Predicting Video Game Success Through Quality Signals and Product Innovativeness

(1)

Predicting Video Game Success Through

Quality Signals and Product Innovativeness

Name: Stephan van Balen Student number: 1604228

University of Groningen

Faculty of Economics and Business MSc Strategy & Innovation

Master Thesis

(2)

2

Abstract

This study investigates the effect of quality signals and product innovativeness on video game sequel performance. 469 video game sequels released between 2000 and 2011 for the PlayStation 2 were studied to find which factors influence a sequel‘s performance. Quality signals were shown to positively influence financial performance, with star power, media adaptations, sequel number, and volume of user reviews contributing most to this result. None of the innovativeness indicators significantly influenced sequel sales, and innovativeness was also found to not have a curvilinear relationship with sales, both contrary to expectations. However, very high innovativeness was shown to influence sales negatively when quality signals are high. Video game developers are recommended not to introduce too much innovativeness to high profile sequel products as this will alienate the existing customer base.

Keywords: product innovativeness, quality signals, video game industry, video games. Word count: 7545

Acknowledgements

(3)

3

1. Introduction

Producing sequels to (successful) products is a strategy that is increasingly popular in the video game industry, as illustrated by the fact that 61% of all major video games released in the United States in 2011 were sequels (147 of 239, Metacritic.com, 2012a). The popularity of sequels stems from its appeal to existent as well as new customers; a sequel can signal a certain quality to these consumers. Quality signals are those signals that potential customers can use to form a perception of quality (Basuroy, Desai, and Talukdar, 2006). Signaling quality to the consumer is important in the video game industry because a video game is what is referred to as an experience good: goods the consumer cannot easily judge before consumption (Nelson, 1970; 1974).

Thus, video game developers inevitably run into the issue of having to convince consumers to buy their product before the customer can ascertain whether it is the best product for them. In order to overcome this issue they have to somehow signal the quality of their product to consumers (Kirmani and Rao, 2000). Through quality signals they try to influence the perceived quality of a product, which Zeithaml (1988) defines as ―the consumer‘s judgment about a product‘s overall excellence or superiority‖ (p.3). In the absence of intrinsic quality indicators, as is the case with experience goods, consumers can only rely on these signals to infer the true quality of the product. In other words, video game developers who are able to effectively signal quality to the consumer can expect to see an improvement in the financial performance of their video games.

While some of these quality signals are important for every video game (e.g. critical evaluation, awards, and competition), other variables are specifically important for video game sequels (e.g. the success of the predecessor). Product innovativeness is important for all videogames, but perhaps even more important for sequels because the consumer has to be convinced that the sequel is a significant improvement upon the parent product. If it is not, and the customer already owns the predecessor, there is no reason for him or her to buy the sequel. On the other hand, if a sequel is too innovative, the customer base may be alienated and prefer to keep playing the predecessor.

(5)

5 a famous character in your video game can increase financial performance, and Zhu and Zhang (2010), who showed that positive user reviews influence financial performance for online and less popular video games. This lack of research on signaling in the video game industry necessitates using film industry literature as a basis, something which Zhu and Zhang (2010) have also done.

Thus, this study helps video game developers understand why certain sequels perform better than others, with a focus on the importance of innovativeness, quality signals and the interaction between those two.

2. Literature review

2.1 Conceptual framework

The focus of this study is on sequel video games because of the large number of sequels produced in the video game industry and the lack of research on the subject. A video game sequel, which is a follow-up to an original video game, is a quality signal in and of itself for consumers, because consumers who enjoyed the original video game will have high expectations of the sequel and can be expected to be more willing to pay for it. Dhar, Sun, and Weinberg (2012) argue that sequels can be seen as a brand extension of the original or parent product, and many studies on the film industry (e.g. Ravid, 1999; Basuroy and Chatterjee, 2008; Moon, Bergey, and Iacobucci, 2010) shows that film sequels outperform their non-sequel counterparts. In fact, it is one of the only variables that practically all research in the film industry consistently finds to have a significant and positive effect on financial performance. As illustrated in the conceptual framework in figure 1, this study defines financial performance as the number of units sold and the revenue of a video game. The other variables in the framework will be discussed in the following paragraphs.

(6)

6 An additional advantage to studying only sequels is the fact that it allows controlling for the success of the parent game, thereby decreasing endogeneity. Endogeneity is essentially a causality loop between the independent and dependent variables (Dhar, Sun, and Weinberg, 2012). This occurs because video games do not randomly spawn sequel games; when an initial game is successful the chance that the sequel is also successful increases because the developer can more easily copy the factors that made the initial video game successful.

2.2 Quality signals of video game sequels

Little academic research has addressed the role of signaling in linkage with performance prediction of video games. This research will therefore draw upon the signaling literature of the film and literature industries. In order to find which variables were most often used in research on signaling or performance prediction in other industries, a large number of these articles were summarized and the most important variables tabulated in table 1. These variables will be discussed more in depth in the following paragraphs.

