Data, Results and Discussions


In order to measure the effect of the addition of the Internet medium, this chapter compares the coefficients of the Bass Diffusion Model on two consumer durables, one introduced prior to the commercialization of the Internet in the 1990s and one after. It is common practice among diffusion researchers to use the coefficients of the Bass Model to compare differences such as time, culture and word-of-mouth influence among countries or regions (Gatignon and Robertson, 1989; Helsen et al., 1993; Jain and Maesincee, 1995; Mahajan and Muller, 1994; Putsis et al., 1997; Sultan et al., 1990; Talukdar et al., 2002; Takada and Jain, 1991; Tellefsen and Takada, 1999).

Mahajan and Muller (1994) used the coefficients of p and q to represent the new-product diffusion patterns associated with the time differences before and after the unification of Europe. They compared the Video Cassette Recorder data before and after 1992 and found the coefficients of the Bass Model did not remain the same, and concluded that the unification in 1992 resulted in faster diffusion of new technologies for member countries. Similarly, the current study uses the differences in the coefficients of innovation p, the coefficient of imitation , q, and the time to peak, t*, as surrogate measures of the diffusion of new-products related to the time differences before and after the introduction of the Internet new medium [2] .

The first set of hypotheses (H1a and H1b) - the more the Internet increases business- directed b2c or c2c communications, the more consumers are likely to become innovators corresponds to a larger coefficient of innovation, p , expected (when compared to pre- Internet time) as depicted in the 'Diffusion Parameter Change' column in Table 13-3. The second hypothesis - the more the Internet increases consumer-directed communications, the larger the increase in its social influence in the diffusion process, and the more consumers are likely to become imitators - corresponds to a larger coefficient of imitation, q , expected (H2 and its diffusion parameter change in Table 13-3). The third hypothesis - the more the Internet increases communications, the larger the increase of imitators when compared to innovators - corresponds to the increasing ratio of the coefficient of imitation, q , over the coefficient of innovation, p (H3 and its diffusion parameter change in Table 13-3). Lastly, the fourth hypothesis - the more use of the Internet medium to eliminate time differences and geographic distances, and to amplify signals of product flaws from consumers to manufacturers, the faster the speed of diffusion - corresponds to the time to peak, t*, becomes shorter (H4a and H4b and their diffusion parameter change in Table 13-3).

Table 13-3: Hypotheses, diffusion parameters and results

Hypotheses

Diffusion Parameter Change

Results

H1a. The more the Internet medium increases business-directed b2c communications, the more consumers are likely to become innovators.

Larger coefficient of innovators, p, expected.

Not supported

H1b. The more the Internet medium increases business-directed c2c communications, the more consumers are likely to become innovators.

   

H2. The more the Internet medium increases consumer-directed communications, the larger the increase in its social influence in the diffusion process, and the more consumers are likely to become imitators.

Larger coefficient of imitators, q , expected.

Partially supported

H3. The more the Internet medium increases communications, the larger the increase of imitators when compared to innovators.

The ratio of the coefficient of imitators, q , over innovators, p, becomes larger.

Supported

H4a. The more use of the Internet medium to eliminate time differences and geographic distances, the faster the speed of diffusion.

The time to peak, t*, becomes shorter.

Not supported

H4b. The more use of the Internet medium to amplify signals of product flaws from consumers to manufacturers, the faster the speed of diffusion.

   

Data

In the selection of post-Internet products for comparison, two choices are available: one, an Internet-dependent (digital goods or goods that have to be purchased online) product, and two, an Internet-independent product. An independent product is chosen so that it is more comparable against a pre-Internet consumer durable diffusion in terms of both products being distributed through traditional brick-and-mortar stores. The Video Cassette Recorder (VCR), one of the more popular new products used among diffusion researchers, is chosen as the pre-Internet new-product, and Digital Versatile Disk (DVD) player for the post-Internet time.

click to expand
Figure 13-2: US DVD player annual sales (1997-2003)

