• Selected
• Entire List

Abstract:
Many markets have historically been dominated by a small number of best-selling products. The Pareto Principle, also known as the 80/20 rule, describes this common pattern of sales concentration. However, by greatly lowering search costs, information technology in general and Internet markets in particular have the potential to substantially increase the collective share of niche products, thereby creating a longer tail in the distribution of sales.

This paper investigates how demand-side factors contribute to the Internet's ¿Long Tail¿ phenomenon. It first models how a reduction in search costs will affect the concentration in product sales. Then, by analyzing data collected from a multi-channel retailing company, it provides empirical evidence that the Internet channel exhibits a significantly less concentrated sales distribution, when compared with traditional channels. The difference in the sales distribution is highly significant, even after controlling for consumer differences. Furthermore, the effect is particularly strong for individuals with more prior experience using the Internet channel. We find evidence that Internet purchases made by consumers with prior Internet experience are more skewed toward obscure products, compared with consumers who have no such experience. We observe the opposite outcome when comparing purchases by the same consumers through the catalog channel. If the relationships we uncover persist, the underlying trends in technology and search costs portend an ongoing shift in the distribution of product sales.

search cost, product variety, concentration, long tail, Internet, electronic commerce

Abstract:
We use a controlled field experiment to investigate the dynamic effects of retail advertising. The experimental design overcomes limitations hindering previous investigations of this issue. Our study uncovers dynamic advertising effects that have not been considered in previous literature. We find that current advertising does affect future sales, but surprisingly, the effect is not always positive; for the firm’s best customers, the long-run outcome may be negative. This finding reflects two competing effects: brand switching and intertemporal substitution. We also find evidence of cross-channel substitution, with the firm’s best customers switching demand to the ordering channel that corresponds to the advertising.

Abstract:
We present findings from a controlled field experiment that allows us to investigate the dynamic effects of retail advertising. The experimental design overcomes limitations that have hindered previous investigations of this issue. Our study uncovers dynamic advertising effects that have not been previously considered in the literature. We find that current advertising does affect future sales but surprisingly, the affect is not always positive; for the firm's best customers the long-run outcome may be negative. We argue that this finding reflects two competing effects: brand-switching and intertemporal substitution. The study also provides evidence of cross-channel substitution, with the firm's best customers switching demand to the ordering channel that corresponds to the advertising.

Advertising, Direct Mail, Field Experiment, Internet, Catalog, long-run demand

Abstract:
Economists typically assume that demand curves are downward sloping. We present evidence that increasing the price of an item from $44 to $49 may increase unit demand by up to 30%. This effect is substantial, has broad application, is easily replicated, and contradicts the downward-sloping property of demand functions. The paper offers a rational-actor explanation for the effect by arguing that price endings may serve an informational role, signaling to customers which prices are low compared to other market prices. In equilibrium customer reliance on price endings is consistent with optimal firm behavior. We explain why firms prefer to use $9 endings on items with low (relative) prices. As a result, $9 price endings lead to more favorable customer price perceptions and increased customer demand. Retailers face a trade-off that regulates how many $9 endings they use. The effect is moderated by proliferation, so that the demand increase is smaller when more products have 9-digit endings. Predictions from the model are tested in three randomized field tests that measure the purchasing behavior of actual customers in two mail order catalogs. Together, the model and data yield three conclusions. First, using a $9 price ending on a product increases demand for that product. Second, $9 endings have a smaller effect when more products have them. Third, the $9 ending effect is further moderated when sale signs also inform customers about relative price levels. A corollary of the second conclusion is that firm profits are concave in the number of items that have $9 endings, and it is this concavity that makes use of $9 endings self-regulating.

