affect panelist demand for organicproducts and their reaction to marketing for organic prod-ucts. We discuss these in turn.
Why Do Consumers Buy Organic VersusConventional Products?
Motivations for buying organics include health reasons,environmental concerns, nutritional value, and taste (e.g..Bourn and Prescott 2002; Fotopoulos and Krystallis 2002;Zanoli and Naspetti 2002) as well as considerations regard-ing ethics and animal welfare (Makatouni 2002). Some con-sumers also acknowledge that social approval plays a rolein their purchase of organic products (Grunert and Juhl1995). Self-reported obstacles inhibiting organic productpurchase are their low availability and distribution, theirprice premium, and consumer lack of knowledge (Bonti-Ankomah and Yiridoe 2006). Consumers often begin withorganics in categories such as produce, meat, and dairy, forwhich they perceive higher benefits from going organic(Oberholtzer, Green, and Lopez 2006; Organic Trade Asso-ciation 2009a).
How Does Product Labeiing Affect Consumers'Responses to Organic Products?Studies in marketing have analyzed the role of product/nutrition claims in consumer food choices. Kozup, Creyer,and Burton (2003) report that consumers' positive attitudestoward products are enhanced when favorable nutritional(e.g., fortified with vitamins) or health (e.g., heart-healthy)claims appear on the packaging. In the context of our study,organic product labeling and certification logos have beenshown to play an important role in stimulating consumerappeal for organics. Using Rokeach's (1968) theory ofvalue and halo effects (Han 1989), Bauer, Heinrich, andSchäfer (2012) report that organic labeling results in ahigher level of perceived healthfulness, hedonism, environ-mental friendliness, and food safety. Janssen and Hamm(2012) hypothesize that because organic products are cre-dence goods, a high degree of uncertainty is associated withthem, and appropriate labeling might mitigate this uncer-tainty. Given these findings, third-party certification issuperior because consumers have greater trust in indepen-dent certifiers than private manufacturers. However, con-sumers do not perceive all types of labels to be the same.Generic organic labels, which typically list the word"organic" on either the brand or the product description, donot elicit the same kind of trust that organic certificationlogos do (e.g., the U.S. Department of Agriculture [USDA]seal). Moreover, Janssen and Hamm (2012) find that well-known and trusted certification logos command the highestprice premiums. Similar findings are reported by Kiesel andVillas-Boas (2007), who find that consumer response in themilk category is higher for certification (USDA) logos than
for organic labels or other markers (e.g., rBGH-free [i.e.,free of the recombinant bovine growth hormone]), espe-cially after the National Organic Program went into effect.In this study, we consider organic products with the USDAseal (certification logo) and organic products without theUSDA seal but with generic organic labels on the packages.
How Do Consumers React to Retail MarketingActions for Organic Products?A handful of studies use revealed data (typically scannerpanel) to analyze how organic consumers react to retailprices. Glaser and Thompson (2000) report large price elas-ticity (between -3.63 and -9.73) for U.S. organic milk inthe late 1990s. In contrast, Kiesel and Villas-Boas (2007)report small price elasticity (between -.001 and -.003) forU.S. organic milk in the 2000s. In the most recent study,based on French data, Ngobo (2011) finds that lower pricesand wide distribution make shoppers less likely to buyorganics up to a point; Ngobo concludes that organic prod-ucts may be a poor fit for the typical marketing actions oftraditional retailers. In summary, the magnitude, and eventhe sign, of organic price elasticity remains an empiricalquestion.
An issue with these studies is their representativenessfor all the shoppers at a mainstream retailer. Relying on apanel of households, they further restrict the panel toaccount for the paucity of organic purchase observations asopposed to the conventional ones. Thus, they focus on thecore organic consumer segment while ignoring the noncoresegment, whose purchase of organics represents a keyopportunity and challenge for retailers. Moreover, most ofthe previous studies analyze a few, mostly similar cate-gories. Finally, they do not analyze the effects of increasingorganic assortment, which is a relatively costly and endur-ing decision for retailers. These limitations impede action-able insights into what marketing actions retailers canundertake and in which categories to increase overall retailperformance (e.g., by increasing organic sales withoutdecreasing conventional sales). This article contributes tothis research stream by quantifying the long-term own- andcross-elasticities of organics and conventional productgroups using store data across 56 categories spanning sevenyears. Next, we develop our hypotheses.
Hypotheses DevelopmentWe develop our hypotheses on the basis of consumers' per-ceived benefits and costs of buying organic versus conven-tional products. Perceived benefits of buying organicsinclude health, nutritional value, taste, animal welfare,ethics, and environmental protection (e.g.. Bourn andPrescott 2002; Fotopoulos and Krystallis 2002; Makatouni2002; Zanoli and Naspetti 2002). However, buying organicproducts represents a cost to a mainstream retailer's con-sumers because organic products are typically (1) moreexpensive than conventional products and (2) more difficultto find in the exact form, fiavor, and quantity the consumerprefers (Michelsen et al. 1999). The former represents anout-of-pocket monetary cost to the consumer, and the latterdenotes a transaction cost. Our key assertion is that con-
What Is Special About Marketing Organic Products? / 33
sumers weigh the potential benefits of organics by thesecosts, which are likely to endure throughout their futurepurchases of organic products. Indeed, assortments andregular prices are "sticky" compared with temporaryactions such as displays, features, and promotions (Pauwels2004). This assertion is grounded in previous literature onorganics.
A regular, diverse, and accessible supply of organicproducts is vital for inducing higher organic sales (Silver-stone 1993). The wider the assortment of organics, thegreater is the likelihood of the availability of specific fia-vors and/or package sizes, which creates more opportunitiesfor customers to buy them (Aertsens, Mondelaers, and VanHuylenbroeck 2009). Reduced distribution would createmore transaction costs, making it less worthwhile for thetypical retailer's consumer to buy them (Campo, Gijs-brechts, and Nisol 2000). Likewise, if consumers perceivethe price of organics as high, they will be less willing topurchase them (Michelsen et al. 1999; Verhoef 2005; Zanoliand Naspetti 2002). Thus, organics capture a larger cate-gory share when their price premium over conventionalproducts is relatively low (Wier et al. 2003).
Would temporary actions such as price promotions, fea-ture, and display have the same effect? They may if theyrepresent a buying incentive as strong as assortment andregular price changes, with less potential for perceivedquality erosion (Delvecchio, Henard, and Freling 2006).However, we assert that temporary actions will be lesseffective for organic products, which represent a moreenduring involvement. Previous research has shown thatconsumers choose organics as a means of achieving impor-tant life values (Makatouni 2002). Therefore, consumers arelikely to consider not just the costs and benefits at the cur-rent purchase situation (for which an organic product maybe available at a low promotional price) but also the futurelikelihood that they can buy a suitable organic product at areasonably low price. In summary, enlarging assortmentsand decreasing (regular) prices, but not increasing tempo-rary promotions, should be more effective for organics thanfor conventional products.
Hi: The long-term own-elasticity of sales to (a) assortmentand (b) regular price is greater for organic products thanfor conventional products.
The extent to which such enduring costs representobstacles to buying organics should depend on the strengthof a consumer's conviction regarding the benefits oforganic products. Although it is typically not cost-effectivefor mainstream retailers to survey all shoppers on this mat-ter, they can infer such conviction from revealed prefer-ences (i.e., the consumer's general purchase patterns oforganic products). Core organic consumers frequently buyorganic products, revealing a higher intrinsic value fororganic over conventional products. Previous research hasshown that such consumers tend to be socially conscious(e.g., show greater environmental orientation) and alsoexhibit a greater concern for their health (Zanoli andNaspetti 2002). Consumers with such values should be lesssensitive to the enduring costs of limited assortment and thehigh price of organic versus conventional products in any
specific category. In contrast, noncore organic consumershave little experience with organics in general. A limitedassortment and/or high regular price may be key deterrentsfor buying a specific organic product. Thus, the lowerintrinsic value of organics should translate into a highersensitivity to regular price and assortment.
H2: The long-term own-elasticity of organic product sales to(a) assortment and (b) regular price is lower for coreorganic consumers than for noncore organic consumers.
Next to their intrinsic preference for organics, con-sumers' sensitivity to organic marketing may also dependon the perceived cost-benefit trade-off in a specific cate-gory. Category-specific costs include the category's expen-siveness and share of the consumer's wallet and the organicprice premium (over conventional products). Moreover,perceived costs of trying new, expensive products are lowerin impulse purchase categories, which should stimulateorganic sales. Perceived benefits from buying organic aregreater for products with higher purchase frequency andproducts that are directly related to taste, environmental,animal welfare, and local farmer concerns (Fotopoulos andKrystallis 2002; Makatouni 2002). Such direct-from-the-farm categories include produce, dairy, meat, and poultryproducts (Davies, Titterington, and Cochrane 1995; Verhoef2005). The health benefits of organic products are also morecongruent with virtue products (connected with self-controlgoals) than with vice products, which provide immediategratification (Wertenbroch 1998). Finally, storable productsare visible longer at home to consumers (and their friendsand family), which increases the salience of organic bene-fits. Note that previous literature has discussed the impactof such category characteristics only on organic appeal andsales (i.e., a main effect), not on consumer response to mar-keting in such categories. We expect that greater organicappeal in a category may also translate into higher con-sumer reactions to organic marketing in that category.
When organic marketing activities succeed in raisingorganic sales, how will this affect the sales of conventionalproducts in the same category? Consumers may simply addthe organic product to their shopping basket (e.g., when anewly introduced organic product adds a salient attribute tothe category) (Boatwright and Nunes 2001). Impulse-buycategories are especially prone to this behavior. In general,however, such "free lunch" for the retailer is unlikely: Con-sumers tend to focus on the perceived value of organic ver-sus conventional products and thus substitute the conven-tional product with the organic product (Durham andAndrade 2005; Kiesel and Villas-Boas 2007). Thus, success-fully promoting organic products should reduce demand forconventional products in the same category.
