How much impact is the Internet really having on advertising and marketing? Is it just another emerging niche medium with some peculiar creative capabilities and constraints? Or might it transform consumer marketing in the same way that network television revolutionized consumer culture and commercial practice four or five decades ago?
Interviews with marketers reveal that few believe the Internet will change their approach to advertising. Most see it as little more than a complement to traditional marketing practices, and don’t expect it to reduce expenditure on broadcast and print media or change the form, pricing, or delivery of advertisements. Their view is probably a reaction to the early hype about the Internet and the World Wide Web, which created unrealistic short-term expectations among marketers and frustration with the inadequacies of the delivery technologies among consumers.
We take a contrary view. We believe that Internet advertising will account for a growing proportion of overall advertising expenditure. Moreover, advertising—and marketing in general—will adopt practices first developed or deployed on the Internet. As the technology improves, the impact of Internet advertising will increase and become easier to measure, and the gap between this new precise, interactive marketing capability and conventional "fuzzy" passive media will widen. Over the next few years, advertising agencies and consumer marketers will be under pressure to change their whole approach to marketing communications.
Marketers will become more accountable for their results, and they will pay more attention to building a total customer relationship. Offering consumers value in return for information will become vital in eliciting their preferences, which in turn will be critical to customizing advertising.* And companies’ entire marketing organizations will be progressively redesigned to reflect interactions with consumers on the Internet.
For ad agencies, fees based on results will become standard. The economics of Internet advertising are likely to make current business models obsolete. New capabilities will be required as creative production speeds up and becomes more closely integrated with marketing activity. A deep understanding of enabling technologies will become a prerequisite for fresh forms of advertising.
Our views on the evolution of Internet advertising and its impact on traditional marketing may seem provocative to some, premature to others. But the intriguing marketing experiments taking place on and off the Internet suggest it is time for consumer marketers to begin looking to networks for new ways of thinking about the marketing theories and approaches on which they have long relied—and to begin capturing the lessons Internet advertising holds for all their advertising practices, online and conventional.
Caution: Changes ahead
Looking at today’s Internet advertising to predict what tomorrow will bring is about as helpful as using a rear-view mirror to watch the road ahead. But a point of view about what online advertising will look like in three to five years’ time can and should influence current management decisions about how to invest marketing communication dollars on the Internet. A number of fundamental forces are currently reshaping Internet advertising: the near-daily emergence of new technologies that improve measurement, targeting, and data interpretation; the strenuous efforts of primarily entrepreneurial marketers to make business use of the Web; and the establishment of patterns in consumers’ use of these new interactive networks. Thanks to the impact of these forces, tomorrow’s ads will differ from today’s in the shape they take, in the metrics available for gauging their effectiveness, and in the pricing structure that governs their purchase and sale.
New shapes
The first and most obvious change in advertising will be in what consumers see on their screens. Ads are likely to change in terms of their content, the type of customization they employ, and their delivery to the consumer.
Content
Aspirations to transcend today’s form of Internet advertising will first be realized in the content of ads. The development of new technologies such as virtual reality and chat, coupled with consumers’ growing preference for material that is directly valuable to them, is driving the emergence of new forms of content. Three main types are on the horizon: experiential, transaction-oriented, and sponsored content.
Virtual reality will make ads even more experiential: customers will feel as though they are test-driving a new car
Experiential content will allow consumers to "experience" the ownership of a product, service, or brand. The best current examples let the user test out a product. Sharp’s Web site offers a personal tour of the Zaurus personal digital assistant in which consumers can input calendar or address information exactly as they would if they used the product in real life. At The Gap’s site, customers can "try on" outfits and mix and match separates from the current range. In the future, technologies such as virtual reality will make ads even more experiential: customers will feel as though they are test-driving a new car, or walking down the aisles of a grocery store.
Transaction-oriented content will invite consumers to make a purchase directly from an ad. Advertising content will become increasingly oriented toward transactions. Indeed, the Internet may already be changing consumers’ buying behavior, particularly for considered purchases such as cars. Prospective car buyers who are looking for product information before making a decision can obtain more information more quickly through the Internet than by any other means currently available. Having done their research in advance, they are more ready to buy at the point when they actually encounter a manufacturer or seller.
The implication for marketers is simple: they need to make it possible for consumers to carry out transactions easily and seamlessly, or risk losing sales to competitors. Consider Casio, which uses Virtual Tag technology developed by First Virtual to enable customers to make purchases from an Internet banner ad. An Internet user can learn about Casio products, purchase a watch on line, and select the means of delivery without ever leaving the banner.
