1
HOW TO SELL FREE STUFF?
by
ARD BUIKEMA
University of Groningen
Faculty of Economics and Business
Msc Business Administration – Business Development
First supervisor: Dr. J.D. van der Bij
Second supervisor: Dr. C. Reezigt
July 2012
Meerhuizenplein 28L
1078TD Amsterdam
0619058905
a.buikema@student.rug.nl
student number: 1656740
2
Abstract
While the effects of piracy have gotten much attention in research already, preventions
are very much underexposed. This study focuses on finding a business model that will
provide consumers with enough value to persuade them to pay for their music rather than
downloading it illegally. By incorporating both business model and adoption literature, the
reasons to adopt an innovation and the necessary and preferred components of business
models are tested. This leads to a full-fledged redesign of the business model in the music
industry and a number of both theoretical and practical suggestions.
Keywords: music, piracy, willingness to pay, business model, adoption
1. Introduction
In the past decade, the landscape of the music market has changed dramatically. The uprising of digital music, which started in the late nineties, has caused numerous changes in the way people listen to, acquire and perceive music. The most important changes in the music industry are the following: the definite changes in consumer demand and a dramatic shift in piracy. Although no general census on the impact of piracy exists yet (e.g. Oberholzer and Strumph, 2007; Karaganis, 2011 vs Blackburn, 2004; Rob and Waldfogel, 2006), this subject has at least garnered substantial attention in academic research.
The shift in consumer demand, however, is still a largely unknown concept. There is not a single study which shows what the customer actually wants and how the music industry should approach them. The goal of this research, therefore, is to find the business model that will convince consumers to pay for their music, rather than download it for free. In order to do so, three questions will be answered. (1) What business models are possible and proposed in literature on the music industry? (2) How do customers respond to the dimensions of the various models? (3) What are the reasons for individuals to adopt innovations?
The recent changes in the music industry set challenges for both executives in the industry and academics. The boxed text on the next page shows a successful example of an experiment with a different business model. Also in academic literature on the effects of piracy, authors have suggested that the music industry should change its business model. Maltz and Chiappetta (2002, p.77) for instance, state: “content creators cannot effectively protect absolute rights to digital
content through legal or technological means – and they likely undermine themselves by trying”.
Sinha and Mandel (2008) agree, as their study shows that positive incentives for consumers will provide better results than negative incentives. So, many people agree that the business model should be changed. What it should be changed into, however, remains a largely unanswered question, and evidence for functional improvement is largely anecdotal.
3
To fill up the literature gap, this article is the first to provide an overview on the business models proposed, and use Chesbrough’s (2002) business model dimensions to classify and test the models with consumers. The adoption literature is widely considered a mature area of research, and vital in explaining the diffusion of innovations. Still, this research is the first to incorporate it to determine the likelihood of success for business models in the music industry.
The outline of this article is as follows. First, the theoretical framework will provide an overview on this subject and more specifically, present the business models proposed so far in literature. Also, adoption theory is presented in that part. The empirical research discusses the questionnaire that was devised, and that part is followed by an analysis and discussion of the results. Lastly a redesign is presented, which is a proposed business model, based on the results of the questionnaire. This model will be compared to the options which are available in the market right now.
2. Theoretical framework
This chapter will start with a short description of business models in general. Furthermore, it will present an overview of new business models that have recently been proposed in literature. Finally, adoption theory will be introduced and discussed.
2.1 Business models
“The term business model is very often used, but not often clearly defined.” (Chesbrough, 2007, p.
12). It seems that researchers find it hard to define what a business model is. Other authors (e.g. Osterwalder, Pigneur, Tucci, 2005; Morris, Schindehutte, Allen, 2003) also notice frequent use of the term, without it being properly understood. Despite their observations, none of these authors provide a clear definition in their article either.
Besides a definition, the dimensions by which a business model can be defined also vary greatly between authors. In an effort to show the lack of consensus, both Osterwalder et al. (2005) and Morris et al. (2003) created an overview in which the problems become clear. Morris et al. (2003) found no less than 24 different items, with fifteen of them receiving multiple mentions. Also the number of components varies greatly, between three and eight are mentioned as the building blocks of a business model. Osterwalder et al. (2005) showed a table with nine business model components on the y-axis and fourteen authors who presented a combination of these combinations on the
x-Radiohead’s experiment
In 2007, Radiohead released its album ‘In Rainbows’ without the use of their record company. Instead, they opted to distribute their album through their website, and having customers pay via a pay-what-you-want principle. While 1/3 of the customers did not pay, 2/3 decided to pay 4 pounds on average, prompting far higher revenues for Radiohead than a normal release would have raised. In addition, in a subsequent CD release, they also managed to sell 5-6 times more CDs than their earlier ones. They also sold 100.000 collector’s box sets. In sum, Radiohead’s experiment was a huge success, and proved that changing their business model was very effective. Also, it showed that people are actually willing to pay for stuff that is also available for free.
4
axis: none of the columns contained the same combination and most rows contained different definitions of the same component as well.
This poses several problems. First, there are problems for academic research, e.g. in determining which conditions make a certain model appropriate, in the fact there is no existence of generic model types, and the fact that there are no standards to evaluate certain models (Morris et al., 2003). Second, it poses a problem to the managers of firms, as they are not provided with tools to capture, analyze, communicate, and, most importantly, change their own business models (Osterwalder et al., 2005).
Although the business model concept is described as an underdeveloped area of research and there is a lack of consensus on its definitions and dimensions, there is agreement on its importance to both companies and the academic world. Osterwalder et al. (2005) describe the potential of the concept in understanding, analyzing and managing the business logic of the firm. They even propose it will foster innovation and readiness for the future. Morris et al. (2003) underline these same functions and add that the business model can be a central construct in entrepreneurship research.
Because of the aforementioned importance of being able to define the business model of a firm, this article will incorporate two components of a business model, which are mentioned by almost all authors: the value proposition and the revenue generation mechanism (Osterwalder, 2005). These components are highly important, as they are directly visible to the consumers (contrary to some other components, which are only important or visible to the firm). The value proposition is the value created for users by the offering based on technology (Chesbrough, 2002, p. 533). The revenue generation mechanism is the method that is used to have the customers pay for that proposition. As this is an adoption study, focused on consumer’s perceptions, these are the components that will serve as the dimensions to describe and test the business models in this article.
