The last few years has witnessed a sweeping increase in use of the Internet. As a result, marketers have already become very active participants in this medium since they have recognized they need to exploit innovative information/communication technologies to be successful in this ever-changing marketplace (McKenna, 1995). The Internet has been the fastest growing medium in the last few years. User demand on the Internet has experienced dramatic increase since its launch. The approximate number of total Internet users is 32 million (Forrester Research, 1996 at URL: http://www.forrester.com). The Internet growth rate in domain and host numbers internationally also supports the notion of dramatic increase in demand on this medium (Cyber Atlas, 1996 at URL: http://www.cyberatlas.com/news.html).

The explosive increase of the Internet can also be demonstrated by the number of downloads of the two most popular Web browsers: Netscape and Internet Explorer. According to Netcape (1996 at URL: http://www.netscape.com/), its Web site receives more than 85 million hits per day. Microsoft Internet Explorer 3.0 has the user demand of more than 2.5 million units-downloading worldwide, which implies that one copy of the unit has been downloaded every second since its launch (Microsoft Internet Explorer, 1996 at UTL: http://www.microsoft.com/ie/press/091196.htm). Advertising expenditure is another indicator of the prosperity of the Internet as a mainstream medium. According to Jupiter Communication (1996 at URL: http://www.webtrack.com/), Web advertising will increase from $312 million in 1995 to $5 billion by 2000.

Although the commercial viability of the WWW is not clear at this point, a recent study by Gupta and Chatterjee (1996) presented at the "Electronic Marketing" conference held at The University of Texas at Austin on October 2-3, 1996, suggests that, in certain product categories such as home electronics, travel, video, books and music, the growth in retail purchasing through the Web has been great from Quarter 1 1995 to the present. The general conclusion of that Conference as a whole was that the phenomenal growth now taking place in the Web environment would lead to commercial success for those using creative techniques in retailing and marketing.

As the demand of the Internet increases, technology supporting the Internet becomes more important. One important technology is a programming language for the Internet which can connect people with information and transmit programs through the Net. As the world becomes more and more connected through the Internet, this Internet-Related Programming Technology (IPT) becomes more and more important. IPT allows Internet users to access and transmit programs and data with greater effectiveness and ease than has been possible in the past. The following section explains in detail what IPT is, how it relates to advertising, and some possible implications for advertising.

Internet-Related Programming Technology (IPT) is the programming language technology of the Internet that can distribute data and programs over the Internet. Two main characteristics of IPT are its "platform-independence" and real interactivity capability. First, from the technological aspect, IPT enables Internet users to easily receive data and programs over the Internet in real time, regardless of different computer operating systems they are using like PC, Mac, Unix, and others. (San Jose Mercury News, 1996 at URL: http://uts.cc.utexas.edu/~tecas/Adjava.html).

Another main characteristic of IPT is that it can provide real interactivity and connectivity between senders and receivers over the Internet. Interactivity has been one of the most important characteristics describing the new media including the Internet. It is also becoming an increasingly important characteristic of marketing as illustrated in the concept of one-to-one marketing (Peppers and Rogers, 1993). There are three ways of defining interactivity: 1) interaction between senders and receivers (Flaherty, 1985; Cook, 1994), 2) interaction between humans and machine (Rice, 1984), and 3) interaction between message and its users (Steuer, 1992; Williams et al, 1988).

Compared to limited interaction between viewers and traditional audiovisual mass media, for example, selective exposures (Klapper, 1960) and zipping and zapping (Westin et al, 1993), the Internet brings more interactivity--users' more control over the pace, time, and messages. IPT can bring even more interactivity to the Internet. It first can make the communication process between senders and receivers of messages easier and more efficient with its ubiquitous platform-independent capability. Senders do not have to compile different program versions for different operating systems and receivers do not have to buy different software/programs for different systems. Second, IPT can also make the interaction between humans and machines and between the message and its users easier and more efficient. IPT can provide users with higher level of control over the messages--selection, modification, and search of the form and content of the mediated messages.