TABLE 1: Variables derived from studies on the film and literature industry Variable Rationale Translation to videogames Sources

Sequel

A sequel is a form of brand extension which will result in a higher perceived quality

A videogame that is a sequel or prequel to, or a remake of, an existing game

Basuroy and Chatterjee (2008); Chang and Ki (2005); Dhar et al. (2012); Moon et al. (2010); Sood and Drèze (2006) Critics‘

evaluation

High critical evaluation will result in a higher perceived quality

These variables can be used exactly as they are used in the film industry

Clement et al. (2007); King (2007); Litman and Kohl (1989)

User rating

High user ratings will result in a higher perceived quality

Chang and Ki (2007); Hennig-Thurau et al. (2006); Moon et al. (2010

Awards Receiving awards will result in a higher perceived quality

Clement et al. (2007); Gemser et al. (2008); Hennig-Thurau et al. (2006); Litman and Kohl (1989); Ravid (1999) Media

adaptation

A license to well-known

intellectual property will result in a higher perceived quality

Hennig-Thurau et al. (2006); Joshi and Mao (2010); Simonson (2009a)

Sequel number

A higher number of intervening sequels between the parent and the focal sequel will result in a higher perceived quality

Basuroy and Chatterjee (2008)

Star power The presence of a star will result in a higher perceived quality

Game with a celebrity in the title (e.g. Tony Hawk's Pro Skater) or on the cover (e.g. Wayne Rooney, FIFA 12)

Ainslie et al. (2005); Gemser et al. (2007); Hennig-Thurau et al. (2006); Sharda and Delen (2006); Ravid (1999)

(7)

7 shown in many theaters) exhibit a positive relation between critical rating and gross earnings, but limited releases show no such relationship. Gemser, Van Oostrum, and Leenders (2006), on the other hand, showed that the size and number of reviews (volume), rather than the rating (valence), are important. Dellarocas, Zhang, and Awad (2007) also show that volume, not valence, can be used to forecast motion picture sales.

Users’ evaluation. Next to critical evaluation, the evaluation of users can be a predictor as well (e.g. Hennig-Thurau, Houston, and Walsh, 2006), because of its signaling function. An assumption could be made that this is also the case for video game consumers, possibly to an even greater extent due to the large amount of time they spend online searching for information (Cook and Coupey, 1998). Research by Zhu and Zhang (2010) in the video game industry shows that this is not necessarily true; their research shows that online reviews are influential only for less popular and online games.

Awards. Awards are also a clear quality signal to consumers because they indicate that a certain product is better than all its contemporaneous competitors. Research on awards in the film industry has shown that only the ‗Best Film‘ award, and its nominations, have any correlation with financial performance (Litman and Kohl, 1989). Furthermore, Gemser, Leenders, and Wijnberg (2008) found that the most important awards in the film industry, the Academy Awards, do not outperform other awards in terms of predictive capability.

Star power. Star power in the video game industry can either refer to a video game character with a broad appeal such as Mario or the endorsement of a celebrity. Binken and Stremersch (2009) find that superstar power of a video game character positively correlates with sales, in particular during the first five months, but they pay no attention to celebrity endorsement. This study will supplement their research by focusing solely on celebrities lending their name (e.g. Tony Hawk‘s Pro Skater) or appearance (e.g. Wayne Rooney for FIFA11) to market a video game. According to Elberse (2007) this type of star power can influence financial performance because it increases the perceived quality of a product. In the film industry star power is used as a predictor in the majority of research, albeit with mixed results. For example, Chang and Ki (2005) find no significant power in star power as a predictor of financial performance at all, while Basuroy, Chatterjee and Ravid (2003) find that star power only correlates with financial performance if the evaluation of critics is negative.

(8)

8 eponymous book). Much like a sequel, a media adaptation is a form of brand extension. The reason this may influence the financial performance of a video game is the fact that the advertising budget and word of mouth of the film or book will to some extent carry over to the video game. Furthermore, part of the existing customer base of the other media can be expected to buy the video game, also improving financial performance. Adapting a book is a very popular strategy in the film industry but most research shows no correlation with financial performance. For example, Hennig-Thurau, Houston, and Walsh (2006) argue that when a film is based on a familiar concept such as a novel, comic, or television series the box office success will improve, but their results show an insignificant correlation. Simonton (2009a) also finds that adaptations do not stand out at the box office. In depth research, however, by Joshi and Mao (2012), shows that book adaptations do perform better at the box office—especially during the opening weekend.

Sequel number. The last of these potential signals is specific to sequel products and comes from Basuroy and Chatterjee‘s (2008) research on the correlation between the number of previous sequels and the consequent sequel‘s performance. Their conclusion is that a high sequel number correlates positively with financial performance because it is a signal of the continued success of the series. Consumers will think that ―a rational studio would not attempt to make so many remakes unless they have received positive response from the market‖ (Basuroy and Chatterjee, 2008, p. 800). This phenomenon is expected to occur in the video game industry as well, although an argument could be made that in the long run the opposite could happen as well. When consumers are faced with a the fifteenth iteration of a video game series they may get tired of playing a similar video game and refrain from buying it. While Basuroy and Chatterjee (2008) did not find this to be true, video game series can run relatively long, making it possible for this to happen more frequently in the video game industry than in the film industry.

The strength of these quality signals is predicted to positively influence the financial performance of a video game, leading to the first hypothesis guiding this research:

H1: A positive relationship exists between the strength of a video game‘s quality signals and its financial performance.

2.3 Innovativeness of video game sequels

(9)

9 2002). This is important for video game sequels, because sequels that are deemed to be too similar to the parent will not be bought by customers who already own the parent and video games that are too innovative may alienate the loyal customer base. This implies a curvilinear relationship between innovativeness and financial performance.

Measuring innovativeness is a complex issue because it is a multi-dimensional concept (Neely and Hii, 1998) and measuring it for video games is perhaps even harder because common measures like patent counts are largely irrelevant in the video game industry. In order to overcome this issue, this study uses multiple indicators of innovativeness to define sequel innovativeness. The indicators of innovativeness used for this study are: titling departure, genre departure, publisher/developer departure, and innovation awards, which will be discussed in the next paragraphs.

