• Selected
• Entire List

Abstract:
We describe industrial market structure using a unique database spanning 31 consumer package goods (CPG) industries, 39 months, and the 50 largest US metropolitan markets. We organize our description of market structure around the notion that firms can improve brand perceptions through advertising investments, as in Sutton's endogenous sunk cost theory. In the data, observed advertising levels escalate (i.e. larger brands) in larger markets while the number of advertised brands within an industry remains stable. Correspondingly, observed concentration levels in advertising-intensive industries are bounded away from zero irrespective of market size. For two industries, we collect historic order-of-entry data. The geographic distribution of entry is found to account for the levels, rank-orders and covariation in the geographic distribution of brand shares, perceived brand qualities and advertising effort. Interestingly, alternative potential sources of geographic asymmetry on both the supply and demand sides do not mimic the geographic patterns of shares. In general, our findings highlight several striking persistent geographic patterns in the industrial market structures of CPG industries.

Endogenous Sunk Costs, Market Structure, Advertising, Entry, Geography

Abstract:
We study persistence in the geographic variation in market shares of branded goods in consumer packaged goods industries across 50 U.S. city-markets. We match scanner data on local market shares and survey data on local quality perceptions for the largest brands in 34 consumer packaged goods industries. These data are then matched with historic information on the year and US city-market in which each brand was first launched. We find that these consumer brands have persistently higher market shares in markets closest to their respective cities-of-origin than in markets farthest from their respective cities-of-origin, where they were typically launched later. For 6 of the 34 industries, we collected more complete historic entry data with which we can determine the local order of entry among the top brands in each of the 50 U.S. city-markets. We find a persistent effect from differences in the order-of-entry of competing brands on their current relative brand shares and quality perceptions across US cities. The historic order of entry also appears to correlate with the current rank-order of brand shares across cities, leading to large asymmetries across markets in brand shares and in quality perceptions. This persistence is particularly striking since many of the brands studied herein originated during the mid-to-late 19th and early 20th centuries, roughly a century prior to the sample period of the market share and quality perception data.

Abstract:
We review the importance of structural modeling for the purposes of simulation of the effects of changes in marketing policy. We pay particular attention to the aspects of structural modeling which present unique challenges for researchers in marketing. We also catalogue some of the advantages that marketing researchers enjoy in the development and estimation of structural models. This paper was orginially commissioned as a comment on Philip Hans Franses' paper, 'On the Use of Marketing Models for Policy Simulation', (forthcoming, Journal of Marketing Research).

Structural models, policy evaluation, marketing instruments

Abstract:
Using aggregate product search data from Amazon.com, we jointly estimate consumer information search and online demand for durable goods. To estimate demand and search primitives, we introduce an optimal sequential search process into a model of choice and treat the observed market-level product search data as aggregations of individual-level optimal search sequences. The model builds on the dynamic programming framework by Weitzman (1979) and combines it with a choice model. At the individual level, the model has several attractive properties including closed-form expressions for the probability distribution of alternative search sets and breaking the curse of dimensionality. Using numerical experiments, we verify the model's ability to identify consumer tastes and search cost from product search data. Empirically, the model is applied to the camcorder online market and is used to answer manufacturer questions about market structure and competition, and to address policy maker issues about the effect of recommendation tools on consumer surplus outcomes. We find that consumer search for camcorders is typically limited to about 10 choice options, and that this affects the estimates of own and cross-elasticities. We also find that the vast majority of the households benefit from the Amazon.com's product recommendations via lower search costs.

Demand Models, Consideration Sets, Optimal Sequential Search, Durable Goods

Abstract:
In direct competition between national brands of consumer packaged goods (CPG), one brand often has a large local share advantage over the other despite that the products are similar. I present an explanation for these large and persistent advantages in the context of local competition on perceived quality or brand image. The main result of the analysis is a relation between varying degrees of horizontal product differentiation and local share advantages. Namely, I find that local share advantages can be sustained especially if competing brands are objectively similar. Conversely, local share advantages can not be sustained if brands are dissimilar. This paper provides two independent intuitions for this result. First, if brands are objectively similar, different levels of investments in local quality perceptions can co-exist in the same market. Specifically, if it pays to do so, early movers will invest in high perceived quality. Late movers have reduced incentives to invest because of subsequent demand sharing and price competition. Second, if the perceived quality advantages are geographically distributed across competitors, the above argument is reinforced by multimarket contact. Even if local "brand building" is free, firms have an incentive to sustain asymmetric market shares because, holding total demand constant, multimarket profits are increasing in share variation, i.e., monopoly power, across regions. This increase is steeper when the products are similar, because price competition looms large.