There are currently three major kinds of IPT: Java, Shockwave, and Active X. Compared to Shockwave and Active X, which are products of the Internet, Java originally had nothing to do with the Internet. Its original attempt was to create a simple object-oriented and graphical interface for users and extend its capability over different products so that they can communicate with each other (Moon, 1996 at URL: http://uts.cc.utexas.edu/~tecas/WhatIsJava.html).

Despite the importance of Internet programming technology, the terminology describing this technology is scarce. This lack of terminology was one of the starting points of this study. The researchers have suggested a new term, Internet-Related Programming Technology (IPT), to embrace all different types of Internet programming languages and output programs. There are two perspectives to classify IPT: (1) by level of generality of Internet programming language itself; and (2) by end-users or end-objects created with IPT. The first perspective, generality, classifies different programming languages into three levels of generality: high, medium, and low. Java and Active X have a high level of generality because they can be used to make any kind of applications for any purpose. For example, Java can make any program from simple applications like animation to complex applications like TECAS Media Planner (URL: http://uts.cc.utexas.edu/~tecas/admedia.html). Additional Information on this application can be found in Leckenby (1995). Shockwave/Director has a medium level of generality because it is used for specific purposes such as graphics and animation but with some variety of outputs. Adobe Acrobat/Exchange has a low level of generality because it does only one job--getting people Web documents from paper documents or word-processed documents.

The second perspective of IPT classification is not looking at the programming language itself but focusing on what end-users can do with the end-products made with IPT. There are two dimensions in this end-users-perspective classification system: (1) interactivity and (2) functionality. Here, interactivity means whether the end-users can interact with the end-products created with IPT, that is, whether the end-users can input something to get some output. Input can be anything from pointing and clicking to actual data input in order to get some pre-programmed or newly-created output. Interactivity can be divided into two levels: high and low interactivity.

The second dimension is functionality or task-fulfillment. It is based on what kind of outcome or function the end-users are getting by using programs made with IPT, that is, with the products made with IPT, what kinds of tasks or functions the end-users can fulfill. Functionality also can be divided into two levels: high and low functionality. With high functionality, the end-users can have a high level of control and creativity over the outcomes or tasks made possible by IPT. Based on the two dimensions with two levels, IPT can be categorized into four different types: Functional-Interactive, Nonfunctional-Interactive, Functional-Noninteractive, Nonfunctional-Noninteractive IPT. Table 1 shows the resulting IFI Grid (Interactivity-Functionality IPT Grid).

Functional-Interactive IPT has high interactivity and high functionality. Here, end-users can interact with IPT-made objects by inputting data to get some outputs while enjoying high control and creativity over the outcomes. The products created by this type of IPT are function-oriented or task-fulfillment programs such as TECAS Media Planner (Figure 1). The second type is Nonfunctional-Interactive IPT with high interactivity and low functionality. The programs created with this type of IPT are non-task fulfillment programs with some interactivity. The example would be interactive animation programs where the object is animated or makes sounds when users click a button. Another example would be online games on the Internet, where users have high level of interaction with the content but can not fulfill any task.

The third type is Functional-Noninteractive IPT with high functionality and low interactivity. The programs created with this type of IPT are mostly operational applications for computer or Internet itself. One example is multimedia applications such as Internet audio applications, fast graphics viewers, and imbedded video clips. Here, users cannot input anything to get some output; rather, the programs run within operational systems automatically when needed. The fourth type of IPT is Nonfunctional-Noninteractive and characterized by low functionality and low interactivity. The programs created with this type of IPT provide no control over objects at all. End-users cannot do anything with this program except just watching one-way, pre-programmed applications. An example would be simple Java applets such as moving letters or billboards, graphic displays, and simple animations.

In terms of control over objects and creativity of outcomes, figure 2 shows another way of differentiating four different types of IPT, from the highest to the lowest: Functional-Interactive, Nonfunctional-Interactive, Functional-Noninteractive, Nonfunctional-Noninteractive IPT.

