I recently retrieved my old dissertations from MA Design for Interactive Media course back in 1995… This one discusses the likely commercialisation of the then-nascent Web – perhaps worth pointing out that this was when the latest cutting-edge browser was Netscape 1.1, Google and Macromedia (now Adobe) Flash didn’t yet exist, and the concept of ‘Social Media’ was still a decade away… The ‘SellNet Project’ referred to was a piece of interactive multimedia that I and two colleagues (Lee Woodard and Iain Jones) developed for our final MA project. It probably still exists somewhere, but whether it would run on today’s computers is another matter – it was built in Macromedia Director (version 3 as I recall) and at a time when anything above an 800×600 screen, 256 colours and half-a-meg or so of RAM was sheer luxury… ;-)
Looking back, I’m pleasantly surprised by how much now seems so prescient (notwithstanding a few moments of naivety!). Fascinating (for me) to re-read it after all these years. I hope you may find the same.
How might the commercialisation of the Internet affect the look and feel of the World Wide Web?
Submitted in partial fulfilment of module ADC4926 of the MA Design for Interactive Media, Middlesex University Centre for Electronic Arts
29th September 1995
This essay is a discussion of possible future developments on the Internet. Focussing in particular on social and commercial issues, it examines how the Internet might become more commercial and what effects this might have on its look and feel. Its main contention is that as a result of the needs of greater on-line commercial activity, the World Wide Web will take on more of the functions of other parts of the Internet, and will move from being a solitary environment to a more social one.
The SellNet Project, undertaken for the final semester of the MA Design for Interactive Media, explored some of the possible future developments of the internet over the next few years, focussing primarily on commercial and social issues. In this dissertation, I will discuss in more depth some of the themes which were developed in the project. Broadly, I will examine how the Internet might become more commercial, and what effects this might have on the way in which we use it. I shall concentrate in particular on how this might change the look and feel of the World Wide Web. It is my contention that, as a result of commercial needs, the World Wide Web will take on much of the functionality of other areas of the Internet and, as a further consequence of this, using the Web will move from being a solitary activity to a considerably more sociable one.
The World Wide Web (hereafter referred to simply as the Web) is just one of many components which go to make up the Internet. Other components include e-mail (the on-line equivalent of the postal service), Usenet discussion groups (to which users `post’ comments for others to see, rather like putting a note on a conventional notice board) and MOOs (where users can engage in near-realtime conversations). These components are mostly text-only, and are accessed via terminal emulation software such as NCSA Telnet. In contrast, the Web is graphically-based, and hence visually much more impressive, and is accessed via separate `web browser’ software, such as Netscape. The fundamental distinction is that the Web is based around Hypertext Transfer Protocol (http), while the other components of the Internet are not.
The Web consists of millions, of `sites’ which reside on computers connected by the telecommunications infrastructure of the Internet. Each site is stored on a server, and contains one or more `pages’ which form the content of the Web. Pages are made up of text, graphics and links to other pages or sites; at present these links comprise a `hot text’ or a `hot image’ which is clicked on in order to download the page to which it points. Web pages do not commonly feature audio, video, or any high level of interactivity.
Underlying the Web is HTML – Hypertext Markup Language – and, on a few sites, VRML – Virtual Reality Markup Language. These are simple languages which are used to encode Web pages to a common standard by adding control codes to the raw contents of the page – i.e. the text and graphics. These control codes then tell the browser software how to format the page on screen – the style and size of different sections of text, which words or pictures are to act as links and where they lead to when clicked, and so on. This helps to reduce the file size of the page, and hence the time taken to download it, since all the software needed for the display and functioning of the page is already contained within the browser. Given the problems of limited and overstretched bandwidth, the fast but nonetheless insufficient data transfer rate of today’s modems, and the capabilities of many of the personal computers currently in use, this is of considerable advantage to users. However, along with poor design skills on the part of many Web authors, this has the disadvantage that all Web pages tend to look very similar, and offer considerably less in terms of interactivity and multimedia content than is available from other computer-based media (for example CD-ROMs).
The use of markup languages makes producing content for Web pages relatively easy. The task is slightly harder than using a word processor, but considerably easier than programming a piece of software. Without wishing to suggest that the Web is `the people’s medium’ and that anyone can have a Web site (in theory this is true, but in practice the cost of access is prohibitive to a great many people), it is noticeable that many pages have not been produced by professional Internet experts, but by `ordinary’ Web users. (I do not wish to get into a discussion of Web demographics, but it must be noted at this stage that the population of Web users cannot be said to represent a cross section of the population in general. To give one example, the fact that educational establishments have free access to the Internet means that a disproportionate number of users are students). This may help to explain why the Web has expanded so rapidly. Of course, it is also extremely easy to produce a bad Web page – one which is unengaging, poorly presented and difficult to use; this, combined with the sheer size of the Web, often makes finding useful information about a specific subject difficult and time consuming.
