Box Office Performance in the Chinese Film Market
Author: Jinlei Hou
Student number: 10696636
Supervisor: dr. F.B.I. (Frederik) Situmeang
MSc. in Business Administration – Marketing Track Institutions: University of Amsterdam (UvA)
This document is written by Jinlei Hou who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document are original and that no sources other than those mentioned in the text
and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Abstract... 1
1. Introduction ... 2
2. The Movie market in China: previous research... 3
2.1 Reviews of online word of mouth influence movies’ box office sales... 3
2.2 Drivers of Movies online reviews... 6
2.3 Variables description and hypothesis...10
2.3.1 Independent variables...10
2.3.2 Dependent variable...11
2.3.3 Moderator variable... 11
3. Model and data...15
3.1 Model... 15
3.2 Data sources... 16
3.2.1 Sample collection... 16
4. The significance test for variables and result...17
4.1 Independent variable correlation text...17
4.2 The significant test for moderator variables...19
4.2.1 Star power... 19
5. Conclusion...23
6. Marketing Suggestions...27
7. Deficiency and further prospects...30
7.1 Deficiency... 30
7.2 Future research prospects...31
Abstract
With the rapid development of movie marketization and Industrialization, the concept "Marketing is greater than the movie" becomes increasingly important. With the increasing development of the Internet, WOM, in this study, online conversation gradually highlights its unique power, becoming an important element affecting the market. This paper focuses on how movie WOM influences consumers’ decision and studying the effect of the WOM on the movie box office sales, revealing the important influence of online conversation for film industry.
Firstly, this paper will review and combine some foreign and domestic research about WOM, combing with the characteristics of the film, dividing WOM into the volume of WOM and the valence of WOM. Then this study will explore its effect on film respectively. In addition to that, as a cultural product, box office is affected by many other factors. This paper introduces the star power of the film and language of the film these two factors as the manipulated variable of WOM. The results show that the volume of customer online review has a positive correlation with box office. But the valence of customer online review has a negative correlation with box office. Neither star power nor languages has impact on box office.
This paper selects 132 movies from March 2015 to end of 2017 as the research samples. This research samples are analyzed by SPSS statistical 22 software. It is concluded that online conversation impacts the movie box office sales. On this basis, this paper put forward relevant suggestions of movie WOM for the market.
Keywords: Volume of online conversation, Valence of online conversation Movie box
1. Introduction
Along with the development of social economy, people may pay more attentions on the entertainment life to enrich their daily life. In China particularly, with increasing average familiar income, it is becoming a popular entertainment way of watching both Chinese and foreign movies in cinemas. Hence, movie industry is a huge market in China today.
In a few years, Web2.0 offers more power to consumer buzz on the network. More and more people across the world attracted towards new media as a way of gathering entertainment, information and culture. Online conversation has two key attributions which are includes volume and valence (Neelamegham and Jain, 1999). Volume means total amount of online reviews. Valence refers to the contents of online reviews, which could be positive or negative in our research (Liu, 2006). Due to the dramatically development in movie industry, more domestic movies and international movies are swarming into Chinese movies market. Box office is an important predictor of movies success degree in the Chinese market. (Huang and Wang, 2008) Online word of mouth (WOM) has a significant impact on product’s revenues. Previous studies have concluded WOM volume is a repaid factor that influences product revenues. (Joonhyuk, Wonjoon, Naveen and Jaeseung, 2012) There are some studies propose that WOM valence significantly influences product sales, while some studies not. On the other hand, there are different online conversations between Chinese movies and international movies. Language and star power are two of the major factors.
Therefore, the goal of this study is to investigate that customer online conversation aff ects box office revenues of Chinese movies and international movies in Chinese mark et. In order to do research on this topic, we empirically explore the effects of both WO M volume and WOM valence on the box office of movies. Language and star power a re two of moderators in this relationship. Most former studies have inspected the affec ts of movie online WOM on box office sales from the opening night or opening weeks (Duan, Gu and Whinton, 2008; Elberse and Eliashberg, 2003). However, this study m akes several contributions from pre-release week to the 10th weeks. First, this thesis will review some previous studies in movies box office sales and online reviews field.
Second, some gaps will discussed from these previous studies and make this research questions. Then, the result and contribution of this research will be discussed at the end.
2. The Movie market in China: previous research
2.1 Reviews of online word of mouth influence movies’ box office
Over the technology development, Web 2.0 has more consumer power than Web 1.0. It helps customers to decide whether watch new movies in cinema based on the online reviews, scores, star power and director popularity. Thus, online conversation has a significant impact on new movie marketing. Anita and Jehoshua (2003) argue that movie attributes and advertising expenditures are influence revenues indirectly through their impact on exhibitors’ screen allocations. Garrette, Leigh and Oliver (2011) indicate that word of mouth is an influence source of information for consumers to make a purchase. In the research of Thorsten, Caroline and Fabian (2015), Twitter effect shows that 600 Twitter users who decided not to see a new movie based on negative WOM. Consumers are make their decisions based on WOM information on Twitter. According to Godes et al, online word of mouth tends to
facilitate information exchange and user-generated content diffusion, such as online conversation through ‘one to many’ or ‘many to many’. Hence, many firms are increasing their spending on online conversation marketing to attract audiences. Duan,Gu (2005) and Andrew (2008) argue that everyone can share their thoughts and opinions with a large number of internet users. Their thoughts and opinions influence others' decisions through online conversation. Dellarocas (2003) indicate that an online review system from customer is a most formidable channel to capture online word of mouth.