Sponsored content will blur the line between editorial matter and advertising. A lot of sponsored content already exists on the Internet—for example, Nissan sponsors weekly soccer tips on Parent Soup in association with the American Youth Soccer Association—but by and large it tends to resemble the "brought to you by ABC" model familiar from traditional media. The emergence of advanced forms of hybrid commercial-editorial content will be driven by consumers’ ability to "tune out" straightforward commercial messages, be they banners, interstitials (ads that pop up while users wait for a requested Web page to appear), or standard forms of sponsorship, and by advertisers’ desire to influence attitudes in more subtle ways.
By way of analogy, consider the growing use of product placement in films and television (James Bond drives a BMW Z3 in his latest movie) as marketers seek to make their offerings stand out from the clutter of ads and break through the cognitive filters that allow consumers to discount ordinary commercials. The network environment offers ample scope for hybrid content: entire sites can be funded and co-managed by advertisers (as with Procter & Gamble and ParentTime), while avatar technologies* bring advertisers into chat rooms. However, the issue of editorial independence and the possibility of consumer rejection or backlash may ultimately set limits on the pursuit of this approach.
Customization
Anyone who has been offered a credit card they already hold can appreciate the need for greater customization or "addressability" in mass-market advertising, and even in direct mail. Indeed, the level of response that advertisers receive largely depends on the accurate and timely targeting of messages, as do the number of transactions and the degree of loyalty that are generated.
The Internet is supposed to enable marketers at last to target their offers to that elusive "segment of one." Yet advertising on the Internet has so far been targeted mainly on the basis of editorial content, just as it is in traditional media. Part of the reason is technical, though the development of tracking software that allows ads to be delivered only to target audiences is overcoming this obstacle. Consumers’ reticence has been a further barrier, but as Internet users grow more willing to provide information about themselves, two types of customized content will emerge.
First, content will be customized by means of information inferred about users. The Ultramatch technology recently launched by Infoseek, to take one example, makes it possible to target those Web users who are most likely to respond to a given ad. Based on neural networking technology, Ultramatch observes users’ behavior when they put out queries and explore subjects, collecting the results in its database. Advertisers using the service can select individuals according to their interests and thus pitch their campaign to a receptive audience. Ultramatch also allows them to ascertain which individuals are responding to ads, and to move the ads to places where they will attract similar users.
Consumers seem to be willing to release information about themselves as long as they are the prime beneficiaries
Second, ads will be customized on the basis of information voluntarily provided by users. The key to making this approach work will be to overcome consumers’ desire for privacy or anonymity by offering them rewards for personal details in the form of special information, discounts, or promotions. On ParentTime, for example, users who enter the ages of their children receive relevant care information as well as Pampers ads geared to those age groups. Experience suggests that consumers are willing to release information about themselves as long as they are the prime beneficiaries. Organizations such as etrust (an initiative sponsored by leading companies to develop electronic commerce) and the Internet Marketing Council take a similar view. The IMC requires marketers to provide a "giveaway" or discount before they can gain certification. This scheme is specifically designed to prevent information provided by consumers from being misused in e-mail.
Delivery
The recent hype about "push" technology on the Internet might suggest that this will be the dominant vehicle for delivering advertising on the Web. We believe the reality will be more integrated, combining today’s "pull" format Web sites with "push" technology such as PointCast to deliver ads to people according to their interests. Triggered banners (ads that appear when certain key words are mentioned) and interstitials are early examples that point the way. Consider how one automaker’s ads are pushed to chatroom participants when the topic of cars comes up, or how a user waiting for content to be downloaded is sent an ad related to that content. Marketers must ask themselves a number of questions: What is the right balance? Where can push technology be exploited most effectively? How much push are users willing to take before they begin to tune out?
As online advertising develops, advertisers will discover that the Internet is the only medium that can deliver certain types of message, such as multisensory and interactive ads. These new forms will allow advertisers to achieve several objectives—some of them unattainable via conventional media—simultaneously (Exhibit 1). They are likely to make Internet advertising more important in the overall marketing mix as marketers capitalize on their unique capabilities. At the same time, our glimpse of the emerging future casts doubt on the merit of current heavy investments in big brand sites that require content to be "pulled," or in banner ads that—like most on the Internet today—merely replicate the forms of advertising that exist in the physical world.
New metrics
The Internet affords marketers an unprecedented opportunity to measure the effectiveness of their advertising and learn about their viewers. The capacity to measure impact sets the Internet apart from other media. Measurements available for television, for example, estimate the total size of an audience; what they don’t do is tell an advertiser how many people actually saw an ad, or what impact it had. On the Internet, by contrast, marketers are able to track click-throughs, page views, and leads generated in close to real time. The result: measurements that are more precise and meaningful than anything available in traditional media.