While there is a small amount of crossover between the Web and the other components of the Internet – for example, e-mail is possible from some Web sites via specially provided electronic forms, and some MOOs have Web counterparts – at the time of writing the Web is still fundamentally separate from the rest of the Internet. It is of particular significance to this discussion that most of the social aspects of the Internet occur in areas other than the World Wide Web – the Web itself is a very unsocial place.
The Internet offers huge opportunities for business. It presents the possibility of reaching tens of millions, and ultimately perhaps billions, of consumers throughout the world, unrestrained by geographical or political borders. The potential for advertising, selling, market research, and the creation of new markets is enormous. However, the trade-off which comes with having access to such a large audience is that the audience is now in control – users can choose, for example, not to see an advertisement, and they can respond much more effectively to violations of their choice.
The primary commercial use to which the Internet is currently being put is that of providing on-line promotion for companies and the products and services they sell in the real world. Unfortunately, many companies seem to view the Internet as just another place in which to sell advertising space – for example, the September 1995 issue of Internet magazine reports that advertisers on the Wall Street Journal’s Web site are being charged $50,000 for six months, while the Microsoft Network is offering space at $7,500 per month. Yet little thought seems to have been paid to the fact that users need to be persuaded to view these advertisements!
So how can companies advertise and promote themselves more effectively on the Internet? Since users do not have to tolerate advertising, conventional techniques are unlikely to be as effective as in other media – new methods are needed. In particular, it is of vital importance to make users want to see an advertisement and visit the company’s Web site. How might this be done? Three methods which have been discussed by others are `deal-making’, `advertising-on-demand’, and `giving something back’.
Don Peppers and Martha Rogers (1993) put forward one possibility – deal-making. Until the development of VCRs and remote controls, TV advertisers had an implicit bargain with consumers: `you watch my commercials, I’ll pay for the television programme.’ Interactivity, they argue, is now making that bargain explicit. Companies advertising in any interactive medium, such as the Internet, might reach more consumers by offering them some reward or incentive for receiving promotional material. For example, a company might enter the user into a prize draw in return for watching a short commercial, or they might offer free or reduced-price products in return for completing a questionnaire, and so on. Peppers and Rogers point out that this sort of deal-making is already happening on the telephone – they cite a U.S. company called FreeFone which pays consumers to listen to advertisements whenever they dial their telephones at around 5 to 10 cents each; another company, HomeFax, offers a free fax machine to anyone willing to receive a limited quantity of unsolicited commercial faxes each week.
At present, perhaps the ultimate way to encourage users to view advertisements or visit pages on the Internet would be to pay for their access time, and indeed this may be about to happen. Internet magazine (September 1995) reports a proposal to produce floppy disks containing a limited version of the Netscape Web browser which would be distributed through conventional media – for example, in cereal packets, or sent in return for coupons etc. These disks would offer the user free Internet access (via a toll-free telephone number) for thirty minutes, in return for automatically sending the user to the sponsor’s home page when they first log-on. After leaving their details users would be free to surf the Web as they wished – after the thirty minutes has elapsed, they might gain more free time by, for example, agreeing to receive promotional e-mail, or taking part in market research.
One characteristic of on-line advertising, then, is likely to be `invitational advertising’ – companies will invite consumers to view their advertisements, rather than virtually forcing them to watch (as in television, for example), as if trying to irritate them into remembering the company’s brand.
Another method of promotion which is likely to be effective on the Internet is `solicited advertising’, or advertising-on-demand. This utilises the fact that, at certain times, consumers need and want information about a company’s products – for example when they are thinking of making a purchase and want to compare different brands. One way a company might take advantage of this would be to ensure they are listed in an on-line Yellow Pages to which users can go when they need information, but this in turn presents the problem of making sure that users know of, and where to find, that site (a possible solution to this will be discussed below). Another way would be for a company to post messages to newsgroups, on-line conferences and so on, giving details of where to find its site. This would have to be done carefully, and only when appropriate, usually in response to a query (`Where can I find information about x?’) – “there is a high bias against anything resembling a sales pitch and using electronic junk mail has already proven arrogant and unsuccessful” (Dinsdale, 1995).
Although this is a perfectly reasonable way of providing information on demand, as Michael Strangelove (1995) points out, it would not be difficult for advertisers to hire people to ask leading questions, and others to answer with veiled plugs for the advertiser’s Web site or other on-line presence. Still other people could be commissioned to write glowing reviews of a company’s products and post them to relevant newsgroups. However, companies using this method must be extremely careful not to gain a bad reputation, which is all too easily done given the Internet community’s dislike of unsolicited e-mail (Ellsworth and Ellsworth, 1994). Because the Internet allows rapid, global communication, “bad PR spreads very quickly and widely on [it]” (Ibid) but potentially so too might a good reputation. If mention of the company and its site could be promoted in on-line discussions then awareness of it would be spread by word of mouth, without the advertiser having to continually monitor the relevant newsgroups, MOOs and so on.