Titling departure. The first possible innovativeness variable is derived from a study by Sood and Drèze (2006). They have studied the importance of sequel titling and they conclude that dissimilarly titled sequels (e.g. Need for Speed: The Run, rather than Need for Speed 18) are rated higher by consumers because they are experienced as more different from the parent film. Thus, the product is perceived as being more innovative.

Publisher/developer departure. A change in publisher, or more importantly developer, can also be an indicator of innovativeness. After all, if Rockstar Games outsources the development of Grand Theft Auto V to another developer (or even sells the IP), many loyal fans of the Grand Theft Auto series may expect the video game to be very different from what they have come to expect from the series.

Genre departure. The few studies that investigate innovativeness of experience goods focus on the combination or introduction of new genres (e.g. Mezias and Mezias, 2000; Jones, 2001; Perretti and Negro, 2007). This approach lends itself to this study because it allows for the comparison of the perceived innovation of a sequel video game compared to the previous video game in the series. Consumers are likely to perceive a video game as more innovative when they read or hear that the developer has switched genres.

(10)

10 Similar to existing research (e.g. Rogers and Shoemaker, 1971; Kunz, Schmitt, and Meyer, 2011) the assumption is made for these indicators that if they effectively influence the perceived innovativeness they can be expected to affect the financial performance of the video game. This leads to the following hypothesis:

H2: A curvilinear relationship exists between a sequel‘s innovativeness and its financial performance.

Finally, this study will investigate the combined effects of innovativeness and quality signals. Studies on brand extension have shown that the success of a brand extension depends heavily on the strength of a brand (e.g. Bapat and Panwar, 2009). Thus, a sequel (the brand extension) that is perceived as being too innovative may sell better or worse depending on the strength of the quality signals of the series (the brand). In other words consumers may be persuaded to still buy a video game that is considered too innovative when it has strong quality signals (Beverland, Napoli, and Farrelly, 2010) or be further dissuaded when it is lacking quality signals. This expected result is illustrated by figure 2, in which the curvilinear relationship between financial performance and innovativeness is displayed in blue. The purple line is the expected relationship between the two when combined with strong quality signals; the negative effects of too much innovativeness are minimized or even made positive. Similarly, in case of weak quality signals, the green line illustrates that the relationship is expected to worsen. This leads to the third and final hypothesis:

H3: The negative effect of too much innovativeness is lessened by strong quality signals or exacerbated by weak quality signals.

(11)

11 2.4 Control variables

To provide a stronger test for the impact of the identified variables this study incorporates several control variables, which will be discussed in the next paragraphs.

Cumulative average selling price. Cumulative average selling price refers simply to the average price a video game was sold for during the entire sales period, which is important to control for as price can be an important deciding factor for consumers. It is particularly important when using cumulative units as a dependent variable, as extreme price promotions, where a video game is sold for a fraction of its regular price, may heavily skew the results.

PEGI Rating. The MPAA rating is controlled for in many studies on the film industry because the R (Restricted) Rating means the film will not be played in many theaters. In the video game industry there are several such ratings associations, the largest one of which is the

PEGI rating system. Their classifications are legally enforceable in the many countries,

meaning some video games, those rated 18 particularly, are likely to face lower sales because younger customers are unable to buy them.

Genre. Controlling for genre is very important in this type of research and is also done in the film industry (e.g. Gemser, Leenders, and Wijnberg, 2008). Apperley (2006) argues that it is even more important in the video game industry because the differences between the genres are much larger than those in film. Apperley (2006) argues for a very broad but small set of video game genres, while others argue for a much larger set of genres (e.g. Wolf, 2001).

Competition. As consumers only have a limited amount of money to spend on video games, competition in any given period is also important to control for. Competition intensity is measured by several different variables in the film industry. Ho, Dhar, and Weinberg (2009) measure ―the total production budget of all movies released in the same week and one week prior to the release of [a] movie‖ (p. 171), Sharda and Delen (2006) link the competition intensity to release timing, and Ainslie, Drèze, and Zufryden (2005) determine competition by the number of films in the cinema that have the same genre or rating.

(12)

12 fewer developers willing to develop video games for the platform. Presumably, this coincides with or is preceded by a decrease in demand. Similarly, platform price will influence the number of people who will consider buying a console to play a particular game, making this important to control for as well.

FIGURE 3: Number of video game sequels released by year on the PlayStation 2

December release. Some periods such as Christmas are shown, in the film industry, to attract a bigger audience which leads to higher box office performance. In the video game industry this is also likely to be the case; the Christmas period is likely to show increased sales due to video games becoming increasingly popular gifts. Binken and Stremersch (2009) show that there is a ―bump in hardware unit sales in December due to the holiday effect‖ (p. 97), which suggests an increase in software sales as well.

Platform exclusivity. Unique to the video game industry is exclusivity; it is a way to create a competitive advantage for the owner of a particular platform by exclusively developing a game for that platform. The owner of a platform, particularly when it is young, has to offer incentives to game developers because the number of potential customers is lower (Eisenmann, Parker, and Van Alstyne, 2009).

