在线口碑和产品销售动态对电影行业的实证调查-学位论文.doc

毕业论文外文翻译
题　　目：网络口碑效应研究——基于消费者购买决策视角的探索　　　　　　　　　　　　　　
一、外文原文
标题：The dynamics of online word-of-mouth and product sales—An empirical investigation of the movie industry
原文：
Introduction
Word-of-mouth (WOM) has been recognized as one of the most influential resources of information transmission since the beginning of human society (Godes and Mayzlin 2004; Maxham and Netemeyer 2002; Reynolds and Beatty 1999). However, conventional interpersonal WOM communication is only effective within limited social contact boundaries, and the influence diminishes quickly over time and distance (Bhatnagar and Ghose 2004; Ellison and Fudenberg 1995). The advances of information technology and the emergence of online social network sites have profoundly changed the way information is transmitted and have transcended the traditional limitations of WOM (Laroche et al. 2005). The otherwise fleeting WOM targeted to one or a few friends has been transformed into enduring messages visible to the entire world. As a result, online WOM plays an increasingly significant role in consumer purchase decisions.
Online WOM presents both challenges and opportunities to retailers. On the one hand ,WOM provides an alternative source of information to consumers, thus reducing retailers’ ability to influence these consumers through traditional marketing and advertising channels. Prior studies show that a variety of aspects of WOM influence retail sales. Some found that WOM dispersion (Godes and Mayzlin 2004) and valence (Chevalier and Mayzlin 2006; Forman, Ghose, andWiesenfeld 2008) have significant effects on product sales, while others found that WOM volume serves as the key driver of product sales (Chen, Wu, and Yoon 2004; Liu 2006). On the other hand, online WOM provides a new venue for retailers to reach consumers and to strategically influence consumer opinions. Anecdotal evidence has surfaced in recent years suggesting that online WOM could be successfully leveraged as a new marketing tool (Dellarocas2003). A unique aspect of the WOM effect that distinguishes it from more traditional marketing effects is the positive feedback mechanism between WOM and product sales. That is, WOM leads to more product sales, which in turn generate more WOM and then more product sales. The positive feedback mechanism indicates that WOM is not only a driving force in consumer purchase but also an outcome of retail sales ( Godes and Mayzlin 2004; Srinivasan, Anderson, and Ponnavolu 2002). Prior studies on WOM have not fully recognized this unique nature of WOM effect and often treat WOM as exogenous, like traditional marketing effects (Chen et al. 2004; Liu 2006). Ignoring WOMs dual roles of precursor and outcome may misplace causality and lead to erroneous results. The objectives of this study, therefore, are to explicitly model the positive feedback mechanism between WOM and retail sales and identify their dynamic interrelationship. We propose a simultaneous equation system to fully capture the dual nature of online WOM and its dynamic evolution in a panel data setting.
We have chosen the movie industry as our research context because industry experts agree that WOM is a critical factor underlying a movie’s staying power, which leads to its ultimate financial success (Elberse and Eliashberg 2003). In addition, the movie industry has by far received the most attention in marketing literature on WOM, which allows in-depth comparison of our results with those of previous studies. We, however, note that movies are a unique type of experience goods and the results from the industry do not necessarily generalize to other retailing sectors. Rather, our goal is to use the movie industry as a context to highlight the importance of considering the dynamics of and the interrelationship between retail sales and online WOM and to demonstrate the validity of the simultaneous equation approach in this setting. We found that both a movie’s box office revenue and WOM valence significantly influence WOM volume. WOM volume in turn leads to higher box , office performance. Our results clarify conflicting results reported in earlier studies with regard the influence of user ratings on box office revenue. We show that user ratings do not directly influence box office revenue. However, they affect box office revenue indirectly through WOM volume.
Online WOM in the movie industry takes many forms, including online reviews, discussion boards, chat rooms, blogs, wikis, and others. In this study, we focus on online user reviews because statistics suggest that user reviews are more prevalent than other forms of WOM communication in the movie industry. Beyond volume, another subtle but important difference between online user reviews and other types of WOM is that user reviews usually reflect user experience and consumer satisfaction, which are mainly viewed as a source of product information (Chen and Xie 2004; Li and Hitt 2008). Meanwhile, other types of WOM, such as discussions in online community sites, reflect more about consumer expectation, which could be heavily influenced by social structure (Gopal et al. 2006; Liu 2006).
The rest of the paper is organized as follows. The next section provides the literature review followed by the discussion of our conceptual framework and research hypo theses. We then describe our sources of data and the empirical model and estimation. Main findings are presented and discussed next, and the paper ends with a discussion of implications, limitations, and future research.
Empirical model specification
The development of our empirical model is guided by the following considerations. First, as we are interested in the drivers of both box office revenue and WOM, we construct a system of two interdependent equations: one equation with daily revenue as the dependent variable (the revenue equation) and the other
with WOM volume as the dependent variable (the WOM equation). We assume that in each time period (i.e., day) the errors in the two equations may be correlated, which implies that factors not included in our model could simultaneously influence both movie revenue and WOM.