The Internet “has traditionally been anti-commercial” (The Coalition for Networked Information, 1994) and few users are likely to visit a company’s Web site unless they stand to gain something from their visit. Thus advertisers must `give something back to the Internet’ by providing information or entertainment which is not in itself promotional, but which rewards the user for visiting the advertiser’s site. This giving back “is the accepted norm on the Internet…. A business that only provides information about their [sic] company and products will be perceived very negatively” (Ellsworth and Ellsworth, 1994).
`Giving back’ might mean providing resources (images, sound files, film clips, software etc.) for the user to download, or providing additional content which the user might find useful or entertaining. For example, Aaron Weiss (1995) cites a service called `What’s On Tonite!’ which provides free daily TV listings via e-mail in return for the user also receiving advertising and promotional material from the service provider; a sports equipment manufacturer’s Web site might not just contain information about the company and its products, but could also contain images and videos of famous athletes, health and fitness advice, links to other related sites, and so on, all of which may be of interest to sports enthusiasts who, as a result, would be more likely to visit, and frequent, the site (an example of such a site can be found at the WWW Tennis Server).
While the three methods discussed above may not result in the tens of millions of visitors who could potentially access a company’s site or see their advertisement, they would result in more accurate targeting of consumers – even if only a hundred people visit the site each day, that’s a hundred people who want to be there (as opposed to the everyday, real world, situation whereby several million people are annoyed at having their favourite television programme interrupted by an advertisement in which they have no interest).
Commercial interests are gradually increasing their presence on the Internet, and especially the Web, but few seem to have woken up to the fact that conventional approaches to advertising are not going to be effective on-line – we have yet to see the above techniques become commonplace. Perhaps the problem is that the structure in which to do this is not yet in place – the Internet has not yet become a market in itself, but rather is still being used merely as a promotional outlet for other markets. How the Internet might become its own market will be addressed below.
In order to make full use of the Internet’s potential for accessing a huge audience, it is not enough for companies simply to set up their Web site, put the its address on all their off-line (i.e. `real world’, conventional) literature and advertising, and expect millions of people to visit. The chances are, the only visitors they will have are people who already know of the company and its products, a few people who happen to stumble across their site while net surfing, and those people who would have responded to their off-line advertising anyway and who happen to have Internet access. While this is no reason not to have a Web site, it is likely that a much greater audience could be reached, and hence far more visitors acquired, if there was some means of methodically advertising a Web site within the Web itself – in other words, meta-advertising. At present, one can have links to a site placed in a number of places such as `what’s new/what’s cool’ pages, Internet Yellow Pages, on-line business directories and so on, but none of these do more than make their visitors aware of the existence of the sites they mention – users are not actively encouraged to want to visit them. On-line advertising of other on-line sites, then, needs to consist of more than just a piece of text or an image linking to the site. Advertisements need to be far more widely placed, so that as many users as possible will see them, yet their placement must not annoy or antagonise users. While it is conceivable that the Internet user base might change to be more tolerant of conventional `in-your-face’ advertising if advertisers are determined to apply such methods to the Internet, even though this would seem to take away some of the freedom that current users enjoy, for the time being, and in the future if such a change does not occur, advertisements must make the users want to see them, and then want to interact with them in order to visit the site which they promote.
It seems to me that the two key elements here are the user’s interaction with the advertisement, and the nature of the site they will reach through that interaction. At present the only interaction available on the Web is to point and click, which is not likely to engage the user to any significant degree. If the user were able to interact with an advertisement in an enjoyable and engaging way before choosing to use it as a link, then they will spend time doing just that and, as a result, will view the advertisement for longer. Perhaps they might also be more likely to visit the site it leads to… Thus it becomes desirable to introduce a greater level of, and more varied, interaction to the Web. Of course, when the user gets to an advertised site, they must not be disappointed! If all commercial sites are careful to ensure that they are engaging, entertaining and useful, and not just full of blatant hard-sell advertising (perhaps by using some of the techniques discussed above), then they will not be at odds with the perceived `free’ (in all senses of the word) nature of the Internet, and hence users will not deliberately avoid them just because they are commercial.
Both advertisements and the Web sites they promote might, in the not too distant future, be made more engaging through the use of interactive multimedia techniques.
The SellNet Project explored a number of ways in which multimedia techniques and greater interaction might be used to engage the user. In particular, it addressed questions of how light, shadow and contrast might be used to get the user’s attention, how sound and motion might be used to draw the user’s attention, and how hiding and revealing information might engage the user.
One of the problems with Web pages at present, and arguably with many information-rich software environments, is that they present the user with an entire, usually fairly busy, screen full of information in one go. There is little to attract the user or to guide their eye, to draw their attention to any particular area or item. A common theme throughout many of the simulated Web pages in the SellNet Project is the filtering of information – the showing only of selected items at any one time.
The Gateway page in the project was an experiment into how this filtering of information might be used to initially get the user’s attention. This involved creating a dark and gloomy page, with very little content visible; individual items of information – in this case small iconic images serving as advertisements and links to other pages – would slowly, each in turn and each in a different location on the screen, become slightly visible for a few seconds then fade away. The visible item, therefore, was starkly contrasted with the rest of the page, and hence much more likely to be noticed by the user. Further, moving the cursor over any area of the screen where an item was located, whether currently visible or not, caused that item to appear and become brighter, as if emerging from shadows, until it was fully visible. Thus if the user actively responded to the item which had caught their attention, or if they simply explored the page with the cursor and stopped on a particular item, they were presented with more information about that item. If their interest had been suitably aroused, clicking on the item took them to the page it advertised.
