What's Pop, and What's Not?

Have you noticed the proliferation of access statistics icons on your favourite bands' Websites? How do you feel about being told you're visitor number 10870 to the Star Wars hate page? Have you wondered why you don't gain weight from all the cookies you seem to be getting from your Internet music retailer? Did you sign a complete stranger's online guestbook? Are you annoyed with the dozens of pop-up windows that keep asking you to 'RATE THIS SITE!!!'? Don't worry: it's not you, it's them. You're witnessing the symptoms of existential angst. In most media, to be seen, read, or heard is everything. To have an audience, preferably a large and loyal one, is crucial: in the mass media's views, audience size and share determines popularity, and popularity attracts private and/or public funding. 'Popular', for these media, doesn't mean much active intervention from the people (in contrast to the way the word is often used in cultural studies): 'popularity' means a solid base of dedicated and continuous users, preferably larger than that of their competitors. Commercial Websites must similarly justify their setup and running costs by the amount of visitors they attract, but not-for-profit and private Webmasters, too, usually need that knowledge to justify and reward the effort that has gone into the site. A Website without visitors might just as well not exist at all.The problem is that on the Web this form of popularity is almost impossible to determine with any accuracy -- despite the multitude of measuring methods you're likely to be subjected to within just an hour of heavy Web browsing. That's not to say that some major sites on the Web aren't quite obviously major sites: the Amazons, CDnows and Yahoo!s of the Web are clearly visited by thousands, even millions of users each day. But for the majority of medium and minor content providers, the situation is far from clear, especially if the attention is focussed on the relative audience shares between a number of comparable services. These providers have a hard time determining whether they're amongst the leading sites in their field, and whether they're known to and enjoyed by a sufficient share of their target audience. Such difficulties largely are a continuation of similar problems in other media, and so it's worth taking a brief tour through the depths of audience measurement elsewhere. Audience research has become an important industry, but what's often overlooked in the endless battle for better ratings is that those ratings are often quite misleading -- the more so the less material a medium appears. While for culture that is linked to material artefacts (books, CDs, videos, newspapers) some relatively credible circulation, sales and unsold returns figures can usually be obtained (although magazines often multiply these figures by a set number to generate more impressive 'readership' figures), there is no direct-feedback way of gauging how many listeners tune in to a particular radio programme, or watch a certain television show. The amount of 'hits' (to borrow a Web term) to a programme cannot be monitored by a station itself; instead, it relies on peoplemeters placed in a selection of supposedly representative households to log such accesses. Additionally, there is the general question of what consumers do with any product, and whether every access to it can honestly be counted towards its popularity: I may buy the weekend newspaper only for the personal ads, disregarding its editorial content; you may channel-surf across the available TV programmes without really watching any of them attentively -- and alternatively, I may make copies of a CD I've bought for any number of friends; and you may tape a radio programme to listen to (repeatedly, even) at a later time. This real-life context of accesses will usually either escape or confuse peoplemeter devices: they may keep a record of what channels the family TV was tuned in to at any particular time -- but what they cannot record was if a viewer has fallen asleep, turned the sound off while talking on the phone, or gone to the kitchen to fix dinner; or indeed if the VCR is at the same time recording another show. Additionally, it is also highly doubtful that households with peoplemeters accurately represent the viewing habits of the wider population: the anecdote that current affairs shows regularly rate extraordinarily well if they include a story about families with peoplemeters is only an obvious example here. The more diverse the range of situational settings for the consumption of a particular medium, the less likely is it that any sample group of consumers can accurately represent the audience as a whole -- and the more we study consumption contexts, the more individualised they appear, as Ang has pointed out for television: "emphasis on the situational embeddedness of audience practices and experiences inevitably undercuts the search for generalisations" which audience research with its scientific approach engages in (164). Above a certain level of situational diversity such generalisations can only find a lowest common denominator which is trivial and largely useless: a certain size of audience may have been tuned in at one time or another, but for how long or with what degree of satisfaction remains unclear. Recent developments in