Prior research focuses on how WOM volume effect on Movies' box office revenue. According to Duan, Gu and Andrew (2008) indicate that rating of online user reviews no significantly influences movies’ box office revenues. But box office sales are large influenced by the volume of customer online reviews. Therefore, online user reviews are important indicator of word-of-mouth that plays a role in driving box office revenues. Dellarocas (2003) also state that online review posting are most powerful pattern to generate online word-of-mouth. According to Duan, Gu and Andrew (2008), Movie industry experts indicate that word-of-mouth is a key factor based on a movie’s ultimate financial success. Liu (2006) argue that actual word of mouth activities are impact on box office revenue for new movies. It is also instrumental in increasing box office sales over all the dramatic release. In addition, Liu (2006) also find out that WOM activities play a very important role during a movie's prerelease and opening week. Movie audiences tend to hold relatively high expectations before releasing. And it becomes more critical in the opening week. WOM information has rapid explanatory power for both total and weekly box office. Thus customer online conversation is an important measure model to forecast new movies’ box office. This
explanatory power influence box office only comes from the volume of WOM, but WOM valence is mixed. Garrette, Leigh and Oliver (2011) also argue that performance are influenced by volume of positive, negative and neutral online communications that are dynamics and endogeneity. Godes and Mayzlin (2004) indicate that around 50% of young Internet users depend on online recommendations to purchase movies, CDs, videos or DVDs. In order to measure volume of online reviews, Godes and Mayzlin (2004) use online user rating as a driving force of product choice. Duan, Gu and Andrew (2008) indicate that both movies’ box office revenue and WOM valence are influenced by WOM volume. But, the rating of online user reviews has no rapidly impact on box office sales of movies. As a result, WOM volume leads to higher box office performance and it reminds the importance of awareness effect, as it also influences the movies’ box office revenues.
Huang and Wang (2008) indicate that the model with high buzz management is more accurate for forecasting box office revenue. Prior studies of movies industry focus on WOM valence and WOM volume have different impacts on different segmentation products of movies market. Movies used to segment into mainstream and non-mainstream products in the market. Jonnhyuk, Wonjoon and Naveen (2012) state that valence of customer online review has a significant effect on box office revenue only in the non-mainstream movies. On the other hand, the effect of online conversation volume on box office revenue is greater for mainstream movies. Duan, Gu and Andrew (2008) state that consumer purchase decisions are effected by online user reviews. Most review sites allow a user to provide star grade of overall rating. The rating could effect other consumers’ perception of product quality.
Previous studies on WOM impact in movie industry show some mixed results.
Elberse and Eliashberg (2003) argue that measurement of online WOM is a key predictor of box office by analyzing the average revenue per screen in the previous week. Neelamegham and Chintagunta (1999) try to find the relationship between online conversation and weekly revenue, but it failed to obtain any rapid results. Dellarocas, Awad, and Zhang (2007) indicate that online review metrics are important predictors of future movie sales through using a modified bass diffusion model to analyze online user reviews effect on forecasting movies’ revenue. Prashant, Madhupa and Ramendra (2015) investigate the impact of consumers' pleasure and satisfaction on customer, likelihood of WOM and positive WOM. They analyze the data by using structural equation modeling. Feng, Yin, Xiaoling, Huawei (2010) use the new product diffusion model and explore how media publicity and word of mouth about a to-be-released new movie drive consumers’ behavior in emerging movie markets. They find that data was collected from the Chinese motion picture industry reveal that to-be-released media appearance and online conversation influence decision-making. Media publicity determines moviegoers’ innovation probability, however online conversations determine innovation and imitation probability.
2.2 Drivers of movie online reviews
Movie online review, the information intensive by moviegoers or potential moviegoers, has become a useful measurement to enhance box office sales today. Ekaterina (2011) demonstrate that movie online review not only throughout movies but also throughout movie actor. Kerrigan (2010) state that the movie actor as a key element to constitute a film’s identity. This is an elements considered by consumers in their selection of films. Nelson and Glotfelty (2012) examine the relationship between
star power and box office revenues. The study uses the box office data from a measure of star power based on the number of visits to an actor’s web page. They indicate that replacing an average star with a top star would increase box office. Farhan Hassan Khan, Usman Qamar, Saba Bashir (2016) use the Sentiment Analysis (also known as Opinion Mining) to analyze many variables such as customers’ sentiments, opinions, and attitudes towards different elements such as movie topics, experts’ reviews and others. During the age of the big data, authors use the statistical methods such as machine learning or the lexical analysis to achieve sentiment classification. The data from seven benchmark datasets that have been used in this research including large movie review dataset, multi-domain sentiment dataset and ‘Douban’ movie review dataset, all of which are available online. Study by De Vany and Walls (1999) indicate that a star feature is a variable in some of econometric analyses of box office performance. Star buzz can increase box office sales in time of opening week. Star buzz also have either positive or negative impacts on anticipation of the movie pre-release. But this study will examine the impact of star power on online reviews during pre-release and later runs. Suman, Subimal and Abraham(2003) state that the popular movie stars improve box office for movies that receive more negative critical reviews than positive critical reviews. Anita and Jehoshua (2003) examine that the variables of movie attributes and advertising expenditures usually influence box office sales in an indirectly way. The study also mention that online WOM for movies is perishable due to time lag is longer between domestic and foreign market performance through foreign exhibitor's screen allocations. In this study, examine the most important factors of language to make impacts on Chinese movies and international movies. According to Daragh and Finloa (2013), customer used make choice of firm is influenced by different languages. Due to customer want
to learn different languages through firm and gain understanding of different cultures. Ramos, Calvão and Anteneodo (2015) think current users and consumers can review and rate products through online services, which provide huge databases which can be used to explore people's preferences and unveil behavioral patterns. They investigate patterns in movie ratings, considering the Internet Movie Database. Jin-Cheon Na, Tun Thura Thet, Christopher S.G. Khoo (2010) investigate that the characteristics and differences in sentiment expression in movie review documents from four online opinion genres. They are respectively blog postings, discussion board threads, user reviews and critic reviews. They collected data from the four types of web sources. The data are about movie review documents. And authors collect a sample of 520 movie reviews were analyzed to compare the content and textual characteristics across the four genres. The study also identified frequently occurring positive and negative terms in the different genres, as well as the pattern of responses in discussion threads. They find that critic reviews and blog postings are longer than user reviews and discussion threads, and contain longer sentences. However, the study only analyzed movie review documents. Similar content and text analysis studies can be carried out in other domains, such as commercial movie reviews, celebrity reviews, company reviews and political opinions to compare the results.