This might be made even more engaging with the addition of audio, by locating sounds (perhaps jingles, theme tunes, or appropriate or catchy effects) associated with individual items in a stereo (or perhaps quadraphonic?) field corresponding to the item’s location on the screen. Further, as well as rolling over objects, users might elicit more information, or be otherwise encouraged to continue their interaction (perhaps by seeing an amusing animation, for example), by `rubbing’ the object in different ways with the cursor, or by dragging it to other areas of the screen, for example.
It is hoped that by feeding the user small morsels of information in this way, and thereby giving the impression that there is a great deal of information for them to find through interacting with the page, the developer can make the user inclined to remain at the page.
In the Box Office page, the example of a video-on-demand site was used to explore some ways in which sound and motion might be used to draw attention to different areas of the page. The screen was divided into quadrants, around a central neutral area, each representing a different genre of film (a real video-on-demand site would undoubtedly have more than four categories of film, but our intention was not to demonstrate a working model of such a site). Within each quadrant, four moving images (in this case very short video clips) were blended into the background image, and played in rotation; each clip lasted only approximately one second – the intention was not to show recognisable film scenes, but to create an impression of movement. That said, however, each group of clips was intended to be representative of a genre, and identifiable as such. At any one time, then, four different video clips would be playing in separate areas of the screen.
This was augmented by the use of audio – a looped sound track was created for each film genre and placed in the area of the stereo field which corresponded to that genre’s location on the screen. Ideally all four soundtracks would have been playing simultaneously at a low volume level, but this intention was compromised by limitations on hard disk access time and the capabilities of MacroMedia Director, such that a generic background soundtrack was used instead, with the individual genre-specific tracks being played as necessary (see below).
The underlying functionality, however, remained essentially the same – the cursor initially located in the central area; if the user’s attention was caught by any of the movies, and they moved the cursor into that area of the screen, then the volume of the associated soundtrack increased, so long as the user continued to move the cursor into that area. It was hoped that the user would do this if the soundtrack attracted their attention, but would move away to other areas of the screen if it did not.
Moving the cursor into a screen quadrant also caused a page of further information (in this case a directory of pay-to-view movies from the relevant genre) gradually to become visible, blending in over the background image. The further into a screen quadrant the user moved, the more this additional information became visible, and the more the soundtrack dominated the page’s audio output, until finally the new page was fully visible and the user was able to interact with it. Moving out of one quadrant and into another before the new page was fully visible caused that page and its soundtrack to disappear, and those associated with the new quadrant to begin appearing. The centre area of the screen acted as a neutral space – moving the cursor into this area effectively returned the user to the main home page, with no secondary pages visible, and only the generic background soundtrack audible.
At the risk of going off at a tangent, the reasoning behind making the second page blend in over the first one, as described above, was to avoid the current situation where one always `goes’ to another page – this has great potential for getting the user lost in cyberspace! The idea was to have the entire Web site present on a single screen, with pages organised as layers which would appear as necessary. Instead of the user `going’ to another page, he or she is presented with a layer of additional content over the top of the home page. Thus the user never has to actually leave the home page in order to view other pages in the site.
The intended result, then, was that the user’s eye would be caught by one of the moving images, they would move the cursor towards it, and be rewarded by hearing a soundtrack and seeing information associated with that area of the page. Ideally sound would also have been used to draw attention into a quadrant, not just to reward it – by having all four soundtracks playing at all times (although quietly), the user would hopefully be drawn not only by the images, but also by whichever soundtrack they found appealing, without having to first move the cursor out of the central, neutral space.
Making a Web page more engaging also increases the chances that the user will stay there for longer. Again avoiding the presentation of a whole screen-full of information in one go, another way to engage the user might be to make him or her reveal elements on an otherwise blank screen by exploring, using the mouse to control alternatives to the cursor, such as a torch beam (as in the `digital junk’ page of the SellNet Project), a movable key-hole, a magnifying glass, or a coin with the page as a virtual scratch-card. Of course, to make such exploration engaging it must either be fun in itself, or must be rewarded in such a way as to make the user want to continue.
The action of exploring itself might be made more engaging again through the use of sound to create an illusion of being in a space. For example, as the user moves around the screen, various sounds might appear to move around the stereo (or ideally quadraphonic) field, depending on the location of whatever is being used to explore the screen. In addition, the sounds and other contents of this illusory space might change over time, thus giving the impression of people or things coming and going. Hence the more the user explores, the more there is to explore and, hopefully, the more time the user will spend at the page…
If the action of exploring is not in itself particularly rewarding, then the user must know that the information they will reveal through their searching is going to be of value or interest – ideally, perhaps, it should be something they know they are going to want once they have found it, but cannot identify until then; in other words they may not be looking for anything specific (if they were, they would want a much quicker way of getting it), but know that, in looking, they will find something they like, need or want (a page’s reputation might become significantly more important in this respect).