Cross-elasticities with conventional products should beasymmetric in favor of higher-priced organic products if, aswe believe, the asymmetric price competition literatureapplies (Allenby and Rossi 1991; Blattberg and Wisniewski1989; Kamakura and Russell 1989; Sethuraman and Srini-vasan 2002; Sivakumar and Raj 1997). In addition, iforganic products bestow intrinsic quality benefits to theconsumer (e.g., provide health benefits, taste better),switching back to conventional products would represent a
34 / Journal of Marketing, January 2013
loss in those benefits, which consumers aim to avoid (Bron-nenberg and Watthieu 1996).
H3: Long-term cross-effects are asymmetric; organic market-ing activities hurt conventional products sales more thanvice versa.
Finally, which type of conventional products shouldexperience most harm from organic marketing activities? Inaddition to the price-tier effect (e.g., Nowlis and Simonson2000), previous literature has shown that brands whoseprices are closer have higher cross-price effects than brandsthat are priced further apart (Sethuraman and Srinivasan2002; Sethuraman, Srinivasan, and Kim 1999). Thus, wemaintain that organic marketing activities will hurt salesmost for brands that are more similar to organic brands interms of expense—first top-tier national brands, followedby second-tier national brands and private labels.
MethodologyOur research questions suggest a methodology for analyzingmarketing effects on sales and aggregate retailer performance(sales revenues and profits), while accounting for potentialmarketing endogeneity. Therefore, we chose the persistencemodeling approach (Dekimpe and Hanssens 1995), whichhas previously been applied to long-term marketing effec-tiveness for conventional products (e.g., Nijs et al. 2001;Pauwels, Hanssens, and Siddarth 2002), offering a basis forcomparison. This approach involves four steps. First, unitroot and cointegration tests investigate whether the perfor-mance and marketing variables are stationary, evolving, orcointegrated (Enders 2004; Johansen, Mosconi, and Nielsen2000). Second, using the test results, we estimate a vectorautoregressive (VAR) model or a vector error correction(VEC) model (Dekimpe and Hanssens 1999). Third, we com-pute impulse response functions, which track the effect of amarketing variable on the performance variables of interestover time (Pesaran and Shin 1998). Fourth, we perform aweighted least squares regression of the estimated long-termelasticities on the product category factors, using the inverseof their standard errors as weights (Srinivasan et al. 2004).
The econometric specifications have been well docu-mented in previous literature (e.g., Trusov, Bucklin, andPauwels 2009). The researcher chooses (1) the endogenousvariables that are explained by the model, (2) the exogenousvariables that may affect the endogenous variables but arenot themselves affected, and (3) the lag length (p), based onthe Bayesian information criterion, which trades off predic-tion accuracy and model complexity. After model estima-tion, we perform the required diagnostic checks on theresiduals (Franses 2005) and report on the explanatorypower of each model. The VAR model has the generalspecification shown in Equation 1 for each category (e.g.,Srinivasan et al. 2004):
the matrix of intercepts, X is the vector of exogenousvariables (seasonal dummies, holiday! dummies, and a timetrend) to control for factors unrelated to marketing, and E isthe full variance-covariance matrix of residuals. To addressour research questions and run validation checks, we esti-mate VAR models with different variables in the Y vector ofendogenous variables, as detailed in Appendix A.
We assess Hi and H3 in our sales model, which connectsorganic and conventional product sales with their respectivemarketing actions. Thus, the endogenous variables in themodel are (1) the logarithm of assortment size, unit price,promotion breadth, promotion depth, and price specials,respectively, for organic and conventional products (mar-keting variables) and (2) the logarithm of volume sales fororganic and conventional products (performance variables).
We assess H2 by replacing the two performancevariables with organic and conventional sales from the coreorganic and noncore organic segments. We assess whichconventional brands are hurt most by replacing conven-tional sales and marketing variables with the correspondingvariables of first-tier national brands, second-tier nationalbrands, and private labels. To avoid overparameterization inthis model, we use price and assortment and promotionbreadth and depth separately as endogenous variables, whileincluding the remaining marketing variables as exogenous.Finally, to analyze store performance, we replace the perfor-mance variables with category profits and store revenues.Further models investigate the robustness of our findings to,respectively, quadratic price effects, social infiuence, differ-ent definitions of "organic" products, and store heterogeneity.
After VAR model estimation, we obtain long-term mar-keting elasticities through generalized impulse responsefunctions (Pauwels, Hanssens, and Siddarih 2002). We cal-culate the "long-term marketing elasticity" (because wehave a log-log model specification) as the cumulative effect(i.e., summing up all significant impulse response coeffi-cients). Note that we do not recalibrate the model for insig-nificant impulse response values, as these are derived fromthe estimated coefficients (Pauwels 2004). In the final step,we use the estimated long-term marketing elasticities in theweighted least squares regression (using the inverse of theirstandard errors as weights) to investigate how they arerelated to the category characteristics (Nijs et al. 2001;Srinivasan et al. 2004):
(2) LTE, = ßoi + ßiVIRTUE + ß2DMP + ßjPRODUCE
(1) Y, =
where Y is the vector of endogenous variables explained byits own past (thus, the term "vector autoregression"), A is
We include in Equation 2 the virtue nature of the product(VIRTUE) and whether the category is of the type corre-sponding to dairy, meat, and poultry or produce (PRO-DUCE). In addition, we include the characteristics relatedto category purchase frequency, storability, impulsivity,category expensiveness, market concentration, categoryshare of consumer wallet, organic price premium (over con-ventional price), organic penetration (current base), and
'The holidays are Easter, Memorial Day, Independence Day,Labor Day, Halloween, and Thanksgiving as well as the weekafter Thanksgiving, Christmas, and New Year's.
What Is Special About Marketing Organic Products? / 35
organic growth rate, which may capture unobserved cate-gory-specific factors. We collectively refer to thesevariables as (the vector) Zj in Equation 2. Table 1 presentsour operationalization of the category drivers.
In addition to persistence modeling, we examine the rela-tionships of interest with the Koyck model (Franses and VanOest 2007), which allows for all same-week effects specifiedin the VAR model and for some dynamic marketing effectsthrough autoregressive and moving average terms. Equation3 shows the Koyck model for organic sales (Org_Volt):
(3) Org_Volt = n -H ô,Org_GP, -I- Ô20rg_PB, -h Ô30rg_PD,
_ Ast, -f
+ ô|oCon
The independent variables in the preceding equation—Org_GPt, Org_PB„ Org_PDt, Org_PS,, and Org_Astt as well
as Con_GP„ Con_PB;, Con_PD,, Con_PSt, Con_Ast,-refer to regular price, promotion breadth, promotion depth,price specials, and assortment for organic as well as conven-tional products, respectively. We estimate the same Koyckmodel specification with conventional product sales as thedependent variable. Maximum likelihood estimation yieldsthe coefficient estimates for the models (Franses and Van Oest2007). Compared with the VAR model, the Koyck model ismore parsimonious but imposes exponential decay (vs. moreflexible dynamic effects, such as wear-in and wear-out), andthe feedback effects among the performance and marketingvariables are absent. We investigated for the presence ofsuch feedback effects with Granger (1969) causality tests.
DataThe data come from a large retail chain in the northeasternUnited States that operates 75 stores. The store-level datacontain purchase transactions for volume sales, actual
Variable
TABLE 1Variable Operationalization for Long-Term Marketing Elasticities Regression Equation
Operationalization
Virtue
Dairy, meat, and poultry
Produce
Category frequency
Storability
Impulsivity
Category expensiveness
Market concentration
Category wallet share
Organic premium
Organic growth rate
Organic penetrationAssortment
The virtue versus vice nature of the product category was labeled according to the classificationin Hui, Bradlow, and Fader (2009), who use three independent judges for this purpose.A dummy variable that takes the value of 1 for the categories of milk, creams, yogurt, eggs,butter, cheese, beef, chicken, and turkey and 0 otherwise.
A dummy variable that takes the value of 1 for the produce categories of tomatoes, oranges,grapefruits, strawberries, peaches, potatoes, apples, carrots, ready-to-eat (packaged) salads,greens (unpackaged salad and others), onions, mushrooms, grapes, lemons, and blueberriesand 0 otherwise.
The average number of times per year the category is purchased. Using the procedure outlinedpreviously, we selected the households that buy in a category (h). For these households, wecalculated the purchases made each year and then average across households and years.A dummy variable indicating whether the product is considered perishable or storable (e.g.,Narasimhan, Neslin, and Sen 1996).
A dummy variable indicating whether a product is typically associated with an impulse versus aplanned purchase (e.g., Narasimhan, Neslin, and Sen 1996).
We first computed the regular price (using the method described in the main text) of eachbrand. We calculated the category-level measure by the market share weighted average of theregular prices of the brands in the category (see Raju 1992).
We measured the category's competitive structure by market concentration, following previouswork in industrial organization and marketing (Bowman and Gatignon 1995), as the sum of theshares of the top three brands in the category.
This variable denotes the relative amount of money a consumer spends on a category and iscalculated from the household basket data. We first randomly selected a sample of 10,000households that have a high loyalty to the chain. For these households, we extracted all theirbasket transactions for 2004-2007. From these baskets, we calculated the dollars spent on thecategory and the total dollar value of the baskets. The total wallet share is then this ratio.The difference of the average organic and conventional price divided by the conventional price.We calculated the percentage growth rate of the organics each quarter as the difference in thecurrent quarter's dollar sales and the previous quarter's dollar sales divided by the previousquarter's dollar sales. We then averaged across the data period.
The dollar sales of the organic products divided by the total dollar sales of the category.
The number of unique SKUs the retailer carries. We began from a chain-level file, whichcontains the comprehensive listing of all SKUs across all stores each week. We inferredassortment additions by identifying new SKUs that have been added to the chain-level file eachweek and which obtained positive unit sales in any of the stores in the subsequent four weeks.To identify SKU deletions, we required that the SKU is dropped from this list and not have anysales in any of the stores in the subsequent quarter.
36 / Journal of Marketing, January 2013
prices, and retailer/manufacturer discounts for each stock-keeping unit (SKU). In addition, we have data pertaining tothe costs (or wholesale prices) for each SKU for the entirechain, which enables us to calculate retailer gross profits.Using this information, we compiled a data set that spans355 weeks (January 2004 to October 2010) across the 49food and 7 nonfood categories (for detailed data descrip-tion, see Appendix B).