Analyses of Web Site for IPT Grid

Understanding the difficulty in obtaining appropriate data for any type of analysis dealing with Web marketing at the current time (Gupta, 1995; Miller, 1995), the researchers personally contacted a renowned Web research company called Jupiter Communication. This firm provided a list of the top 50 Web publishers based on their Web advertising revenue. The first analysis was done with this sample of the top 50 sites (Table 2) to test the viability of the newly-developed typological system of IPT. The researchers examined these 50 sites to check whether each site used one of the four types of IPT in IFI Grid. Table 3 shows that total 24 sites use IPT: zero Interactive-Functional IPT, one Nonfunctional-Interactive, zero Functional-Noninteractive IPT and 23 Noninteractive-Nonfunctional IPT. The result indicates that IPT is quite common among popular Web sites (48%) but most of them are in the level of Nonfunctional-Noninteractive. These sites use simple animations without providing any control over objects for end-users.

The second analysis was conducted with the sample of 220 IPT sites: 200 Java sites and 20 Shockwave sites. The sample of 200 Java sites was drawn from one Java collection site called Gamelan (URL: http://www.gamelan.com) which was the first site to begin a collection of Java applets and applications. The researchers randomly selected 200 Java sites out of total 3,732 Java collections in Gamelan.

A sample of 20 Shockwave sites was randomly selected from Macromedia's Shockwave collections of total 151 sites (URL: http://www.macromedia.com/shockwave/). The researchers examined the sample of 220 IPT sites to categorize them into IPT Grid. Table 4 shows that all 220 IPT sites could be categorized into the IPT Grid, with four Interactive-Functional, 123 Nonfunctional-Interactive, 28 Functional-Noninteractive and 65 Nonfunctional-Noninteractive IPT sites (X2 = 146.44*, p < 0.01).

In this section, the relationship between IPT and advertising is examined. This section consists of the analyses of advertisements made with IPT in the top 50 Web sites. The researchers first examined the top 50 Web sites to check whether they have IPT-created Web advertising. The researchers defined Web advertising as paid or free advertising on the Web which is placed on independent site. The researchers did not consider advertising banners themselves as advertising, instead the linked site by clicking the banners as real advertising. All advertising banners in the top 50 sites linked users into the separate advertising sites. The researchers examined these separate-advertising sites linked by the banners to check whether the advertising sites were using IPT or not.

Analyses of Ads made with IPT

The result shows that 20 out of the top 50 sites have at least one advertisement using IPT and that total 32 advertisements are using IPT in these 20 sites since one site can have more than one IPT-using advertisement. Table 5 shows the product types of these 32 IPT-using advertisements. It indicates that computer-related products take majority with 50 percent. Other product types are travel- and telephone-related products with 12.5 percent and 9.4 percent respectively (X2 = 13.25*, p < .05). This result is quite understandable when we consider the fact that advertising medium is the Internet.

The researchers also examined interactivity of IPT-using advertisements in the top 50 Web sites since interactivity is the most important benefit Web advertising can provide for users. As mentioned previously, there are three types of interactivity: (1) interaction between senders and receivers, (2) interaction between human and machine, and (3) interaction between message and its users. Based on these three interactions, the researchers define interactivity as having three dimensions: (1) manipulation, (2) feedback, and (3) information search. Manipulation or interaction between human and machine literally means clicking and pointing for input and output. Feedback or interaction between senders and receivers is defined as the presence of the possible channel of feedback from receivers to senders and vise versa.

Information search has been emphasized as one of the main characteristics of interactivity (Raman and Leckenby, 1995; Heeter, 1989; Yang, 1994). Here, the researchers define information search or interaction between message and users as the presence of content-search method on site. A high level of interactivity includes all three above-mentioned dimensions of interactivity: manipulation, feedback and information search. Medium level of interactivity means manipulation and feedback without information search, and low level of interactivity means manipulation only.