2.2 Business model innovation
Chesbrough (2007) stresses the importance of business models, and business model innovation in particular, to companies. He states that technology in itself has no economic value, as it is the business model that actually makes it commercially interesting. Chesbrough (2007) challenges both companies and academics to take on business model experimentation, as they will otherwise fail to capture value from their innovations. The problems to business model innovation can be that the right business model is unknown to the company, or that the right business model is readily recognized, but its development is prevented (Chesbrough, 2007). In the former case, information that is not consistent with the dominant logic of the firm, will be filtered out. This causes innovative ideas to disappear. The latter case occurs, when innovations are obstructed by people within the company. Amit and Zott (2001) explain how, for instance, a field sales president will take exception to online sales experiments. First, because (part of) his resources will be allocated to another division. Perhaps even more so, because he might eventually lose his job. In addition, Chesbrough (2007) found that innovation could be obstructed, because companies will allocate capital to the most profitable uses (which will be the established rather than the emerging technology in the beginning), depriving the emerging technology of much needed resources.
Both the scenarios that can prevent business model innovation from happening, could be relevant in the music industry. But actually, the innovation itself does not even belong to the industry. The
5
important innovation is the internet (or distributing music through it), while the music industry is given the chance to appropriate value from it. It is clear that the old business model is no longer working, but it is not all clear what the new model will be (Chesbrough, 2007). This is why business model experimentation is so important. Luckily, some experimentation has already taken place. The next part will present the business models that were proposed in literature and that were already put in practice.
2.3 New business models
Although it was usually not the main goal of past research to come up with a new business model, some authors have suggested solutions to the call for (a) new business model(s). Note that a broad perspective on business models is taken in this overview. This reflects the broad perspective on business models as can be found in literature on piracy. However, all five models will be described using the two dimensions mentioned in the paragraph above (some models focus more strongly on the value proposition and some on the revenue generation method - which does not pose a problem, as these dimensions will be tested separately).
1. Music and advertisement
Some people argue recorded music should be free. There are two approaches, advertising through music and music as advertisement.
Music as advertisement
Before widespread internet piracy was there, artists toured to promote their albums. Casadesus-Masanell and Hervas-Drane (2010) however, suggest the music industry should make its money by live performances and merchandising. In order to create as much advertisement for the tours and merchandise, artists should give away their music online. This will help their popularity and diffusion through society. Also, Casadesus-Masanell and Hervas-Drane (2010) suggest online file sharing could be used as a tool to monitor demand. Labels could program live performances according to the popularity of their artists at a certain place and time in the file sharing business. This means that when artists give away their music, they can track the diffusion of their music in order to determine their popularity in a certain place at a certain time. The industry should be careful with this strategy, however. Blackburn (2004) found that the availability of free music will hurt sales of ex-ante popular artists. According to his case, ex-ante unknown artists could benefit from this business model, as he states they will find their sales rising through illegal file sharing. This means they (unknown artists) would also benefit from giving away their music for free.
Advertising through music
Another approach taken is that of giving away music in order to gain the biggest audience and consequently making money by incorporating advertisements in the distribution. Quite a number of subscription services (such as Spotify) also have a free advertised model. They let customers listen to music for free, but interrupt music every now and then to show advertisements. So this is much like the radio model, only now the customer can choose which music he wants to listen to. It is also suggested artists could make their money by using product placements in their performances, for instance by having Nike pay them their income if they wear Nike clothes in public
(Casadesus-6
Masanell and Hervas-Drane, 2010). Blackburn’s (2004) findings are applicable here as well so the industry should be careful not to hurt their own sales using this approach (e.g. people could stop buying CD’s entirely if they could obtain the music online for free). Subscription services such as Spotify tackle this problem by offering the so-called freemium model. This means they will offer customers a free option, which is advertised. However, they will also offer a premium option which contains several advantages (such as high quality music and not having to listen to commercials) and customers are charged a fee for this option. More on this will be discussed later on.
2. Windowing
Bhatia, Gay, and Ross Honey (2003) proposed a business model based on the movie industry’s windowing strategy. The movie industry has had great success deploying this strategy, which revolves around gradually opening new windows with offers, which are increasingly cheap and accessible. Figure 1 shows
their proposed windowing strategy for the music industry, which is developed to capture as much value from the market as possible. It starts off with touring and exposure on the radio, continues to selling CDs and subsequently tries to capture more value on the digital market. Finally, left-over CDs are combined and sold at very low prices. Karaganis (2011) showed
that this strategy is not really working for the movie industry anymore, so this is probably not a really good option for the music industry either. The movie industry is narrowing the windows to counter the effects of piracy, so it would be strange if the music industry would start widening windows. 3. Added value proposition
The majority will pay for a product, when they see greater value in it than in a similar free product, and the majority of users will be frustrated by technical issues in using new technologies (Maltz and Chiappetta, 2002). The industry could help them by offering a content site with excellent customer experience and a customer friendly design. By ensuring the customers protection and offering extra services, all user-friendly, the industry could outduel the illegal content sites.
While most of the articles on the music industry focus on whether or not piracy has an effect on the industry, Sinha and Mandel (2008) contributed by trying to find the way to eradicate piracy. They started by asking respondents input for a list of features that would increase their willingness to pay (WTP – corresponding to their intention to pirate). Then, in a subsequent study, the features were presented to respondents and they were asked how much more likely the feature would make the
7
respondent use the legal website. Consumers turned out to have a higher WTP for each of the features proposed. Lastly, Sinha and Mandel (2008) presented a third study, in which they presented the consumers with three scenarios, ranging from low functionality on a legal content website, to high functionality. It appeared that the intention to pirate decreased, as functionality on the legal site increased. So, by creating value around music offerings, the industry could persuade the customer to pay rather than pirate.
4. Celestial jukebox/ Flat fee
Many people, both in- and outside the industry, have been wanting a so-called celestial jukebox (Bhatia, Gay and Ross Honey, 2003). The idea is you have a subscription so you can access any song at any time, any place and on any device you want. Multiple examples of this model are already in place. Since May 2010, Swedish Spotify has also set foot in the Netherlands (see boxed text on the left). Although it is not exactly the celestial jukebox as mentioned before, the features are impressive and they are expanding. Also, Spotify has made a deal with KPN (an internet service provider (ISP) and mobile network operator). This might be the beginning of a model that has been proposed in relation to the celestial jukebox. Seidenberg (2010) is one of the people who discusses this idea. Rather than fighting piracy, the industry would make it legal to share all music online, and then charge every customer a fee, through their ISPs. The fees can be collected and distributed to the artists/ labels, according to the amount of playbacks.
5. iTunes
The iTunes Music Store (hereafter referred to as iTunes) is often mentioned as the example that shows people are willing to pay for downloading music (Sinha and Mandel, 2008). The model of iTunes is quite straightforward, they offer as many songs/ artists as possible, and you can download this music and pay for it per song or per album.