Predecessor controls. As mentioned previously, similar to previous studies on sequel and re-release performance (e.g. Hennig-Thurau et al., 2007), this study will control for the success of the predecessor—which will also help minimize endogeneity. Similar to Basuroy and Chatterjee (2008) this study will control for the revenue of the parent video game, an indicator of the loyal customer base a video game series has, and time in between the focal release and

its predecessor, which should have a negative relationship with financial performance

(13)

13

3. Methodology

3.1 Sample statistics

This study draws upon a database of all video games released in the United Kingdom between January 1st 2000 and August 20th 2011, spread over 8 platforms. Of these 8 platforms, the PlayStation 2 was selected because its entire lifetime is encapsulated in these 11 years and it has the largest number of released video games of all the platforms in the database. The database contains 2226 PlayStation 2 games, 917 of which (41%) were determined to be sequels by three independent raters. The agreement between these raters was moderately high (Cohen‘s Kappa: 0.57), which is acceptable (Landis and Koch, 1977). 519 of these 917 video games had a predecessor that was also released on the PlayStation 2, a criterion which was added to control for platform effects where a customer has to buy a new console in order to play a sequel. A further 50 games were excluded because they were released within half a year of their predecessor, something which may indicate video games that are not true sequels given the time generally required to develop a new video game. The final sample thus consists of 469 video games. Data available for these sequels includes sales data, critical evaluation, user evaluation, age rating, genre, competition level, and release date among others. Data not included in the database, such as whether games have celebrity endorsement or are media adaptations, was gathered from a variety of sources.

3.2 Variables and data collection

Dependent variables. This study chooses two dependent variables: cumulative units and cumulative value (or revenue). Cumulative units will be used as the primary dependent variable while cumulative value will be used to improve robustness. Both values have been log-transformed, similar to studies on financial performance of films (e.g. Basuroy and Chatterjee, 2008), in order to prevent outliers from skewing the results.

Control variables. As mentioned previously, cumulative average selling price is simply the average selling price in pounds. The cumulative average selling price is used instead of the average selling price to control for price promotion periods during which disproportionately many games were sold for a low price.

PEGI rating will be represented by five dummy variables for the age ratings: 3, 7, 12,

(14)

14

Genre will be represented by fourteen dummy variables, one for each genre. The database

employs a large list of genres supplied by platform owners but because many of these genres are only represented by a very small number of games these genres were merged into fourteen metagenres based on their overlap. These fourteen genres are: action, fighting, graphic adventure, music, platform, puzzle, racing, real game, RPG, shooting, sim, skateboarding, sports, and war. Again, one of these genres will be randomly excluded.

To measure competition this study follows Ho, Dhar, and Weinberg (2009) by measuring sales in the period before and after the release of the focal product. Video games are in stores longer than films are usually in cinemas, and accordingly the timeframe used in this study is longer; competition will be measured by the sales of video games released from 4 weeks before the focal release until 4 weeks after release.

Platform age is measured by the number of days between the day of the release of the

platform and the day of the release of the focal video game. Platform price is measured by the price in pounds of the PlayStation 2 at the time of the sale of the focal video game.

This study will use the same method as Binken and Stremersch (2009) to control for

seasonality; a dummy will signify a December release (1) or not (0). Platform exclusivity is

also represented by a dummy variable signifying exclusivity (1) or not (0).

Finally, to minimize endogeneity the revenue of the parent video game in pounds, the

time in between the focal release and its predecessor in days, and the price difference between the focal video game and its parent in pounds are controlled for.

Independent variables: quality signals. The critics’ evaluation is drawn from Metacritic, a website with a very large video game database, which creates a Metascore that attempts to ―capture the essence of critical opinion‖ (Metacritic, 2012) and has been used as an indicator of critical evaluation in various other studies (e.g. King, 2007; Simonton, 2009b). It basically provides an average grade that critics have given to a particular piece of entertainment. This Metascore will be used to define valence while the volume is the number of critic grades used to come up with the Metascore. Both will be included in the regression analysis.

Metacritic will also be used for amateur evaluation, because it also offers the possibility for users to rate video games. Again, both volume and valence will be retrieved.

(15)

15 categories: video games can either be nominated for or win a Best Game Award, or they can be nominated for or win a Best PlayStation 2 Game Award.

TABLE 2: Overview of the selected awards Award

British Academy of Video Game Awards Spike Video Game Awards

Academy of Interactive Arts & Sciences‘ Interactive Achievement Awards Game Developers Choice Awards

IGN Reader‘s Choice Golden Joystick Awards

Star power is measured with a dummy variable, which was created by manually going

through all the video games in the sample and checking for an endorsement by a superstar. This superstar could be either a celebrity whose name is in the video game title (e.g. Tony Hawk‘s Pro Skater) or a celebrity who endorses the video game on the box cover (e.g. Rooney, FIFA 12). This variable was retrieved by three independent raters, and the interrater reliability for these three raters was substantial with a Cohen‘s kappa of 0.73 (Landis and Koch, 1977).

Media adaptation is also measured with a dummy variable, which indicates a sequel

based on other media. This study defines media adaptations as those video games that are directly based on products released on other media, not all video games based on a familiar concept, because they are expected to benefit from the advertising budget and word-of-mouth of the other media. This variable was retrieved by the same three independent raters, this time with moderate agreement (Cohen‘s kappa: 0.57).

Sequel number will indicate which iteration of a series the focal video game is. For

most games this is relatively straightforward, but included in the sample are also expansion packs. Expansion packs are smaller expansions to a video game which are generally sold for a fraction of the price of a full video game and the original game is often required to run the expansion pack. Expansion packs are numbered by linear interpolation, placing them in between the video game to which they are an expansion pack and the real sequel to that game. For example: Dynasty Warriors 4 is the fourth game in the Dynasty Warriors series and is thus marked with a ‗4‘; Dynasty Warriors 4: Extreme Legends is the first of two expansion packs and therefore gets a ‗4.33‘; Dynasty Warriors 5 is the actual sequel to Dynasty

Warriors 4 and thus gets a ‗5‘. If Dynasty Warriors 4 had only one sequel, it would be ‗4.5‘.