We analyzed all food and nonfood categories in whichthe retailer had at least two organic SKUs. All food organicSKUs have the USDA seal, which is permitted for twoclasses of products: 100% Organic and (at least 95%)Organic (the seal is not permitted for two other classes:Made with Organic Ingredients and Less than 70% OrganicIngredients). In the case of nonfood products, USDA label-ing is mostly absent, so we use the Organic and Made withOrganic Ingredients classification instead.
Because the focal retailer follows a chain-level strategywith respect to assortment, pricing, and promotions, wedecided to aggregate the data from the store level to thechain level, as well as from the SKU level to the productgroup (organic vs. conventional) level in each category. Weoperationalize weekly assortment as the number of uniqueSKUs the retailer carries. Price is operationalized as priceper unit calculated for each SKU and then share weightedby the SKU market share using constant weights for eachstore (Pauwels and Srinivasan 2004). We define "promotionbreadth" as the percentage of SKUs that are promoted in agiven week for a given store. To obtain the promotion depthvariable, we first calculated the per unit dollar discount,which is the difference between the promotion price and theregular price. We define the regular price as the average ofthe previous four nonpromoted prices (see Hendel andNevo 2006). We then obtain the promotion depth as a per-centage by dividing the unit dollar discount by the unitregular price. Price specials capture the feature and displayactivities, and we set them to I if a particular SKU is on aprice special; we then aggregated them to the category levelusing constant weights for each store. We aggregated all thevariables to the chain level using store sales as weights.
In our analysis, we considered all organic and conven-tional SKUs. The focal retailer has actively marketed organ-ics since 2004. The organic products are stocked both nearthe conventional products and in specially designated sec-tions of the store. (The retailer uses a mix of integration/separation strategies with respect to organics.) Moreover,store features contain advertising for both organic and con-ventional products, and displays are used for both types ofproducts throughout the store. Thus, consumers are exposedto the marketing activities of both organics and their con-ventional counterparts.
Marketing activity differs substantially not only fororganic versus conventional products but also across cate-gories at the focal retailer. We find that product assortmentfor organics is 25% of that for conventional products, onaverage, with their largest assortment (both absolute andrelative to conventional products) being in processed foodcategories (e.g., cereal, crackers, jams/jellies) and thesmallest in produce. The gross prices for organics arehigher than those of conventional in each category, but this
price premium varies substantially. Notably, conventionalproducts experience more promotional activity: promotionbreadth, depth, and frequency of price specials. Again, thedifferences between categories are striking; for example,organic products are on price specials quite often for dairy,produce, cereal, salad dressing, and jams/jellies. Retail mar-gins are, on average, 25.18% higher for organics; this dif-ference is the highest for frozen pizza and the lowest forgrapes. Average annual growth rates for organic productsare approximately 17.85% versus 2%-3% for conventionalproducts, in line with U.S. trends during the analysis period(Organic Trade Association 2009b). Organic penetrationranges from a high of 27.71% (greens: salad/others) to alow of .35% (cheese), while organic premium ranges from ahigh of 182.30% (eggs) to a low of 5.72% (tomatoes). Ingeneral, organic premiums are high for dairy, meat, andpoultry categories (e.g., eggs, beef, milk) and low for pro-duce categories (e.g., tomatoes, lemons, grapefruits), whichare dominated by generic rather than branded organics.
Although store-level data have the benefit of coveringall purchases, they do not allow us to distinguish consumerswho frequently buy organic products (the core organic seg-ment) and those who do not (the noncore organic segment).To this end, we obtained panel data, which record individ-ual transactions covering 95.4% of all purchases made atthe retailer. The panel data cover the same period and cate-gories as the store-level data. We grouped consumers intocore and noncore segments according to their individualorganic purchase histories. Our two classification alterna-tives are (1) a median split based on organic volume pur-chases in the category over the entire data duration and (2)a median split based on overall organic purchases at theretailer during the past 12 months (with four purchases asthe threshold). We also used a nested logit model that con-sists of category incidence and product choice to differenti-ate core and noncore organic consumers (see Appendix C).We randomly selected 700 consumers in each category whomade at least two purchases of organics for analysis. Fromthe intercept term of product choice, which can be interpretedas consumers' organic intrinsic preference, we classifiedthem as belonging to core (higher than the mean intercept)or noncore (lower than the mean intercept) segments. Usingthese classifications and relevant variables, we conductedthe VAR analysis separately for each segment. We calcu-lated the long-term assortment and price elasticities on thebasis of the segment-level VAR estimates. On comparison,we found that the elasticities obtained through the segment-level analysis using the nested logit and organic volumepurchases are similar.
ResuitsWe first relate the key variables in a median split fororganic products sales share, the correlation matrix of mainvariables, and the analysis results of the Koyck model. Wethen discuss the analysis pertaining to the Granger causality,unit root, and cointegration tests. After reviewing modelspecifications and estimation, we present the substantivefindings of the VAR models, which relate back to thehypotheses.
What Is Special About Marketing Organic Products? / 37
Figure 1 compares organic premium, organic growthrates, category expensiveness, and frequency of purchase ina median split by organic sales share in the category. Cate-gories with above-median organic sales share have a lowerorganic premium (Al.06% vs. 51.84%) but a higher annualpurchase frequency (9.89 vs. 7.28). Table 2 reports both theoverall and the specific correlations between top-tier andsecond-tier national brands, private labels of conventionalproducts, and organic products. As the table shows, conven-tional top-tier brands have a higher correlation in sales andmarketing actions with organic products than second-tiernational brands and conventional private labels. This is con-sistent with our classification of organic products as top tierin the category.
We begin with the Koyck model results (Table 3). Wefocus on the coefficients of interest: own- and cross-elasticitiesof organic versus conventional products for the five ana-lyzed marketing actions. First, organic products have ahigher own price elasticity (-3.00) than conventional prod-ucts (-1.95), and the same holds for own assortment elastic-ity (2.63 vs. 1.69). In contrast, own promotional elasticiesare not significantly different for organic than conventionalproducts. Second, the cross-elasticities indicate that promo-tional breadth and depth on organic products hurt conven-tional product sales more than vice versa. Other differencesare not significantly different from zero. We next investi-gate dual causality in our data to gauge the need for a morecomplicated VAR model.
Granger Causality TestsWe focus on the Granger causality among organic and con-ventional product sales and retail marketing actions. First,the results (available upon request from the authors) showthat virtually all marketing activities Granger-cause salesfor the intended products (e.g., organic price on organic
sales). Second, organic marketing actions Granger-causeconventional product sales in 44% of categories, while con-ventional marketing actions Granger-cause organic productsales in 70% of categories. Third, organic marketing activi-ties Granger-cause conventional marketing in 14% of allcases (43% for organic price Granger-causing conventionalprice), while conventional marketing activities Granger-cause organic marketing in 16% of all cases (52% for con-ventional price Granger-causing organic price). Thus, theretailer shows some evidence of coordinating marketingacross product groups. Finally, we find several cases of per-formance feedback (i.e., sales are Granger-causing market-ing for the same product group). In summary, the Grangercausality tests confirm the dual causality loops amongorganic marketing, conventional marketing, and retailerperformance captured by the VAR model.
Modei Specification Choices and ModelEstimation
Conventional product sales are evolving in only 3.6% ofcases, consistent with previous research (Nijs et al. 2001;Srinivasan et al. 2004). In contrast, organic sales are evolv-ing in 8.9% of all cases. Moreover, sales are trend station-ary (i.e., only stationary after we account for a deterministictime trend) for 33.9% of organic cases (16.07% of conven-tional cases). Such a time trend may capture gradual gainsin awareness/appeal of organics because of factors outsidethe retailer's control (e.g., health concerns). In all cases oforganic sales evolving or trending, the sales series is grow-ing, while 41.6% of the conventional sales with evolutionand time trend are declining. Thus, our data reflect thestronger growth in organic versus conventional productsales observed in the business press but also indicate thatsuch growth is not self-evident: Most organic sales seriesare mean stationary.
FIGURE 1Category Characteristics by Median Split of Organic Sales Share
51.84
47.06• Below median organic sales share' ^ Above median organic sales share
25.58
Organic Premium (%) Organic Growth (%)
9.89
Category Purchase FrequencyExpensiveness
38 / Journal of Marketing, January 2013
TABLE 2Correlation Results
A: Overall Organic and Conventional Sales (Share Weighted)
ConVol
ConRP
ConPB
ConPD
ConPS
ConAst
OrgVol
OrgRP
OrgPB
OrgPD
OrgPS
OrgAst
Con VolCon RPCon PBCon PDCon PSCon AstOrg VolOrgRPOrg PBOrgPDOrg PSOrg Ast
-.621
.47
.401
.32
.31
.191
.35
.38
.48
.211
.58
.33
.32-.03
.331
-.23.10
-.20.15.22
-.201
.18
.19
.42
.24
.35-.18-.701
-.41-.20-.09-.15-.06-.08
.09-.041
-.31-.16-.04-.07-.01
.06
.45-.25
.291
-.29-.30-.23-.09-.09-.13
.16-.21
.56
.451
-.03.18
-.14-.10-.08
.17
.65
.51
.07
.15
.121
B: Overall Sales for the Different Type of Conventional and Organic Products (Share Weighted)
TTNB TTNB TT NB ST NB ST NB ST NB PL PL PL Org Org OrgVol RP Ast Vol RP Ast Vol RP Ast Vol RP Ast
TT NB Vol T~TTNBRPTT NB AstST NB VolST NB RPST NB AstPL VolPLRPPL AstOrg VolOrg RPOrg AstNotes: Con = conventional, Org = organic, Vol = sales volume, RP = regular price, PB = promotion breadth, PD = promotion depth, PS = price
specials, Ast = assortment, NB = national brand, PL = private label, TT = top-tier, and ST = second-tier.