Contrary to the huge success in other countries (iTunes has by far the biggest market share in the digital music market), iTunes has never caught on in the Netherlands. Only 4 percent of Dutch
Spotify
Spotify is a software program which allows you to stream music directly to your computer. This means you can look up almost every song, artist or album you like, and then listen to it directly, without having to download it. The quality of the music is high and there is almost no time lag in the playback. Besides offering you virtually all the music you want, Spotify also allows you to create playlists and share those with friends. There are 3 variants of the Spotify service:
Spotify open – lets you stream music for 20 hours each month and you don’t have to pay for it. Every once in a while you will have to listen to/look at a commercial in between songs.
Spotify unlimited – allows you to listen to music as long as you want and you don’t have to listen to commercials. You are charged 5 euros each month for this service. Spotify premium – has all the benefits of unlimited. In addition, you also have offline playback (which means you do not need an internet connection to play the songs in your playlists – you could, for example, create a playlist at home and then bring your laptop on a holiday and listen to the music. Also, you can use Spotify premium on your mobile phone, so you can always listen to the music you want. The fee for this service is 10 euros each month.
8
youngsters buy music online, that’s the lowest number of a research in 33 countries (www.nu.nl, 28-10-2010).
The reason iTunes has not caught on in the Netherlands, probably has a lot to do with relative advantage and price perception (simply put: the improvement of this service over the current service and how much value customers put on this – these constructs are defined later in this chapter). Contrary to the United States, downloading music, which was illegally uploaded, is not illegal. This is why almost every teenager in the Netherlands grows up downloading music for free. According to www.nu.nl (28-10-10), 37 percent of Dutch youngsters is downloading music without paying for it. With no moral objections to downloading and a perceived price of zero, chances of people paying 1 euro to download a single song are very slim.
2.4 Review of business models
Table 1 displays a summary of the business models discussed. As can be seen, this study will test four of the models with consumers.
Table 1: Review of business models
Business model Motivation for in- or exclusion Include?
Music as advertisement
While this could help create some extra revenue, it will cannibalize sales for many artists when music is freely distributed.
No Advertising
through music
This could provide some extra value, when combined with another model (e.g. the freemium model)
Yes Re-segmenting
the market (windowing)
The movie industry is increasingly narrowing the windows to avoid piracy (Karaganis, 2011), so this will probably not help the music industry.
No
Added value proposition
Difficult, as exclusive material rarely remains exclusive. Could be interesting though, and can be combined with any other strategy.
Yes Celestial
Jukebox/Flat Fee
According to research and executives in the industry, this could be the model of the future.
Yes iTunes Most profitable digital music service in the world, but has not caught
on in the Netherlands, only 4% of Dutch youngsters buy music online, lowest number in research of 33 countries
Yes
The biggest difference between advertising trough music, the celestial jukebox, and iTunes is their method of payment (free/time-investment, flat fee, or pay per song, respectively). The biggest comparison is they all imply some form of added value, as many customers will have to be convinced to choose the legal over the illegal alternative. The added value proposition provides that leverage over the illegal alternatives. So the first two of Chesbrough’s (2002) six functions are clearly apparent: the value proposition and the revenue generation system. The different models in this article will be tested using these dimensions. The first part differentiates several options to add value; the second part differentiates the methods of payment. After testing, the results will be used to propose a redesigned business model, which is also compared to the models described above.
2.5 Adoption
In order to bridge a gap in literature, Moore and Benbasat (1991) described an instrument which could be used to study both adoption and diffusion of innovations, by measuring potential adopters’
9
perceptions of innovations. They test the adoption of new technology by individuals, by using Roger’s characteristics (relative advantage, compatibility, observability, complexity and trialability) of an innovation (which affect its rate of diffusion).
Moore and Benbasat’s (1991) instrument is one of the included models in the article of Venkatesh and Morris (2003). Despite the maturity of the user acceptance research area, prior to Venkatesh and Morris (2003), no serious attempt at comparing similarities and differences of all eight prominent models (as identified in the article) had been done. Up to their article, therefore, researchers had to make a choice between one of the models or combine constructs from the different models. Venkatesh and Morris (2003) filled this need for a unified view by describing the development of a unified theory of individual acceptance and use of technology (UTAUT). This model combines eight important user acceptance models to create UTAUT, and is shown to outperform each of the eight individual models.
Despite the substantial contribution of the UTAUT model, this article will use the instrument of Moore and Benbasat (1991). First, because there is a complete and tested instrument available, which is not yet the case for UTAUT. Second, because UTAUT is developed using tests that measured respondents reactions to new information systems in the period they already started using it. Moore and Benbasat (1991) targeted their research at the perceptions of individuals towards certain innovations, which can also include people who have yet to use an innovation. This last argument is vital, as this article is trying to determine what the business model should be, so it cannot be tested by any users right now.
The instrument comprises of the following eight constructs:
Relative Advantage: the degree to which using the innovation is perceived as being better than its
precursor.
Compatibility: the degree to which using the innovation is perceived as being consistent with the
existing values and past experiences of potential adopters.
Ease of use: the degree to which a person believes that using the innovation would be free of effort
(adapted from Davis, 1989 – this definition was not explicitly stated in the article by Moore and Benbasat).
Result demonstrability: the degree to which the results of using the innovation are observable to
others.
Trialability: the degree to which using the innovation may be experimented with before adoption. Image: the degree to which using the innovation is perceived to enhance one’s image or status in
one’s social system.
Voluntariness: the degree to which using the innovation is perceived as being voluntary, or of free
will.
10
To develop this list, Moore and Benbasat (1991) mostly relied on Rogers’ five characteristics of innovation (relative advantage, complexity, compatibility, observability and trialability). Image was added to Rogers’ five variables, as multiple researchers (e.g. Tornatzky and Klein, 1982) argued it was different enough to be included as a separate factor. Voluntariness was added, since it was considered more than just a binary variable. Furthermore, Rogers’ observability was split in result demonstrability and visibility, as the two appeared to have a higher validity and reliability when separated. Lastly, complexity was substituted with Davis’ (1989) ease of use. The constructs were renamed the perceived characteristics of innovations (PCIs).
Tornatzky and Klein (1982) also identified five other characteristics: cost, communicability, divisibility, profitability, and social approval. Moore and Benbasat (1991) did not include these characteristics. Communicability, divisibility and social approval are closely related to other constructs, namely observability, trialibility and image (in that order). Contrary to Moore and Benbasat (1991), profitability and cost are relevant for this research, as this article deals with individuals rather than organizations. Profitability will not be included here, as the model is not known yet. Cost however, will be included, but in line with other literature, it will be named price. The concept of price is also mentioned as part of relative advantage (Van Slyke, Ilie, Lou and Stafford, 2007). However, the use of relative advantage by Moore and Benbasat (1991) is much closer to Davis’ (1989) construct usefulness, which is a totally different concept than price. Fair pricing is mentioned as a very important driver for adoption by important researches in the music industry (Sinha and Mandel, 2008; Maltz and Chiappetta, 2002; Seidenberg, 2010; Plowman and Goode, 2009).