Daehoon and Eenjun (2013) think the rapid development of social network services make it convenience for people to express their thoughts or opinions on various subjects, such as movies. They propose a method for mining public opinion from user comments from social networks and they predict whether the movie will be a box office hit based on customers’’ opinion. Timothy (2007) states that the film market characterized by a very large number of consumers. Information and advice would be
expected to make a significant impact on the film demand. McKenzie (2009) used a range of variations such as distribution, production, and customers’ signals. It is shown that film success responds to previewing, advertising, critical reviews and box office. Charles (2007) offers a method that can identify and measure the impact of online conversation. Charles (2007) also states that information appears to affect consumer behavior quickly, with the length of a movie's run mattering more than the number of prior admissions. Wei Xu, Zhi Liu, Tai Wang and Sanya Liu (2013) research the sentiment recognition technology on online Chinese micro movie reviews and propose an ensemble-learning algorithm based on random feature space division method. David A. Reinstein, Christopher M. Snyder (2005) pointed out an inherent problem in measuring the influence of expert reviews on the demand for experience goods. That is a correlation between good online reviews and high demand may be spurious, induced by an underlying correlation with unobservable quality signals. Using the timing of the reviews by two popular movie critics, relative to opening weekend box office revenue, they apply an approach to circumvent the problem of spurious correlation. After purging the spurious correlation, the measured influence effect is smaller though still detectable. Positive reviews have a particularly large influence on the demand for dramas and narrowly‐released movies (David A.Reistein, Christopher M.Snyder, 2005).
Pradeep, Sladjana, Helen and Henry (2016) compared film online reviewers' evaluations with the online community evaluations. The study finds that community members evaluate movies differently than film reviewers. The results also reveal that community evaluation have more predictive power than film reviewers' evaluations. Wen-Chin Tsao (2014) states that movie reviews by professional critics and ordinary
consumers can significantly influence the behavioral intentions of moviegoers and investigates the influence of consumer expectation and online reviews, movie selection and evaluation by moviegoers. And the results indicate that, without considering the interaction effect, potential moviegoers attach greater importance to consumer reviews than they do critical reviews, and that consumer reviews influence their movie selection as well as their post-viewing evaluation. The influence of negative consumer reviews on movie selection is stronger than that of positive consumer reviews. In contrast, positive consumer reviews are more influential in the evaluation of movies than negative reviews. Consumer expectations are found to moderate the influence of consumer review valence on movie selections and subsequent evaluations. In addition, moviegoers with lower expectations toward a movie had more inclined to be influenced by consumer online reviews and expert opinions.
2.3 Variables description and hypothesis
Compare with previous studies, current study focuses on how the online review volume and valence of customer online reviews affect box office in Chinese market. This paper focus on customer online conversation impacts on movie box office in prerelease week, opening week until the 10th weeks. In this study, research will focus on the impacts of customer online reviews to influence the box office in Chinese market. Besides, star power and language are two factors that will be used to measure the relationship between online conversation and movie box office.
2.3.1 Independent variables
The independent variables of online conversation identified into two types of data in this research. The current study examined two independent variables called volume
and valence: the total amount of customer online review and the total average rating of movies. The 132 movies released between the May of 2015 and the end of 2017. This study collects information of each movie includes: Movie ID, Name of movie, languages, volume of online review per week, score of movie by each user, average score of movie, post date and a series kind of properties. Then, the currently study calculates the weekly amount of online reviews for volume and weekly average rating for each movie for valence through Excel. For the valence, the reviews of star are showed by percentage. For example, there are 13.8 percentages reviewer give the 5 stars to the movie “The Fate of the Furious 8”.
According to Duan, Duan et al. (2008)) measured the online conversation with “Yahoo! Movies” in the same way with the currently study’s approach. This research of movie data was selected from “movie.douban.com”, which is the biggest movie review website in China.
Hypothesis:
H1a: the volume of customer online conversation influences the box office. H2a: the valence of customer online conversation influences the box office.
2.3.2 Dependent variable
In addition, the movie’s box office revenue data was collected from one week before movie’s release date until the 10th weeks. All of movies’ release date is based on time of china in Chinese market. There are canned data of weekly movie’s box office revenue were collected by the website of “www.cbooo.cn”1.
2.3.3 Moderator variable
Apart from the influence of movie properties, other factors will also affect the box office sales. Foreign researchers studied the problem that which factors influence box office sales earlier than Chinese researchers. Studies showed that a series kind of elements such as the type of movie, post date, language, star power, the film production costs and advertising costs can effect the box office sales as well. Because of the specialty of the film industry, some elements such as the film production costs and advertising costs are trade secrets, uneasy to gain accurate information. But those factors will not influence the conservatism of other factors, so in this study, they are not considered. Other factors such as type of the movie post date are also not considered, because this study select movie during a certain period, so the post dates of movies are very close, and are not significant. As to the type, because in this study, we focus on studying the relationship between online conversation and box office sales, type of movie can hardly affect online conversation in China. In other words, the number volume or valence of online conversation will not vary if the type of movie is different. In this study, the following factors will be tested to examine the influence of box office sales.