Abstract:
Choice-based conjoint analysis (CBC) is used widely in marketing for product design, segmentation, and marketing strategy. We propose and test a new "polyhedral" question-design method that adapts each respondent's choice sets based on previous answers by that respondent. Individual adaptation appears promising because, as demonstrated in the aggregate customization literature, question design can be improved based on prior estimates of the respondent's partworths - information that is revealed by respondents' answers to prior questions. The otherwise impractical computational problems of individual CBC adaptation become feasible based on recent polyhedral "interior-point" algorithms, which provide the rapid solutions necessary for real-time computation. To identify domains where individual adaptation is promising (and domains where it is not), we evaluate the performance of polyhedral CBC methods with Monte Carlo experiments. We vary magnitude (response accuracy), respondent heterogeneity, estimation method, and question-design method in a 4x23 experiment. The estimation methods are Hierarchical-Bayes estimation (HB) and Analytic-Center estimation (AC). The latter is a new individual-level estimation procedure that is a by-product of polyhedral question design. The benchmarks for individual adaptation are random designs, orthogonal designs, and aggregate customization. The simulations suggest that polyhedral question design does well in many domains, particularly those in which heterogeneity and partworth magnitudes are relatively large. In the comparison of estimation methods, HB is strong across all domains, but AC estimation shows promise when heterogeneity is high. We close by describing an empirical application to the design of executive education programs in which 354 web-based respondents answered stated-choice tasks with four service profiles each. The profiles varied on eight multi-level features. With the help of this study a major university is revising its executive education programs with new formats and a new focus.

Abstract:
We propose and test new "polyhedral" question design and estimation methods that use recent developments in mathematical programming. The methods are designed to offer accurate estimates after relatively few questions in problems involving many parameters. With polyhedral question design, each respondent's questions are adapted based upon prior answers by that respondent to reduce a feasible set of parameters as rapidly as possible. Polyhedral estimation provides estimates based on a centrality criterion (the "analytic center" of the feasible parameter set). The methods require computer support but can operate in both Internet and other computer-aided environments with no noticeable delay between questions.

We evaluate the proposed methods using two approaches. First, we use Monte Carlo simulations to compare the methods against established benchmarks in a variety of domains. In the simulations we compare polyhedral question design to three benchmarks: Random selection, efficient Fixed designs, and Adaptive Conjoint Analysis (ACA). We compare polyhedral estimation to Hierarchical Bayes estimation for each question design method. The simulations evaluate the methods across different levels of respondent heterogeneity, response accuracy, and numbers of questions. For low numbers of questions, polyhedral question design does best (or is tied for best) for all domains. For high numbers of questions, efficient Fixed designs do better in some domains. The best estimation method depends on respondent heterogeneity and response accuracy. Polyhedral (analytic center) estimation shows particular promise for high heterogeneity and/or for low response errors.

The second evaluation employs a large-scale field test. The field test involved 330 respondents, who were randomly assigned to a question-design method and asked to complete a web-based conjoint exercise. Following the conjoint exercise, respondents were given $100 and allowed to make a purchase from a Pareto choice set of five new-to-the-market laptop computer bags. The respondents received their chosen bag together with the difference in cash between the price of their chosen bag and the $100. We compare the question-design and estimation methods on both internal validity (holdout tasks) and external validity (actual choice of a laptop bag). The field test findings are consistent with the simulation results and offer strong support for the polyhedral question design method. The preferred estimation method varied based on the question design method, although Hierarchical Bayes estimation consistently performed well in this domain.

The findings reveal a remarkable level of consistency across the validation tasks. They suggest that the proposed methods are sufficiently promising to justify further development. At the time of the test, the bags were prototypes. Based, in part, on the results of this study the bags were launched successfully and are now commercially available. Sales of the features of the laptop bags were consistent with conjoint-analysis predictions.

Abstract:
Web-based customer panels and web-based multimedia capabilities offer the potential to get information from customers rapidly and iteratively based on virtual product profiles. However, web-based respondents are impatient and wear out more quickly. At the same time, in commercial applications, conjoint analysis is being used to screen large numbers of product features. Both of these trends are leading to a demand for conjoint analysis methods that provide reasonable estimates with fewer questions in problems involving many parameters.

In this paper we propose and test new adaptive conjoint analysis methods that attempt to reduce respondent burden while simultaneously improving accuracy. We draw on recent "interior-point" developments in mathematical programming which enable us to quickly select those questions that narrow the range of feasible partworths as fast as possible. We then use recent centrality concepts (the analytic center) to estimate partworths. These methods are efficient, run with no noticeable delay in web-based questionnaires, and have the potential to provide estimates of the partworths with fewer questions than extant methods.

After introducing these "polyhedral algorithms" we implement one such algorithm and test it with Monte Carlo simulation against benchmarks such as efficient (fixed) designs and Adaptive Conjoint Analysis (ACA). While no method dominates in all situations, the polyhedral algorithm appears to hold significant potential when (a) profile comparisons are more accurate than the self-explicated importance measures used in ACA, (b) when respondent wear out is a concern, and (c) when the product development and marketing teams wish to screen many features quickly. We also test a hybrid method that combines polyhedral question selection with ACA estimation and show that it, too, has the potential to improve predictions in many contexts. The algorithm we test helps to illustrate how polyhedral methods can be combined effectively and synergistically with the wide variety of existing conjoint analysis methods.