3. Empirical research
3.1 Set up of questionnaire
To test the four models that were identified in the previous section, an instrument was developed (see appendix A). This starts with a general description of a music service which allows the customer to browse a large catalog of virtually all music, which he has at his disposal instantly. The four models are tested as follows. The added value proposition is mostly included in the basic service. This service aims to provide a better experience to the customers than he would otherwise enjoy when downloading content illegally. Next to that, a couple of features, which possibly add more value, are tested. The iTunes model, music as advertisement and the celestial jukebox model are treated as the same service, and the differences that will be tested are the payment methods.
The instrument used in this research, consists of four parts. The first part is devised to test the features that should be included in the business model. The features tested are all derived from literature. Table 2 shows the ten features that were found by Sinha and Mandel (2008) to increase WTP. Six of these items are included (with 3 of them combined), so 4 features are presented in the instrument.
The recommendation engine suggests alternative songs bases on users preferences and historical data. Creating playlists is a feature that gives people the chance to list their favorite songs and share these with their peers. High quality music speaks for itself and finally, the exclusive content is already described in table 2.
11
Table 2: Proposed features
Feature Include as feature?
(1) An extensive music catalog No, basic service
(2) Free samples of music No, only interesting in case of payment per unit
(3) Easy navigation on the site No, basic service (4) Recommendation engine Yes
(5) Easy-to-burn compilation CDs Yes, reformulated to creating/ sharing playlists, burning CDs is irrelevant
(6) Extremely fast downloading No, irrelevant and outdated (7) Rare recordings, exclusive
artists, new releases
Yes, combined with (9) & (10): exclusive content
(8) Extremely high digital quality Yes
(9) Live concerts on website Yes, combined with (7) & (10): exclusive content
(10) Availability of additional downloads (ringtones, video)
Yes, combined with (7) & (9): exclusive content
Testing the preferred payment method is done in the second part. As shown in the previous chapters, there are three distinct possibilities:
Flat fee payment (monthly subscription) Pay per unit (Album or a single song) ‘Pay’ by watching/ listening to commercials
The second part also tries to capture the concept of fair pricing, by asking the respondents to indicate their perceived fair price for each of the three payment methods (regardless of their choice for a certain method).
For the first and second part, items that had to be rated were measured on a 7-point scale, which is consistent with Sinha and Mandel (2008). Respondents were asked to indicate how much more likely the availability of a particular feature or method would make them use the service (with 1 corresponding to ‘not more likely’ ranging up to 7, which indicates ‘much more likely’).
The instrument that Moore and Benbasat (1991) devised is used in part 3 of the questionnaire. Of the eight available constructs, six will be measured. Voluntariness is left out, as this adoption is strictly individual and free of will. Visibility is also left out, since whatever business model is chosen, visibility will be the same, as we’re dealing with a pure software service (the results are what can be experienced, and these are measured with result demonstrability). For the six constructs that remain, the short list items are chosen from the instrument, which leaves 13 items in this part. These items will all be rated on a 7-point scale, with 1 indicating the respondent strongly disagrees, and 7 indicating strong agreement. These scales were also used in the instrument of Moore and Benbasat (1991).
The fourth part of the questionnaire is designed to obtain information about the demographics of the sample.
12
3.2 Sample
The sample was obtained in early 2011 and consists of 64 Dutch citizens (30% female), 98% of them aged between 20 and 30 (average age was 25.4). At the time, they were either studying (43%) or working (57%). Therefore, in the sample, students and young people are much overrepresented, which means the study lacks external validity (Van Aken, Berends and Van der Bij, 2007). This means findings cannot be generalized to the whole society, but at best to the population of students and young professionals.
4. Results of Analysis
The analysis is done in four parts. First, the four features introduced in table 2 are compared. Second, the same execution is done for payment methods. Next, groups are compared on features and payment methods. In the fourth part, the influence of Rogers’ variables on WTP is tested.
4.1 Features
Paired T-tests were used to determine if the averages scores for a certain feature were significantly higher than for another feature. To be able to do this, scores for a feature have to be approximately normally distributed. Q-Q Plots were used to determine that all features were indeed approximately normally distributed. With a mean score of 5.5625 (on the seven point scale), high quality music (HQ) proved to be significantly more valued than recommendation engine (4.8281), creating playlists (4.6563) and exclusive content (4.2344). The latter showed significantly less value than the other features (see table 3).
Table 3: Paired Samples T-Test for comparison of mean differences between features
Paired Differences t df Sig.
(2-tailed) Mean Std. Deviati on Std. Error Mean 95% Confidence Interval of the Difference Lower Upper HQ – Recommendation engine ,73438 1,70193 ,21274 ,30924 1,15951 3,452 63 ,001 HQ music – Playlist ,82258 1,77897 ,22593 ,37081 1,27435 3,641 61 ,001 HQ music – content 1,32813 2,17529 ,27191 ,78475 1,87150 4,884 63 ,000 Recommendation engine – Playlist ,09677 1,76217 ,22380 -,35073 ,54428 ,432 61 ,667 Recommendation engine – Exclusive content ,59375 2,25880 ,28235 ,02952 1,15798 2,103 63 ,039 Playlist – Exclusive content ,48387 1,89686 ,24090 ,00216 ,96558 2,009 61 ,049 4.2 Payment methods
Q-Q plots showed that all payment methods were approximately normally distributed. Next, the paired t-test showed that all three payment methods scored significantly different from each other
13
(see table 4), with flat fee scoring highest (average score of 4,9219 on the 7-point scale), payment per unit lowest (3.2188), which leaves ‘paying by commercials’ in the middle (4.1719).
Table 4: Paired Samples T-Test for comparison of mean differences between payment methods
Paired Differences t df Sig.
(2-tailed) Mean Std. Deviation Std. Error Mean 95% Confidence Interval of the Difference Lower Upper
Flat rate – Per unit Flat rate – Commercials Per unit – Commercials
1,70313 2,38833 ,29854 1,10654 2,29971 5,705 63 ,000
,75000 2,91139 ,36392 ,02276 1,47724 2,061 63 ,043
-,95313 2,62122 ,32765 -1,60789 -,29836 -2,909 63 ,005
4.3 Fair pricing
Respondents indicated a fair price of 6-10 euros for the flat rate service on average (they were given five options: 0-5; 6-10; 11-15; 16-20; 21 euros and up. There were also five options for payment per unit: 0-.25; .25-.5; .5-.75; .75-1; 1 euro and up. This yielded an average option score of 1,4063, so it seems the fair price should be around .25 euros. Finally, respondents were asked how many songs they should be allowed before another commercial is presented, and the score was 4.0313, indicating 1 commercial after 4 songs.