(16)

16 Escape), similar to (e.g. Grand Theft Auto: Vice City following Grand Theft Auto 3), or completely different from (e.g. Amplitude following Frequency), the previous video game in the series. The second indicator, genre departure, will be represented by a dummy variable indicating change (1) or no change (0) in comparison to a video game‘s predecessor. The third and fourth indicators are also represented by dummy variables, signifying departure (1) or no departure (0) in developer or publisher. The fifth and final innovativeness indicator concerns

innovation awards, which is represented by two variables: one count variable for winning an

innovation award and one for being nominated for an innovation award.

Procedure for measuring interaction. To find which factors measure innovativeness a factor analysis was performed, which which will add to the simplicity of the model as well. A principal component analysis (PCA) was conducted on all innovativeness variables with oblimin rotation, because this allows the components to correlate. This yielded 3 components with an eigenvalue of more than 1. However, because publisher departure and innovation awards (consumers probably do not perceive winning many innovation awards as a bad thing) were considered less important the component that included the other three variables was chosen to represent innovativeness. The variables which loaded highly on this component were entered into another PCA, which yielded an innovativeness index (table 3).

TABLE 3: PCA innovativeness TABLE 4: PCA quality

The innovativeness variable will be squared to create (innovativeness)², both of which will be used to find whether a curvilinear relationship exists between innovativeness and financial performance, testing hypothesis 2. More specifically, a curvilinear effect exists when the correlation between innovativeness and financial performance is positive and the correlation between innovativeness² and financial performance is negative (Aiken and West, 1991).

A quality variable was also created through factor analysis, and this PCA yielded 4 components with an eigenvalue over 1. However, the one with the highest eigenvalue, on which the critical and user evaluations loaded highly, was selected to represent quality. Entering these variables into another PCA (table 4) yielded a quality index. This index will be used to test hypothesis 1.

(17)

17

(innovativeness)² quality will then be created through multiplying innovativeness² by quality,

and this variable will determine whether quality signals influence the relationship between innovativeness and financial performance, testing hypothesis 3.

3.3 Analysis

Using the variables described above, this study tests six different models. Model 1 includes only the control variables, the second model attempts to find which quality signals are important, and the third model adds product innovativeness to the analysis. The fourth, fifth and sixth models attempt to find the interaction effect between innovativeness and the quality signals by adding the quality, innovativeness, and interaction variables.

4. Results

Table 3 provides an overview of the descriptive statistics of the variables in the sample. Genre and age rating are not included in this overview because they are represented by multiple dummy variables. Only those variables which initially correlated significantly with cumulative units at the p < 0.05 level are included in the table, and those same variables are used for the regression analysis. The results for cumulative revenue are not shown because they were practically identical to cumulative units. A complete correlation table for cumulative units can be found in appendix A.

TABLE 3: Descriptive Statistics

_N _{Minimum Maximum} _Mean _{Std. Deviation}

Cumulative units 469 3 2352406 100229,14 200852,546

Cumulative ASP 469 6,3200 39,7700 22,835096 7,3878822

Competition 469 1 91 47,19 21,282

Platform age (days) 469 325 3991 1838,54 767,768

Platform price 389 159,1213 422,6455 215,178938 39,9422369

December release 469 0 1 ,03 ,158

Platform exclusivity 469 0 1 ,36 ,481

Predecessor units sold 469 122 2352406 137615,17 219786,168

Price difference 469 -25,9400 12,6600 -3,529851 5,9618879 Time in between 469 183 2951 560,28 347,273 Critic valence 469 0 95 49,13 36,060 Critic volume 469 0 80 18,33 17,290 User valence 469 0 10 5,12 3,891 User volume 469 0 789 22,98 53,887 Star power 469 0 1 ,15 ,357 Media adaptation 469 0 1 ,13 ,341 Sequel number 468 1,33 22,00 5,1416 3,97722

Best game awards won 469 0 2 ,01 ,122

Titling departure 469 0 2 ,52 ,537

Genre departure 469 0 1 ,05 ,216

(18)

18 Table 4 shows the results of all 6 models for cumulative units. For model 1, only the control variables were entered. These control variables together explain a large share of the variance in financial performance (R² = .807). Cumulative average selling price and predecessor unit sales are both significant positive contributors at the p < .001 level while platform age, platform price, December release, platform exclusivity, and time in between are all significant negative contributors (p < .05). Predecessor units sold (β = .688), platform age (β = -.254), and time in between (β = -.163) contribute most to this model.

Model 2 added quality signals and significantly improved the explained variance (∆F = 4.562, p < .001). Media adaptation contributed positively (p < .001), and so did star power, sequel number, and user volume (p < .05). Media adaptation (β = .081), volume of user scores (β = .077), and sequel number (β = .062) contribute most to the improvement.