The history of the DVD player is quite similar to that of the VCR. The VCR started with a battle of two systems, VHS vs. Betamax. It took the industry a long time before VHS was adopted as the industry standard. At the beginning, Matsushita Electric, Toshiba and Time/Warner held their Super Disc (SD) technology, while Sony and Philips held their Multimedia CD (MMCD) technology. Under pressure from the industry, the two groups formed a DVD Consortium to develop a single standard that became known as the DVD Forum. The first DVD player was commercially launched in the U.S. in March 1997. Sales of the DVD player have been the fastest in terms of penetration rate as compared to other media players. It took the DVD player five years to reach sales of 30 million units, the CD player eight years, and the VCR 13 years (www.business2.com). I collected DVD player U.S. sales data from the Consumer Electronics Association and plotted the annual incremental and accumulated sales in Figure 13-2.

The Internet and DVD Player Diffusion

When the DVD player was launched in mid-1997, Internet usage was at approximately a 16% penetration rate or 50 million units in the U.S. (how many online, n.d., www.uai.ie). The Internet medium has diffused the fastest when compared to all previous media - the radio took 38 years to reach 50 million people, and the TV 13 years (Rangaswamy and Gupta, 2000). The accelerated growth beyond 1997 is depicted in Figure 13-3. By 2001, over 50% of the population in the U.S. had access to the new medium.

click to expand
Figure 13-3: Internet diffusion in the US (1997-2003)

Source: www.nua.ie

Parallel with the popular use of the Internet, related DVD Web sites began to emerge on the Internet to include dvdtalk.com, dvdtracks.com, dvdreview.com, dvdreview.net, dvdtown.com, dvdjounal.com, dvd.ign.com, totaldvd.net and many more. Together with the development of these Web sites is also the increasing supply of movies recorded on DVD, which were once only recorded on videocassette tapes. Indirectly, movie reviews mushroom on the Internet to include filmcritics.com, imdb.com, yahoo.com and others.

Perhaps the most notorious is Netflix launching its online DVD rental services in 1997 that eliminated the late-fee business model of Blockbuster, and eventually lured Walmart to participate in the same service by 2003 (www.wired.com, Dec. 2003). All these forces play a role in DVD player diffusion.

While the examples cited above may belong to marketing issues such as marketing mix, marketing strategy or channeling , e.g., Netflix may belong to a distribution channel phenomenon . This chapter argues that the new medium plays a major role directly or indirectly in facilitating these marketing activities. Therefore, the new, medium-enabled marketing communication (both business-to-consumer and consumer-to-consumer) is the strong plausible factor that accounts for the differences of diffusion pattern between the VCR in the 1980s and the DVD player in the 1990s, as expressed within the parameters of the Bass Model.

Results

The results of the DVD player diffusion in the U.S. from 1997 to 2003 are reported in Table 13-4. A comparison with published VCR diffusion results from Helsen et al. (1993), and Lilien et al. (2000) is listed in Table 13-4.

Table 13-4: DVD player and VCR diffusion results
 

p

q

t* Peak time (in years)

q/p (imitation over innovation)

Data interval

Data range

VCR (Helsen et al., 1993)

0.006

0.365

11.07

60.83

Annual

n/a

VCR (Lilien et al., 2000)

0.025

0.603

5.068

24.12

Annual

1985-1996

DVD Player (current study)

0.003

0.442

11.10

147.33

Annual

1997-2003

This chapter has hypothesized that the coefficients of p and q would be larger for post- Internet, new-product diffusion in H1a, H1b and H2. The DVD player's coefficient of innovation, p , is 0.003, smaller than all of the previous VCR results of 0.006, and 0.025. The coefficient of imitation, q , is 0.44, smaller than the Lilien et al. (2000) result of 0.60, and larger than the Helsen et al. (1993) result of 0.365. Both the DVD player's coefficient p at 0.03 and the coefficient of q at 0.44 are below the mean average of p at 0.03 and q at 0.57 as suggested by Sultan et al.'s (1990) meta-analysis. Hypotheses 1a and 1b are not supported. The coefficient of innovation, p , does not become larger by adding the Internet medium. Hypothesis 2 is supported with the coefficient of imitation, q, becoming larger when compared to the result by Helsen et al. (1993).