4.4 Groups compared
For this part, respondents are divided in groups based on their answers in the fourth part of the questionnaire. Six groups were created initially: a division based on gender, education, occupation (student vs. employed), usage of music (light vs. heavy), age, and usage of platforms (CD and DVD). Analysis of Variances (ANOVA) tests showed almost no significant differences between the groups. The following conclusions could be drawn though:
WO students score significantly higher on the creating playlists feature.
Students score significantly higher on fair number of commercials (i.e. their tolerance for the amount of commercials is lower).
Heavy music users score significantly higher on payment method flat rate.
DVD users score significantly higher on exclusive content than non DVD users. They also show a lower tolerance for commercials.
Since literature (Plowman and Goode, 2009; Sinha and Mandel, 2008) showed important differences between light and heavy music users, extra groups were created within the answers of the fourth part of the questionnaire to differentiate between these groups in more detail. Table 5 on the next page shows the results.
14
4.5 Perceived characteristics of innovation
The questionnaire consists of thirteen items to measure the PCIs (three for ease of use and two for relative advantage, compatibility, image, result demonstrability and trialability). A factor analysis was conducted in order to determine if items loaded too high (>.50) on more than one factor or did not load strong enough (<.40) on any factor. This was done using a principal component analysis with VARIMAX rotation. When all thirteen items were entered in the analysis, only five instead of six factors appeared and several items loaded too highly on multiple factors. Thus, all items were deleted and re-entered one by one, in order to determine which items need to be kept to obtain six factors, with as much items remaining as possible. Table 6 shows the items that remained after the analysis, so that no construct had to be deleted and all factor loadings can be considered excellent (>.71; Moore and Benbasat, 1991). Descriptive statistics for these items are shown in table 7.
Table 6: Principal component analysis with VARIMAX rotation
Items Components 1 2 3 4 5 6 Image 1 ,941 Image 2 ,924 Trialability 1 ,937 Trialability 2 ,918 Compatibility 1 ,936 Compatibility 2 ,777 Ease of use 1 ,892 Relative advantage 2 ,965 Result demonstrability 2 ,922
Next, the Cronbach’s Alpha was determined for image (.910), trialability (.902), and compatibility (.818). As all alpha’s are well above the targeted minimum reliability scores (.7), the items can be
Table 5: Music usage and spending compared (see questionnaire for exact definitions of groups)
Group Group division Results
# Music hours/week (1) 0-14, (2) 15 and up
Group (2) scores significantly lower on payment method per unit
# Concerts/year (1) less than 2, (2) 2 and up
Group (2) scores significantly higher on payment method per unit.
# Downloads/month (1) less than 10, (2) 10 and up
no significant differences # CDs/month (1) No CDs,
(2) any number
Group (2) scores significantly higher on exclusive content feature
# DVDs/month (1) No DVDs, (2) any number no significant differences € Current spending/ month (1) less than 10, (2) 10 and up
Group (2) scores significantly higher on high quality, recommendation engine and flat rate payment method
€ Future spending/ month
(1) less than 10, (2) 10 and up
Group (2) scores significantly higher on fair flat rate price
15
added together and treated as one construct. Since the other three constructs only had one item left, these items were treated as if they were the construct.
Table 7: Descriptive statistics for PCIs
Construct Mean Standard Deviation
Cronbach's Alpha
Compatibility 4,492 1,501 0,818
Ease of use 5,641 1,036 N/A (only one item)
Image 2,938 1,456 0,910
Relative advantage 4,656 1,360 N/A (only one item) Result demonstrability 5,125 1,256 N/A (only one item)
Trialability 6,211 1,038 0,902
Lastly, all six constructs were entered into a regression analysis in order to determine their effects on the dependent variable, future spending (as a measure of WTP). The only construct to have a significant impact on future spending is compatibility, with a positive coefficient of 2.830 (significant at the 5% level). See table 8 for regression results.
Table 8: Results of regression analysis
Constructs Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta (Constant) -,077 12,399 -,006 ,995 Compatibility 2,830 1,344 ,343 2,105 ,040 Image -1,096 1,274 -,129 -,860 ,393 Trialability -,708 1,732 -,059 -,409 ,684 Relative advantage -1,114 1,417 -,111 -,786 ,435 Ease of use ,943 1,978 ,079 ,477 ,636 Result demonstrability 1,429 1,724 ,125 ,829 ,411
a. Dependent Variable: Future spending
5. Results of Redesign
Following from the analysis, all features and payment methods can be ranked, and the importance of the PCIs can be determined. This also makes it possible to evaluate the merits of the various models. After discussing all these findings, these will be incorporated in the redesign of the business model. This redesign is presented to the music industry, as a solution to the call for a new business model.
5.1 Features
Regarding features, two important observations can be made. First, the fact that all features scored above 4 on the 7-point scale (indicating they would make it at least a bit more likely a customer would use the service), suggests that the added value proposition is indeed a valid model to have the customers choose a legal alternative. Second, has become quite clear which feature the customers
16
prefer. The ranking was clearly determined, so high quality is by far the most important feature the industry should focus on (also amongst most important features in the study of Sinha and Mandel, 2008). Contrary to Plowman and Goode (2009), no relation between light or heavy music users and their need for high quality was found. However, this might be due to homogeneity of the sample and/or the sample size. The only significant difference was found within the spending groups. Bigger spenders have a bigger need for high quality.
Recommendation engines and creating playlists are solid features that could be added, which is also consistent with Sinha and Mandel (2008). The exclusive content feature found less support than in Sinha and Mandel’s (2008) study, however, this feature might not be for everybody. People who buy DVDs and CDs value this feature significantly higher than others, so this might be offered as an extra to those.
5.2 Payment methods
As for the payment methods, the clear winner is the flat fee payment method. This method is even more important to heavy music users, which is quite logical, since they would benefit the most. Good news for the industry is that the current average price for this method (10 euros) is within the price range that respondents indicated to be fair (6-10 euros). People who indicated the higher future spending amounts even indicated a significantly higher flat rate price.