TABLE 4: Linear regression models 1-6 for cumulative units

Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 General controls

Genre dummies Included Included Included Included Included Included Age rating dummies Included Included Included Included Included Included Cumulative ASP (£) .136 (.030)*** .089 (.031)** .086 (.032)** .105 (.032)** .104 (.032)** .109 (.031)*** Competition .041 (.025) .033 (.025) .036 (.025) .039 (.025) .040 (.025) .044 (.025) Platform age -.241 (.037)*** -.279 (.039)*** -.278 (.040)*** -.220 (.037)*** -.216 (.037)*** -.208 (.037)*** Platform price (£) -.219 (.047)*** -.217 (.048)*** -.213 (.048)*** -.196 (.047)*** -.198 (.047)*** -.193 (.047)*** December release -.060 (.026)** -.066 (.025)** -.064 (.025)** -.054 (.026)** -.052 (.026)** -.048 (.026) Platform exclusivity -.056 (.021)** -.050 (0.22)** -.050 (.022)** -.048 (.021)** -.048 (.021)** -.044 (.021)* Predecessor controls

Predecessor units sold .619 (.024)*** .591 (.024)*** .592 (.024)*** .620 (.024)*** .621 (.024)*** .625 (.024)*** Price difference (£) -.086 (.022)*** -.074 (.021)** -.076 (.022)*** -.077 (.022)** -.079 (.022)*** -.084 (.022)*** Time in between -.154 (.022)*** -.140 (.022)*** -.140 (.023)*** -.154 (.022)*** -.153 (.022)*** -.161 (.022)***

Quality signals

Star power .041 (.020)* .041 (.020)** Media adaptation .072 (.021)*** .072 (.021)** Best game award -.035 (.023) -.041 (.025)* Sequel number .056 (.024)* .057 (.024)** Critic valence -.021 (.036) -.021 (.036) Critic volume .046 (.037) .045 (.038) User valence .024 (.032) .024 (.032) User volume .062 (.028)* .063 (.028)** Innovativeness Titling departure -.016 (.023) Genre departure .014 (.020) Innovation award win .016 (.013)

Interaction variables Quality signals .067 (.023)** .067 (.023)** .102 (.026)*** Innovativeness -.014 (.022) -.030 (.026) -.030 (.026) (innovativeness)² .016 (.014) .024 (.014) (inno)² * quality -.043 (.016)** R² 81% 82% 82% 81% 81% 81% F-change 71.104*** 4.562*** .587 4.578** 1.289 7.487** Observations 469 469 469 469 469 469

(19)

19 Entering the innovativeness variables for model 3 did not yield significant results nor improve the model. For model 4 the individual quality and innovativeness variables were removed and the variables derived from the factor analysis added, which improved the explained variance significantly (∆F = 4.578, p < .01). Only the quality index significantly contributed in this model (p < .001).

In model 5 (innovativeness)² is added, which does not significantly correlate with financial performance, nor does it improve the model. (innovativeness)² quality, added in model 6, is a significant negative contributor at the p < .05 level and improves the model significantly (∆F = 7.487, p < .01).

For figure 4 the sample was split by the mean of the quality variable, resulting in two curves: one for low quality signals and one for high quality signals. This figure illustrates that video games with too much innovativeness, which also have strong quality signals, perform worse financially. For video games with low quality signals the slope is also negative, but insignificant.

FIGURE 4: Three-way interaction between quality, innovativeness, and cumulative units

5. Discussion and conclusions

5.1 Discussion

This paper investigated the role of product innovativeness and quality signals in a sequel context. The goal was to analyze why certain sequels perform better than others and to develop a model to help video game developers create successful sequels. These goals were accomplished: the final model explains a very large percentage of the variance (R² = 81%) by analyzing a large number of quality signals as well as product innovativeness. In doing so, the

0 2 4 6 8 10 12 14 16

Low Innovativeness High Innovativeness

(20)

20 study contributes to our knowledge of video game success, and in particular sequel success. The most important results will be discussed in the next paragraphs.

Quality signals. Media adaptation and sequel number both perform in line with previous research (Basuroy and Chatterjee, 2008; Joshi and Mao, 2012), contributing positively to the explained variance. These results confirm the idea that developing sequels to successful series is a good strategy, particularly if they are developed simultaneously with related media. This study also shows that video games featuring stars perform better. This study only focused on celebrities lending their name or appearance to a video game, supplementing rather than confirming similar results by Binken and Stremersch (2009) who studied video game characters. Number of user scores is the only evaluation variable to correlate positively with financial performance, but this is likely because the more copies of a video game are sold the more users have the possibility of rating that video game. That neither the critics‘ nor the users‘ valence correlates with financial performance matches the conclusion by Zhu and Zhang (2010) that evaluation is not in and of itself a good predictor. The composite score for quality also correlates positively with cumulative units, confirming hypothesis 1: strong quality signals lead to improved financial performance.

Innovativeness. Neither innovation awards nor genre departures correlated with financial performance in any way. Title departure initially correlated negatively with financial performance, but did not yield significant results once the control variables were entered into the regression. These results suggest that innovation is relatively unimportant when developing video game sequels—customers seem to not want change. It seems that once a developer has developed new intellectual property (IP), there is no need to innovate within this IP to remain successful. Of course, this may not be the case for video games in general, lending credence to Tschang‘s (2007) conclusion that video game developers need to balance rationalization (focusing on existing IP) and creativity (creating new IP).

(21)

21 innovativeness in this way because there is no loyal customer base to be disappointed. Another reason for this result may be the fact that the measures of innovativeness used in this study are relatively strong indicators; a game can innovate to a large extent, in for example story or graphics, without a genre or title departure. Those smaller innovations, which were not measured by the variables used in this study, may still be welcomed by consumers of high quality series.

