Simultaneous Vs. Sequential Group-Buying Mechanisms

MANAGEMENT SCIENCE Vol. 59, No. 12, December 2013, pp. 2805–2822 ISSN 0025-1909 (print) ISSN 1526-5501 (online) http://dx.doi.org/10.1287/mnsc.2013.1740 © 2013 INFORMS Simultaneous vs. Sequential Group-Buying Mechanisms Ming Hu, Mengze Shi, Jiahua Wu Rotman School of Management, University of Toronto, Toronto, Ontario M5S 3E6, Canada {[email protected], [email protected], [email protected]} his paper studies the design of group-buying mechanisms in a two-period game where cohorts of consumers Tarrive at a deal and make sign-up decisions sequentially. A firm can adopt either a sequential mechanism where the firm discloses to second-period arrivals the number of sign-ups accumulated in the first period, or a simultaneous mechanism where the firm does not post the number of first-period sign-ups and hence each cohort of consumers faces uncertainty about another cohort’s size and valuations when making sign-up decisions. Our analysis shows that, compared with the simultaneous mechanism, the sequential mechanism leads to higher deal success rates and larger expected consumer surpluses. This result holds for a multiperiod extension and when the firm offers a price discount schedule with multiple breakpoints. Finally, when the firm can manage the sequence of arrivals, it should inform the smaller cohort of consumers first. Key words: promotion; group buying; coordination game History: Received August 4, 2011; accepted February 27, 2013, by J. Miguel Villas-Boas, marketing. Published online in Articles in Advance July 19, 2013. 1. Introduction commitment through escrow payment systems. Most A group-buying mechanism is a scheme designed of the representative group-buying websites that to help coordinate a group of interested buyers so became popular in the late 1990s, including Mercata, that they can reach their common purchase goals. Mobshop, and Letsbuyit, either ceased operating or In a typical group-buying mechanism, no transac- changed their business models a few years later tion will take place unless the total number of (Kauffman and Wang 2002). Interestingly, despite committed purchases exceeds a specified thresh- the failure of these pioneering group-buying sites, old within a certain time period. Various forms of a decade later another generation of social buying group-buying mechanisms have been observed on websites like Groupon and LivingSocial emerged. Led a wide range of occasions and for a long time. by the market leader, Groupon, these newcomers typ- One version is street performer protocol (SPP), an ically offer “a deal a day” tailored to each local agreement between an artist and a group of poten- market (Wortham 2009). The market enthusiasm for tial users. The artist does not start the creative online group buying peaked when Groupon declined work until the potential funders have pledged a a $6 billion offer from Google (Weiss 2010). required amount of support. Renowned musicians This paper investigates the information manage- like Beethoven and Mozart used such arrangements ment strategy for group-buying mechanisms, specif- to ensure that enough tickets were sold for their ically, whether or not the sponsor should ask concerts. The recent emergence of social media has participants to make decisions without knowing the popularized these funding concepts and generated decisions of others. We take the perspective of third- similar concepts like threshold pledge systems and party group-buying platforms like Groupon and crowd funding. Websites like Kickstarter.com allow Kickstarter and investigate the impact of alternative artists, museums, and entrepreneurs to post project information management mechanisms on deal suc- proposals and seek funding from interested donors. cess rates. Knowing how to improve the success rates In another version of the group-buying mechanism, a is important because group-buying firms typically number of buyers pool their purchases, often through earn revenues from successful listings only and not the facilitation of a third party, to obtain a quan- all group-buying deals succeed. For example, from tity discount from sellers. Online group-buying web- the launch of Kickstarter.com in April 2009 to March sites first appeared in the late 1990s, as part of the 2011, a total of 20,371 projects were proposed. Among wave of innovative online market-based mechanisms. them, 7,496 projects attracted enough funds, lead- Usually the consumers had to make the purchase ing to a success rate of 43% (blog.kickstarter.com). 