-.631
.56
.391
-.23.23
-.161
.21
.12-.17-.581
-.20-.22
.14
.47
.331
-.18.16
-.13-.23
.21-.201
.14
.11-.10
.20
.20-.21-.611
-.16-.10
.09-.21
.25
.16
.53
.141
-.30.27
-.24-.15
.20-.17-.13
.20-.131
.28
.22-.20
.12
.19-.18
.10
.12-.11-.701
-.25.24.23
-.18.16.19
-.08.13.17.65.51
1
Cointegration is present among organic marketing andperformance in 2.67% of cases and among conventionalmarketing and performance in .89% of cases. This rareoccurrence of a long-term equilibrium between marketingand performance is consistent with previous research (e.g.,Bronnenberg, Mahajan, and Vanhonacker 2000; Nijs et al.2001; Srinivasan et al. 2004). We use the VEC specificationin the case of cointegration and the VAR specification eitherin levels or first differences otherwise. The optimal laglength of 1, selected by the Bayesian information criterion,yields a good model fit for all models/categories (averageR2 = .80, and adjusted R2 = .78) and is superior to that ofthe Koyck model in all cases. For example, in the case oforganic and conventional volume sales, the adjusted R-squares for the VAR and Koyck models are .84 versus .72and .81 versus .69.
Substantive Findings from Impulse ResponseAnalysisWe focus on the long-term elasticities (Pauwels, Hanssens,and Siddarth 2002). We observe that only 1.04% (organic)
and .59% (conventional) of marketing-sales effects arepermanent .2
Long-term own sales elasticities. Table 4 displays theaverage elasticities across the 56 analyzed categories, eitherwithout weighting ("simple average") or weighting thecategory results by the respective category's contribution tothe overall store revenues.
In support of Hj, the sales elasticities for assortment andregular price are significantly higher for organic than con-ventional products. In contrast, organic products do notenjoy higher sales elasticities for promotional activities.Figure 2 shows the impulse response graph for a representa-tive category, tortilla chips. Note that product assortmenteffects show a similar pattern over time for organic and
2To compare them with cases that show only temporary effects,we calculated the net present value of the permanent effect byusing a weekly discount rate of .15%, after discussion with theretailer. We then added this net present value to the immediate andadjustment effects to calculate the total, long-term elasticity of theperformance variable to the marketing action. We verified that oursubstantive results remain the same when ignoring these infre-quent and small permanent effects.
What Is Speciai About iVIarketing Organic Products? / 39
Variable
TABLE 3Organic and Conventional Long-Term Elasticities Using Koyck Model for Overali Saies
A: Own Sales Elasticities
Percentage of95% CI of Categories
Organic Conventional Difference Difference Significant
AssortmentRegular pricePromotion breadthPromotion depthPrice specials
2.63-3.00
1.08.59.68
1.69-1.95
.64
.50
.75
.94-1.05
.44
.09-.07
(.68,(-1.39,
(-.09,(-.32,(-.25,
1.21)-.71)
.97)
.50)
.10)
62.571.421.414.223.2
B: Cross-Sales Elasticities
Variable
OrganicMarketing on
Conventional Sales
ConventionalMarketing onOrganic Sales Difference
95% CI ofDifference
Percentage ofCategoriesSignificant
AssortmentRegular pricePromotion breadthPromotion depthPrice specials
-.161.795
-3.135-1.861-.937
.105
.568-1.735
.693-.836
-.27.227
-1.40-2.550-.10
(-.56,(-.12,
(-1.79,(-2.34,
(-.31,
.02)
.57)1.00)
-1.83).11)
33.939.289.275.025.0
Notes: The values in bold indicate that zero does not belong to the 95% confidence interval. Thus, the elasticities are significantly different fromeach other at the 5% level. CI = confidence interval.
conventional products, with a long wear-out of approxi-mately 30 weeks. The key difference is the size of theeffect: three times as large for organic as for conventionalproducts. In contrast, consumer reactions to regular pricechanges differ in both magnitude and pattern for organicversus conventional products. A regular price decrease onlysignificantly benefits the average conventional tortilla chipproduct for 1 week, whereas it benefits the average organictortilla chip for 28 weeks (four months). This pattern is con-sistent with our argument that consumers perceive a moreenduring commitment to organic products; lowering theregular price induces higher sales for several months.
How does the higher regular price elasticity for organicsdiffer across segments? Table 5 reports our results for thesegment-level model based on organic loyalty, which wedefine using a median split of organic volume purchases.We obtained similar results when using other operational-izations. In support of H2, we find that the elasticities forthe enduring activities are higher and significantly differentfor the noncore than the core segment.
As for category-specific costs and benefits, first, oursecond-stage analysis indicates a higher sensitivity toorganic promotions for products with high purchase fre-quency, of a virtue nature, and that come directly from thefarm (produce, dairy, meat, and poultry). Second, productstorability and impulsivity increase consumer sensitivity toproduct assortment. While deep promotions induce higherconsumer response in storable and impulse-buy categories,regular price reductions are less effective. Third, expensivecategories show lower consumer sensitivity to regular priceand price specials for organic products. This indicates thatthe higher price of organics is not such an obstacle for buy-ing organics in categories in which prices are generallyhigh. Likewise, regular price sensitivity is lower in cate-gories with a high organic price premium, indicating a
greater willingness to pay for organics. Plausibly because ofthis high willingness to pay, consumers are more responsiveto assortment additions in categories with high price premi-ums. Fourth, high organic penetration is associated withhigher sensitivity to assortment and price, consistent withour finding of high sensitivity to these enduring actions bynoncore consumers (who tend to constitute a larger part oforganic buyers in categories with high organic penetration).Finally, retailers obtain higher organic sales benefits withpromotions and price specials when organics are alreadygrowing strongly in the category.
Long-term cross-sales elasticities. We report the crosssales elasticities in the bottom panel of Table 4. In supportof H3, we observe an asymmetry: Organic marketing activi-ties hurt conventional product sales more than vice versa.These differences are significant at the 5% level for promo-tion breadth and promotion depth. In Table 6, we reportelasticities across the different combinations of conven-tional products and organics. As expected, we find that pro-moting organic products hurts top-tier national brands themost, followed by second-tier national brands and privatelabels of conventional products.
Overall retailer performance. In light of the large own-and small cross-elasticities of enlarging the organic assort-ment, this activity seems desirable if the retailer wants toincrease sales of organics without hurting conventionalproduct sales. Even in the case of cannibalization, thehigher unit margin on organics may still increase overallcategory and store performance. Table 7 reports the resultsof our overall performance model.
Considering the long-term performance elasticities, wefind that organic assortment and organic and conventionalpromotion breadth have a significant, positive effect ongross category profits. The long-term elasticities in this case
40 / Journal of Marketing, January 2013
FIGURE 2Sales Response to Assortment and Price
Changes for Organic Versus ConventionalProducts in the Tortiila Chips Category
1.21.0
.8
.6
.4
.20
-.2
.05
.0
-.05
-.10
-.15
-.20
-.25
— Organic assortment— First standard deviation (+/-)
1 3 5 7 9 1113 15 17 19 2123 25 27 29 3133 35
.10
.05.0
-.05-.10-.15-.20-.25-.30-.35-.40
'Z^z^^^^^^-'^^-—'^'
— Organic regular price— First standard deviation (+/-)
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35
.451
.40
.35
.30
.25
.20
.15
.10
.050
-.05
\\
— Conventional assortment— First standard deviation (-H/-)
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35
— Conventional regular price— First standard deviation (+/-)
1 3 5 7 9 1113 15 17 19 2123 25 27 29 3133 35
are .47, .41, and .20, respectively. Compared with the sameactions for conventional products, we find a significantcategory margin advantage for higher assortment and morefrequent promotions on organic products (last column ofTable 7). Do these activities also improve overall storerevenues? We find few significant effects, consistent withprevious promotion studies for conventional products (e.g.,Srinivasan et al. 2004). The two exceptions are enlargingthe organic assortment and decreasing organic regularprices in the category, which yield a long-term elasticity on
store revenues of .19 and -.36, respectively. Again, theseactions have higher store revenue elasticity for organicproducts than for conventional products. Regarding cate-gory-specific factors driving overall retailer performancevariables, we find that the gross category margin elasticityis significantly higher in produce categories for changes inorganic assortment, regular price, and price specials.
Long-term elasticities and organic product labeling.Given previous research on the importance of organic prod-uct labeling, we compare our results on products with theorganic USDA seal with those labeled "organic" without theUSDA seal and with those labeled "natural" (i.e., productsthat do not contain any artificial flavoring, color ingredients,chemical preservatives, or artificial or synthetic ingredients).We find that our substantive results hold when we combineorganic products with and without the USDA seal as organicproducts in the analysis, with a similarly high powerexplaining sales (this model: R2 = .815; the main model forfood categories: R2 = .790). However, using only the organicproducts without the USDA seal yields a lower explanatorypower (R2 = .695) and effect estimates that, while direction-ally similar in many cases, are substantially lower thanthose for the organic product with the USDA seal (our mainmodel). Finally, using "natural" products yields an evenlower model fit (R2 = .515) and shows hardly any significantdifferences in own and cross-elasticities between naturaland conventional products (without any organic or naturalclaim). From these results, we conclude that organic label-ing is quite important across categories, consistent with con-sumer decision-making literature on organic labeling.
Additional analyses. We performed several additionalanalyses to gain further insight into our main results. Wechecked for quadratic price effects (as in Ngobo 2011) andfound that they failed to improve model fit. We included asocial influence variable from sampled consumers' self-reports of such influence on their organic buying decisions.We found that the social influence variable is significantand improves the model fit, implying that it is a driver oforganics. However, our substantive results held after wecontrolled for social approval.
Our final robustness checks consider aggregation bias,store trade area characteristics, and changes over time. First,to determine whether our results are sensitive to aggregationbias, we estimated the models on a store-by-store basis andcomputed the weighted mean using stores' sales as weights;we found that the results are substantially similar. Second,to check the sensitivity of the results to store trade areacharacteristics, we conducted the same analyses for anotherchain operating in a geographically distinct area catering to adifferent clientele.^ We found similar results. Third, to checkfor changes over time, we estimated our main models onsemiannual periods and fit a local trend model through thetime variation in the estimates. No discernible over-timepattern appears, while regular price elasticity varies
chain operates 90 stores in the mid-Atlantic United Statesand does not compete with our focal chain. The clientele for thischain differs in terms of income, education, and ethnic composi-tion and average store size.