Control variables. Several of the results of the control variables are worth noting as well. A December release negatively impacts financial performance, which directly contradicts the conventional wisdom of the importance of releasing your product around Christmas and results in previous studies (e.g. Binken and Stremersch, 2009). It may be that the high level of competition in December negates the seasonality effect. Platform exclusivity also has a negative relationship with financial performance, which is surprising in a study that includes only one platform. It suggests that making a sequel platform exclusive will cause that sequel to perform worse even on the platform it is exclusive to. Finally, by far the largest contributor to the explained variance of the model is predecessor sales, reinforcing the argument that developing sequels to successful games is a solid strategy.

5.2 Managerial implications

The video game industry‘s yearly revenue of $67 billion (Gaudiosi, 2012) coupled with the large number of sequels that are developed each year make this study incredibly relevant to video game developers. The lack of research on which factors make video games successful and the fact that many video game developers still have trouble deducing what makes video games successful (Tschang, 2007) only strengthen this relevance. This study helps developers understand why certain sequels perform better than others, make predictions about sequel success, and improve their understanding of when and how sequels should be developed and marketed.

(22)

22 at all. Finally, video game developers also use the model devised in this study to fine-tune sequels to already successful video games.

5.3 Limitations and future research

A limitation to studying only sequels is that the results cannot easily be generalized for video games in general. However, this study does show that some success factors for sequels are different from those for video games in general, underscoring the importance of controlling for sequels when researching video game performance. Another limitation is that the effects of some variables were measured through relatively simple measures. For example, future research could be more fine-grained by expanding sales data to include first week sales, media adaptation to include popularity of the media that was adapted, and award data to include the type of award.

Another limitation revolves around the way innovation was measured in this study, using departures from the predecessor. Future research should focus more on the newness of certain innovations to the market, rather than the newness compared to a predecessor. Furthermore, the innovativeness variable could have been measured more objectively by measuring the newness of certain game elements such as the engine, characters, or storyline or by having industry experts rate a video game‘s innovativeness. Finally, more longitudinal studies are required to discover the importance of innovativeness in the long run. It may be that a video game sequel that is deemed too innovative rejuvenates the series such that its successor profits from the innovations.

Finally, the database employed for this study lacked certain variables that could have improved the model further, such as marketing and production budget. It seems likely that these variables would be strong predictors of financial performance, as they have proven to be in the film industry (e.g. Chang and Ki, 2005; King, 2007).

6. References

Aiken, L.S. and West, S.G., 1991. Multiple regression: Testing and Interpreting Interactions. 1st ed. London: Sage.

Ainslie, A., Drèze, X., and Zufryden, F., 2005. Modeling Movie Life Cycles and Market Share.

Marketing Science, 24(3), 508-517.

Apperley, T.H., 2006. Genre and game studies: Towards a critical approach to video game genres.

Simulation and Gaming, 37(1), 6-23.

(23)

23

Basuroy, S., Chatterjee, S., and Ravid, S.A., 2003. How Critical Are Critical Reviews? The Box Office Effects of Film Critics, Star Power, and Budgets. Journal of Marketing, 67(4), 103-117. Basuroy, S., Desai, K.K., and Talukdar, D., 2006. An Empirical Investigation of Signaling in the

Motion Picture Industry. Journal of Marketing Research, 43, 287-295.

Basuroy, S. and Chatterjee, S., 2008. Fast and frequent: Investigating box office revenues of motion picture sequels. Journal Of Business Research, 61(7), 798-803.

Beverland, M.B., Napoli, J., and Farrelly, F., 2010. Can All Brands Innovate in the Same Way? A Typology of Brand Position and Innovation Effort. Journal of Product Innovation

Management, 27(1), 33-48.

Binken, J.L.G. and Stremersch, S., 2009. The Effect of Superstar Software on Hardware Sales in System Markets. Journal of Marketing, 73, 88-104.

Chang, B. and Ki, E., 2005. Devising a Practical Model for Predicting Theatrical Movie Success: Focusing on the Experience Good Property. Journal of Media Economics, 18(4), 247-269. Chintagunta, P.K., Gopinath, S., and Venkataraman, S., 2010. The Effects of Online User Reviews on

Movie Box Office Performance: Accounting for Sequential Rollout and Aggregation across Local Markets. Marketing Science, 29(5), 944-957.

Clement, M., Proppe, D., and Rott, A., 2007. Do Critics Make Bestsellers? Opinion Leaders and the Success of Books. Journal of Media Economics, 20(2), 70-105.

Cook, D.L. and Coupey, E., 1998. Consumer Behavior and Unresolved Regulatory Issues in Electronic Marketing. Journal of Business Research, 41(3), 231-38.

Damanpour, F., Walker, R.M., and Avellaneda, C.N., 2009. Combinative Effects of Innovation Types and Organizational Performance: A Longitudinal Study of Service Organizations. Journal of

Management Studies, 46(4), 650-675.

Dellarocas, C., Zhang, X., and 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. Dhar, T., Sun, G., and Weinberg, C.B., 2012. The long-term box office performance of sequel movies.

Marketing Letters, 23, 13-29.

Eisenmann, T.R., Parker, G., and Van Alstyne, M.W., 2009. Opening Platforms: How, When and Why? Platforms, Markets and Innovation, 131-162.

Elberse, A., 2007. The Power of Stars: Do Star Actors Drive the Success of Movies? Journal of

Marketing, 71, 102-120.

Field, A., 2009. Discovering statistics using SPSS. 3rd ed. London: Sage.