Might it also be the case that, having invested some of their time in exploring a page in order to find an item of interest to them, users are then more likely to pay a small fee to download that item in order to fully view, hear or use it? The answer might be `no’ if the user did not enjoy having to search for what they wanted, but what if this exploration was made to be sufficiently engaging that the user did not perceive their time as having been wasted? This might be analogous to those real-world junk stores which one sometimes visits purely for the sake of browsing on the off-chance of coming across something interesting – often the browsing is more fun than the finding…
The key point underlying the above two sections is that simply by adding the interaction of rolling over an area of the screen, and by being able to control the visibility of individual on-screen elements, Web pages can be made significantly more engaging to the user. Incorporating additional graphical elements (such as animated images or movies, alternative cursors, and blending, fading and other ink effects, as in the actual form taken by the pages of the SellNet Project), can enhance this still further.
The use of interactive multimedia in Web pages will, of course, require considerably higher bandwidth than is currently available. One only has to consider how much longer it takes to download Web pages which have been developed for Netscape 1.1, with its increased graphical capabilities – in particular the ability to add a textured background to pages – to realise just how essential this is going to be. However, we can fairly safely assume that bandwidth is going to increase exponentially over the next few years for many reasons, not least of which is the demands of commerce – namely the need to allow more people to download more information more quickly.
In addition to facilitating the use of multimedia techniques such as those described above, the advent of greater bandwidth will also encourage two other developments on the Web which will be of value to on-line commerce. While not impossible at present, greater bandwidth will make it more practical to create a self-contained `digital market’ on the Web. The second possibility is that the Web will become a social environment, rather than a solitary one as it is at present. This will not just benefit Web users, but will also be of value to business in a number of ways, which will be discussed below.
At present, the Internet is still being used primarily to advertise and promote real-world companies and products. Placing an order on-line, using a credit card, has the same effect as if the order were placed by telephone or regular mail – the product is delivered to the purchaser’s real-world address. While this has the advantage of making it easier for consumers to order products from companies in distant parts of the world, where the cost of ordering by long-distance telephone call might be prohibitive, this is not the only way in which the Internet can be used to make money. Once a secure and reliable form of on-line currency has been developed (and DigiCash may be well on the way to being just that), then it will be possible for a self-contained digital market to develop on the Internet – in addition to advertising real world companies and products, Web sites will be able to sell digital information (such as news, stock market prices etc.), resources (such as images, movies, software etc.) and services, in particular, time- and effort-saving agents. This market is self contained since its products never leave the digital domain. There will be no packaging, handling or shipping charges involved and so, theoretically at least, the whole cost of doing business should be considerably reduced, and hence profits increased.
Ideally this market should operate on the high-volume-low-margins principle, with a great many visitors each being charged only a few pence (or better still, fractions of a penny) for each item downloaded or service used. One must bear in mind that “most of the millions of files available via the Internet were offered voluntarily to the public” (Ellsworth and Ellsworth, 1994). To charge too much for information is likely to have one of two consequences – either people will find the same information (if it is available) or equivalent information elsewhere, either for free or at a more acceptable cost, or an individual or group will pay to download desirable information once, and then make it available for free on another site. Either way, the original retailer loses sales.
What sellers of digital information and resources need to consider is that they will not so much be charging for the product as for the time the user will save as a result of buying it from their site. For anyone seeking a particular type of information or resource (for example, research on a particular topic, or images of a particular theme), one of the most frustrating aspects of the Internet, and especially the Web, is having to sift through irrelevant or poor quality information before finding anything of use. Even the best on-line search engines do not provide any real indication of the quality, relevance, or usefulness of the addresses they return – they only reduce the number of sites one has to investigate. Charging for information can probably be justified (and hopefully thereby made acceptable to Internet users) if it reduces or removes the time and effort involved in finding it – if the user knows that all of the information and resources located at a particular site have been carefully selected by someone else then, assuming the user agrees with that person’s taste (again, reputation becomes important here), he or she might be more willing to pay a small fee to download something from that site, rather than spend time searching for it elsewhere.
Where a site contains information or resources which are not available elsewhere on the Internet (for example, the site might have been set up by an artist as a place to display his or her work, or by a programmer as an outlet for his or her software), charging could be used as a partial means of overcoming the problems of copyright infringement in the digital domain. Paying to download an item might give the user the right to use it as they wish, but not to redistribute it, thus the owner receives a small royalty and the purchaser can legitimately use the product. Such products might include encrypted information about their owner as part of their source code, so that if the original owner came across their work being distributed from another site (against their wishes), they would be able to prove it did not originate there. Who exactly they would prove this to is another matter, since the issue of law enforcement on the Internet is a grey area. However, the Internet community is already to a large extent self-policing, as exemplified by the use of flaming against the senders of junk mail, and any owner with a legitimate grievance (such as the site not complying with their request to remove their work) might have considerable success in getting the site boycotted by other users (there already exists an Advertiser’s Blacklist at http://math-www.uni-paderborn.de/~axel/BL/blacklist.html).