(1) Star power
Litman (1989) argues that star power was measured by ranking in International Motion Picture Almanac. The currently study refers to star power was measured by ranking of star searching volume index from website of index.baidu.com. This index refers to the one of main actors searching average volume index from a week before the movie performance date till the next two months for each movie. This study selects one main actor and uses the average searching volume index data as star power.
H2b: Star power influences the relationship between the volume of customer online conversation and the box office sales.
H2c: Star power influences the relationship between the valence of customer online conversation and the box office sales.
(2) Languages
The moderator of languages is divided into two types: Chinese & other languages. This study can get this data through collection of data from movie’s properties in website of “movie.douban.com”. In this study, we use ‘1’ to represent Chinese, ‘0’ to represent other languages. This study picked up 66 Chinese movies and 66 foreign movies at random during the May of 2015 to the end of 2017.
H3a: Language influences the box office sales.
H3b: Language influences the relationship between the volume of customer online conversation and the box office sales.
H3c: Language influences the relationship between the valence of customer online conversation and the box office sales.
Table 1 Variables Definition Table
Variables Symbol Description Data sources
Dependent variable Box office sales (Million)
Box office The overall revenue of movie from one week before movie’s release date until the 10th weeks
www.cbooo.cn
Independent variable Volume of customer online conversation
Volume The sum of people who write comments
movie.douban.com
Valence of customer online conversation
Valence Average rating of movies movie.douban.com
Moderator variable
Star power Star The average of star searching volume index from the one week before the performance date till the next two months
Index.baidu.com
3. Model and data
3.1 ModelAccording to previous studies on online conversation and box office sales, the volume of customer online conversation and the valence of customer online conversation are two factors influencing the box office. Yong Liu (2006) demonstrated that WOM affects customers’ behavior mainly though two ways. On one hand, the volume of online conversation will influence the cognition degree of products among customers. The more people know the product, the easier people will buy it. On the other hand, the valence of online conversastion, which can be categorized as positive or negative, will also affect attitudes of customers. This study use customer online conversation are including volume and valence as independent variables, box office sales as dependent variables, languages and star power as moderator variables, analyzing how those variables influence the box office sales and the interaction effect between moderator variables and online conversation of volume and valence. The relation model is shown in figure 1：
3.2 Data sources 3.2.1 Sample collection
This study select 132 movies from May of 2015 to the end of 2017at random as research samples, collecting each movie’s box office revenues, volume of customer online conversation, valence of online conversation, language and star power respectively for analyzing. In order to investigate this thesis topic, the sample of this study was collected from three sources from Chinese websites. The currently study random chooses 132movies, are both 66 Chinese movies and 66 foreign movies, for collecting online conversation data from the website of movie.douban.com. On the other hand, the dependent variable of box sales refers to weekly box office revenue is collected from the China's official website of the box office2. According to Alex’s ranking of visit, Baidu.com is a website which has the largest number of visitor per day in China. This study uses the index of star searching to measure the star power in the Chinese market from index.baidu.com. Hence, star power was measured by ranking of star searching’s volume index.
In order to research database analysis, the currently study uses combine with method of data mining, database research and grounded theory to collect useful data from the sample sources. The currently study uses quantitative methodology to calculate and analysis sample data through SPSS 22.0 for Windows. There are some strengths and limitations of methods in the research design. First, the strength of this paper in the method of collect data is it has a very large number of movies. Sample of movie’s properties are clear and easier to collect in the website flexibility. Random selection from sample of movies can enhance the accuracy of the result. This research design 2The official Chinese box office websitewww.cbooo.cn
also has advantage of low costs by websites allow to free collect data. Then, is potentially limited to data of foreign movies are collected from similar cultures and preferential country, such as American movies. Additionally, Williams, Cote and Buckly (1989) argue that people cannot rule out the possibility of the measures suffers from common method variance because data are self-report. Hence, this study also existing this problem same with them.
4. The significance test for variables and result
In this chapter, the correlation and regression analysis will be discussed below. The correlation analysis of modal 1 is used to quantify the association between two independent variables and a dependent variable. The test of modal 2 will be analyzed the relationship between two continuous variables with two moderator variables.
4.1 Independent variable correlation text
This paper use volume of online conversation and valence of online conversation from customers as two dimensions of word of mouth. First and foremast, the correlation between independent variables and dependent variable will be examined, analyzing the correlation between various dimensions of online conversation and movie box office sales.
Table 2 Descriptive Statistics of Independent Variables
Variables N Minimum Maximum Mean SD
Volume Valence 132 132 3,681 3.1 501,065 9.2 102,757 6.699 99,371 1.565 Notes: Valid N=132
In table 2, descriptive statistics was used to analyze the number of the customers’ online review and the rating of online comments of the movies. For these 132 samples, the maximum number of online review is 501,065. And the minimum number of online review was 3,681. The average number was 102,757, which indicates that for different movies, the difference of the number of online review was relatively large. The maximum rating of movie was 9.2, the minimum rating was 3.1, and the average was 6.699, which means that for different films, the acceptance of customers’ was different. To sum up, the average online rating of the sample movies was about 6.7. Table 3 Correlations and Regression between independent variables and dependent variable
Table 3 Correlations and Regression between Independent Variables and Dependent Variable
Model Variables B Stad.Error Beta t Pvalue
1 (Constant) 644.289 181.418 3.551 .001 Volume Valence .003 -99.025 .000 29.298 .768 -.415 7.783 -3.287 .000 .001 Notes: N=132. Dependent variable: Box office sales
From Table 3, according to the Correlation of dependent variables and independent variables, it is easy to know that the B value were 0.003 for volume and -99.025 for valence. Which means that the two independent variables are relevant to the dependent variable. And also indicates that the volume of online conversation has positive correlation with box office sales, but the valence of online conversation and box office sales are negative correlation. What’s more, their P values both were less than 0.05 (P value of volume=0.000, P value of valence=0.001), reflecting that the volume and valence of online conversation have significant correlation with box
office sales of movies. Hence, hypothesis H1aand hypothesis H1bare supported.