We close with suggestions on how polyhedral algorithms can be used in other preference measurement contexts (e.g., choice-based conjoint analysis) and other marketing problems.

Abstract:
It is puzzling that firms often knowingly continue to invest in product development projects even after receiving damning customer feedback. We argue that bad products are hard to kill because firms face an inherent conflict when designing managers’ incentives. Rewarding success encourages managers to forge ahead even when demand is low. To prevent managers from ignoring signs of low demand, the firm must also reward decisions to kill bad products. However, rewarding failure effectively undermines the rewards for success. The inability to resolve this tension forces the firm to choose between paying an even larger bonus for success or accepting continued investment in low-demand products. We explore the boundaries of this argument by evaluating different motivations for rewarding success, and comparing how the timing of demand information affects the outcome.

product development, managerial incentives, moral hazard, adverse selection, information acquisition

Abstract:
In February 1995 Continental Airlines introduced an incentive scheme that promised monthly bonuses to all 35,000 hourly employees if the company achieved a firm-wide performance goal. Conventional wisdom suggests that free riding will render such schemes ineffective. We study the impact of the Continental scheme by comparing the change in performance at airports where workers were eligible for the scheme and airports where they were not. A combination of cross-sectional and time-series data enables us to control for both airport differences and intervening industry or firm changes. The results offer support for claims that the incentive scheme raised employee performance despite the apparent threat of free riding. To explain why the scheme may have been effective we argue that, despite its size, Continental is able to exploit some of the benefits enjoyed by small firms. In particular, the organization of employees into autonomous work groups enables it to induce mutual monitoring among employees within each work group. Moreover, interdependencies between airports magnify the impact of each work group's performance on overall firm performance, so that firm level measures are sufficient to motivate each group to choose high effort.

Abstract:
We compare bargaining costs with internal and external suppliers using a unique data set describing internal and external transactions of the same categories of parts at a single firm. The findings confirm that direct bargaining costs are higher with external suppliers, at least in part because there is more to bargain over. We also observed higher indirect bargaining costs with external suppliers. Information that may hinder contractual negotiations is often suppressed or delayed. To enforce these restrictions, all communication with external suppliers passes through procurement personnel, greatly hindering coordination and contributing to the determination of which parts are made internally.

Abstract:
We compare the magnitude of bargaining costs within and between firms. The results are derived from a unique dataset comparing internal and external transactions for the same categories of parts at a single high-technology firm. They confirm that direct bargaining costs are higher with external suppliers, at least in part because there is more to bargain over. The need to negotiate price and formal contracts typically leads to longer ex ante negotiations and an increased likelihood of ex post renegotiations when circumstances change. We also observed higher indirect bargaining costs with external suppliers. The introduction of procurement specialists to external supply relationships disperses information and decision-making more widely across the organization. Moreover, information that may hinder contractual negotiations is often either suppressed or delayed and, because engineers are unable or unwilling to enforce these information restrictions, all communication with external suppliers passes through procurement personnel. The data suggests that these differences greatly hinder coordination and contribute to the determination of which parts are made internally versus externally.

Abstract:
We use a dataset describing ownership of productive assets in the carpentry trade to evaluate several factors influencing the allocation of asset ownership between an employer and his employees. The findings suggest that the allocation involves a tradeoff between two incentive effects influencing how the employee uses the asset and what the employer decides it should be used for. In particular, the allocation of ownership hinges on whether an asset is easily lost or stolen, which favors employee ownership, and whether the employer's task assignment affects the asset's depreciation, which favors employer ownership. There is also evidence that more expensive assets and assets that are shared by more than one employee are more likely to be owned by the employer. The results suggest that a general theory of asset ownership should be able to account for at least these effects.

Integration, theory of the firm, ownership,assets

Abstract:
We use a dataset describing ownership of productive assets in the carpentry trade to evaluate several factors influencing the allocation of asset ownership among employers and employees. The findings suggest that the allocation involves a tradeoff between two incentive effects influencing how the employee uses the asset and what the employer decides it should be used for. There is also evidence that more expensive assets and assets that are shared by more than one employee are more likely to be owned by the employer. The results suggest that a general theory of asset ownership should be able to account for at least these four effects.