the mass media have only increased the diversity of access situations, however. First, there is the ongoing expansion in available media channels. Where in Australia there used to be only a handful of television networks, for example, the introduction of pay-TV has added dozens more channels, few of which are available to all viewers; and where there used to be only a few daily newspapers, the rise of carrier media such as the World Wide Web now means that readers can make the New York Times or the Süddeutsche Zeitung rather than the Sydney Morning Herald or, heaven forbid, the Courier-Mail their preferred morning paper, if they so desire. Such developments further underline the point that for example "the boundaries of 'television audience', even in the most simple, one dimensional terms, are impossible to define. Those boundaries are blurred rather than sharply demarcated, precarious rather than absolute" (Ang 154). This raises the general problem of defining the exact boundaries of a media market, and the channels through which this market is accessed by producers and consumers. A cultural product's 'popularity', if expressed in the number of accesses to the product, can only possibly be measured with any degree of accuracy at the bottlenecks through which products must pass into and out of the market: for material goods, this is the distribution process, where the number of products (newspapers, books, CDs, etc.) shipped can be listed against the number of unsold products returned, and circulation figures can be calculated. (Whatever the means of measurement at these bottlenecks, it is clear that the measurement itself must be automatic, and cannot rely on the users themselves: survey-based audience research results are questionable ab initio, since they are drawn only from that part of the audience that is willing to participate, and thus rule out those users which may variously be less active or less interested, or conversely more suspicious or more active -- and thus too busy to fill in a survey.) For less 'material' cultural products, the bottlenecks reside in the equipment needed to send and receive them: radio and TV sets, for example -- but as we have seen, this bottleneck can be bypassed with the help of sound and video recorders, and new media forms such as the Internet, which provide additional access channels to the older media; it is also a bottleneck that is less accessible to researchers than that on the distributors' side. How many peoplemeters are there next to PCs with TV tuner cards? How should accesses to online editions be figured into the circulation numbers of newspapers? Ironically, unlike electronic broadcast media the Internet does appear to offer a way to directly measure audience access to content, of course: as a 'pull' medium which requires the user to request content individually rather than the provider to send programming indiscriminately, such individual accesses (predominantly to Web pages) can be monitored. But for the same reason that peoplemeter statistics are fundamentally inaccurate, so are Web counter data: accesses ('hits') don't equal readers, since Web browsers may jump elsewhere without having read a whole page, and since proxy servers may access a page once, but redistribute that page to any number of clients. Again, the situational context of access cannot be monitored with such relatively simplistic measures -- and it can be argued that the range of diversity for Web access situations is even greater than it is for other electronic mass media; while TV access (with any degree of attention), for example, remains largely in recreational settings, engaged Web access spreads from these to offices, laboratories, libraries, and cafés. Ironically, unlike electronic broadcast media the Internet does appear to offer a way to directly measure audience access to content, of course: as a 'pull' medium which requires the user to request content individually rather than the provider to send programming indiscriminately, such individual accesses (predominantly to Web pages) can be monitored. But for the same reason that peoplemeter statistics are fundamentally inaccurate, so are Web counter data: accesses ('hits') don't equal readers, since Web browsers may jump elsewhere without having read a whole page, and since proxy servers may access a page once, but redistribute that page to any number of clients. Again, the situational context of access cannot be monitored with such relatively simplistic measures -- and it can be argued that the range of diversity for Web access situations is even greater than it is for other electronic mass media; while TV access (with any degree of attention), for example, remains largely in recreational settings, engaged Web access spreads from these to offices, laboratories, libraries, and cafés. Ironically, unlike electronic broadcast media the Internet does appear to offer a way to directly measure audience access to content, of course: as a 'pull' medium which requires the user to request content individually rather than the provider to send programming indiscriminately, such individual accesses (predominantly to Web pages) can be monitored. But for the same reason that peoplemeter statistics are fundamentally inaccurate, so are Web counter data: accesses ('hits') don't equal readers, since Web browsers may