4.2 The significant test for moderator variables
Considering there are many other factors probably influence box office sales, to better test how customers’ online conversation affect box office sales, this paper select star power and languages as moderator variables to boost research. In order to verify whether the model is established, these moderator variables need to accept significance test.
4.2.1 Star power
Though sorting and analyzing the sample movies, many movies were found possessed famous stars who have worldwide influence and reputation. So in this study, star power was considered as an important moderate factor which is may affect box office sales. This paper picked up two main actors for a movie to represent the star power of the movie, through index.baidu.com. The index of these stars was calculated by customers’ searches. This study used the averaged data within two months when the movie is on show as star power to analyze the influence of star power on box office sales.
Table 4 Descriptive Statistics of Star Power
Variable N Minimum Maximum Mean SD
Star Power 132 135 216,187 14,448.33 99,371
Notes: Valid N=132
According to Table 4, the maximum of star power was 216,187, the minimum was 135 and the mean is 14,488.33
Table 5 Correlations and Regression between Star Power and Box Office Sales
Model Variables B Stad.Error Beta t Pvalue
1 (Constant) 644.289 181.418 3.551 .001 Volume Valence StarPower .003 -99.025 .000 .000 29.298 .001 .768 -.415 -.017 7.783 -3.287 -.207 .000 .001 .837 Notes: N=132. Dependent variable: Box office sales
Table 5 demonstrates that the correlation B value of star power and box office sales was 0.000. It shows that star power does not have positive or negative correlation with box office sales, meanwhile, P value=0.837>0.05, can also prove that star power does not have significant correlation with box office sales. So hypothesis H2a is not correct. It can be explained by today’s movie market, because of high degree of entertainment, whether the actors were famous can hardly be defined. Meanwhile, a glance at Chinese and foreign actors is easy to discover that though some actors were not main actors of a movie and did not receive an important award in the industry, their high visibility and high popularity should not be neglected.
Table 6 Correlation and Regression between Independent Variables and Dependent Variables with Moderator of Star Power
Model Variables B Stad.Error Beta t Pvalue
2 (Constant) 1341.85 243.998 5.499 .000 Volume Valence StarPower SP_Vol SP_Val .004 -176.91 .005 .000 .001 .000 33.963 .003 .000 .002 .848 -.618 .344 .254 .084 9.418 -5.209 1.654 2.283 .437 .000 .000 .101 .024 .663 Notes: N=132. Dependent variable: Box office sales
According Table 6, the P value=0.024 which means the star power is a significant moderator to influence the relationship between volume of online conversation and box office sales. Then, the hypothesis H2b is correct. But the regression coefficients B value and t value are 0.000 and 2.283 respectively. Then, star power has less meaningful to impact the relationship between the volume of online conversation and box office sales. On the other hands, star power is also not useful moderator to influence the relationship between the valence of online conversation and box office sales because the P value equal 0.663. It is above the 0.05. So the hypothesis H2c is inconsistent.
4.2.1 Language
This paper selected 132 movies from various areas including mainland of China, Hong Kong, America, British, Japan, etc. So it is hard to define the language. In this study, all languages of the movies were divided into two parts: Chinese and other languages. Chinese was valued 1 and other languages were valued 0. This study collected 66 Chinese movies and 66 foreign movies. It is showed in the Figure 2:
Figure 2 Bar Chat of Languages Distribution
Table 7 Correlations and Regression between Languages and Box office sales
Model Variables B Stad.Error Beta t Pvalue
1 (Constant) 854.080 214.297 3.985 .000 Volume Valence Languages .003 -118.62 -146.51 .000 30.998 81.583 .744 -.415 -.167 7.783 -3.827 -1.796 .000 .000 .075 Notes: N=132 Dependent variable: Box office sales
The output of modeling correlations and regression for box office sales and language is showed as Table 7. Due to the B value was -146.51, the positive correlation between the languages and box office sales. Table 7 also showed that P value=0.075>0.05, it can draw the conclusion that the influence of language on box office sales is NOT significant. Hypothesis H3ais false.
Actually, in China, with the rapid development of movie market, more and more foreign movies were imported, and more and more Chinese were attracted by foreign movies such as some Hollywood movies, for example, Transformers, had made a very high box office sales in China. But apart from those Hollywood movies, other foreign movies did not seem to have so many audiences. On the contrary, domestic movies are no matter it is big-budget movies or low-budget movies, almost all have satisfied box office sales, which may be related to language difference.
Table 8 Correlation and Regression between Independent Variables and Dependent Variables with Moderator of Star Power
Model Variables B Stad.Error Beta t Pvalue
2 (Constant) 1341.85 243.998 5.499 .000 Volume Valence Language Lang_Vol Lang_Val .004 -176.91 -298.84 -.001 281.179 .000 33.963 80.753 .001 65.662 .848 -.618 -.341 .130 0.398 9.418 -5.209 -3.701 -1.453 4.282 .000 .000 .000 .149 .000
Notes: N=132. Dependent variable: Box office.Languages divided to Chinese (1) and others languages (0).