Second, recognizing that interactions between consumers’ movie-going behavior and WOM can go beyond the concurrent term (Elberse and Eliashberg 2003), we develop a system of dynamic equations. That is, in the revenue equation, we include
not only the contemporaneous term of daily WOM volume, but also multi-lag terms. Likewise, in the WOM equation, multi-lag revenue terms are also incorporated. Such a specification also helps identify both equations for the simultaneous equation system since the lagged terms are exogenous variables in either equation. In addition, following extant research, we use alog-linear formulation (e.g., Elberse and Eliashberg 2003; Liu2006) in our model. The log-linear formulation is consistent with
theoretical models of a multistage consumer decision-making process, where sales of a movie can be viewed as a series of conditional probabilities applied to the consumer base. A log transformation converts the relationship into a linear form for empirical estimation. Moreover, log transformation smoothes the distribution of variables in the linear regression, and the estimated coefficients of the log-linear form directly reflect the elasticity of independent and dependent variables. Third, to control for any movie idiosyncratic factors that could influence revenue and/or WOM, such as budget, marketing, star, and others (Basuroy et al. 2003; Elberse and Eliashberg 2003; Liu 2006), we include fixed effects in the model by adding movie-specific dummy variables. Fixed effects capture any noontime varying factors, including intrinsic movie characteristics, critic reviews, and other exogenous determinants. In addition,
fixed effects estimation also allows the error term to be arbitrarily correlated with other explanatory variables, thus making the estimation results more robust.
Implications, limitations, and future research
Our model specifies the dual causal relationship and reveals the positive feedback mechanism between online WOM and product sales. Our findings strongly support the value of considering the endogeneity of WOM and its interdependence with consumers’ consumption behavior. The notably different results obtained from 3SLS (the statistically more robust method) and OLS suggest that extant research using simple regression techniques may have drawn biased conclusions about the direction and magnitude of the effect of WOM. Our results validate our assertion that the volume of online user reviews has an intertwining relationship with retail sales.
Our findings also bring important extensions to previous research (Basuroy et al. 2003; Eliashberg and Shugan 1997; Liu 2006) on the relationship among WOM volume ,WOM valence and box office sales. Previous research has been focusing on
the direct impact of WOM volume and valence on box office revenues and find that most of the explanatory power comes from WOM volume not WOM valence (Liu 2006). Our study extends this approach by considering the interaction between WOM valence and WOM volume. We find that while WOM valence does not directly affect revenue, higher WOM valence indirectly increases box office revenue by generating higher volume of WOM.
The contributions of this research to retail literature on WOM are multifaceted. From the methodology perspective, we bring to light the importance of separating the effect of WOM as both a precursor and an outcome of sales. Our results also highlight
the importance of using a dynamic system and high-frequency data in studying the effect of WOM in the digital environment. From the managerial perspective, we show that WOM valence and WOM volume play different roles in influencing product sales. We also show that time-series changes in WOM valence influences WOM volume which leads to higher product sales. Our findings support the idea that the online WOM process has a significant impact on sales, suggesting that businesses should
embrace and facilitate WOM activities.
There are a number of opportunities to extend the current research. One important and interesting extension of our research will be to investigate the consumer decision process under the influence of WOM information, especially in the digital
environment. In addition, not all WOM is equal. Consumers need to distinguish the “true” and “honest” opinions from all kinds of feedback and recommendations on the web. Under such circumstances, how consumers choose their information source find trusted information sources will be of particular interest for future research.
Online user reviews are only one type of consumer-generated media. The recent explosive growth of popular online social communities (e.g., www.YouT, www.F, and www.D) has generated a renewed interest in the Internet as a new medium for content generation and distribution. Different from online review sites we explored here, online social communities encourage interaction between users, which potentially changes the dynamics of WOM distribution. The modeling approach used in this research therefore may not be sufficient to these contexts. Further study to characterize and identify the effect of online WOM in the new medium would be beneficial to our understanding of the effect of online consumer-generated media on marketing and retailing strategies.
Our analysis is, by necessity, restricted to online users who choose to post reviews and to post them on Yahoo! Movie. Thus, our estimates are conditioned on such a user population. While such a restriction does not necessarily bias the panel data estimation results, they should be interpreted as applying to a self-selected set of online users. In addition, this paper studies only the relationship between the postre lease WOM and sales. However, WOM certainly has existed before a movie’s release, and movie studios have made various prerelease marketing efforts to promote the WOM(Liu 2006).A careful survey of Yahoo! Movie indicates an important and interesting difference between prerelease WOM and post release WOM. We note that prerelease WOM activity centers in Yahoo! Movie discussion boards, and the postings mainly reflect consumer expectations; meanwhile, post release WOM activities center in Yahoo! Movie’s user review sites, where postings mainly reflect consumer satisfaction and product experience. This difference indicates that there exist potentially different mechanisms for prerelease WOM versus post release WOM. Thus, an important extension to the current research would be to study the dynamic relationship between prerelease and post release online WOM and to differentiate their influences.