jump elsewhere without having read a whole page, and since proxy servers may access a page once, but redistribute that page to any number of clients. Again, the situational context of access cannot be monitored with such relatively simplistic measures -- and it can be argued that the range of diversity for Web access situations is even greater than it is for other electronic mass media; while TV access (with any degree of attention), for example, remains largely in recreational settings, engaged Web access spreads from these to offices, laboratories, libraries, and cafés. Cultural producers can still take some information from their access statistics, of course -- no matter how inaccurate the figures, a thousand hits per day are still better than ten, and while page reloads and browsing durations may indicate technical problems or extraneous distractions just as much as attentive engagement, such data too may be useful to some extent. Web publishers may even try to compare their figures with those of other Websites which they regard as competitors in the field. It has become impossible, though, to claim market and audience shares with any degree of accuracy: when the total size of the audience cannot be determined, no percentages can be calculated; ratings-based systems will fail. This is a major shift especially for the entertainment industry, where ratings battles have become notorious; it is a shift directly related to the unregulated, unlimited nature of the online market, where no limits on the number of competitors exist or can be enforced (a situation markedly different from that in the practically closed TV and radio markets in many countries), and it is a shift which may lead to some deal of paranoia on the part of the established media outlets: on the Web, there is always a danger that upstart competitors could snatch a share of the market (a development, moreover, which wouldn't show early on in any ratings figures). While popularity ratings weren't an exact science at the best of times, then, they are becoming hopelessly inaccurate as media and audiences change -- not just in the case of the Web, but (as we gradually move towards a much-anticipated media convergence) in the case of many others as well. Few media forms will remain unaffected by these developments: as 'pop' music fragments into multitudes of sub-genres, for example, each with their own radio stations (terrestrial as well as online), publications, record labels, CD shops, or even online distribution schemes, does it still make sense to speak of 'popular' music? As we gain access to a global media market with Thai newspapers, Brazilian radio stations, and German TV programmes only a click of the mouse away, is there still a point to local or national ratings figures? Such questions haven't necessarily stopped ratings users from relying on them in the past, of course -- Ang's critique of TV audience ratings was published in 1991, but the ratings appear no less important to TV stations now than they did then. Ang expected this: "television institutions ... are likely to continue the quest for encompassing, objectified constructions of 'television audience' -- as the continued search for the perfect audience measurement technology suggests" (155). For newer media like the Web, though, this troubled experience with audience measurement in television and elsewhere, and the many impracticalities of accurately measuring Web audiences, may serve to tame the desire for similarly "conveniently objectified information" (Ang 152) on audience participational patterns -- information which fails to take note of the context of such participation -- before that desire develops into a TV-style obsession with one's own popularity as expressed through ratings and audience sizes. Indeed, once the novelty of Website access statistics has worn off, perhaps this is where we return to a different conception of 'popularity'. As the mass media splinter into collections of specialty channels, as the audience differentiates into individuals belonging to and moving through any number of interest groups in the course of a single day, with each group gradually gaining access to their own channels, and as many-to-many media give certain people (though not everybody) the ability to communicate without the need to subject themselves to mediation by any existing media institution, perhaps the translation of 'popular' as 'from the people' is once again on the ascendancy. And at the very least, as the ratings' accuracy continues to deteriorate, so will their relevance and importance, and cultural producers may feel less strongly the need to appeal to the lowest common taste denominator. That can't be a bad thing. References Ang, Ien. Desperately Seeking the Audience. London: Routledge, 1991. Citation reference for this article MLA style: Axel Bruns. "What's Pop, and What's Not? Measuring Popularity in the Many-to-Many Age." M/C: A Journal of Media and Culture 2.4 (1999). [your date of access] . Chicago style: Axel Bruns, "What's Pop, and What's Not? Measuring Popularity in the Many-to-Many Age," M/C: A Journal of Media and Culture 2, no. 4 (1999), ([your date of access]). APA style: Axel Bruns. (1999) What's pop, and what's not? Measuring popularity in the many-to-many age. M/C: A Journal of Media and Culture 2(4). ([your date of access]).