What Is Speciai About Marketing Organic Products? / 41
ta
îO
M
m
E
no
ioUTJno'Ê(0O)
"S (0 *;(U 0) c
mH_ 0)O u
Ü S
O) Q
O. (0
CO '
II '
a (0
co
•oB c.c mO) <0
I
oCOCO
0000
' "
COo
COen1
mC3)
enoCM
CD
m
*1
CO
oCO
T—
CO
rtm
en
oCOCO
<c
ooinN
o>o>1
Si
CMCD
T
00en'
in•T—
T
CDCOT—
inenT
o
'^
enr--
in
-3.
a>o
lar
pn
3
S'GC
Sen
.,_in
' ^
CO00
CMCD
CT)CM
COCO1
CMO
CJ)
. f—
CDCM
1
inCD
''"
CDin
rea
.a
Otio
n
EoQ.
inCOf^
CD
• *
CMCD
-
O
mCMCM
C3)1—
1
CMCD
CMCT)CM
COO1
CO
enCD
'5.(D
Otio
n
E2
CL
O
oinCM
ooCO
CMCO
m1
CO
CO
enh^1
CO
oen'
inCO1
5
CMen
CO
to
speci
oQI
"5 0) *;V)
S O)
^ « 3
Ü <l>
ino>
ted
.c
iffe
a
nce
i tD
'S CO
IIilo u
il • i lco ^
•o
B cx: (0
coO)c
(5
u
(0O)
o
ales
COa>E3P
ion
c0)
c0
Ü
pie
ECO
ted
^^O)
XR
ci
c10aS
cn
.a
I
.1- .1- CO CO
K r .*CM in in •CO 00 CO I . CM
Obb
.64
CMCM
CM
.44
in"w
T
.90
o"o>1
.91
0
COen
I I1 I
en c:5 co CMCM co . 1 - en
•<* CM co i n r--.en in co CD co
. - - T - CM CM
CM i n 1 ^ • ^ co7 - CJ> co CD h -
I I Î T '
in CO 00 T- CO00 CM en in CO
'T ' 1'
f CO CO C3Í T fCM . 1 - CO CO O
IT) CM CM CO COCO c:5 en .<t CM
CM
in 7- Tt co cot CJ) o co • *
co I CM co "*
co o h- CD NT- CM o h- CD
i'TT
in r - - 1 - co T-o o o in in
I I
O o
ilii: a. 0. CL
E
*2
Eo>V(0a>
(0
HI .0
m n
o
coiSUJ
•4- 0)
o uÜ so> Q
.c a>
.2» S5 it> Q
Q.
i•a
I cO) a>'S S
.11co ^
l
co c\i
00 c:>CT) c o
CD ir ico
co rin o
in co. 1 - CT)
CD CDK en00 cO
co
p inÖ CM'
•r^ CO
CO ino r
d coen T-
co enq CM
1(D OO W)
' C CO0- <
R
©
'S©0
©TO
onfi
Ü
inen
the
0
ac
belo
'ScCO
8• D
0
N
'co
©
'Sx>g
• 0
EgCO©
©
-
CO
©CO
-ent
©IB S
U(
CO
©
'S©Üc
ifide
8II
y.C
I
©>
a.©
'cCO0
'c
'cô©CO
CO
©
ativ
D)
C
Q
[ive
.^CO
a(D
CO©
egor
i
•.s coro o•go
IICO L_
'S §_© /1Í
42 / Journal of Mariceting, January 2013
È
Too
oü
(O
u"EnO)
os(O
no
i<o o
O
X(O
Si cQ. re
Ö5 ^•o
.c reO) 0)
•o
2m"rëo
re
ai
Pou
aE ,S) '
.c reO) a>
1 =
(U
m(O0)
(O
LU«(O
so
(O
•aEm"(5o
E
.Si c
.c reO) 0)
.2I
CJ) 1 - co • ^
od CD c i T^i n co 00 i~-
CD
co co
en loin co
K 00cp cj>
i00 COtn c\j
coo co
I - ^ CMCD 1 -
1 OO •--I O ) CTM
T-^ C M
1 O in
r iri
n N coo 00 oo
OvJ ^ NO OJ en
j co N.o r-- O
9r
CJ)
CX3 IT) CI5•1- ^ O
r;cor -1-
ill•c « oo c
OJ
a
ii n
eDco—;,
»
rÖCD
CO
>
Ü )
OCio
icht
hie
s
oO)CD'S
(Uuc(0E
rfor
i
CL
eta
DC
oo
(M
q
7 T
C\J
î T03
co
o ou o T-CD 1 -
oo en en ojCD q 0C3 h*-
CM OJ
eM i n OuCD OC) q CD
\' \' '
inCD
oo
Ö 2Q.JC Q-
.9 .9 y "o .9 £
ICDo
I0)
1
ü
«s
o
(U
oܕo
(0u"E!>
o
o O) • ;
tilc a> cS « .2»fe" «o. ü
• o
O)
inO)
a>oc
ere
0)Q.
Ecô
• D
S.cO)
re0)
c(0
t
5i c.c reO) 0)
tía
o>ni
Ü
mO
com«Suî)(TI
n S
ales
E
1
XICM
COCO
—
nCM
oen
oco
1
.20
.18
(
roCM
CMCM
CD
i nCM
1
.20
ssor
tmen
co
C3)
nOvl
o1
i n
m
1
.22
co
1
coCM
CO
1
CMCO
i nCD1
coCM
I —
1
.28
O)
1
(1)c
eaul
ar o
r
oc
S)
n
toOJ
0
,—Ovl
ro
1
o
;;.
00
o
oCM
en
ro
1
coT —
1^co
.18
.cT I
rea
c
rom
otio
n
a.
com
o
MO
1
i n
co
1
.19
.,-
OJOJ
CM
1^OJ
O
1
CDCM
ano
.26
o1—
rQ(1)
T-
rom
otio
n
o
ino
o
moo
1
o
o1
.11)
004
(
co
cooo
CDO
i no1
CMOvl
o1
.19
900
1
- co
co
rice
spec
CL
3L.<1)>a>
OCVo
co(ASU4..*a¡m
n S
ales
EO
w
m
T—
CM
noo
o
oo
00o1
.05
.05
(
coo
CDO
10OU
OJ1
i no
T —
.07
CD
CG)
CO<
M3)•O
O
rCM
0
1
O
co
i ncoT
.13
-.17
(
i n
oOvl1
CD
r--enT
C)1
CD
CDCO
1
Cl)C
N
q
o
MO
1
i nCM
coo
1
mq
o
Y —
o1
ooCM
O
oo CD
.07
(
oOJ
CDO
i n
T —
1
CM
i no
.22
h -o
rea
.c:
egul
ar p
rro
mot
ion
LL CL
.09
(
o
T —
mo
1
i n
o
.13
CDO
rQa)
T
rom
otio
n
û .
o
o
r>o
o
oo1
.10
.002
(
enocooo
0 0
o
CMO1
Ovl
O
003
- coCC
rice
spec
Q.
What Is Special About Marketing Organic Products? / 43
between -3.17 and . 01 for organic products and -1.75 and-2.2 for conventional products. Thus, we find no evidencethat our findings are sensitive to these potential issues.
Discussion and ImplicationsBuying organics represents a rather enduring commitment,which involves both transaction costs (finding organicproducts with the right flavor, size, and so on) and out-of-pocket costs (the price of organics). Therefore, loweringtransaction and tnonetary costs by increasing the organicassortment and decreasing organics' regular prices repre-sents the enduring marketing actions most likely to induceconsumers to buy organics. Our findings offer new insightsinto the ongoing debate on the place of organic products onthe (mainstream) retail shelf. We distinguish implicationsfor shopper response, manufacturer strategy, retailer mar-keting strategy, and policy makers and organics advocates.
First, we find that shoppers react differently to enduringmarketing actions for organic versus conventional products.We observe these differences for both organic products withand without the USDA seal but not for "natural" products.Thus, labeling and branding do seem to play an importantrole, as both the USDA seal and organic branding yieldsales benefits. Our distinction among consumer organicusage levels reveals that increasing assortments and reduc-ing the regular price for organics is especially effective fornoncore organic consumers but also stimulates purchase bycore organic consumers. Such core organic consumersshould therefore not be taken for granted: Although theyhave much experience buying and consuming organics, thegreater perceived benefits do not mean that they will buyorganics at any cost. A strategy of first getting consumers"hooked" on organics with low prices and then increasingprices seems ill-advised in light of our findings.
Second, manufacturers of top-tier national brands havethe most to lose from organic product growth and thusshould be the first in line to either develop their ownorganic products with brand names that are distinct fromtheir own or acquire organic brands of smaller companies.The latter is advised for minimally processed products (e.g.,produce, milk, yogurt, cereal), for which we found a highersales impact for small/independent than large manufacturerbrands in an additional analysis.
Third, mainstream retailers should consider increasingassortment and lowering regular prices, especially for thenoncore organic segment, but also for the core organic seg-ment. The highest return for such actions materializes forproducts with high purchase frequency, of a virtue nature,and that come directly from the farm (produce, dairy, meat,and poultry). In contrast, regular price reductions are lesseffective than deep promotions in storable and impulse-buycategories. Thus, retailers can keep the regular price a bithigher in such categories, while offering deep promotionsto induce impulse buying and stockpiling.
Our study also has implications for policy makers, sus-tainability proponents, and advocates of organic products.In contrast with Ngobo (2011), we find that typical actionsof mainstream retailers, such as broadening assortments andlowering prices, substantially increase sales of organics.However, these are enduring actions for the retailer, whichthus are unlikely to be swayed by temporary subsidies. The
full supply system should be considered: It is easier forretailers to increase organic assortment and reduce regularprices if manufacturers of organic products do the same.Transparent certification is important in this regard: TheUSDA seal increases consumer response to marketingactions for organic products.
ConclusionWhat makes marketing organics special in mainstream U.S.retail settings? Not as much as Ngobo (2011) implies:Reducing price and increasing price promotions and assort-ment strongly increase organic product sales in our large-scale analysis over 56 categories. Moreover, reducing priceson organic products hurts conventional product sales morethan vice versa, consistent with the asymmetric price com-petition literature (Sethuraman and Srinivasan 2002). How-ever, marketing organic products is special to the extent thatretailers need enduring actions (assortment and regularprice) to overcome the perceived costs of going organic,especially for shoppers with currently low intrinsic valuefor organic products. Increasing organic assortment is alsosuperior to increasing conventional assortment in terms ofcategory margin and store revenues.