Table 6 shows that if interactivity is defined as providing all three dimensions, 24 out of 32 IPT-using ads (75%) are considered to be interactive. It also shows that 90.6 percent is considered to be interactive when defining interactivity as manipulation and feedback and that 100 percent is interactive when defining interactivity as manipulation only. The result suggests that IPT-using ads have a high level of interactivity taking full advantage of the two-way interactive nature of the Internet.

IPT can be also used in managing advertising more efficiently. A good example of using IPT for advertising management is TECAS Media Planner created by Leckenby (URL: http://uts.cc.utexas.edu/~tecas). Figure 1 shows TECAS Media Planner.

Analyses of Top 50 Web Sites

The researchers divided the top 50 Web sites into three groups based on their advertising revenues: large (the top 17 sites), medium (the next top 17 sites), and small (the remaining 16 sites), and examined each group to check whether they have IPT-using advertisements. Table 7 shows that 13 out of 17 large-size Web sites (76.5%), 5 out of 17 medium-size Web sites (29.4%), and 2 out of 16 small-size Web sites (12.5%) have IPT-using advertisements. It indicates that large-size Web sites are most likely to have IPT-using advertisements (X2 = 15.26*, p < 0.01). The result implies that IPT-using advertisements are quite common among popular sites and that IPT has an impact on advertising already.

Table 8 shows the relationship between business types of Web sites and use of IPT on site. There is no relationship between these two variables (p > 0.9); that is, use of IPT is quite spread equally over different business types of Web sites. It suggests that the characteristics of Web sites are not a determining factor of usage of IPT on a site. It is now known that IPT is quite common and omnipresent among popular Web sites, regardless of business types of the sites. Why is it common and what can IPT do for advertising?

First, IPT can make advertising itself more fancy and attention-getting. Simple animation created with IPT like moving letters and animated billboard is a good example. Actually, it is very common among popular Web sites; 23 out of 24 IPT-using sites (95.8%) in the top 50 Web sites are using IPT for this purpose. Second, IPT can make advertising more interactive, for example, virtual car advertising using Shockwave (URL: http://www.macromedia.com/) and mBED (URL: http://www.mbed.com/). Consumers can open doors and trunk, look at the interior, change the interior and exterior colors, sound a horn, and more. Any type of interactivity can be made more possible by IPT so that both advertisers and consumers can take full advantages of two-way interactive nature of the Internet. Without interactivity, advertising on the Web is nothing more than traditional one-way advertising in existing mass media. Therefore, IPT's interactivity-generating characteristics manifests the large implication of IPT to advertising

Third, IPT can provide more benefits for both advertisers and consumers. Software sites with IPT are examples. Since IPT makes it possible for programmers to write one platform-independent software for all Internet users, regardless of operating systems they are using, software sites can attract many Internet users into their sites, which in turn attract advertisers to subsidize the sites.

In other words, Internet content becomes software and advertising supports Internet content. As mentioned before, IPT can provide more interactivity which is the most distinctive feature of the Internet. Interactivity is what both advertisers and consumers seek. Therefore, through IPT, Internet content becomes more interactive, which may draw more audiences and advertisers.

Analyses of IPT Sites

The researchers examined the sample of 220 IPT sites randomly drawn from Gemalan site (URL: http://www.gamelan.com). Table 9 shows the relationship between the types of IPT and advertising. The results illustrate that 28 out of total 220 IPT sites (12.7%) have advertising. It also shows that 3 out of 4 Functional-Interactive IPT sites (75%), 19 out of 123 Nonfunctional-Interactive IPT sites (15.4%), none of 28 Functional-Noninteractive IPT sites (0%), and 6 out of 65 Nonfunctional-Noninteractive IPT sites (9.2%) are subsidized by advertisers. There are only three Web sites (out of the 220 sites) which are the top 50 Web sites. According to Jupiter Communication (URL: http://www.webtrack.com/adspend/4q95.html), advertising dollars are densely concentrated in the top ten Web sites taking around 75% of total Web advertising revenues. Therefore, the fact that only 28 out of total 220 sample sites (12.7% ) have advertising is quite understandable since most samples are not within the top 50 Web sites based on their advertising revenues.