The method where customers ‘pay’ by listening to commercials could be used as an alternative to people who don’t want to spend any money. This could prove a valuable model next to the flat fee offer, as many people in Sinha and Mandel’s (2008) study indicated they would not pay for music at all. The industry might consider reducing the number of commercials a bit though, as respondents indicated a tolerance of one commercial after every four songs, which seems to be lower than what is currently the norm in the industry.
This study clearly shows why iTunes is not that much of a success in The Netherlands. The respondents indicated they are not very eager to use the service in this study if they have to pay per unit. Also the fair price they indicated (somewhere between 0 and .5 euros) is much lower than the current average price (which lies around 1 euro). The people that would likely spend the most money (the ones listening to music the most), even indicated significantly less willingness to use this payment method.
5.3 PCIs
Of the PCIs, compatibility showed to be the only significant one. Compatibility is often found to be one of the most important drivers for adoption (Tornatzky and Klein, 1982; Karahanna, Straub and Chervany, 1999). However, these studies (and many others; e.g. Venkatesh and Morris, 2003) also suggested relative advantage to be important, but that finding has not been confirmed by this study. Moore and Benbasat (1991) conclude however, that compatibility is still really close to relative advantage, so more research would be necessary to determine the empirical differences between the two. Also, the timing of the innovation is very important. Castano, Sujan, Kacker and Sujan (2008), showed that in the near future affective cost and switching cost are the most important uncertainties, whereas performance becomes more important in the distant future. Switching cost uncertainty is associated with learning how to use an innovation (Hoeffler, 2003). This seems very
17
much like a combination of compatibility and ease of use, so if respondents perceive this innovation important in the near future, they might value compatibility higher than (e.g.) relative advantage in that case.
Lastly there were only minor differences found between the created groups. Previous studies indicated that variables such as age, experience, gender and voluntariness were important moderators (Venkatesh and Morris, 2003). However, because of homogeneity and a small sample size these variables could not show considerable differences based on demographics. These same sample characteristics and the fact that the sample is not entirely random cause some internal validity problems (Van Aken et al, 2007). However, the 2.830 coefficient of compatibility is still representative of a huge impact on future spending. For every point added in compatibility, future spending will grow by 283% based on these findings.
5.4 Business models
As mentioned above, the failure of iTunes in the Netherlands could be explained by these results. People consider prices too high and are not willing to pay per unit. Besides that, iTunes does offer some features they might enjoy. The celestial jukebox however, will draw a substantial amount of customers. Not only does it contain the features the customers want, it is also payable by the preferred payment method, at a price which is perceived fair. Spotify, the famous example of a celestial jukebox, also incorporates the advertising through music model. This also seems a valuable model to quite some consumers, especially those who indicated they would not pay. Consistent with earlier findings (e.g. Sinha and Mandel, 2008), the added value proposition proves to be interesting to lots of consumers, judging by the fact that they indicated a higher willingness to adopt the service with certain features in place. As Spotify contains (to a great degree) the right dimensions of the desired business model, it will serve as the foundation the redesign will be built on.
5.5 The Redesign
Redesigning the Spotify business models will be done along the two dimensions of Chesbrough (2002): the value proposition and the revenue generation mechanism.
Dimension 1: The value proposition
Concerning the value proposition, Spotify has almost everything the customer wants. High quality music, options to create and share playlists, and a recommendation engine. Besides that, most of the platforms used by the customers are compatible with Spotify’s services. The problem here is that some options are only available to the customers who pay for the service. This means that the non-paying customers are not entitled to use the service on anything but a computer (which has to be online) and they will not be provided with the best
music quality. Chances are slim that they will endure listening to commercials if they are not provided with an advantage compared to downloading the music for free. To make sure consumers see this added value, four recommendations can be made for Spotify.
Figure 2: High quality for all users (kbps is an indication of quality)
18
1. High Quality music should be for everybody. Make sure everybody gets to enjoy high quality
music, rather than just premium customers (see figure 2). This was the number one feature respondents indicated to increase their willingness to pay for this service.
2. Provide access on all platforms. Concerning the platforms Spotify is available on, compatibility
could be an issue. At the time of the data collection, over half of the participants in the questionnaire made use of mp3 players to listen to music. An important feature which will be missed, because of that, is the option to download the songs. Although this feature is available in Spotify, it is not really well incorporated and it is relatively expensive. Spotify should make sure users get access to the mp3 versions of the songs. Follow-up research, however, should research whether Spotify should focus their attention on other mobile devices (see figure 3). Smartphones are often used as a replacement for music players, and Spotify offers users an app to access their account on their phone. As smartphone penetration has begun an enormous spurt (from 22% penetration on the Dutch phone market in Q1 2010 to 52% in Q4 2011, www.marketingfacts.nl, 2012), mp3 players might soon be out of scope. Even then, there is work to do. Currently only premium users have access to Spotify on their cell phone. This should be changed.
3. Deliver the promise of ‘All the music, all the time’. Even though Spotify has a vast
amount of available songs, there are still numerous albums and even complete artists which cannot be found in the catalogue, forcing
consumers to
download the music illegally or buy CDs at local music stores – which is very much inconvenient. Also, it is not uncommon to find that a certain album or song is available in the catalogue but ‘not in your region’. Therefore Spotify should aim to provide consumers with virtually all music available. Making sure everybody gets to use the music on whichever device they like, will also greatly increase compatibility of the service. Since this is their tagline (see figure 3), this is the least customers will expect.
4. Communicate advantages and remain a step ahead. There are some
features Spotify could exploit in the area of compatibility as they have an advantage over downloading music for free (see figure 4). First of all, since they have a streaming music service, it generally works faster for users than actually downloading songs. This is compatible with the growing desire of customers to listen to music instantly. Also, with Spotify, people can be sure
Figure 3: Deliver as promised
Figure 4: Keep expanding this list
19
that they will not download any viruses or low quality music. Both of these issues are quite common for people who download music. They could spend quite some time avoiding these issues themselves, but obviously it would be a good move of Spotify to lend them a hand. Lastly, people often find it interesting to download full discographies, and Spotify can produce these in an instant. Spotify should capitalize on these advantages and actively communicate them, in order to convince consumers to use their service. Towards the future, this should also be Spotify’s intention: to remain a step ahead of pirates. Not just incorporating their innovations, but being proactive in offering the consumer added value compared to illegal alternatives. Various alternatives come to mind, for instance, offering a full entertainment package with audio and video, creating a TV app, and so on. Dimension 2: The revenue generation system
In order to increase attractiveness of the service, three changes should be made to the existing revenue generation system.
1. Two methods of payment is enough. The two preferred methods of payment a flat fee
subscription and ‘paying’ by commercials, are in place. The option to buy individual songs or albums is not likely to catch on with consumers so Spotify should consider dropping this feature. There are two clearly identifiable market segments, light and heavy music users. Generally, light music users will favor the free proposition with commercials, whereas heavy users will choose one of the flat rate propositions.