In light of recent contradictory findings, we note thelarge (negative) price elasticity of organic products in eachof our 56 analyzed product categories. What might explainthe difference between some of the results of our study andthose of previous research? First, we base our analysis ondata that also include transactions by customers who(almost) never buy organic products, whereas other studiesfocus solely on the core organic consumer. Indeed, we findthat the core organic consumer is not as price sensitive asthe noncore organic consumer. Second, we use more recentdata in which the organic penetration is approximately 5%.Thus, we believe we are capturing a more current state oforganic demand and supply characteristics. Industry reportsconfirm that organic sales have remained strong during therecession, mostly because manufacturers and retailersdecreased organic prices iSupermarket News 2011).
Limitations of the current study include the absence ofdata on competing retailers' marketing, actions by suppliersof organic products, category advertising, and consumerperceptions of the store and its organic offering. As in anyeconometric study, our focus was on the sign and size (i.e.,the "what" and "how much") of consumer purchase actions,not on the "why." Further research should unravel the moti-vations behind these observed actions and generalize ourfindings to other retail settings. For example, we find littleevidence that marketing organics can increase store reve-nues at the studied retailer. A more focused repositioning,even fully converting to organic products (e.g.. WholeFoods), may be needed to achieve this.
This study represents an important step forward inresolving the "balancing act between the old and the new"for conventional retailers iProgressive Grocer 2008). Ouranalysis implies that organic products are compatible withconventional retailers and marketing acfions. With theinspiration from high-profile role models (e.g., MichelleObama starting a White House organic garden) and thepractical support of mainstream retailers, the future oforganics looks healthy indeed.
44 / Journal of Marketing, January 2013
><QZ
a.Q.<
ted
(Q
E
Stil
UJ(0
del
o
OUJ
>
o
ptio
ncr
i
moQ
(0
0)
Q
0)
iria
b
>
geno
us
oX
UJ
w
s1M
O
ogen
End
iptio
i
k.
Ctfg;
1E
ber
Num
odel
m
lain
Res
ul!d
fo
r M
odei
§reco (ô"D 2CD o> O)
^ CD
Sto
re-
56 c
ateDE
es,
ti
E
dum
nd h
olid
i
re"rëcen "Ure cCD CD
coS
CL c
Ô ^
O) o
ôo<S"5 <^^
O O O> >
and
een
Org
betw
en
o
1ç
eral
l sa
les
> oOO
1 -
comcOre"re
Pan
el
CDE
en"
1
dum
re'
"5
al a
i
cocoreeD
CO
1co
înt
Ol
Ml'
O)
Cor
e
>
c
: an
d C
o
Ö
1.Qenco"G2"cenCD
Sal
CM
enQ)
D)BreCJ
"OcCD
COco"û-CL c
°Q-Q_ ^
O) o
oô_
o o>
O)ep
inic
s
o
icore
o
core
and
rfo
r
ore"re• o
"3)
1
Sto
re-
1—
a.
Qdii'a
cd
x:G
:>
S'
-5
^CL
ô"
CO—-
>
tier
een
toD
-be
tw
cg
tera
ct
"S
°
oZen
o
:eg38
cal
_iQ.Om"
CLO
ST
PS
,
QCL11
GQCL
CL"
ÛL
côQ:
a."
| I
(eith
e
>
s, a
nd
0
_i
c
ß(T.
CQ
Z• $
s. s
econ
d-
ce
o
Ecßui
bran
di
O)
o
Org
Cû3,
Org
P
Q._ lQ.O
_QCL
Q"
°-co"
°-
Wor
)
CL"
O<¿
CP
L,
iîCD
sortm
ï
Dric
e an
d
°
1 tw
o
eu
O
for
prc
oc
D.
—
RP
,TT
A
ocoCL
m"ngOp:
0-
Ü.
d
d
CO
rrn
•.¿COCDa>
eo
mod
el
_ iCL
<
CP
L
CL"
CL
HCO
HCO
_>,
rereQ.
w
"re
"rëc| - o
en i ;
io en"
< ÊE'E
OC to
o .cC/5 ca. re
)
ore"coTU
•el
Sto
re-
0E
es, t
i
E
dum
m
nd h
olid
i
re"rëcoenre0co
S"ig
en
en0
Oßai
reo
T3c0
..oCO ^CL c
Ö ""
"5)0"
! mar
sto; 'san
i
: Sto
re
>
iSiE
perfo
r
i . .
eral
l re
taile
a
Q £
O) o
~ - CD"m ^
O(S
Ö8>
o
Org
and
Cvi
s
"Sen
en0
o
B"reü
0E
es,
ti
E
dum
re"
1•gre
re0co
0o"coen"0
0
k_
0o-Î2.
n V
ol
oO
:Org
\
>
itwee
n C
C0i _
COcoCL
O)
oQCL
ô
qPB
,CL
o
o
QCLco
CÛ ^n n^ c
Ö O "
EÔ>
a.CDo>
tic p
rice
O
S
t0a.
eral
l sa
les
è
o-£
ed w
i
3,
i C
on a
nal
re
Con
Orq
P
•§en
en0
o0
56 c
al
0E
es,
t
E
dum
nd h
olid
i
re"rëcoenreco
n V
ol
O
1cCQ
Soeo.¿•c0
co"
<
oQCLO)
o
Í P
B,
O
nrO) O)
OO>
and
een
Org
1enco
tera
ct
c
eral
l sa
les
O
co
>
5ur-
0
influ
e
"rëö
/ith
se
5
N>CCCrecoo
coQ.C
o"
ooQQ.C
Ö
Con
en
1 co
nsum
e
••-1
0
CO
O<
B^re
a>
0E
es,
t
E
dum
m
nd h
olid
i
re"rëcocore0co
p
:Orq
>
>
and
een
Org
betw
coc
tera
ct
c
eral
l sa
les
O
0
used
.
c0
CL
<
O
ô;P
B,
Örr
:Orq
l
>
rodu
cts
nie p
CDo
es o
f
a
n w
ith a
ll t
oO
—<
c côE"R2 8
0 "cCO 0
co3_c
uo;
£0
<^
a.
0
Con
rgan
ic (
a
0
re0c
Org
alu
ding
100
st 9
5%)
Ü rec 0
3CO0
0ßai
49 c
al
0E
es,
t
E
dum
nd h
olid
i
co
"rëc
sre0co
"?P
:Orf
l^
>
and
een
Org
1cg
tera
ct
c
eral
l sa
les
a
CX)
C0
7ÍCL
<
O)
Ö
0
nr
:Orq
l
>
re
0
re
• >
n fo
r O
rq v
0
coû.UO
;
0<c
0* _
fin
0C
onja
nic
"^0
• a
redi
ents
ai
ing
sed.
eo0
0ßai
49 c
al
0Eeo"
1
dum
re"
"5•0
re"rëcgre0co
n V
ol
Ö"5
:Na
t\•e
en o
nly
BS -aen en
ri -2
oU
S[
tera
ci
c ç
redie
nts
"-er
all
sale
s
O) >
.e 0
O)
c0
zco"Q."co CO
Z Ü.
Q""COCL Z
"re Q"z CL
tPB
,,
Nat
z £nr -5»
:Na
tFRP
, r
>
00
ä an
det
!:u
ral
prod
u
re
en0
0ßai
56 c
al
0E
ies,
t
E
dum
re"
"5•acco
"rëc
re0co
00"enen"3c0
re r
e
0
(si
m V
ol
00
"5
0
Q.
g
i
;e)
in
c
E
1S
eral
l sa
les
0
• oc0
ç
mar
gd
stor
e-t
rere
îdC
o
re
:wee
n O
rgre
M
CÛ
a)
wei
ij:
( 0CDc/)QJ
ww 0fl) "^uQ.Il
C/)Û_0)
1
Vol
=
"c
asso
rtmei
II
j z "
- n
tion
<
0ccr0Q.
IlQ
¿
E0Q.Il
mQ.
ice
ular
pr
req
IIû.ce
1 var
iabl
e:
uc
1èoCc.Il
coni
nce
varia
bh
co
1eu0II
Q.Ú
1
ig
eo
Ü
efl
rodu
cîv
entio
na=
cor
MUO:
0
Juct
s
ea.
mic
l u
0
0
ands
,
r.
|BU
natio
ii-t
ier
§II
en
bran
ddu
cts.
r na
tiona
lat
ural
pro
i:o
p-tie
M
abel
, TT
a>
PL
= p
rivat
al b
rand
.na
tion
cII
"èfl
T3Ceo
Îcu
iona
l priv
at«
: con
vent
CP
I =
What Is Special About Marketing Organic Products? / 45
^1(Q
m
ND
IXA
PP
Eic
sis
t
re
ary
SS
um
m
re
'SQ
Sp
ec
a>u' EQ.
O
om
oti
Q.
co
oti
2Q.
'ice
Q.ta
ros:
ncy
a>3t r0)u.
epth
(<
Q
ead
ffl
nit)
10)
o
o
o
S
(0co<
Ë3
Ü
U
Ü
OO)
coÜ
^^
; • ^ T-; CO K CM cq ir)
CVJinCOCVJ i - C M C J - i - ' - t i - C M C o '
•<t -"t oq
m-r-i-T-i— OJt— CO-r-CJCMi- OJ-r-
meo t-~. co coh~coco co corrq c M c j c M c D C \ j c \ ) c q i r ) c o i n c n t r ) ^ n
qCM
otor3-.- CM
co o 1 - co Io CM s . ••- I
) C 5 5 C D c d c M i r ) t N C DCMC\JT-OJC^JC^JCM i - o co Lo s. m
• • - T - CM CM
t i C M o O * o o m a ) Tcoi—coqq-i—cpojcocvjco
LO CM CMco
mNCMCM
1 - co co II co 1 - r-~
oKincniNJooococjîT - C O - ' — T - I - ^ C M
O> CVJ co T~ T— O5i n • * 1 - - ^ I--.. h-
T - CM
S2
co
co . ï
CD
CD -c 2
¿ co m aB-^'ccô •
co
I co <S>c C31
I I Í E | Í E ^ | ^ ^ | C O
Ü
CD
?fflO
«5n.