The commercial potential of saving Internet users’ time and effort is also likely to lead to the development of new on-line services, and in particular, automated agents. Example of these services and agents, some of which were explored in the SellNet Project, include the netBus, `gateway’ advertising sites, link placers and promotional agents.
When one is looking for Web sites related to a particular topic, one does not always know where to find such sites, or even what relevant sites exist. A search engine would be the usual starting-point but, as has already been mentioned, these have their drawbacks. There is potential, then, for agents which take users to preselected sites, vetted for their quality, relevance to a topic, interest, or other criteria. This is the role of the netBus.
Much like real-world buses, different netBuses would, for a small fee (bearing in mind that the use of search engines could also be charged for), take the user on a tour of various Web sites. They would start from dedicated `bus terminal’ sites, but it would be possible to board a bus at any point along its route, whenever it arrived at a page (there would have to be provision made within the browser to indicate whenever any bus arrived at any Web page); likewise it would be possible to get off the bus at any point if the user wanted to leave the tour and remain at a particular page.
Each bus might be provided, or sponsored, by individuals or companies who would then either profit from the bus fares charged or, in return for providing a free bus service, could include in its route their own sites or other sites in which they had a commercial interest. Thus buses could provide both a service to users, and advertising for their sponsors. Bus routes, and hence bus identity, might be organised by personal interest (for example, `Joe Bloggs’ Favourite Web Sites’ or `The Cosmopolitan Tour of Web Sites Every Woman Must See’ and so on), or by subject matter (for example, universities might provide a tour of relevant Web sites for each of the courses they teach). The providers of bus terminal sites would also profit by charging individual bus providers for using their site. Such terminals would be desirable since they would provide users with somewhere from which they know they will be able to get on a bus, the alternative being to surf around the Web until they come across one by chance. Terminals themselves could be provided or sponsored by companies or other interests; this would probably also benefit users, since it would afford some possibility of giving them an idea of what sort of netBuses they are likely to find at a given terminal – for example, there would probably be a marked difference between those available at a terminal sponsored by The Sun , and those at a terminal provided by The Daily Telegraph (of course, not all Web users will have heard of either newspaper…).
NetBuses might not only provide a convenient starting point for researchers, but also for newcomers to the Web who want to get a feel for what sort of sites exist.
At present, there is no methodical way by which a company can meta-advertise its Web pages elsewhere on the Web beyond having links to it placed in on-line business directories and hoping (or trying to persuade) other people to include links on their own pages. What is needed is a way in which companies can be sure of getting links to their sites placed on as many other sites as possible. The development of an on-line digital market offers a solution.
Among the on-line services it would be possible to charge for would be the provision of space in which to place advertisements and links (I use these terms interchangeably here since either without the other might not be particularly effective, excepting, perhaps, hot-text links within text documents). The cost of maintaining a Web site might be partially offset by devoting part of its space to advertising (not necessarily on a separate page – advertisements relevant to the site’s subject area(s) could be placed in-line at the appropriate points). Entire Web-sites might be devoted to advertising; if each of these were organised in some way, then they would also serve as gateways to other areas of the Web, thus benefiting users by providing another route for research, exploration or surfing – users would know they can go to a gateway related to their interest to find relevant advertisements and links.
By knowing that at least some sites were willing to accept advertising and links in return for a fee, companies could set about promoting their sites in a more methodical manner. This however would take time, since one might have to apply to many thousands of other sites to have a link placed. This might be facilitated by the creation of a new type of agent – a link placer. The advertiser would provide it with information, for example, details of what kinds of pages they wanted to have their link placed on, how much they were willing to pay and so on; the agent would then search the Web for suitable sites and request the placement of its client’s link; if the request was accepted, the client might automatically pay the required fee, or if the asking price was greater than the client’s specified amount, or if some other condition needed to be negotiated, the agent could provide a return e-mail address by which the site administrators could contact the advertiser. If Web sites accepting advertising were also equipped with an agent which could accept or reject requests for placements, then a large part of the advertising process could become completely automated. For example, if a new site containing information about a particular subject came on-line, then in principle, other sites containing related information could automatically add links to the new site. All for a fee, of course…
Another type of agent which might be developed would also facilitate advertising on the Internet. It has already been said that promoting mention of a site in newsgroups and MOOs, and responding to requests for information can be a good way of encouraging visitors to that site. Rather than having to monitor newsgroups and MOOs itself, a company could employ an agent to do this, searching for key phrases in queries and responding with preprogrammed messages. Of course, agents could also be used to ask leading questions, prompting the desired reply from a companion agent. They could also `drop in’ to newsgroups and MOOs, pretending to be real people, and announce the existence and location of a new site, or a special offer at an existing site, or whatever, although this might be in danger of being too much like unsolicited junk-mail, with all the consequences that would entail…
If used sensibly, promotional agents could offer another cheap and efficient means of automated advertising, thus saving the advertiser time and effort and hence money.