2. Change the propositions that are offered. As described before, Spotify currently offers three
propositions (two flat rate propositions and one ‘free’ proposition, see figure 5 on the page below). It is probably smarter to stick to one flat rate proposition for the bulk of the users. The usage differentiation could remain, as the heavier music users are generally the ones who will favor the flat rate payment, whereas the ones who favor the version featuring commercials are lighter users. However, if the business model is sound, than it should also allow for an unlimited free proposition. If free users listen to more music, they will also listen to more commercials, which should lead to higher revenues. This last argument is vital, as it also applies to the changes proposed in the previous sections, making sure the free and flat rate propositions offer practically the same service.
An extra flat rate proposition can be created around the exclusive content feature. This proposition will cater to the needs of CD and music DVD buyers of nowadays, offering exclusive content such as live performances or unreleased material. This is not necessary to win the crowds, but it could attract some heavy music users, the audiophiles, and generate some more revenue.
3. Equate current prices with perceived fair prices. The price of 9,99 euro Spotify charges for the
service falls within the range consumers indicated to be fair (6-10 euros). Prices per unit are way too high, but as mentioned before, this feature should be dropped anyways. Consumers indicated that on average one commercial (lasting 15-30 seconds) after every 4 songs would be fair to them. This seems to be a bit less than the current number of commercials, so Spotify should consider some adjustments here. More research is needed to determine the price for the exclusive proposition. Redesigning the business model along the two dimensions as suggested, will provide a great offer to consumers. This could therefore have a big impact on their choice to pay for their music, even though it can be obtained for free.
20
6. Discussion
This section consists of three parts; theoretical implications, practical implications and finally, limitations of the research and proposals for future research.
6.1 Theoretical implications
By redesigning Spotify’s business model, this article found and answer to the research question. In order to satisfy the customers, research was done to test their preferences for certain dimensions of the business model. Also, the reasons for customers to adopt an innovation were uncovered. No adoption research had been undertaken in this area so far. Therefore, this article is a leap forwards in finding empirical evidence for directions the music industry should be heading. Rather than calling for new business models without backing up the need for this model, and without presenting a tested proposal of that model, this study is the first to present a model which is demanded by consumers. Furthermore, this study demonstrates that Moore and Benbasat’s (1991) set-up for measuring PCIs is also of great use outside of an information systems context. The items still proved to be solid measurements of the various constructs. An important finding here is that compatibility proved to be the most important predictor of future spending by far. This is interesting as often relative advantage is the strongest predictor. This underlines an area of adoption research that still needs more development. Lastly, this is a first attempt in practically describing business models along the value proposition and revenue generation mechanism components. These components proved useful in testing existing models with consumers. By measuring these components, this is a first step in devising instruments to start testing business models. This should help this area of research out of its status quo, and develop itself towards the mature state that is needed.
6.2 Practical implications
The most obvious contribution of this research is the redesign. Its design is based on consumer needs and, most importantly, compatibility. Compatibility showed to be a very important predictor of WTP. The value proposition and the revenue generation mechanism were also tested and this information can be very helpful to the industry when they design their new business models. Contrary to what the music industry is doing right now, holding on to ancient business models which are no longer
21
working, this model incorporates the benefits of illegal models and then adds value in order to persuade the customer to pay. It was also made clear why some models are less successful in The Netherlands (e.g. iTunes), so these should be adapted by the industry. It would be in line with another lesson learned: in order to be innovative business model wise, experimentation is vital. A couple of reasons for not being innovative were mentioned, and all could be applicable to the music industry. So a practical conclusion would be that the music industry should not try to keep innovations from happening. Instead, they should experiment themselves, and reward third parties who do so with a license agreement. This will benefit them eventually, as this will lead to the business model that makes people pay for free stuff.
6.3 Limitations and proposals for future research
As mentioned a couple of times before, the sample is a limitation of this article. Because of the small sample size, the homogeneity of the sample and the fact that it was not randomly selected causes both internal and external validity problems. In a subsequent study, the sample should be enlarged and be more heterogeneous. Analyzing the data presented quite some robust findings. However, if the sample would have been bigger and less homogenous, this number could have been increased. Findings could then be generalized to a wider public.
This article contains important adoption and business model literature, but it is not exhaustive. It focuses on Moore and Benbasat’s (1991) research model, whereas Venkatesh and Morris (2003) found their UTAUT model to explain adoption better than any of the individual models they combined. Further research could be undertaken to see if findings remain robust using a different part of adoption theory. Network effects were also a part of adoption theory that is left out. Although network effects are not as apparent as in other industries, they might certainly play a role. The playlists feature works best when you have many people to share them with. Van der Bij et al. (not published yet) showed that the relation between compatibility (the most important PCI in this article) and the intention to adopt, is moderated by network externalities. Also, Chou and Shy (1990) found that network effects may exist without network externalities. By increasing the number of users, more value adding services could be made available to the customer of a certain product. Facebook is a good example, as it now offers an integrated service with Spotify.
This study was specifically focused on the consumer and not so much on the industry itself. In a subsequent study, the business model could be developed further, by including barriers and opportunities of the industry. For instance, the choice was made to only include two components of a business model, whereas most categorizations count more components. Including more dimensions, such as the profit margin and the value chain for the organization, would further help determining the viability of the different models. In future research, an instrument could be developed to test a set of dimensions. Next, these dimensions could be tested to compare them in terms of viability. This article made an important step in empirically testing business models with consumers and presenting a redesign for the industry. The future research could build on this by taking all components of business models in scope and by providing a ranking of business models. This could be helpful, as business models can coexist and complement each other. And even if they fail to succeed in the marketplace, it will be beneficial for the greater good, as business model experimentation is the key to successful business model innovation.
22
References
Amit, R., Zott, C. (2001). “Value creation in e-business.” Strategic Management Journal, 22, 493-520. Bhatia, G. Krishan, Richard C. Gay, and W. Ross Honey (2003). “Windows into the Future: How Lessons from Hollywood Will Shape the Music Industry.” Journal of Interactive Marketing, 17 (2), 70– 80.
Bij, Van der J.D.., Keizer, J., Podoynitsyna, K., Song, M. & Wouters, M., (not published yet). “Moderating effect of network externalities on the relationship between perceived innovation characteristics and the intention to adopt.”
Blackburn, D. (2004). On-line Piracy and Recorded Music Sales. Harvard University, Department of
Economics.