CD
o3COCO
co03
lî
CO
euoco
CLo .yCL Ü.
o oCL a.
46 / Journal of iVlarketing, January 2013
oX V
il
o EË 2
co .2
.a £ Cç
ill
oc O)
s
2 a»n Q
U
O
O
CQ
^1o s
u
o
0)N
Ort
(0(0
4)
E
z
Ü
oO)
00
OJ
a)c\iooc\ir--oooco ooo5a>r--o
CO coooincoojc\icDO ^ O ^ C J O O C D
C\joocûCT)0to T- 1- m T-
cu
u)coooC d CM
T3O)en
CO
Q .
COQ)
03 i-í (/) «.u
ó cn¿ - "-*T' i " " ^ -n íA
tr ce
CD
C
en CO.03 (u g .
3 ><
w w 551
What Is Special About Marketing Organic Products? / 47
Appendix CElasticity Caiculations for Core and
Noncore Consumers Based onOrganic intrinsic Preference
We formulate the organic buying process using a nestedlogit model. In this model, the consumer first decides to buyin the category and then decides to buy either the organic orthe conventional product.
Category incidence
Consumer h at time t is faced with the decision of whetherto purchase in category k or not (i.e., category incidence),the utility of which is given as follows (e.g., Bucklin andLattin 1991):
(Al) + Gj + e4
where Inv^t, IncVal^t, and Ch are the inventory, inclusivevalue, and the consumption rate, respectively. We calculateinventory and consumption rates using the following for-mula (e.g., Ailawadi et al. 2007):
(A2) -Ch,,and
= Invht 3
where q t denotes the quantity purchased, Cht is the con-sumption of the household h at time t, and f is the parame-ter associated with the consumption flexibility.
Product Choice
Given incidence, consumer h at time t is faced with thechoice of buying product i (organic or conventional), theutility of which is given by the following:
(A3) ihi = ßiih + ßaihRPit + ßsih Assorti, +
where ßni, is the intrinsic preference of the consumer topurchase the organic or the conventional product and RPjtand Assorti, ^re regular price and assortment of the respec-tive organic and conventional products. We use the standardidentification restrictions to estimate the model using theno-buy and conventional category as the base case in thecategory incidence and product choice equations.
Estimation
We incorporate individual-level parameters whenever pos-sible and use the simulated maximum likelihood estimationtechnique with Halton draws (Train 2003) for estimation. Werun this estimation for 30 draws for each of the individual-level parameters.
We randomly select 700 consumers in each category toestimate the aforementioned model. Note that the con-
TABLE C1Elasticities for the Core and Noncore Segments Based on Nested Logit and Number of Organic Voiume
Purchases
Category
Price ElasticityOveraii Price ElasticityMilkYogurtPasta sauceCrackersPizza: FrozenCerealOatsCarrotsStrawberriesLaundry detergent
Assortment ElasticityOveraii Assortment ElasticityMilkYogurtPasta sauceCrackersPizza: FrozenCerealOatsCarrotsStrawberriesLaundry detergent
OverallElasticity
-3.15-3.45-2.48-2.31-1.97-2.98-2.85-2.37-3.68-4.01-1.52
2.313.123.292.322.192.151.621.741.371.351.44
Nested Logit
CoreSegmentElasticity
-2.21-2.37-1.87-1.61-1.31-2.02-1.92-1.48-2.76-2.79-1.15
1.321.461.871.081.551.06.98
1.27.70.87.78
NoncoreSegmentElasticity
-3.90-4.35-3.03-3.23-2.59-3.96-3.82-3.17-4.75-5.49-2.27
3.073.954.662.694.032.312.543.091.891.962.09
Organic Volume Purchases
OveraiiElasticity
-3.22-3.57-2.59-2.25-2.07-2.96-2.90-2.43-3.83-4.19-1.75
2.403.253.432.212.232.421.741.831.591.391.61
CoreSegmentElasticity
-2.27-2.29-1.84-1.50-1.43-2.08-2.01-1.62-2.74-2.90-1.21
1.351.841.901.02.92
1.291.091.31.77.95.97
NoncoreSegmentElasticity
-3.97^ .58-3.35-3.31-2.67-3.85-3.78-3.25-4.99-5.61-2.25
3.124.324.972.784.243.083.033.262.152.332.38
48 / Journal of Marketing, January 2013
sumers in each category are distinct because different con-sumers exhibit different purchase patterns across the cate-gories. On obtaining the estimates, we classify consumersby their organic intrinsic preference parameter (ßnh) in eachof the 56 categories. Thus, consumers with above-mean val-ues of ßiih constitute the core segment, and those withbelow-mean values are the noncore segment. We obtain theindividual-level parameters by making draws from themean estimates and standard deviations (Train 2003) andaveraging them. We use 10,000 draws for this purpose.
Using this classification and relevant variables, we con-duct the VAR analysis separately for each segment across
all 56 categories. We calculate the price and assortment
elasticity using the impulse response functions. We com-
pare our original approach of classifying consumers simply
by their purchases (as we describe in the main part of the
article) with the more involved method of classifying them
by their organic intrinsic preference parameter (as we
describe herein). Table Cl reports the overall elasticity
(simple average across 56 categories used for analysis) as
well as elasticities for some prototypical categories. As the
table shows, the elasticities using either method do not sig-
nificantly differ from one another.
REFERENCESAertsens, Joris, Koen Mondelaers, and Guido Van Huylenbroeck
(2009), "Differences in Retail Strategies on the EmergingOrganic Market," British Food Journal, 111 (2), 138-54.
Ailawadi, Kusum L., Karen Gedenk, Christian Lutzky, and ScottA. Neslin (2007), "Decomposition of the Sales Impact ofPromotion-Induced Stockpiling," Journal of MarketingResearch, 44 (August), 450-67.
Allenby, Greg M. and Peter E. Rossi (1991), "Quality Perceptionsand Asymmetric Switching Between Brands," Marketing Sci-ence, 10(3), 185-204.
Bauer, Hans H., Daniel Heinrich, and Daniela B. Schäfer (2012),"The Effects of Organic Labels on Global, Local, and PrivateLabels: More Hype Than Substance?" Journal of BusinessResearch, (published electronically January 4), [DOI: 10.016/j.jbusres.2011.12.028].
Beverage Industry (2010), "Protein Appeals to the Masses," 101(2), 14-20.
Bijmolt, Tammo H.A., Harald J. van Heerde, and Rik G.M. Pieters(2005), "New Empirical Generalizations on the Determinantsof Price Elasticity," Journal of Marketing Research, 42 (May),141-56.
Blattberg, Robert and Kenneth Wisniewski (1989), "Price-InducedPatterns of Competition," Marketing Science, 8 (4), 291-309.
Boatwright, Peter and Joseph C. Nunes (2001), "Reducing Assort-ment: An Attribute-Based Approach," Journal of Marketing, 65(July), 50-63.
Bonti-Ankomah, Samuel and Emmanuel Yiridoe (2006), "Organicand Conventional Eood: A Literature Review of the Economicsof Consumer Perceptions and Preferences," research report.Organic Agriculture Center of Canada, Nova Scotia.
Bourn, Diane and John Prescott (2002), "A Comparison of theNutritional Value, Sensory Qualities and Eood Safety of Organ-ically and Conventionally Produced Foods," Critical Reviewsin Food Science and Nutrition, 42 (1), 1-34.
Bowman, Douglas and Hubert Gatignon (1995), "Determinants ofCompetitor Response Time to a New Product Introduction,"Journal of Marketing Research, 32 (February), 42-53.
Brandweek (2009), "Despite Economic Dip, Organic Food SalesSoar," 49 (23), 6.
Bronnenberg, Bart, Vijay Mahajan, and Wilfried R. Vanhonacker(2000), "The Emergence of Market Structure in New Repeat-Purchase Categories: The Interplay of Market Share andRetailer Distribution," Journal of Marketing Research, 37(February), 16-31.
and Luc Wathieu (1996), "Asymmetric Promotion Effectsand Brand Positioning," Marketing Science, 15 (4), 379-94.
Bucklin, Randolph E. and James M. Lattin (1991), "A Two-StateModel of Purchase Incidence and Brand Choice," MarketingScience, 10 (I), 24-39.
Campo, Katia, Els Gijsbrechts, and Patricia Nisol (2000),"Towards Understanding Customer Response to Stockouts,"Journal of Retailing, 76 (2), 219^2.
Ceslfo (2011), "Ifo Industry Analysis: Organic Produce Still HasHigh Growth Potential," (April 8), (accessed July 23 2012),[available at http://www.cesifo-group.de/portal/page/portal/ifoHome/e-pr/elpz/_generic_press_item_detail?p_itemid=15495283].
Chain Store Age (2009), "Forty Under 40: Small Chains Flourish,"85(1), 22-27.
Ciu, Lina (2008), "Growing Naturally in Alabama: Needs andPossibilities," doctoral dissertation. Agricultural Economicsand Rural Society, Auburn University.
Davies, Anne, Albert J. Titterington, and Clive Cochrane (1995),"Who Buys Organic Food? A Profile of the Purchasers ofOrganic in Northern Ireland," British Food Journal, 97 (10),17-23.
Dekimpe, Mamik and Dominique Hanssens (1995), "The Persis-tence of Marketing Effects on Sales," Marketing Science, 14(1), 1-21.
and (1999), "Sustained Spending and PersistentResponse: A New Look at Long-Term Marketing Profitability,"Journal of Marketing Research, 36 (November), 397^12.
Delvecchio, Devon, David H. Henard, and Traci H. Freling(2006), "The Effect of Sales Promotion on Post-PromotionBrand Preference; a Meta-Analysis," Journal of Retailing, 82(3), 203-213.
Dimitri, Carolyn and Lydia Oberholtzer (2009), "Marketing U.S.Organic Foods: Recent Trends from Farms to Consumers,"Economic Information Bulletin, EIB-58. Washington, DC:Economic Research Service, United States Department ofAgriculture.
Drug Store News (2008), "As Green Movement Grows, OrganicProduct Standards Emerge," (April 21), 53.