In the table 8, it is very clear to show that language does not influence the relationship between volume of online conversation and box office sales, since the P value was 0.149. Also the moderator of language effects negative relationship between volume of online conversation and box office sales. Hypothesis H3b is not correct. But language has a significant relevance between the valence of online conversation and box office sales. Because the B value showed that there are large positive correlation between valence of online conversation and box office sales. Also the P value less than 0.05, which is means H3cis supported.
5. Conclusion
With the deeper development of movie market, movie becomes more and more industrialization and Marketization. It is no longer a movie but becoming a kind of competition and a way of profits among producers. The pivot of the competition is movie box office sales. Movie box office is the reflection of a movie, revealing its reputation, its overall revenue, customers’ consumption results, etc. Also, movie box
office sales can embody customers’ interest, their decision-making and their buy-up behavior, which can be important for both the market and movie producers. According to consumer purchase decision theory we learn that customers’ buying decision can be influenced by various kinds of factors. Not only customer itself but also the properties of products or service can affect customers’ decision. Same, box office sales of a movie can be influenced by many factors, for example, the type of movie, the quality of movie, propagation, customers’ preference, etc.
Within movie industry there is a saying that is “At first, a movie depends on propagating, later, it relies on word of mouth”. It shows that how important WOM is for a movie. With the ever increasing development of Internet, video websites such as douban.com 、 shiguang.com not only provide information about movie, but also provide a platform for audiences to discuss and comment the movies. Such online conversation are not only help people gain some information about a movie but also influence people’s purchase decision making.
This paper reviewed literature and theories about WOM, online conversation, WOM influence customer purchase decision, based on it, combining with the character of movie this industry, this paper divided online conversation into volume and valence two parts to do the research, analyzing how the volume and valence of online conversation will influence movie box office sales. Meanwhile, this paper imported language and star power two factors as moderator variables. But after testing their significance influence, star power was found not significantly affect movie box office sales and so was deleted. After that, modified model of the relationship between online conversation and movie box office sales was obtained. To be specific, the
volume of online conversation and the valence of online conversation have significant correlation with movie box office, at the same time, the moderator variable language affects the volume and valence of online conversation and thus influences box office sales. This paper selected 132 movies from both China and other countries as sample, using SPSS analyze and verify the hypothesis, drawing conclusion as follows:
The volume of online conversation is significantly positive correlated with movie box office sales. This paper collected data from movie.douban.com, gaining volume and
rating of movies for analyzing and drawing conclusion as previous studies, that the volume of online review is positively correlated with box office sales. The more the reviews are, the more attention the movie gets, and more people will engage in discussing and commenting the movie. That means larger the volume of online conversation about a movie, the more customers can recognize it. It is easy to understand that for products or service, the more people know it, the better the sales.
The valence of online conversation is significantly negative correlated with movie box office sales. Normally, the website of ‘Duban Movie’ provides such a function for
audience to comment and give a rating for a movie directly, the overall rating represents the direct feedback from audience, reflecting an average quality of the movie and audience’s acceptance. When people do not know a movie and is suspicious about the propagation of the movie, they will turn to WOM information for reference and thus make a decision. But, the analysis results of online rating and movie box office sales show that the higher the rating, in this paper, we use valence to represent, not means the box office sales higher. This is the counterintuitive phenomenon. But this counterintuitive phenomenon is actual occur in the Chinese
movie market. According to the Flanagin’s experimental result (2013), customer give the rating could be influenced by social information from others. Customer used to give the rating tends to commonage after watch the movie. On the other words, experts of online conversation have very strong influence to customer in the niche movies area. There are some high ratings of niche commercial movie, but their performances are very low. For example, the movie of ‘Mr. Donkey’ got the 8.3 grade of valance, but his performance is low for the box office sales. Due to this movie spend lots of costs to advertise and bought more ‘Internet water army’3 to score this movie online. However, this movie did not attract real audiences to concern it. On the other side, the Movie of ‘Tiny Times 1.0’ has the number of rating over the twenty-three ten thousand but the grade of this movie was only 4.7. Even though the 30% audiences gave the one star of rating to this movie, but it still has very high performance of box office (five hundred million RMB). This high performance box office is not only the crazy fans to provide, but also from some customers go to cinema because they are curious about the low rating.
The influence of star power is not significant. This paper studied the average star
power of 132 movies from China and foreign countries. It is difficult to define whether the star has high reputation or high popularity since some actors have high popularity without gaining an important award while some actors have high reputation in the industry but few people admired them and have few fans. Such situation probably influences the overall box office sales of a movie. For example, Jim Carrey never awards the prize of Oscar. But his movie always got very good performances of box office. In addition, Denzel Washington award five times Oscar
prize and his movie does not have high box office sales. So star power cannot significantly affect movie box office sales.
Language to some extent influence movie box office sales. This paper regarded
language as a moderator variable, study how this variable will affect the intersection relationship between volume of online conversation, valence of online conversation and box office sales. Language is not a useful moderator between volume of online conversation and box office sales. But it is to some extent to influence the valence of online conversation and box office sales. In the Chinese movie market today, movie of foreign languages are not translated to Chinese directly. More and more customers tend to watch the original movie from other countries with Chinese subtitle. But Chinese movie are easier to understand and make customers’ good feeling to rate it.