Garcia, R. and Calantone, R., 2002. A critical look at technological innovation typology and

innovativeness terminology: a literature review. Journal of Product Innovation Management, 19(2), 110-132.

(24)

24

Gemser, G., van Oostrum, M., and Leenders, M.., 2007. The Impact of Film Reviews on the Box Office Performance of Art House versus Mainstream Motion Pictures. Journal of Cultural

Economics, 31, 43-63.

Gemser, G., Leenders, M.., and Wijnberg, N.M., 2008. Why Some Awards Are More Effective Signals of Quality Than Others: A Study of Movie Awards. Journal of Management, 34(1), 25-54. Hennig-Thurau, T., Houston, M.B., and Walsh, G., 2006. The Differing Roles of Success Drivers

Across Sequential Channels: An Application to the Motion Picture Industry. Journal of the

Academy of Marketing Science, 34(4), 559-575.

Hennig-Thurau, T., Henning, V., Sattler, H., Eggers, F., and Houston, M.B., 2007. The Last Picture Show? Timing and Order of Movie Distribution Channels. Journal of Marketing, 71, 63-83. Ho, J.Y.C., Dhar, T., and Weinberg, C.B., 2009. Playoff Payoff: Superbowl advertising for movies.

International Journal of Research in Marketing, 26, 168-179.

Jones, C., 2001. Coevolution of entrepreneurial careers, institutional rules and competitive dynamics in American film, 1895–1920. Organization Studies, 22, 911-944.

Joshi, A. and Mao, H., 2012. Adapting to succeed? Leveraging the brand equity of best sellers to succeed at the box office. Journal of the Academy of Marketing Science, 40, 558-571. Kirmani, A. and Rao, A.R., 2000. No Pain, No Gain: A Critical Review of the Literature on Signaling

Unobservable Product Quality. Journal of Marketing, 64(2), 66-79.

King, T., 2007. Does film criticism affect box office earnings? Evidence from movies released in the U.S. in 2003. Journal of Cultural Economics, 31, 171-186.

Kunz W., Schmitt B., Meyer, A., 2011. How does perceived firm innovativeness affect the consumer? Journal Of Business Research, 64(8), 816-822.

Landis, J.R. and Koch, G.G., 1977. An application of hierarchical kappa-type statistics in the assessment of majority agreement among multiple observers. Biometrics, 33(2), 363-374. Litman, B.R. and Kohl, L.S., 1989. Predicting financial success of motion pictures: The 80s

experience. Journal of Media Economics, 2, 35-50.

Mezias, J. M. and Mezias, S. J., 2000. Resource partitioning, the founding of specialist firms, and innovation: The American film industry, 1912–1929. Organization Science, 11, 306-322. Metacritic, 2012. How We Create the Metascore Magic. Metacritic, [online]. Available at: <

http://www.metacritic.com/about-metascores> [Accessed 2 January 2013].

Moon, S., Bergey, P.K., and Iacobucci, D., 2010. Dynamic Effects Among Movie Ratings, Movie Revenues, and Viewer Satisfaction. Journal of Marketing, 74(1), 108-121.

Movieinsider, 2012. Sequel Movies 2010. Movieinsider, [online]. Available at: <http://www. movieinsider.com/movies/sequel/2010/> [Accessed 21 January 2012].

Neely, A. and Hii, J., 1998. Innovation and business performance: A literature review. Unpublished manuscript.

(25)

25

Nelson, P., 1974. Advertising as Information. Journal of Political Economy, 82(4), 729-754.

Perretti, F. and Negro, G., 2007. Mixing genres and matching people: a study in innovation and team composition in Hollywood. Journal of Organizational Behavior, 28, 563-586.

Ravid, S.A., 1999. Information, Blockbusters and Stars: A Study of the Film Industry. Journal of

Business, 72(4), 463-492.

Rogers, E. M. and Shoemaker, F., 1971. Communication of Innovations; A Cross-Cultural Approach. Romijn, H. and Albaladejo, M., 2002. Determinants of innovation capability in small electronics and

software firms in southeast England. Research Policy, 31, 1053-1067.

Sharda, R. and Delen, D., 2006. Predicting box office success of motion pictures with neural networks.

Expert Systems with Applications, 30, 243-254.

Simonton, D.K., 2009. Controversial and Volatile Flicks: Concurrent Dissension and Temporal Instability in Film Critic Assessments. Creativity Research Journal, 21(4), 311-318. Sood, S. and Drèze, X., 2006. Brand Extensions of Experiential Goods: Movie Sequel Evaluations.

Journal of Consumer Research, 33(3), 352-360.

Tschang, F., 2007. Balancing the Tensions Between Rationalization and Creativity in the Video Games Industry. Organization Science, 18(6), 989-1005.

Venkatraman, N. and Lee, C., 2004. Preferential linkage and network evolution: A conceptual model and empirical test in the U.S. video game sector. Academy of Management Journal, 47(6), 876-892.

VGReleases, 2012. January 2010 Release Dates: Video Game. [online], VGReleases. Available at: <http://vgreleases.com/ReleaseDates/All-1-2010.aspx> [Accessed 21 January 2012]. Wooldridge, J.M., 2009. Introductory Econometrics: A Modern Approach. 4th ed. Mason: South-

Western Cengage Learning.

Zeithaml, V.A., 1988. Consumer Perceptions of Price, Quality, and Value: A Means-End Model and Synthesis of Evidence. Journal of Marketing, 52, 2-22.

(26)

Predicting Video Game Success Through Quality Signals and Product Innovativeness