Durham, Catherine A. and Diego Andrade (2005), "Health vs.Environmental Motivation in Organic Preferences and Pur-chases," paper presented at the American Agricultural Eco-nomics Association Annual Meeting, Providence, RI (July24-27).
Enders, Walter (2004), Applied Econometric Time Series. NewYork: John Wiley & Sons.
Food Marketing Institute (2009), Shopping for Health. Arlington,VA: Food Marketing Institute.
Fotopoulos, Christos and Athanasios Krystallis (2002), "Purchas-ing Motives and Profiles of the Greek Organic Consumer: ACountry Survey," British Food Journal, 104 (9), 730-65.
Franses, Philip H. (2005), "Diagnostics, Expectations, and Endo-geneity," Journal of Marketing Research, 42 (February),27-29.
What Is Speciai About Marketing Organic Products? / 49
and Rutger van Oest (2007), "On the Econometrics of theGeometric Lag Model," Economics Letters, 95 (2) 291-96.
Glaser, Lewrene K. and Gary D. Thompson (2000), "Demand forOrganic and Conventional Beverage Milk," paper presented atthe Western Agricultural Economics Association Annual Meet-ings, Vancouver, BC (June 29-July 1).
Granger, C.W.J. (1969), "Investigating Causal Relations byEconometric Models and Cross-Spectral Methods," Economet-nca, 37 (3), 424-38.
Grunert, Suzanne C. and Hans J0m Juhl (1995), "Values, Environ-mental Attitudes, and Buying of Organic Eoods," Journal ofEconomic Psychology, 16 (10), 39-62.
Han, Min C. (1989), "Country Image: Halo and Summary Con-struct.'" Journal of Marketing Research, 26 (May), 222-29.
The Hartman Group (2008), The Many Faces of Organic Goods.Bellevue, WA: The Hartman Group.
Hendel, Igal and Aviv Nevo (2006), "Sales and Consumer Inven-tory," The RAND Journal of Economics, 'il (3), 543-61.
Henderson, Ty and Neeraj Arora (2010), "Promoting BrandsAcross Categories with a Social Cause: Implementing Effec-tive Embedded Premium Programs," Journal of Marketing, IA(November), 41-60.
Hui, Sam K., Eric T. Bradlow, and Peter S. Fader (2009), "TestingBehavioral Hypotheses Using an Integrated Model of GroceryStore Shopping Path and Purchase Behavior," Journal of Con-sumer Research, 36 (3), 478-93.
Janssen, Meike and Ulrich Hamm (2012), "Product Labellingin the Market for Organic Food: Consumer Preferences andWillingness-to-Pay for Different Certification Logos," FoodQuality and Preference, 25 (1), 9-22.
Johansen, Soren, Rocco Mosconi, and Bent Nielsen (2000),"Cointegration Analysis in the Presence of Structural Breaks inthe Deterministic Trend," Econometrics Journal, 3 (2), 1-34.
Kamakura, Wagner A. and Gary J. Russell (1989), "A ProbabilisticChoice Model for Market Segmentation and Elasticity Struc-ture," Journal of Marketing Research, 26 (November), 379-90.
Kiesel, Kristin and Sofia B. Villas-Boas (2007), "Got OrganicMilk? Consumer Valuations of Milk Labels After the Imple-mentation of the USDA Organic Seal," Journal of Agricultural& Food Industrial Organization, 5 (4), 1-38.
Kotier, Philip (2011), "Reinventing Marketing to Manage theEnvironmental Imperative," Journal of Marketing, 75 (July),132-35.
Kozup, John C , Elizabeth H. Creyer, and Scot Burton (2003),"Making Healthful Food Choices: Influence of Health Claimsand Nutrition Information on Consumers' Evaluations of Pack-aged Food Products and Restaurant Menu," Journal of Market-ing, 67 (April), 19-34.
Makatouni, Aikaterini (2002), "What Motivates Consumers toBuy Organic Food in the UK? Results from a QualitativeStudy," British Food Journal, 104 (3-5), 345-52.
Michelsen, Jahannes, Ulrich Hamm, Els Wynen, and Eva Roth(1999), "The European Market for Organic Products: Growthand Development," Organic Farming in Europe: Economicsand Policy, Vol. 7. Stuttgart, DE: University of Hohenheim.
Narasimhan, Chakravarthi, Scott A. Neslin, and Subrata K. Sen(1996), "Promotional Elasticities and Category Characteris-tics," Journal of Marketing, 60 (April), 17-31.
Ngobo, Paul-Valentin (2011), "What Drives Household Choice ofOrganic Products in Grocery Stores?" Journal of Retailing, 87(1), 90-100.
Nijs, Vincent, Mamik Dekimpe, Jan-Benedict Steenkamp, andDominique M. Hanssens (2001), "The Category-DemandEffects of Price Promotions," Marketing Science, 20 (1), 1-22.
Nowlis, Stephen M. and Itamar Simonson (2000), "Sales Promo-tions and the Choice Context as Competing Influences on Con-sumer Decision Making," Journal of Consumer Psychology, 9(1), 1-16.
Oberholtzer, Lydia, Catherine Green, and Enrique Lopez (2006),"Organic Poultry and Eggs Capture High Price Premiums andGrowing Share of Specialty Markets," Research Report LDP-M-150-01, United States Department of Agriculture.
Organic Trade Association (2009a), Organic Industry Survey.Greenfield, MA: Organic Trade Association.
(2009b), U.S. Families' Organic Attitudes and BeliefsStudy. Greenfleld, MA: Organic Trade Association.
Pauwels, Koen (2004), "How Dynamic Consumer Response,Competitor Response, Company Support and Company InertiaShape Long-Term Marketing Effectiveness," Marketing Sci-ence, 23 (4), 596-610.
, Dominique M. Hanssens, and S. Siddarth (2002), "TheLong-Term Effects of Price Promotions on Category Incidence,Brand Choice and Purchase Quantity," Journal of MarketingResearch, 34 (November), 421-39.
and Shuba Srinivasan (2004), "Who Beneflts from StoreBrand EntiyTMarketing Science, 23 (Summer), 364-90.
Pesaran, Hashem H. and Yongcheol Shin (1998), "GeneralizedImpulse Response Analysis in Linear Multivariate Models,"Economics Letters, 58 (1), 17-29.
Progressive Grocer (2008), "Outside the Bubble," 87 (2), 44-47.(2009), "Evolving Organics," 88 (7), 58-62.
Raju, Jagmohan (1992), "The Effect of Price Promotions on Vari-ability in Product Category Sales," Marketing Science, 11(Summer), 207-220.
Roheim, Cathy A. and Ronald D'Silva (2009), "Illustration of theU.S. Organic Agricultural Produce Price Premiums: Implica-tions for Eco-Labeled Sea-Food Price Premiums," workingpaper, URI Sustainable Seafood Initiative, University of RhodeIsland.
Rokeach, Milton (1968), "The Role of Values in Public OpinionResearch," Public Opinion Quarterly, 32 (4), 547-59.
Sethuraman, Raj and V. Srinivasan (2002), "The AsymmetricShare Effect: An Empirical Generalization on Cross-PriceEffects," Journal of Marketing Research, 39 (August), 379-86.
, , and Doyle Kim (1999), "Asymmetric and Neigh-borhood Cross-Price Effects: Some Empirical Generaliza-tions," Marketing Science, 18(1), 2 3 ^ 1 .
Silverstone, Rob (1993), "Organic Farming: Food for the Future?"Nutrition and Food Science, 93 (5), 10-14.
Sivakumar K. and S.P. Raj (1997), "Quality Tier Competition:How Price Change Influences Brand Choice and CategoryChoice," Journal of Marketing, 61 (July), 71-84.
Srinivasan, Shuba, Koen Pauwels, Dominique Hanssens, andMamik Dekimpe (2004), "Do Promotions Benefit Retailers,Manufacturers, or Both?" Management Science, 50 (5),617-29.
Supermarket News (2011), "Moving Ahead," (November29),12-17.
TABS (2012), "Organic Products on the Rise," (January 30),(accessed July 17, 2012), [available at http://www.tabsgroup.com/2012/01 /organic-products-on-the-rise/].
Thompson, Gary D. (1998), "Consumer Demand for OrganicFoods: What We Know and What We Need to Know," Ameri-can Journal of Agricultural Economics, 80 (5), 1113-18.
Tondel, Fabien and Timothy Woods (2006), "Supply Chain Man-agement and the Changing Nature of the US Organic ProduceMarkets," paper presented at the AAEA Annual Meeting, LongBeach, CA (July 23-26).
Train, K. (2003), Discrete Choice Methods with Simulation. Cam-bridge, MA: Cambridge University Press.
Trusov, Michael, Randolph E. Bucklin, and Koen Pauwels (2009),"Effects of Word-of-Mouth Versus Traditional Marketing:Findings from an Internet Social Networking Site," Journal ofMarketing, 73 (September), 90-102.
Verhoef, Peter C. (2005), "Explaining Purchases of Organic Meatby Dutch Consumers," European Review of Agricultural Eco-nomics, 32 (2), 245-67.
50 / Journal of Marketing, January 2013
Wertenbroch, Klaus (1998), "Consumption Self-Control by Organic Agriculture: Sustainability, Markets and Policies.Rationing Purchase Quantities of Virtue and Vice," Marketing Paris: OECD Publishing, 257-71.Science, 17 (4), 317-37. Zanoli, Rafaele and Simona Naspetti (2002), "Consumer Motiva-
Wier, Mette, Lars G. Hansen, Laura M.Anderson, and Katrin Mil- tions in the Purchase of Organic Food: A Means-Endlock (2003), "Consumer Preferences for Organic Foods," in Apptoach," British Food Journal, 104 (8), 643-53.
What Is Special About Marketing Organic Products? / 51
Copyright of Journal of Marketing is the property of American Marketing Association and its content may not
be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles for individual use.
![1,2 2,3 2,4 1 1,2,3,4,5,nsmn1.uh.edu/trlee/2017 pdfs/ContentServer.pdf · acids [9–13]. Significant progress has been achieved in magnetic biosensor technology since the concept](https://img.pdfslide.net/doc/110x75/5e23120680138b721830098b/12-23-24-1-12345nsmn1uhedutrlee2017-pdfs-acids-9a13-signiicant.jpg)