L'image d'un magasin influence-t-elle le niveau de dépense des ménages?

Store image has been largely studied in the marketing literature. While past studies have shown that store image influences store patronage intentions, very little research has examined its effects on actual purchase behaviors. Extending research on store image, we use basket size as a dependent variable, account for marketing actions and consumers’ perceptions, and control for both consumer and store heterogeneity. Using scanner panel data and longitudinal survey data from two markets over three years, we find evidence that a model with store image (preference data) does improve model fit statistics thus providing a better understanding of consumer in-store spending than a model based on scanner data variables alone. Some of the relationships are consistent with prior studies (e.g. price image, locational convenience). However, other relationships are different. For example, we find that, contrary to Van Heerde et al (2008), produce quality has a negative influence on basket size. Contrary to studies based on self-reports (e.g. Arnold et al. 1983), we find that the perceived variety of the assortment has a negative influence on spending. The consistency of the results across the two markets suggests that store image perceptions are a valuable input to consumer spending models.

Comment vendre les produits bio en supermarché? Le rôle du marketing mix

The market for organically produced goods has grown at a double-digit rate in the last two decades. Yet, while there is a growing literature on the economics of the organic food products industries, the marketing literature has provided only few insights into the commercialization of organic products. Most of the studies have concerned the consumers’ motives and demographic profiles with mixed findings. The present study examines the effects of marketing actions on the demand for organics in a grocery supermarket context. Using predictions from the literature on the effects of marketing-mix variables, we model and test the impact of five variables: (1) brand ownership (store vs. national brand), (2) price level, (3) feature advertising, (4) display activity, and (5) distribution breadth, and competitive effects from conventional products. We allow for heterogeneity across consumers and product categories. Contrary to prior studies, we estimate model parameters on actual purchase data from two markets over three years. We find that consumers are less likely to buy organic store brands, organic brands that are on promotion, and widely distributed. Price, which serves as a cue for the quality of organics, has an inverted U-shaped effect on the demand for organic products. Consumer (e.g. age, income) and category characteristics (e.g. concentration, expensiveness) moderate the effects of marketing mix variables.

Les données relatives à la satisfaction des clients peuvent-elles aider les analystes financiers à réduire leurs erreurs de prévision?

(Recherche menée avec JF casta et Olivier Ramond et financée par le marketing Science Institute)
Résumé managérial

A growing number of papers have been reporting significant associations between customer satisfaction and various firm performance metrics such as profitability and Tobin’s q. However, while marketers generally agree on its impact, it is unclear whether investors use information on customer satisfaction when they predict firm performance. Much of the evidence on the effects of customer satisfaction comes from direct analyses of its influence on financial performance, bypassing the analysis of whether and how this information translates into stock pricing and valuation. An important way customer satisfaction efforts could translate into the stock pricing is through the financial analysts’ forecasts and recommendations. Analysts play an important role as information intermediaries for the investors. Financial analysts aggregate complex information for other market participants (e.g. macroeconomic data, business plans, and possibly non-financial information) and provide (1) earnings forecasts, (2) price targets, and (3) buy-sell-hold recommendations.
However, the extent to which analysts use non-financial information has received limited attention despite some frustration over traditional financial statements. In this study, we examine the value relevance of customer satisfaction to financial analysts, where value relevance refers to the usefulness of customer satisfaction to analysts when preparing their earnings forecasts. We assembled a dataset of companies studied in the American Customer Satisfaction Index (University of Michigan), which also appear on the Institutional Brokers Estimate System (I/B/E/S) files. By combining these sources, we were able to analyze the forecast errors of over 1300 analysts following 89 companies in 22 industries. We developed a Latent Factor Regression Model, which allows us to estimate the mean effect of customer satisfaction but also the analyst heterogeneity and industry heterogeneity through latent factors and factor scores.

Our findings indicate that customer satisfaction is negatively associated with analysts’ forecast errors. The influence of customer satisfaction varies across analysts and industries. Specifically, we found in industries such as the Internet Software & Services (e.g. Ebay), Managed Health Care (e.g. Aetna), Integrated Telecommunications Services (e.g. AT&T), customer satisfaction has the largest negative impact on earnings forecast errors. In other industries, such as the Broadcasting & Cable TV, the Airlines, and the Electric Utilities, customer satisfaction has a weaker impact. The structure of these industries provides the explanation. Industries, which are less sensitive to customer satisfaction, such as Electric utilities, are in concentrated markets. The former industries are differentiated markets where repeat purchase is sensitive to customer satisfaction. These results are robust even if we examine the influence of the unanticipated component of customer satisfaction. An increase in customer satisfaction has a negative association with the forecast errors and a decrease in customer satisfaction has a positive association with earnings forecast errors. In sum, our findings suggest that by neglecting customer satisfaction information, analysts have deprived themselves of an important proxy of non-financial information.

Les Prix à terminaison 9 sont-ils réellement efficaces?

Nous avons réalisé une étude destinée à quantifier l'effet des prix à terminaison 9 sur le comportement réel des clients en magasin à travers plusieurs catégories de produits. La recherche a été réalisée dans le cadre d'un partenariat avec la société MarketingScan (www.marketingscan.fr) avec deux collègues Patrick Legoherel & Nicolas Gueguen Voici le résumé, l'article complet est disponible auprès du premier auteur:

Retailers largely adopt nine-ending prices and these prices have attracted greater attention from researchers in marketing. Despite this increased interest, very few empirically based studies have tried to quantify the effects of nine-ending prices on consumer actual behaviors. In this article, we investigated the effects of nine-ending pricing on consumer purchase quantity. We distinguished between different types of nine-ending while controlling for the rounded prices and other marketing-mix variables. We accounted for category and household unobserved differences and examined the extent to which category structure moderates the impact of nine-ending prices. We conducted our analysis on 45,017 SKUs (from 5,160 brands) in 117 product categories in 12 stores bought by 1869 households (yielding over 1315189 observations). We find that the effects of nine-ending prices depend upon the way the nine-ending price is framed (e.g. 0.99 vs. 1.39 vs. 9). A ninety-nine-ending price conveys a more negative image than a 9-rounded price. We also identify a set of conditions where nine-ending prices are more effective. For example, we find that ninety-nine-ending prices are less effective in promotional categories but more effective in concentrated categories. We discuss the implications of our research for retailers and researchers.