6. Marketing Suggestions
Under the background of movies industry, the box office success largely represents the success of the film, so the film producers never slow down the step to chase for the box office, but with the deepening of the marketization of the film and the high speed development of the Internet, how to better use the network environment to market the movie word-of-mouth gradually become film producers’ concerns. Based on the conclusions of this paper, relevant suggestions and countermeasures are put forward in this paper:
Improving the production level of the film itself, so as to obtain the high box office with high quality. As a cultural product, the film brings a spiritual satisfaction to the consumers, and the quality of the film itself is the essential factor for its high box office and good reputation. So the producers must first pay attention to the film itself,
pursuit the breakthrough and improvement of the movie content, theme and making quality, in order to increase the film appeal, set up the brand value of the film, finally obtain high box office.
Creating the effect of online conversation, and improving consumers’ attention for films. Movie is a typical experiential consumption, consumer cannot accurately predict the viewing experience before watching movies, only the information which is related to the movie can help them to make the decision whether to watch a movie or not, so any topics which are caused by the film can form certain reputation effect, and improve the movie’s popularity and attention of the audience for movies. And the high popularity and high attention are premises of obtaining a high movie box office. For movie producers and distributors, in order to improve the film publicity, they can manufacture a range of topics to build word-of-mouth effect for film, and improve the film appeal.
Pay attention to the marketing value of and gain competitive advantages. With the rapid development of the Internet era, the spread of word-of-mouth information is faster, and situations are more diversified. The influence of online word-of-mouth for consumers’ purchasing decision also becomes more profound. Currently, some film word-of-mouth websites, which are represented by “Douban Movie”4 and “Mtime”5, have a higher credibility and an authority, and impart more influence on audience than advertising. Movie producers and distributors should attach great importance to the online conversation marketing value, and make the online conversation as an important part of the movie marketing, thus increasing the film appeal, and gain a competitive advantage.
4Douban Movie website:https://movie.douban.com/ 5Mtime Movie website:http://www.mtime.com/
Pay attention to the inherent characters of and play an active role of online conversation in all aspects. Customers’ online review mainly includes the number of online reviews and the value of online reviews, and these two factors greatly affect the influence of the Internet word-of-mouth for the viewer. Movie producers and distributors shall make full use of these two properties, as far as possible play their positive roles, and encourage more people to participate in film discussions on the network platform, and guide them to make more positive evaluation. Movie producers and distributors can cooperate the relevant film discussion websites to encourage people to post their comments and feelings of the film as much as possible on the website. For example, during the release of a movie, they can cooperate with related movie reputation websites and host the film discussions and feeling exchanges on-line meetings, give certain material rewards to the movie audience who actively participate in the film discussion, or own unique insights and positive comments in the film. The rewards include posters, signature photos related to the film, souvenirs, or the website's permissions, etc. The purpose is to encourage more people to participate in the film discussions, and to guide the audience to spread more positive word-of-mouth, then to attract more audiences into the cinema and to achieve higher ticket sales.
In the senior squad, the famous actors are strong backings for the box office guarantee. In the case of cost permitting, can appropriately rise the box office through combined effects between senior squads. But what we need to note that in the high point of quantities’ regression, it sees the marginal effect is importantly diminishing, and is not significant. Recently, the CCTV news channel pointed out that more than 50% of the production costs are taken by domestic big stars in "Actor's sky-high salary" special report. It will lead to huge risks pressure for the other aspects of film and television
production. The producers can improve the film publicity and marketing ability to promote box-office receipts, seize the opportunities which are brought by the “Internet plus” to make good strategy ahead of time, then construct a new three-dimensional distribution system, which can enhance the popularity by manufacturing network hot topic, improving fans’ online interactions, making full use of the mobile terminal flow resources. Based on the situation that network word-of-mouth has a growing influence on people viewing choices, some movie ratings websites should strengthen the construction of film evaluation system, prevent the phenomenon of malicious scrubbing, at the same time, introduce professional film critics and senior fans to play roles of opinion leader and guide more objective and insightful comments. At the same moment, the producers can also provide valuable lessons for future production by referring to the popular film review, then adjusting the marketing strategy and making up for the short coming by.
7. Deficiency and further prospects 7.1 Deficiency
This study selects 132 movies from May of 2015 and the end of 2017 at random as samples. Because the number of samples is limited and the types of samples are different, the research conclusion has limitations. In further study, the research sample should be expended, and should be random, to strength the accuracy and persuasiveness of the research conclusion.
Band it is difficult to gain real-time box office data, this study use historical box office data to do the research. Actually, though literature review and analysis, it is not difficult to find that at different period when the movie was posted, the degree box office sales influenced by online conversation was different. Although this study
tested the influence of online conversation during the post time, the influence in different period time was neglected.
This study gets data mainly from movie.douban.com, because of its authority in Chinese movie market. But online conversation has anonymity which probably lead to some moviemakers improperly propagate their movie, for example, give a high rate for their own movie but a low rate for competitors’. On the other hand, such behavior was limited and few movie producers would do. So the conclusion still has its persuasiveness.
Many other factors probably influence the box office sales, however, in this study, according to the gained data, apart from online conversation, only two factors were considered, having its limitations.
7.2 Future research prospects
This study divided online conversation into two dimensions, volume and valence. In fact, the property of online conversation, which means the content of customer online review, is also an important constitution of online conversation, but it has strong subjectivity, and is hard to quantify, so this study did not include it. In further study, the property can be considered as a factor and add into the research of online conversation.
Because the study included data that film was rolled off the assembly line, can not completely match the box office, in future research, real-time data of online conversation and box office revuences can be collected to analyze the relationship between online conversation and box office sales much more accurate.
This study according to online conversation, explored its influence on box office sales, however, box office sales reflected by customers’ purchase behavior directly. On the
contrary, a movie’s box office sales can affect online conversation as well. It is another problem can be research and analyze in further study.