The three examples above have shown that there is a considerable scope for new on-line commercial opportunities, which it will be possible to develop once a secure form of digital currency is in operation. Perhaps the real potential for on-line commerce may not lie so much in using the Internet as a giant mail-order catalogue, or as a new vehicle for advertising real-world products, but in treating it as a self-contained digital commodity – by charging for the services, information and resources it can provide, many of which will only come into existence in response to the needs of such a digital market.
Another development which will be facilitated by a significant increase in bandwidth is the merging of the Web and other parts of the Internet, in particular MOOs. This will allow the sorts of social interactions found elsewhere on the
Internet to occur on the Web, also prompting a number of developments which will be of value to on-line commerce.
At present, MOOs are social environments, but the Web is not; the Web is graphical, but MOOs are not. While a number of MOOs do have Web counterparts (for example, The Sprawl and BioMOO) these are just graphical mirrors of their Telnet (or equivalent client software) counterparts – the social functions are still not present on the Web since one must log in to the MOO on via separate client software and keep this window open along with the Web browser window. Ideally, Web browsers should have MOO functionality incorporated into them so that social interaction can take place anywhere on the Web; Ubique’s Sesame client is a step towards this, but the advent of higher bandwidth will allow the Web to go beyond the social possibilities of current MOOs by permitting even greater scope for interactions, projecting a personality into cyberspace, and teaming up with other users.
The merging of the Web and MOO functionality will benefit both Web users, by making the Web a more social environment (users would be able to see and interact with other visitors on any Web page), and MOOers, by providing a graphical user interface and greater interaction than can be achieved in a text-only system (this was investigated throughout the SellNet Project, including in particular, gestural interaction with iconic representations of other users, and how users might form and travel in groups with others). This would also benefit commerce since users would now not only be able to recommend Web sites, but also to take people directly to them, rather than just giving the URL. Thus reputation becomes more important, and word-of-mouth promotion more effective.
Further, the limitation on promotional agents being able to drop into MOOs in order to announce the location of their client’s site (due to the risk of comparison to unsolicited junk mail) might be overcome since such agents could do their advertising on Web sites, which are considerably more public than present MOO environments (an analogy might be the difference between someone making an announcement over a loud-hailer in a town high street, and someone doing the same inside a shop). Any sites preferring not to receive visits from promotional agents ought to be able to deny them access; indeed a commercial spin-off from this might be the development of `conference sites’ – secure sites which can be hired by groups (such as MOOs or companies) in order to have meetings on-line without being disturbed by other visitors, including promotional agents. The pages within these sites would be specifically set up for social interaction, offering more functionality in this respect than would be available from the browser at other Web sites (since limitations on screen space if nothing else would lead to a compromise between the provision for social interaction and the provision for experiencing and interacting with the actual Web pages – the most obvious example being that the display of other visitors to a site necessarily takes screen space away from the display of the site itself).
Increasing interest in on-line shopping, combined with the inability to socialise on the Web at present (ignoring Ubique’s Sesame for the moment) has lead to suggestions that “people [may] no longer use shopping as a means for socialisation” (Bosley, 1995). However, the future addition to Web browsers of functions allowing for social interactions will accommodate “today’s shopping mall reality… there is a whole social factor involved, which is partly characterised by the see and be seen idea” (Ibid). This desire to see and be seen can be capitalised upon to the advantage of both Web users and advertisers through the increased possibilities for projecting a personal identity into cyberspace offered by a graphical user interface.
One of the advantages of the Internet at present is that many of the prejudices which exist in the real world do not exist on-line – “a person’s gender, education, age, ethnic origins, appearance, wealth, or social situation are not readily apparent” (Ellsworth and Ellsworth, 1994) so “people have to judge you on your words rather than on your T-shirt, hair, gender, or body odour. You are what you type” (Hand, 1994). However, one only has to spend a small amount of time in a MOO to discover that Internet users (or at least those who use MOOs) don’t seem to want this sort of anonymity, and will often go to some lengths to provide a description of themselves for other MOOers to see. Perhaps, though, the crucial factor is that they do not have to give a description of their real-world selves, but can instead create a completely fictitious character to act as their on-line presence. Any socialising facilities added to the Web and its browsers should allow for the projection of a personal identity into cyberspace, not just because of the desires of some users, but also because this would be advantageous to on-line commerce.
Higher bandwidth, and the Web’s use of graphical user interfaces, will allow for the graphical representation of visitors by icons, photographs or whatever images users chooses to give themselves. Ultimately on-line video-conferencing might be a standard feature, available to all Web users, although many might prefer not to use it since it would return them to the situation of presenting their real selves into cyberspace, and all the problems of prejudice which may ensue. The projection of personal identity on the Web might be further assisted by the use of `tags’.
