|
Article Excerpt
The free-riding problem occurs if the presales activities needed to sell a product can be conducted separately from the actual sale of the product. Intuitively, free riding should hurt the retailer that provides that service, but the author shows analytically that free riding benefits not only the free-riding retailer, but also the retailer that provides the service when customers are heterogeneous in terms of their opportunity costs for shopping. The service-providing retailer has a postservice advantage, because customers who have resolved their matching uncertainty through sales service incur zero marginal shopping cost if they purchase from the service-providing retailer rather than the free-riding retailer. Moreover, allowing free riding gives the free rider less incentive to compete with the service provider on price, because many customers eventually will switch to it due to their own free riding. In turn, this induced soft strategic response enables the service provider to charge a higher price and enjoy the strictly positive profit that otherwise would have been wiped away by head-to-head price competition. Therefore, allowing free riding can be regarded as a necessary mechanism that prevents an aggressive response from another retailer and reduces the intensity of price competition.
Key words: free riding; sales service; selling cost; channel conflict; retail competition; competitive strategy; game theory
History: This paper was received August 31, 2005, and was with the author 6 months for 2 revisions; processed by Florian Zettelmeyer.
1. Introduction
Imagine a college student planning to purchase some audio equipment. Not knowing precisely which product fits her needs best, she might visit a Tweeter store, recognized for its high level of customer service, and ask for the assistance of knowledgeable salespeople. The salespeople expend time and effort to help the college student by listening to her situation carefully, letting her try many different products, and finally identifying the product that fits her needs best. Because Tweeter must incur all these selling costs, they will be reflected in the price of the audio equipment. Knowing this, the student goes to a local discount audio store, which does not offer any sales service, and purchases the same audio equipment that Tweeter recommended, but at a lower price.
What this frugal college student has done is "free ride" off of Tweeter's selling effort, or service, to resolve the matching problem between her needs and the available audio products. (1) She enjoyed the opportunity to try some equipment and Tweeter's expert advice, an experience most discount retailers cannot offer, and then obtained the discount retailer's lower price. This free-riding problem can occur whenever the presales activities needed to sell a product, such as providing informed sales advice to consumers, can be conducted separately from the actual sale of the product. Therefore, it is possible for one retail store to engage in the presales activity necessary to sell the product, but for a different, lower-priced store to make the final sale. As the selling costs incurred for such presales service have grown increasingly significant, the free-riding problem has become more and more substantial for many retailers. (2)
Conventionally, free riding in a retail market should hurt the profit of the retailer that provides the service (for a survey in a distribution channel environment, see Antia et al. 2004, Coughlan et al. 2001; also see Carlton and Chevalier 2001). The cost of free riding for the retailer that offers presales service is clear, but its benefits, if any, may be more difficult to discern.
Traditionally, retail stores have used various devices to keep consumers from free riding, such as territorial restrictions or exclusive dealership agreements for a certain manufacturer (Coughlan et al. 2001). The idea behind these devices is that free riding hurts retail profits; therefore, retailers must make it harder for customers to compare products and prices from different retail outlets to reduce the amount of free riding. Also, the main focus of existing literature has been the negative impact of free riding on a retailer's incentive to offer service and the way retailers and manufacturers prevent free riding (for general references, see Carlton and Perloff 2000). However, those tactics to prevent free riding can be implemented only through channel coordination with manufacturers, and without upstream channel coordination, most retailers are vulnerable to the free-riding problem. To prevent free riding, retailers try to lower their prices, but many are limited in their ability to do so because of the costs of offering presales service. Therefore, such retailers would have to eliminate their sales service (Carlton and Chevalier 2001). Nonetheless, many retailers still continue to provide sales service and thrive despite the free-riding problem. Are they simply doomed to vanish in the near future? How can they cope with the threat of free riding and still provide costly service? This article tries to offer some answers to these questions.
An interesting insight gained from this analysis is that both free-riding and service-providing retailers can benefit from free riding. The basis for this counterintuitive result is that the service provider can use its service offer to attract many people to the store and effectively lock in some customers because visiting another retailer would entail additional shopping costs. This argument is appealing, yet incomplete. If a retailer could lock in all customers who visit the store, all retailers would engage in intense price competition upfront to attract more customers. However, by allowing free riding, a retailer that provides service can induce free-riding competitors to avoid competing severely on price. That is, the competitor (free rider) has less incentive to compete with the service-providing retailer on price because many customers who visit the service provider eventually will switch to the free rider to enjoy its lower price. This induced soft strategic response enables the service provider to charge a higher price. Allowing free riding can therefore be regarded as a necessary mechanism by the service provider to prevent an aggressive response from another retailer.