Castano, R., Sujan, M., Kacker,M., & Sujan, H. (2008). “Managing Consumer Uncertainty in the Adoption of New Products: Temporal Distance and Mental Simulation,” Journal of Marketing
Research, 45 (June), 320–336.
Chesbrough, H. (2010). “Business Model Innovation: Opportunities and Barriers.” Long Range
Planning, 43, 354-363.
Chesbrough, H., ( 2007). “Business model innovation: it’s not just about technology anymore.”
Strategy & Leadership, 35 (6), 12-17.
Chesbrough, H.,& R. S. Rosenbloom (2002). "The Role of the Business Model in Capturing Value from Innovation: Evidence from Xerox Corporation's Technology Spin-Off Companies." Industrial and
Corporate Change 11(3), 529-555.
Chou, C. & O. Shy, 1990, “Network Effects without Network Externalities,” International Journal of
Industrial Organization, 8, 259–270.
Davis, F.D. (1989). “Perceived usefulness, perceived ease of use, and user acceptance of information technology.” MIS Quarterly, 13(3), 319-340.
Dranove, D., & Gandal, N. (2003). The DVD-vs.-DIVX Standard War: Empirical Evidence of Network Effects and Preannouncement Effects. Journal of Economics & Management Strategy, 12(3), 363-386. Hoeffler, Steve (2003), “Measuring Preferences for Really New Products,” Journal of Marketing
Research, 40 (November), 406–421.
Karaganis, J. (2011). “Media Piracy in Emerging Economies.” Social Science Research Council.
Karahanna, E., Straub, D.W., & Chervany, N.L. (1999). “Information technology adoption across time: A cross-sectional comparison of pre-adoption and post-adoption beliefs.” MIS Quarterly, 23(2), 183-213.
Keesan, J. (2008). “Let it be? The challenges of using old definitions for online music practices.”
23
Lange, M.A., (2001). “Digital Music Distribution Technologies Challenge Copyright Law.” Boston Bar
Journal. 45 (14), 1-8.
Maltz, E. & Chiappetta, V. (2002). “Maximizing Value in the Digital World.” MIT Sloan Management
Review, Spring 2002, 77-84.
Moore, G.C., & Benbasat, I. (1991). “Development of an Instrument to Measure the Perceptions of Adopting an Information Technology Innovation.” Information Systems Research, 2(3), 192-222. Morris, M., Schindehutte, M., Allen, J. (2005). “The entrepreneur’s business model: toward a unified perspective.” Journal of Business Research, 58, 726-735.
Oberholzer-Gee, F., & Strumph, K., (2007). “The Effect of File Sharing on Record Sales: An Empirical Analysis”. Journal of Political Economy, 115( 1), 1-42.
Osterwalder, A., Pigneur, Y., Tucci, C.L. (2005). “Clarifying business models: origins, present, and future of the concept.” Communications of the Association for Information Systems, 16, 1-25.
Plowman, S. & Goode, S. (2009). “Factors affecting the intention to download music: quality perceptions and downloading intensity.” Journal of Computer Information Systems, 84 (summer), 84-97.
Rob, R., & Waldfogel, J. (2006). “Piracy on the High C’s: Music Downloading, Sales Displacement, and Social Welfare in a Sample of College Students.” Journal of Law and Economics. 49 (April), 29–62. Seidenberg, S. (2010). “The record business blues.” ABA Journal, 96 (6), 4-8.
Sinha, R.K., & Mandel, N. (2008). “Preventing Digital Music Piracy: The Carrot or the Stick?” Journal of
Marketing, 72 (January 2008), 1-15.
Tornatzky, L.G., & Klein, K.J. (1982). “Innovation characteristics and innovation adoption-implementation: A meta-analysis of findings.” IEEE Transactions on Engineering Management, 29(1), 28-43.
Van Aken, J.E., Berends, H. & Van der Bij, J.D. (2007). “Problem Solving in Organizations”. Cambridge
University Press.
Van Slyke, C., Ilie, V., Lou, H. & Stafford, T. (2007). “Perceived critical mass and the adoption of a communication technology.” European Journal of Information Systems, 16(3), 270-283.
Venkatesh, V., Morris, M.G., Davis, G.B.& Davis, F.D., (2003). “User acceptance of information technology: toward a unified view.” MIS Quarterly 27(3), 425.
Websites
http://www.nu.nl/internet/2366401/nederlandse-jongeren-kopen-weinig-muziek-online (visited on 28-10-2010)
http://www.marketingfacts.nl/berichten/smartphonebezit-fors-gegroeid-meer-dan-de-helft-nederlanders-overstag (visited on 19-03-2012)
24
Appendix A: Questionnaire
Thank you very much for participating in this research. This questionnaire aims to find information regarding the preferred dimensions of a new service in the music industry. The first part of the questionnaire concerns the preferred features the service should contain; the second part is about the payment method; the third part is set up to find out some perceptions about the service; the fourth and final part contains general questions about the participants. The service will be described below:
Imagine a music service which allows you to browse a catalog of virtually all music available. You can listen to the music on the internet (using your computer or any other connected device, e.g. a cell phone or an iPod) so you have instant access to every song you want. Also you’re allowed to download the music to your computer so you can listen to it whilst being offline, or transfer it to another device such as your cell phone or an mp3 player. The service is very much user friendly, works fast and comes with a number of handy features (these will be described below and you will have the chance to rate the importance of them to you). You will be charged for this service, either by paying money or by spending time looking at (and/or listening to) commercials.
Part 1: Features
This part concerns the preferred features for the service described above. You don’t have to consider how you are charged for the service at this point. You will be asked to rate four features. Think of these features as independent options which may or may not be interesting to you.
Feature 1: Quality of music
This feature is about providing you with the music you desire in the highest quality available. Highest quality applies to both the playback and the recording of the file. This means you will not hear any bugs or distortions, and the best recording is used (e.g. in case of classical music, the best choir and orchestra available will be used).
Feature 2: Recommendation engine
This feature provides you with excellent opportunities to discover new music which you might like. It consists of two parts. Whenever you are listening to a song the service comprising this feature will recommend other music you might enjoy, as it is related to the song or artist you are listening to. Also, the service will analyze the music you have listened to in the past. Based on this analysis, the service will automatically create a suggestion on what music you might like to listen to as well.
Feature 3: Creating/sharing playlists
This feature is devised mainly so you can enjoy compilations of various songs by yourself and together with friends. The feature allows you to create a playlist (using any song you want). This playlist will be available to yourself at all times for playback. Also, you can share this playlist with your friends, and they will be able to listen to this playlist instantly as well. Naturally, you can thus also receive playlists which your friends have devised, which you can start listening to immediately. Think of this service as the modern equivalent of mixtapes or compilation cd’s, with the advantage