出处：Wenjing Duan,Bin Gu, Andrew B.The dynamics of online word-of-mouth and product sales—An empirical investigation of the movie industry[J]. Journal of Retailing ,2008(6): 233-242

二、翻译文章
标题： 在线口碑和产品销售动态——对电影行业的实证调查
译文：
介绍
口碑自人类社会开始以来，已经被公认为最有影响力的信息传递的资源(Godes and Mayzlin 2004; Maxham and Netemeyer 2002; Reynolds and Beatty 1999).然而，传统的人际间口碑交流只有在有限的社会接触范围内才会起作用，其影响随时间和空间的推移迅速减弱(Bhatnagar and Ghose 2004; Ellison and Fudenberg 1995)。信息技术的进步和在线社交网站的出现，深刻地改变了信息传播的方式，并已超过了传统口碑的界限(Laroche et al. 2005)。另外，在一个或几个朋友间传播的短暂口碑，已经被看作是能透过此看见世界的信息。因此，在线口碑在消费者购买决策中发挥着日益重要的作用。
在线口碑对零售商来说，既是机遇也是挑战。一方面，口碑为消费者提供了可供选择的信息源，从而降低了零售商通过传统市场营销和广告渠道影响消费者的能力。先前的研究表明，口碑的各个方面会影响到零售的销量。一些研究发现，口碑量(Godes and Mayzlin 2004)和价数(Chevalier and Mayzlin 2006; Forman, Ghose, andWiesenfeld 2008)对产品销量产生重大影响，而另一些研究发现，口碑量被看作是产品销量的主要驱动力(Chen, Wu, and Yoon 2004; Liu 2006).。另一方面，在线口碑为零售商提供一个接触到消费者和战略影响消费者意见的场所。近几年出现的传闻证据表明，在线口碑已作为新的营销工具被成功应用(Dellarocas2003).。不同于传统的市场营销效果，口碑效应的独特之处在于口碑和产品销量的积极反馈进程，也就是说，口碑会提高产品销量，反过来产生更多的口碑和更多的产品销量。积极的反馈机制表明，口碑不仅是消费者购买的驱动力，而且是零售销量的结果( Godes and Mayzlin 2004; Srinivasan, Anderson, and Ponnavolu 2002)。此前对口碑的研究没有充分认识到口碑效果的独特性，而是把口碑看成是像传统营销效果一样的外因(Chen et al. 2004; Liu 2006)，忽略口碑的双重先导角色和可能因误置角色而导致错误结果的影响。因此，这项研究的目的是要明确口碑和产品销量模式的积极反馈作用，并确定他们之间的动态关系。我们提出了一个联立方程系统，以充分诠释在线口碑的双重性质以及它在数据库设置中的动态演进。
我们选择了电影产业作为研究范畴。因为业界专家一致认为，口碑是能令电影持续发挥其功效的一个关键因素，从而实现最终财务上的成功(Elberse and Eliashberg 2003)。此外，电影界迄今为止在文学营销口碑上受到重视，这使得我们的成果能和之前的研究进行深入的比较。然而，我们注意到电影作为一种特殊的体验产品类型，其产业的结果并不一定能够推广到其他零售行业。相反，我们的目标是利用电影产业作为一个内容来强调考虑零售销量和在线口碑的动态和他们之间关系的重要性，并说明设置的联立方程方法的正确性。我们发现，一部电影的票房收入和口碑效价明显地影响着口碑量，口碑量反过来会带来更高的票房和更好的成绩。我们的研究结果澄清了早期研究中用户评分对票房收入影响的博弈。我们表明，用户评分不会直接影响票房收入。但是，他们会通过口碑间接影响到票房收入。
此外，我们的研究还证实了在线口碑不仅是先导，而且是产品销量的结果。我们表明， 忽视口碑的双重性质会导致错误的结果。
电影产业的在线口碑有多种表现形式，包括在线评论，讨论版，聊天室，博客，维基和其他。在此研究中，我们侧重于在线用户评论，因为数据显示，在电影产业中，在线用户评论比其他口碑交流更加普遍。除了量之外，在线用户评论和其他形式的口碑之间的重要微妙区别在于，用户评论通常反映了用户的经历和满意度，这被看作是产品的信息源之一(Chen and Xie 2004; Li and Hitt 2008).。与此同时，口碑的其他类型，比如较能反映消费者的预期的在线社区网站的讨论，深受社会结构的影响(Gopal et al. 2006; Liu 2006)。
本文的其余部分组织如下。下一节将提供支撑我们概念框架和研究模型的文献。然后，我们描述我们的数据来源和经验模型及其预测。之后，将描述和讨论主要的调查结果，以及讨论的意义、限制及未来的研究。
实证模型详述
我们的实证模型发展遵循以下考虑。首先，由于我们对电影票房收入和口碑驱动力感兴趣，我们构建了两个相互依存的系统，一个是每日