Table 10 shows the relationship between interactivity of IPT and the presence of advertising. It indicates that 22 out of 127 high-interactivity IPT sites (17.3%) and 6 out of 93 low-interactivity IPT sites (6.5%) are subsidized by advertisers (X2 = 5.71*, p < .05). This result implies that IPT sites providing high interactivity are more likely to be subsidized by advertisers than IPT sites providing low interactivity. It supports the argument that interactivity is the most important feature of the Internet; that is, low-interactivity IPT sites might not attract both the Internet users and advertisers, which might end up with no subsidization from advertisers.

What is an example of each type of IPT site subsidized by advertisers? Two examples of Interactive-Functional IPT site subsidized by advertisers are the "Dart" site (URL: http://www.netaccent.com/kkotak/dartdem1.htm) which features a fully functional on-site Internet-centric email program completely implemented in Java and Corel Office for Java (URL: http://officeforjava.corel.com/index.htm). An example of a Nonfunctional-Interactive IPT site is the "Riddler" online game (URL: http://www.riddler. com/), which features many online games. Many advertisers subsidize this site. An example of a Nonfunctional-Noninteractive IPT site subsidized by advertisers is "C|net" site (URL: http://www.cnet.com/), which features simple animations such as moving letters and changing billboards. Advertisers also sponsor this site.

There are 28 Functional-Noninteractive sites in the sample of 220 IPT sites, but advertisers do not currently subsidize all sites. The probability is that Functional-Noninteractive IPT sites are not providing interactivity, which might not attract both the Internet users and advertisers and in turn might end up with no subsidization from advertisers.

The researchers had several research problems to study at the beginning of the study. One of them was the lack of terminology describing the Internet programming environment. This paper attempts to address this problem by generating the new term IPT (Internet-Related Programming Technology), and by building the typology for IPT--IFI Grid (Interactivity-Functionality IPT Grid).

Another issue concerns the lack of research on the relationship between IPT and advertising. This paper analyzed the top 50 advertising Web sites and a random sample of 220 IPT sites to examine the relationship between IPT and advertising. Finally, this paper gave some reasons for the potential impact of IPT on advertising.

IPT makes it possible to distribute programs over the Internet. More and more, IPT-created programs are being developed which provide genuine interactivity and usefulness as helpers for everyday tasks such as media planning in advertising. Then, who will pay for these programs used over the Internet? Advertising can be used to subsidize the cost of this type of Internet content in the same way it subsidizes the content of magazines, newspapers, television, and so forth. In other words, Internet content becomes software. In this way, IPT has brought advertising practitioners and others into a close-knit relationship with computer programmers and providers of software content. This makes IPT important now and probably even more important in the future to those interested in advertising (Leckenby, 1996 at URL: http://uts.cc.utexas.edu/~tecas/ Adjava.html).

In conclusion, this study supports the importance of IPT in Internet advertising and the medium itself. The most important role of IPT is to facilitate more interactivity, which is the most important and distinctive feature of the Internet. IPT can provide more interactivity for the medium, advertisers, and audiences and can satisfy all different groups in the Net by providing more benefits for each group. Advertisers can subsidize interactive sites, and audiences can get free interactive programs from the sites. That is, advertisers pay for Internet content (software) and audiences enjoy it free or with small dollars. The role of IPT in Internet advertising may become more and more important as the audiences for such content grow. Advertising academicians and practitioners need to be knowledgeable about these new technologies and their potential impacts as a basis for the rapidly changing nature of the field of advertising.

Table 1

IFI Grid (Interactivity-Functionality IPT Grid)

Web sites with IPT=24; Web sites with no IPT=26; N=50

Table 4

N=50, X² = 15.26*, p< .01

Table 10