Asset Ownership, Theory of the Firm

Abstract:
During the summer of 2005, the three domestic U.S. automobile manufacturers offered a customer promotion that allowed customers to buy new cars at the discounted price formerly offered only to employees. The initial months of the promotion were record sales months for each of the three firms, suggesting that customers thought that the prices offered during the promotions were particularly attractive. In fact, many customers paid higher prices following the introduction of the promotions than they could have, had they bought in the weeks just before. We fail to find evidence that the simultaneous increase in prices and sales is due to advertising, decreased financing costs, industry trends, or other explanations. We conclude that the most likely explanation is that the promotion changed customers' beliefs about current versus future prices, convincing them to purchase during the promotion rather than delay in anticipation of future discounts. We present several scenarios that could lead to such beliefs.

automobile pricing, price signal, promotion, search, information

Abstract:
When a multi-channel retailer opens its first retail store in a state, the firm is obligated to collect sales taxes on all Internet and catalog orders shipped to that state. In this paper, we assess how opening a store affects Internet and catalog demand. We analyze purchase behavior among customers who live far from the retail store but who are obligated to pay sales taxes on catalog and Internet purchases. A comparable group of customers in a neighboring state serves as a control. We show that Internet sales decrease significantly, but catalog sales are unaffected. Further investigation indicates that the difference in these outcomes is partly attributable to the ease with which customers can search for lower prices at competing retailers. We extend the analysis to a panel of multi-channel firms and show that retailers who earn a large proportion of their revenue from direct channels avoid opening a first store in high-tax states. We conclude that current U.S. sales taxes laws have significant effects on both customer and firm behavior.

Sales Taxes

Abstract:
Explaining why prices do not immediately adjust to business cycles remains one of the most important unresolved questions in macroeconomics. When researchers ask managers for an explanation, their responses reveal a reluctance to vary prices for fear of "antagonizing customers." However, there is no empirical evidence that antagonizing customers through price variation reduces firm revenue or profit. We use a 28-month randomized field experiment involving over 50,000 customers to investigate this question. The experiment reveals how customers react if they buy a product and later observe the same retailer selling it for less. We find that customers react by making fewer subsequent purchases from the firm. The effect is largest among the firm's most valuable customers: those whose prior purchases were most recent and at the highest prices. Average revenue earned from these customers falls by over $90. The effect spills over to other products sold by the firm and represents an apparent "boycott"; many customers simply stop purchasing from the firm. The boycott persists throughout the 28-month measurement period and lowers overall firm profits by approximately $155,000. After ruling out alternative explanations we demonstrate the robustness of the findings by replicating the results with a separate firm in a different product category.

Customer anger, Price stickiness, Monetary policy, Price adjustment

Abstract:
During the summer of 2005, the Big Three U.S. automobile manufacturers offered a customer promotion that allowed customers to buy new cars at the discounted price formerly offered only to employees. The initial months of the promotion were record sales months for each of the Big Three firms, suggesting that customers thought that the prices offered during the promotions were particularly attractive. In fact, such large rebates had been available before the employee discount promotion that many customers paid higher prices following the introduction of the promotions than they would have in the weeks just before. We hypothesize that the complex nature of auto prices, the fact that prices are negotiated rather than posted, and the fact that buyers do not participate frequently in the market leads customers to rely on "price cues" in evaluating how good current prices are. We argue that the employee discount pricing promotions were price cues, and that customers responded to the promotions as a signal that prices were discounted.

Abstract:
In order to push a customer and market orientation deep into the organization many firms have adopted systems by which internal customers rate internal suppliers on some measure-- often satisfaction. The internal supplier receives a larger reward for a higher rating. We examine incentive systems based on such rating systems and show that gainsharing between the rater (internal customer) and the ratee (internal supplier) normally will occur. We show that, for two common internal customer-internal supplier incentive systems, the firm can select parameters for the reward functions such that this gainsharing can be factored out and such that both the internal customer and the internal supplier choose the actions that are the same that a risk- neutral firm would choose to maximize profit if it had to reimburse these employees for their costly efforts. Some risk is transferred from the firm to me employees and the firm must pay for this, but in return the firm need not observe either the internal supplier's actions or the internal customer's actions. The incentive systems are robust even if the firm guesses wrongly about what employees perceive as costly and about how employee actions affect profit.