Before offering a formal rationale for this argument with a simple model, we briefly review some associated literature. The free-riding retailing problem has been studied and is well understood in standard economics and marketing literature (e.g., Carlton and Chevalier 2001, Carlton and Perloff 2000). Telser (1960) and Singley and Williams (1995) suggest that free riding increases the price disparity between the service-providing retailer and the free rider. As a consequence, more customers free ride off one retailer's service and buy the product from the other retailer, who charges a lower price. Free riding thereby can dissuade retailers from offering sales service. Several studies in channel conflict literature also suggest that free riding decreases the levels of retail service, such as presale service, customer education about product attributes, and salesperson training, in a multichannel distribution setting (see Antia et al. 2004 for a survey about free riding in multichannel conflicts). (3)
The work of Wu et al. (2004) is a notable exception, which shows analytically that retailers might benefit from providing service even in the presence of the painful free-riding problem. Two segments of customers exist: one with a positive search cost and another with zero search cost. A retailer needs to establish itself as a service provider to make a positive profit by attracting customers and locking in all the customers who have a positive search cost, even if the zero-search-cost customer segment engages in free riding. (4) However, these authors focus on the negative impacts of free riding and regard it as a pure cost to be managed; they do not recognize its benefit to the retailer. In contrast, this article focuses on the strategic role of free riding in reducing the intensity of price competition and evaluates its benefit to the service-providing retailer.
The current work also relates to sales management literature (Churchill et al. 1999, Godes 2003, Wernerfelt 1994a). We follow Wernerfelt's (1994a) conception that sales assistance improves the quality of the match between a customer's need and a product. In markets in which the customer's situation determines his or her needs but the number of possible situations is very large, the map that identifies the best match between the customer's situation and a product might be known only to the salesperson. In the opening vignette, the college student does not know precisely what to buy among so many alternatives; should she buy a single-crystal silver Silway MK II or the Orca limited MK II for her new JVS FS-D5 audio system?s In this context, more than 140 different items are available for audio cable alone, which makes it very difficult to find the best matching product without knowledgeable sales advice.
The structure of this article is as follows. We present a formal model of free riding in [section]2. We analyze the free-riding model in [section]3 and compare it with a benchmark model without free riding, which leads to the main results in [section] 4. In [section] 5, we endogenize sales service and extend the model to the case of multiple free riders in [section] 6. The conclusions appear in [section] 7.
2. Model
Customers
Customers are assumed to purchase either one or zero units of a product. All customers receive the same utility v if the product is a good match with their situation; otherwise, they receive v utility. For simplicity, we set utility v equal to zero. Also, v is common and known to customers, but whether the product fits their situation or specific needs is unknown (Wernerfelt 1994a).
The total market size is normalized to one, and there are two segments of customers: informed and uninformed. Informed customers (a) have sufficient knowledge about the product and know precisely whether it fits their situation or specific needs. In contrast, the uninformed fraction (1-[alpha]) is not sure about the match between the product specification and their situation, and therefore needs the retailer's sales service to resolve this matching uncertainty. (6) The relative size of each segment is common knowledge among both customers and retailers. However, it is important to note that the results we obtain herein do not rely on this two-segment assumption. That is, we do not exclude the [alpha] = case, which would imply that there are only uninformed customers. (7) We return to this point subsequently.
Customers incur shopping costs when they visit the store. Within each segment, customers are heterogeneous in terms of their shopping cost, t ~ U[[t.bar], [bar.t], where [t.bar] > and [bar.t] t + 1, and the density is 1. These shopping costs, which are private information to customers, represent a disutility or opportunity cost for the customer's time spent shopping. Even though customers may incur the same amount of time or cost to visit the store, the opportunity cost for the time varies according to the customer's income level (Narasimhan 1984). In general, as a customer's income level increases, the value attached to successive units of time becomes higher.
Retailers
Two retailers indexed by i [member of] {1, 2} compete for the same customers. Both retailers sell the same product A at prices [P.sub.1] and [P.sub.2], and the unit cost of the product, c > 0, is the same for both retailers. The product A is a good match with a customer's situation with the probability m (<1). The matching probability m is common knowledge to customers and assumed to be strictly less than one. Therefore, it is possible that the product will not match the customer's situation even after sales service is incurred. (8)
The model also includes two crucial features about the role of the selling effort and the asymmetry of retailers in terms of their service provision. First, uninformed customers must consult a service-providing retailer if they want to resolve their matching uncertainty before they purchase. In line with Wernerfelt (1994a), customers need expert service to identify the best match between their situation and a product or must physically inspect the product before they purchase it.
Second, only one retailer (Retailer 2) offers sales service; the other (Retailer 1) does not (we subsequently endogenize this sales service choice). This assumption implies that only Retailer 2 incurs a selling cost, equal to k, for each customer who visits the store. Specifically, we model the cost of free riding as the selling costs that a retailer incurs to serve a customer who may or may not purchase a product (Shin 2005). For example, a local audio store must expend time and effort to help customers identify the...
|