Reference:
Anita Elberse, Jehoshua Eliashberg, (2003)Dynamics for Sequentially Released Products in International Markets: The Case of Motion Pictures. Marketing Science 22(3):329-354.
Charles C. Moul. Journal of Economics & Management Strategy, (2007), Vol.16 (4) C.Dellarocas, The digitization of word of mouth: promise and challenges of online feedback mechanisms, Management Science 49 (10) (2003)
1407–1424
Kerrigan, F. (2010), Film Marketing, Butterworth Heinemann, Oxford.
Daragh O'Reilly Finola Kerrigan, (2013),"A view to a brand: introducing the film brandscape", European Journal of Marketing, Vol. 47 Iss 5/6 pp. 769 – 789
De Vany, A.S. and Walls, W.D. (1999), “Uncertainty in the movie industry: does star power reduce the terror of the box office?”, Journal of Cultural Economics, Vol. 23 No. 4, pp. 285-318.
Duan, Wenjing, Bin Gu, and Andrew B. Whinston. "Do online reviews matter?—An empirical investigation of panel data." Decision support systems 45.4 (2008): 1007-1016.
D. Godes, D. Mayzlin, Using online conversations to study word of mouth communication, Marketing Science 23 (4) (2004) 545–560 Fall.
Randy A. Nelson, Robert Glotfelty. Journal of Cultural Economics, (2012), Vol.36 (2), pp.141-166
Daehoon Kim, Daeyong Kim, Eenjun Hwang, Hong-Gu Choi. New Review of Hypermedia and Multimedia, (2013), Vol.19 (3-4), pp.259-272
David A. Reinstein, Christopher M. Snyder. The Journal of Industrial Economics, 2005, Vol.53 (1).
Word-of-Dellarocas, C., Zhang, X. M., & Awad, N. F. (2007). Exploring the value of online product reviews in forecasting sales: The case of motion pictures. Journal of
Interactive marketing, 21(4), 23-45.
Feng Wang, Yin Zhang, Xiaoling Li, Huawei Zhu. Journal of Interactive Advertising, (2010), Vol.11 (1), pp.50-62
Flanagin, A. J., & Metzger, M. J. (2013). Trusting expert-versus user-generated ratings online: The role of information volume, valence, and consumer
characteristics. Computers in Human Behavior, 29(4), 1626-1634.
Farhan Hassan Khan, Usman Qamar, Saba Bashir. Knowledge-Based Systems, 2016. Garrett P. Sonnier, Leigh McAlister, Oliver J. Rutz, (2011) A Dynamic Model of the Effect of Online Communications on Firm Sales.
Huang, M., & Wang, F. (2008, October). Using Online WOM to Forecast Box Office for Movies Coming Soon. In Wireless Communications, Networking and Mobile Computing, 2008. WiCOM'08. 4th International Conference on (pp. 1-4). IEEE.
Hennig-Thurau, Thorsten ; Wiertz, Caroline ; Feldhaus, Fabian
Journal of the Academy of Marketing Science, (2015), Vol.43(3), pp.375-394 [Peer Reviewed Journal]
Joonhyuk Yang, Wonjoon Kim, Naveen Amblee, Jaeseung Jeong, (2012) "The heterogeneous effect of WOM on product sales: why the effect of WOM valence is mixed?", European Journal of Marketing, Vol. 46
Jordi McKenzie. Journal of Cultural Economics, (2009), Vol.33 (4), pp.279-299 Jin-Cheon Na, Tun Thura Thet, Christopher S.G. Khoo. Online Information Review, 2010, Vol.34 (2).
Karniouchina, E. V. (2011). Impact of star and movie buzz on motion picture distribution and box office revenue. International Journal of Research in
Marketing, 28(1), 62-74.
Litman, Bary. R. and Kohl, L.S.(1989). Predicting Success of Theatrical Movies ： The 80’s Experience ． Journal o f Media Economics, 2, 35-50.
Timothy King. Journal of Cultural Economics, (2007), Vol.31 (3), pp.171-186
Prashant Mishra, Madhupa Bakshi, Ramendra Singh. Australasian Marketing Journal (AMJ), (2015)
Pradeep Kumar Ponnamma Divakaran, Sladjana Nørskov, Helen Hasan and Henry Linger. Information Technology & People, 2016, Vol.29 (1), pp.120-145
Wen-Chin Tsao. Electronic Commerce Research, 2014, Vol.14 (4), pp.559-583
Williams, L. J., Cote, J. A., & Buckley, M. R. (1989). Lack of method variance in self-reported affect and perceptions at work: Reality or artifact. Journal of Applied Psychology, 74(3), 462468.
Ramos Marlon, Calvão Angelo M, Anteneodo Celia. PloS one, 2015, Vol.10 (8), pp.e0136083
Ramya Neelamegham and Dipak Jain,Journal of Marketing Research,Vol. 36, No. 3 (Aug.1999), pp. 373-386
Suman Basuroy, Subimal Chatterjee, S. Abraham Ravid (2003) How Critical Are Critical Reviews? The Box Office Effects of Film Critics, Star Power, and Budgets. Journal of Marketing: October 2003, Vol. 67, No. 4, pp. 103-117.
Thet, T. T., Na, J. C., & Khoo, C. S. (2010). Aspect-based sentiment analysis of movie reviews on discussion boards. Journal of information science, 36(6), 823-848
Yong Liu (2006) Word of Mouth for Movies: Its Dynamics and Impact on Box Office Revenue. Journal of Marketing: July 2006, Vol. 70, No. 3, pp. 74-89