However, the influence of imitation over innovation ( q/p ) is much larger in the post- Internet environment with a ratio of 147 times to that of the VCR at 24 (Lilien et al., 2000) or 60 (Helsen et al., 1993) times, indicating support for Hypothesis 3. This result concurs with the many theoretical projections of the power of the Network Externalities effect and the proliferation of online communities to foster the imitation effect in the Internet environment. The anticipated year for the market to peak for DVD Players is 11 years, the same as that of the VCR reported by Helsen et al. (1993), but longer than the five years reported by Lilien et al. (2000). Thus, hypotheses 4a and 4b are not supported.

Discussion

The purpose of this study has been to investigate the theoretical framework of whether the Internet has expanded word-of-mouth network effects. Although the DVD player diffuses fastest when compared to other media players in the past, and its diffusion pattern differs from that of the VCR, there is the 'other products' effect (Shocker et al., 2004) that might have also contributed to its diffusion other than the introduction of the new medium.

Product Contingency Effect

One of the immediate products that have affected DVD player sales is the availability of DVD discs. The contingent effect of DVD disc sales on DVD players is similar to that between prerecorded Compact Discs and Compact Disc Players (Bayus, 1987), or computer hardware and software (Mahajan and Muller, 1996). Although product contingency has been defined as the purchase of a secondary product (e.g., repeat purchase of Compact Discs or computer software) being dependent on the purchase of a primary product (e.g., computer hardware or CD players), Bayus (1987) theorized that the purchase decision could be based on the consideration of the hardware/software bundle and the interaction of the two is allowed. In other words, consumers could buy the software first, prior to the purchase of the hardware.

In the case of DVD players, without the availability of a wide range of movies recorded in DVD format, the diffusion at the beginning of 1997 remained slow as depicted in Figure 13-2. It took four years from 1997 for DVD player sales to reach above 10 million units, a 10% penetration of all households in the U.S. (at approximately 100 million households in the total population of 300 million or so). In fact, by 2000, there were only about 60% of the movies available in VHS format recorded in DVD format (Karaca-Mandic, 2004). Therefore, the slow diffusion of the DVD player and DVD discs between 1997-2001 may also explain the low value of the coefficients of both innovation, p , and imitation, q , in the empirical result of this study.

Product Complementary Effect

Recent studies from economics also confirmed the relationship between DVD player and DVD disc sales in terms of complementary network effect. Inceoglu and Park (2003) found a 1% increase in software availability increases 0.68% DVD player demand, and Karaca- Mandic (2004) found a 1% increase in new DVD release during May 2001 would increase DVD player sales by 0.5%. In the reverse, a 1% increase in the DVD player installed base in May 2001 increased the number of new DVD releases by 0.19%. These empirical results suggest the complementary effects between the two products should increase the total market size .

While price elasticity of the two products sheds light on the forecasting of the DVD market in general, this approach has neglected the possibility of other hardware that competes with the DVD player. In fact, Bucklin and Sengupta (1993) found the co- diffusion of scanners and UPC codes to be an interactive complementary effect in which the demand for the complementary innovation (UPC code) has a separate adoption track than that of the main product ( scanners ). This is because having a UPC product code label on the merchandize does not necessarily mean the vendor will use the scanner to enter the sales. Similarly, DVD discs do not have to be played in a DVD player. They can also be played on a personal computer or on a laptop.

Product Competition Effect

This leads us to the third product effect of competition. Personal computers or laptops are possible competitive hardware against DVD players. Unlike VHS tapes that can only be played using a VCR, DVD discs can be played in both DVD players and personal computers. However, DVD-ROM Drives were initially external only until 2002 when manufacturers began to assemble them internally within the computer (http:// www.plextor.com). In that sense, DVD players were unlikely to be under the competitive pressure of personal computers in the late 1990s.

