استفاده از فعالیت به اشتراک گذاری P2P برای بهبود تصمیم گیری کسب و کار : اثبات مفهوم برای برآورد محصول چرخه عمر
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|1471||2005||7 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Electronic Commerce Research and Applications, Volume 4, Issue 1, Spring 2005, Pages 14–20
Estimating the life-cycle or duration of a product can be an important input into a firm’s decision-making related to production and marketing. In the music industry, online Peer-to-Peer (P2P) networks have attracted millions of potential music consumers and have had substantial impact on the music business. In this paper, we investigate the possible use of P2P information in estimating product “shelf-life,” in particular the duration of a music album on the Billboard 100 chart. We identify and track the music albums that appear on the Top 100 of the Billboard Charts, spanning a period of six months. We show that P2P sharing activity can be used to help predict the subsequent market performance of a music album.
Estimating the life-cycle or duration of a product can be an important input into a firm’s decision-making related to production and marketing. This is of special significance for entertainment firms that deal with multiple products with short life-cycles, such as music and movies. In standard analysis of such issues, researchers have had little difficulty structuring the objective functions – businesses are profit maximizing and can manage their interaction with consumers. And then along came free peer-to-peer (P2P) networks that provided distinctly “non-commercial” means of exchange. In fact, recent studies have observed high level of free-riding among users in these networks  and . A recent article in Fortune highlighted the non-business approach of KaZaA and its originators, Janus Friis and Niklas Zennstrom, that has so greatly impacted entertainment companies. The KaZaA developers apparently proceeded without a business plan, taking a “just go and do it approach” with subsequent failed attempts to work out licensing deals with major entertainment companies. KaZaA, despite being a virtual non-business, has become the “top search term on Yahoo” [ Fortune, 2004] and, together with its P2P counterparts such as WinMx and Grokster, continues to vex the entertainment industry. But we suggest that businesses might actually leverage P2P networks as information sources to make better production and marketing decisions. In the material that follows, we present our initial investigations into the gathering and potential industry use of P2P activity information. P2P application software such as KaZaA and WinMX are extremely popular and commonplace among potential music purchasers. As explained later, we developed a custom software application that directly observes and takes “snapshots” of P2P music file-sharing activity. Using this data, we are able to directly address whether measures of such sharing activity might be useful in business decision making. In the work presented here, we focus on estimating product “shelf-life,” in particular the duration of a music album that appears on the Billboard 100 chart. Could entertainment firms utilize information on P2P file sharing to better determine the success (measured in chart duration) of music albums? Before discussing data collection or data analysis, we think it prudent to consider characteristics of the P2P “data venue.” In the P2P music sharing setting, incentives are rather different than the normal business/consumer market. Without a profit-maximizing (or similar) objective, what incentive is there for support or customer service tools such as enhanced search tools? In fact, as one would expect, the search options allowed on P2P networks are quite limited. In such environments, searches for digital goods such as music are primarily directed searches. That is, a consumer gathers information from offline as well as online sources about music items (album and artist names) and then searches for a particular item to sample. Certainly, some searches may be broader (e.g., based on music genres), but information overload is likely to quickly occur since there is no support for search aids such as “ranked list of relevant results” in these networks since there are no incentive to provide such services. This deters a “browsing based” search behavior on a particular genre or artist, for example, since a random presentation of search results quickly increase the search cost for a consumer. Hence, the directed nature of the search implies that consumers are more likely to sample those music items for which “information availability” is high (from offline and online sources). This indicates a generally higher sampling and sharing activity for well-known albums that also likely have higher sales based on consumers’ information awareness. With these two factors (most searches are likely to be directed searches and heavily influenced by information availability), we begin our discussion of data collection followed by presentation of initial P2P data analysis. It has been observed that while radio airplay measures the advertising effort for given music albums , airplay does not closely predict consumer interest in such albums  and . In fact, anecdotal evidence points to misjudgment of consumer interest and related promotional activities of new artists and albums by record companies (see, for example  and . Given the increasing interest in research on products such as music , , , , ,  and , we posit that sharing information on P2P networks may be used to predict consumer interest and subsequent sales for music albums. Our proof of concept approach to investigating the possible value of P2P information in business decision begins with consideration of the following three research questions related to albums that appear on the Billboard charts: (i) Can sharing information on online networks during initial weeks on chart be a valid predictor for survival duration on the charts? (ii) Does such early sharing information offer predictive ability beyond such factors as debut rank on the charts? (iii) Finally, is there any relationship between the predictive ability of early sharing information and album visibility or “album information availability”? We also provide a cursory “proof of concept” relating to whether continuing P2P sharing might be helpful to business in determining the life-cycle for albums that have already survived for some period of time on the Billboard chart. The issue is posed in the following question: iv) For albums continuing on the chart, can subsequent sharing activity levels help predict how much longer an album will remain on the charts?
نتیجه گیری انگلیسی
We began by suggesting that the ability to predict product life-cycle has significant value to the firm. The earlier a firm can estimate product life-cycle, the better since the firm can either avoid continuing costs for short-cycle products or make strategic decisions in support of longer cycle products. The difficulty, however, is that there often is scant early information on which firms can build reasonable estimates. While the digital good industry does not typically have the same lead time issues as traditional manufactured products, early knowledge on the likelihood of product success and product life-cycle remain important elements in the profit equation. Here, we have provided an initial analysis of new information – P2P sharing activity – which we suggest may be helpful to estimation of product life-cycle of digital goods such as music. We demonstrated the importance of this new information relative to the previously existing information – rankings on the Billboard Top 100 chart. It is important to stress the preliminary nature of our analysis and results. We did develop a detailed data set on music sharing activity for music appearing on the Billboard Top 100 charts. We did find indications of the relevance of P2P sharing activity. We did provide tentative positive responses to our research questions. But we would argue that our analysis should be viewed as a “proof of concept” demonstration. The results are suggestive, not definitive. The current research is an aid in helping us shape a rigorous and comprehensive research study. This would include a detailed investigation through rigorous estimation of sophisticated models (e.g., hazard models) of the impact of sharing activity on the lifecycle of albums. This also leads to the development of advanced forecasting tools to predict lifecycle from actual consumer activity. These stochastic predictive models can be designed and focused for other digital goods such as digitized movies, books and video games. Our results also provide tantalizing insights on how the design of P2P networks can be enhanced, to the benefit of both consumers and music companies. The fact that sharing activity is a good predictor only for albums with high information availability speaks directly to the poor design of current P2P systems. As such, these applications function well only as repositories of music; repositories that are useful to consumers only if they know what they are looking for. Finally, we note that new and unknown artists and record companies can actively enter the P2P arena and develop enhanced “search and find” functionalities (e.g. ranked lists based on sharing characteristics) that help consumers explore and experience new music products – possibly through emerging fee-based P2P services. Design of such mechanisms remain an active and fruitful area of research in P2P systems.