2805 Hu, Shi, and Wu: Simultaneous vs. Sequential Group-Buying Mechanisms 2806 Management Science 59(12), pp. 2805–2822, © 2013 INFORMS Zhang and Liu (2012), who studied requests for cohort, it is better to inform the potentially larger microloans listed in Prosper.com, reported a success cohort of consumers later because the confidence of rate of 12.3% among 49,693 listings. Without careful consumers arriving earlier can be boosted by a subse- analysis, the firm’s decision does not appear straight- quent stronger cohort who does not suffer discount- forward because of the uncertainty about the number ing in belief due to information asymmetry. of consumer arrivals and their individual valuations. To further understand the dominance of sequen- Looking forward, it can be beneficial to post the num- tial mechanisms over simultaneous mechanisms, we ber of sign-ups if a large cohort of consumers with extend our model in a number of directions. First, we high individual valuations turn out in the early stage, extend our model to one with more than two peri- but it can be detrimental if the first cohort of con- ods, where the threshold of group buying is equal sumers turn out to be small and have low individual to the number of arrivals. Such a model allows a valuations. sequential mechanism to have different reporting fre- To investigate the influence of information manage- quencies. We find that the mechanism with the high- ment mechanisms on deal success rates, we develop est reporting frequency, that is, updating after each a two-period model where two cohorts of consumers period, leads to the highest success rate. Because all arrive at the deal sequentially. The two-period model of the consumers have to commit to purchases for is a stylized capture of the fact that earlier arrivals the deal to be on, the intuition behind the backward are faced with more uncertainty in the deal’s suc- induction described earlier still applies when we trace cess rate than later arrivals. The firm being stud- along the sole sign-up trajectory that leads to the suc- ied chooses between a “sequential mechanism” where cess. Second, we extend the single-level scheme to the firm posts the number of sign-ups at the end a multilevel pricing schedule where the participants of the first period, and a “simultaneous mechanism” can receive a greater benefit when the total number where the firm does not post the first-period out- of sign-ups reaches a higher level of threshold. Our come. Somewhat surprisingly, our analysis shows analysis shows that, all else being equal, the sequen- that the deal’s success rate is always higher under tial mechanism still always yields higher success rates the sequential mechanism. To see the reason for than the simultaneous mechanism. Third, when the this result requires a backward-inductive reasoning, firm faces a capacity constraint, the expected number starting with the second period and then moving of sign-ups may be lower under the sequential mech- back to the first period. A sequential mechanism anism than that under the simultaneous mechanism. increases the ex ante expected sign-up rates of the This is because the capacity constraint introduces second cohort of consumers by eliminating the uncer- negative externality as the chance to receive service tainty facing them. The increased expected sign-up decreases when more people have signed up to the rates of the second cohort enhance the confidence deal. Knowing that a large number of people have of the first cohort of consumers, thereby increas- already signed up can simply depress the expected ing the ex ante expected sign-up rates of the first values of a group-buying deal. cohort. This result underscores the importance of modeling and investigating the dynamics of sign-up 1.1. Literature Review behavior under group-buying mechanisms. The result Our paper is related to research on private provi- also offers a potential explanation for why firms like sion of public goods. Like the group-buying deals, Groupon and Kickstarter display the updated number provision of public goods requires successful coor- of sign-ups along with the minimum number required dination among the private contributors. In the lit- to unlock the deals. erature on private provision of public goods, most When implementing the information management papers examine the simultaneous mechanism (e.g., mechanism, a firm may control the sequence of con- Palfrey and Rosenthal 1984, Bagnoli and Lipman sumer arrivals by, for example, contacting different 1989, Tabarrok 1998). Work on sequential mechanisms groups of consumers separately. Since the sequen- includes Varian (1994), Gächter et al. (2010), and tial mechnism always leads to a higher success rate Romano and Yildirim (2001). Varian (1994) shows that over the simultaneous mechanism, it is better off for a simultaneous decision-making mechanism leads the firm to have cohorts of consumers arrive sequen- to more supplied public goods than a sequential tially rather than simultaneously. Interestingly, our mechanism where the outcome of early decisions is analysis also finds that the firm should first inform revealed to those participants who make decisions the cohort with a potentially smaller number of con- later. In Varian (1994), one player can free ride on sumers, and then the potentially larger cohort of con- another’s contributions under the sequential mecha- sumers.

Load more