Instead, the greater threat to the diffusion of the DVD player was the previous generation of a movie-playing device - the VCR. From 1997 to 2001, VCR and the VHS tapes are dominant. Finally by 2001, 90% of VHS movies were also released on DVD (Karaca-Mandic, 2004). At the same time, the incremental annual growth of DVD player sales reached 10 million and beyond for the first time (Figure 13-2). In a sense, 2001 is the year when the DVD player and DVD discs bundle began to show signs of replacing the VCR and the VHS tape market. This delayed replacement effect could possibly be one of the reasons why the coefficient of innovation, p , and the coefficient of imitation, q , of the DVD player remain small.

Product Substitution Effect

The replacement of the previous technology by a new innovative product is defined as substitution (Bayus, 1992; Norton and Bass, 1987). Marketing diffusion researchers have investigated substitution effect in terms of same-time competition, successive generation substitution (Norton and Bass, 1987), and next generation substitution (Bayus, 1992). The relationship between the DVD player and VCR is one of previous- and next- generation. It is not a successive substitution such as the one between Pentium I and Pentium II. The DVD player is not a second generation of the VCR. In fact, the DVD player is a more user -friendly innovation with less features than the VCR (e.g., the blinking clock in the VCR no one could ever get rid of, or the complicated programming functions). Bayus (1992) found in the case of the past- and next -generation relationship, e.g., between the LP record player and Compact Disc player, that the price decreases faster for the successive product generation.

These substitution studies however have been silent on the changes of the diffusion patterns, specifically , the coefficient differences of innovation, p , or imitation, q , in successive- or next-generation product diffusion. It is conceptually plausible to suggest that the next-generation p is likely to suffer a smaller coefficient as shown in the current study because innovators from the previous generation may hesitate to be the innovators in the next-generation product. The hesitation may come from inertia burdened by a large collection and the continual production of the older generation related products.

Early Adopter Fatigue and the Decrease of Innovation Effect

Apart from the aforementioned four types of product relationships and their effects on the coefficient of innovation, p , early adopters fatigue is also proposed as the reason for less innovators in the diffusion process among recent technological products. Namiranian and Hopkins (2002) suggested that consumers in developed countries might be suffering from 'early adopter fatigue' or 'cooling interest' toward new technologies. Their paper specifically investigated cellular phone diffusion in 12 countries, inclusive of both developed and newly developed countries. They found early adopters in North America, Australia and the United Kingdom are less enthusiastic about the idea of new communication technologies than early adopters in Europe and Asia. Although their results are not a direct comparison to the current study, they nevertheless hint the possibility of early adopters' fatigue.

Contingent, Complementary or Competitive Effect between Products and Services

Other than product effect, DVD player sales could also have been affected by the availability of rental services of DVD discs, movie-on-demand by cable operators, satellite TV services, or TIVO programming and many more. The introduction of Netflix Internet-based online movie rental service in 1997 can be considered as a complementary factor for DVD player diffusion. It would be interesting to watch the continuation of the diffusion after Walmart, which has over 4,000 retail stores and a substantial customer base, introduced the same service in December 2003. At the same time of the launching of the service, Walmart synchronized it with the promotion of a DVD player at the lowest ever retail price of USD $29.99 (http://www.millenniumdesktop.co.uk). This strategy of lowering the price of the hardware (DVD player) substantially in order to further the software sales (in this case the DVD disc rental service) is common between contingent products (Bayus, 1982).

Although the retail giant Walmart has joined the DVD movie rental service, the industry is not without external competition. The immediate threat is the sophisticated technology of cable companies that can digitally deliver movie-on-demand services. TV viewers are allowed to watch movies at their own time at a fixed monthly fee. For newer movies, digital delivery through the cable box costs anywhere from $2.99 to $3.99. This could dampen further DVD player sales.

Does the Internet Eliminate Information Asymmetry?

Our empirical results have indicated a smaller p , a larger ratio of q/p while the time to peak remains unchanged. These results seem to suggest that the Internet medium decreases the number of innovators, imposes a larger imitation than innovation effect, and the time to peak does not differ between VCR and DVD Player diffusion.