Tags are the digital equivalent of branded merchandise, in particular clothing where the brand identity is clearly visible. The fact that some people want to be seen with such products, and will pay premium prices to do so, is perhaps one of the greatest achievements of modern marketing, and there is no reason to think that Web users would not want to do the same on-line. The use of tags would be an effective way of bringing the see and be seen idea to the Web.
Tags would consist of small graphical icons which would be attached to the user’s personal icon in some way – depending on the nature of the browser being used, these might, for example, be visible in a window which opened whenever another visitor chose to take a closer look at someone. Tags might be acquired (either for free or in return for payment) from Web sites by any user wishing to be associated with that site (in other words, a user’s tags would be a way of their saying to the rest of the on-line world `these are my favourite sites’). They could also be used to indicate hobbies, interests, membership of real- or virtual-world societies and so on. A whole on-line market might even be developed around them.
Just as in the real world, where one can wear a different (or no) brand-name on different occasions, Web users might have different sets of tags, and choose which ones, if any, other users could see at any given time, perhaps depending on the nature of their Web visit on that occasion (for example, if someone was doing some serious research, they might present themselves in a different way from when they were surfing for fun). Thus users could control what and how much information about their identity (real or fictitious) they made available to others. They might choose to remain anonymous, or even to be completely invisible to other users.
The use of tags would also make it easier to accurately target advertising. For example, agents might be developed which examined the tags of visitors to a given page, and then invited them to receive advertising which they might find of interest. Of course, it would not be advisable to present visitors with actual advertisements immediately, but individuals (or groups) might be asked if they would like to receive one, perhaps in return for some reward, as discussed above – if they responded positively, the agent would then transmit the advertisement.
The use of tags would also give further impetus to on-line meta-advertising – after all, showing off a designer label only has the desired effect if everyone who sees it knows its significance; “it’s no fun to pay $200 for a pair of shoes, or $40,000 for a car, if your friends haven’t heard of the brand” (Peppers and Rogers, 1993). Thus not only would tags make meta-advertising more important, they would also facilitate it – the more people who `wear’ a given tag, the more the brand is advertised.
The increased possibilities for presenting a persona in cyberspace offered by a graphical user interface, and higher bandwidth, might benefit both users and on-line commerce, in particular by introducing the concept of see and be seen to the Web.
The use of multimedia techniques on the Web, the higher bandwidth this will require, and the other developments discussed above, all call into question the position of markup languages such as HTML or VRML. To cope with the greater levels of interactivity and other multimedia elements, these languages will have to become significantly more flexible and powerful. But once bandwidth is sufficiently high, there is no reason why Web pages could not be constructed from the same programming languages that are used in off-line multimedia software, such as C++. This may even become desirable due to the flexibility offered by such languages – up to a point, the medium would become unimportant, since anything which could be coded on a computer could be used as a Web page. Web browsers would then simply become shells which operated over the top of the individual page’s software. At present, this is being explored in Sun’s HotJava browser.
However, while this approach would allow much greater variety amongst Web pages by doing away with the hypertext transfer protocol, thereby permitting virtually any programmable interaction or presentation to be incorporated, it would also disenfranchise a great many people since pages would now be considerably harder to create. This problem might be overcome with the development of an easy-to-use, object-oriented, multimedia-capable, Web-page authoring tool, ideally at a cost which would put it within reach of all Web users. It is debatable whether such a tool could be considered an extension of HTML or VRML, or whether the advent of higher bandwidth would make such markup languages obsolete.
The potential of the Internet to reach a huge worldwide audience almost inevitably means that on-line commerce will increase. Many advertisers have already `gone on-line’, even if they have not done it particularly well. At present, most commerce on the Internet consists of on-line shopping – using the Internet to pay by credit card for good which are then delivered in the real world. In the near future, a secure form of digital currency will almost certainly be developed; this will permit the creation of a self-contained, on-line digital market. On-line resources and information will be sold; time- and effort-saving agents will be developed and their use charged for. More effective and methodical meta-advertising will also become possible, as will the use of interactive multimedia techniques to increase the user’s engagement in the sites they visit. The increased bandwidth needed for this will be made available. The Web will take on more of the functionality of other parts of the Internet, in particular those of MOOs, in order to allow more social interaction within a single on-line environment. This will be of benefit to users by making Web surfing a less solitary activity, and to commerce by introducing the see and be seen concept, greater scope for word of mouth advertising, and by facilitating more accurate targeting of advertising.
In summary, increased commercialisation of the Internet has the potential to bring greater variety of presentation and interaction the Web, to make finding information easier and quicker, and to make the Web a more social environment, all at minimal cost to the user. This will make the World Wide Web look and feel very different and, I believe, considerably improved, compared to the Web at present.
Magazines and Journals
Internet, Issue 10, September 1995, EMAP Computing, ISSN 1355-6428
Hand, C. 1994, Meet Me In Cyberspace, Computer-Mediated Communication magazine, Volume 1, Number 5, September 1, 1994, page 4.
The Coalition for Networked Information (1994) Electronic Billboards on the Digital Superhighway – A Report of the Working Group on Internet Advertising, http://www.cni.org/projects/advertising/www/adpaper.html