Abstract:
We propose a new method to model consumers' consideration and choice processes. We develop a parsimonious probit type model for consideration and a multinomial probit model for choice, given consideration. Unlike earlier models of consideration ours is not prone to the curse of dimensionality, while we allow for very general structures of unobserved dependence in consideration among brands. In addition, our model allows for state dependence and marketing mix effects on consideration.Unique to this study is that we attempt to establish the validity of existing practice to infer consideration sets from observed choices in panel data. To this end, we use data collected in an on-line choice experiment involving interactive supermarket shelves and post-choice questionnaires to measure the choice protocol and stated consideration levels. We show with these experimental data that underlying consideration sets can be successfully retrieved from choice data alone and that there is substantial convergent validity of the stated and inferred consideration sets. We further find that consideration is a function of point-of-purchase marketing actions such as display and shelf space, and of consumer memory for recent choices.Next, we estimate the model on IRI panel data. We have three main results. First, compared with the single-stage probit model, promotion effects are larger and are inferred with smaller variances when they are included in the consideration stage of the two-stage model. Promotion effects are significant only in the two-stage model that includes consideration, whereas they are not in a single-stage choice model. Second, the price response curves of the two models are markedly diferent. The two-stage model offers a nice intuition for why promotional price response is different from regular price response. In addition and consistent with intuition, the two-stage model also implies that merchandizing has more effect on choice among those who did not buy the brand before than among those who already did. It is explained why a single-stage model does not harbor this feature. In fact, the single-stage model implies the opposite for smaller or more expensive brands. Third, we find that the consideration of brands does not covary greatly across brands once we take account of observed effects. Managerial implications and future research are also discussed.

Consideration, choice, probit models

Abstract:
Using transactional data, we estimate a structural model of demand and supply of sport utility vehicles (SUVs). Consumers in our model choose among SUVs at specific dealerships. By expanding the product space to include the location of point of sales, we can study the size and shape of market areas - defined as the geographic area where demand for an alternative is highest - for each car model, dealership, and manufacturer. By combining demand estimates with profit optimizing behavior of manufacturers, our model is able to provide profit projections and shifts in market areas from managerial decisions such as the relocation of a dealership or the removal of a car model from the market. We empirically apply our model to the case of SUVs in San Diego using a dataset that contains transaction information about locations of dealers and consumers, manufacturer prices, and retail prices. We find high disutility for travel to retailers, which geographically limits preferences to nearby alternatives. We show that most dealers have their own private 'backyard' of demand that is shared with a small set of other alternatives. As predicted by the literature on spatial competition, the size and shape of the market area is strongly dependent on the competitors location. In the majority of cases the highest spatial density of demand is not at the location of the dealer but at locations that are furthest from direct substitutes. We find that discounting prices by 10% leads to market area expansion of 5 miles.

automobile industry, spatial competition, structural models of demand and supply

Abstract:
In this paper, we propose a method to visualize online consumer search in so-called product search maps. Manufacturers can use these maps to understand how consumers search for competing products prior to choice, including how information acquisition and product search is organized along brand-, product attributes-, and/or price related search strategies. The product search maps also inform manufacturers about the competitive structure in the industry and about the contents of consumer consideration sets. Our proposed method first defines a product search network, consisting of the products and links that designate if a product is searched conditional on searching other products. We next model this network using a stochastic, hierarchical and asymmetric multidimensional scaling framework and decompose the product locations as well as the product-level influences using product attributes. The advantages of the approach are two-fold. First, we simultaneously visualize the positions of products and the direction of consumer search over products in a perceptual map of “search proximity.” Second, we explain and relate the dimensions of the map using observed product attributes. We empirically apply our approach to consumer search for digital camcorders at Amazon.com and provide a number of managerial implications.

brand networks, asymmetric MDS, product search, hierarchical Bayes estimation

Abstract:
This paper empirically investigates the determinants of supermarkets' assortment composition and assortment size decisions. We define measures of assortment similarity and use a unique cross-sectional store-level data set to analyze how assortment composition and assortment size are related to underlying factors that describe local store clientele, local competitive structure, and the retail-organization's corporate structure. We then examine whether cross-store variation in assortment is most strongly related to factors describing demand or supply. From an analysis covering two important product categories, cereals and colas, we find that corporate- and ownership-structure of retail stores are the most important drivers of differences in assortment composition. We also find that supermarkets tailor their assortment by considering relevant local demographic dimensions, but this effect is an order of magnitude smaller than that of corporate ownership structure. We conclude that marginal assortment decisions are mostly cost driven, and are made at the corporate level. We discuss our findings in the context of manufacturers designing distribution policies.

Retailing, assortment, supermarket, dyadic data

Abstract:
We analyze a multimillion dollar, three-year field study sponsored by five firms to assess whether DVRs impact consumers’ shopping behavior for advertised and private label goods. A large sample of households received an offer for a free DVR and service and close to 20% accepted. We observe each household's shopping history for 48 consumer packaged goods categories during the 13 months prior and the 26 months following the DVR offer. We fail to reject the null of no DVR treatment effect on household spending on advertised branded or private label goods, either 1 or 2 years after the DVRs are shipped. Our predicted DVR effect is tightly centered around 0, suggesting the data may have sufficient power to identify a true null effect. Using advertising exposure information for seven of the brands in the study, we offer suggestive evidence that ad-skipping occurs for a relatively small fraction of the total television content viewed. Other potential explanations for the lack of a DVR effect are also discussed.

Digital Video Recorder, advertising, field study, brand, consumer packaged goods