Information Advantage

At the core of the diffusion theory is the division of adopters into several groups (Rogers, 1995), and in the case of the Bass Model two groups of innovators and imitators. The major difference between the two groups is the timing of the acquisition of information. In the context of new-product adoption, the first adopters or the innovators have always had an advantage over imitators in terms of obtaining information first. And of course being able to obtain information first is related to many other factors such as cosmopolitanism, financial advantages and so on. We argue, however, that the diffusion theory was based on a pre-Internet time with the tenet that 'information is a scarce commodity.'

In Internet time, when information is abundant and available at the same time at any place at minimal cost, the innovators' information advantage disappears, and therefore we will see a blurring of the line between innovators and imitators. This could plausibly explain a smaller coefficient of innovation, p, as seen in the empirical results of DVD players when compared to VCR diffusion. In other words, within the Bass Model, the number of adopters at time zero has become smaller.

Information Dissemination

Although most research in marketing communication has focused on advertising messages, it is no secret that the most effective dissemination of information is through word-of-mouth in new-product diffusion (Feick and Price, 1987). Goldenberg and colleagues (2001) went further and verified with empirical results that information dissemination is dominated by word-of-mouth rather than advertising. They found weak-tie information dissemination as effective as strong-tie information dissemination. This confirms that Internet-based word-of-mouth (sometimes no-tie at all) is effective in propelling marketing communications. The new medium magnifies the word-of-mouth effect with its one-to-many-person human network effect, and also the many-node-to- node, computer-based network effect.

With the use of a new medium in new-product communication, the increased information- dissemination, word-of-mouth effect coupled with the diminishing innovator information advantage effect is expressed in this chapter's findings of a larger ratio of the coefficient of imitation, q, over the coefficient of innovation, p .

Network Externalities and the Delay of Diffusion

While the ratio of q/p in the Bass model has no direct mathematical relationship with the time to peak, t*, in cross-product comparisons (Mahajan et al., 1990), the increased network externalities may be associated with the delay of new-product diffusion. Gupta et al. (1999) introduced the relationship between High Definition TV (HDTV) and the relevant programming as one example of indirect network externalities and concluded that the HDTV would diffuse slower than originally expected due in part to the lack of programming. An indirect network externality is defined as the greater the availability of complementary products, the more attractive the capital goods are for consumers. Similarly, Goldenberg et al. (2004) argue that network externalities may in fact have a 'chilling effect' on new-product growth. That is, instead of having a positive effect that fuels faster adoption, they illustrated through the use of a Cellular Automata Threshold Model in simulation and concluded that network externalities has a harmful effect that delays sales growth.

In this chapter, manufacturers did not gear their release of movies in DVD format up to par with existing VHS tapes until 2001 (90%) when DVD disc sales surpass that of VHS tapes for the first time (DVD Entertainment Group , 2002). In retrodiction, one can argue that the capital good (the DVD player) was therefore delayed for four years to bring the time to peak, t* , to 11 years, the same as that of the VCR.

[2] The Bass Model is a descriptive and predictive model classified as semi-behavioral (Leeflang et al., 2000). It is not a causal model per se . The behavioral assumptions are explicit. Particularly relevant to this chapter is the fact that the behavioral assumption of one social system is assumed to be relatively constant through time as suggested by Weber (1946) on the foundation of sociology. In the context of this chapter empirically anchoring on the Bass Model, the strong factor that could possibly change the diffusion pattern across time within one social system is the marketing communication process that affects the coefficients of innovation, p , and imitation, q . Subsequent model such as the Generalized Bass Model (1994) includes the effect of marketing mix in the diffusion process and concludes that when advertising rate of increase and price decrease are both constant, the measurement reduces to the original Bass Model.




Contemporary Research in E-marketing (Vol. 1)
Agility and Discipline Made Easy: Practices from OpenUP and RUP
ISBN: B004V9MS42
EAN: 2147483647
Year: 2003
Pages: 164

flylib.com © 2008-2017.
If you may any questions please contact us: flylib@qtcs.net