The emergence of these new metrics will affect not only ads themselves, but also the way that marketers and agencies develop them. First, more precise measurements will yield better insights into the effectiveness of advertising spend. It will be easier to identify ads that don’t work, and to find out why. Advertisers will also start to expect the content of ads to be renewed more frequently in response to audience reaction. A new product from Infoseek offers a hint of things to come. Copy Testing in a Box is a tool that combines the immediate feedback of the Internet with sophisticated targeting technology to allow marketers to refocus their Internet campaigns to the most responsive customer segments within a matter of days.
Second, advertisers will be able to assess the impact of their ads earlier in the spending cycle. As a result, they will have the flexibility to launch and roll out a campaign in such a way that it can be changed before most of the money is committed. This will affect the very process of creating Internet ads, and perhaps spur advertisers and agencies to devise new ways of organizing around it.
New pricing
Whereas marketers tend to have fairly uniform objectives in traditional media, such as shaping attitudes in television or obtaining responses in direct mail, the Internet, as we have seen, allows them to pursue several different goals simultaneously. In the same way, the standard types of pricing used in traditional media, such as CPM (the cost of exposing a message to a thousand viewers of TV or readers of print), will give way on the Internet to pricing that varies as widely as the objectives of the ads themselves. Indeed, the technology can support several pricing mechanisms at once: pay per click-through, lead, transaction, dollar spend, or conventional CPM. This kind of variegated pricing is already appearing in the marketplace: P&G has pushed for pricing per click-through; CD Now pays Web sites commissions on the transactions they generate; and Destination Florida pays according to leads generated. Similarly, DoubleClick is introducing an advertising network, DoubleClick Direct, whose rates are based on results, and has already signed up clients including Alta Vista and GTE’s Internet service.
Because of these factors, pricing for Internet advertising is likely to be multi-tiered, based on results, and tied to marketers’ objectives. At least three pricing mechanisms will coexist: pricing by exposure, response, and action (Exhibit 2).
Pricing per exposure—for instance, via a rate card based on CPM—will prevail for ads placed on the Internet to generate awareness of a product or brand. Over time, this form of pricing should become more refined. As measurability and metering improve, advertisers will want to pay only for impressions on their target customers, while publishers will eagerly search for ways to extract premium exposure rates. The result is likely to be the establishment of an additional tier of "effective" CPM rates.
Pricing per response will establish itself as the standard for simple consumer responses such as click-through. Prices will vary according to the types of user a site attracts and how much advertisers are willing to pay for access to them.
Pricing per action is similar, but more elaborate. A site publisher might charge an advertiser more for a consumer who downloads a piece of software or provides some demographic information, say, than for one who merely clicks on a banner. We believe that the ability of Web publishers to charge advertisers for the true value they receive is likely to make the difference between profit and loss. The price for a lead generated, for instance, could reflect the prospect’s potential lifetime value; if it did, sites would charge automotive OEMs and white goods manufacturers different prices for prospect leads. As a result, a fee per action or sales commission is likely to emerge as a major pricing mechanism for Internet advertising over time.
Results-based pricing gives marketers the opportunity to shift some of the risk of failure to sites or agencies
How quickly and how far these models take hold in the near term will depend on how risk is shared between marketers, agencies, and sites. Results-based pricing gives marketers the opportunity to shift some of the risk of failure to sites or agencies. Publishers and broadcasters in traditional media have usually been loath to take on this kind of risk. However, Internet publishers should find risk sharing attractive if it is appropriately priced, as it could boost the advertising revenues on which their success depends.
Pricing in general is fraught with issues. Will site publishers demand a degree of control over the creative execution of ads to ensure quality, for instance? We believe that the sharing of risk in Internet advertising will ultimately be determined by the prevailing balance of power, which will vary from advertiser to advertiser and site to site, and shift over time. Large, well-known, "safe" advertisers may be able to secure results-based pricing more easily than others, particularly at times when site publishers are struggling to make their economics work. It will be in the best interests of marketers, site publishers, and even agencies to prevent the lowest common denominator setting the industry’s pricing standard. To settle for a simplistic, unsophisticated, "one size fits all" pricing scheme would mean leaving a lot of money on the table. The widespread acceptance of multi-tiered, performance-based pricing will make the Internet both distinctive and highly lucrative as an advertising medium.
The spillover effect
The changes now taking place in the shape, measurement, and pricing of advertising on the Internet may seem dramatic enough in themselves, but we believe they will have a much broader impact on marketing practices in general. This spillover effect will occur for four reasons.
The greater measurability of Internet advertising will prompt marketers to reevaluate all their investments in media
First, new ways of advertising on line will inspire new creative approaches elsewhere. Second, the Internet will prompt marketers to reevaluate their use of traditional media. Third, Internet advertising will help marketers to improve their understanding of consumers’ needs, preferences, and product usage. Finally, once marketers get a taste for the measurability of Internet ads and the tailored pricing it enables, their expectations of the effectiveness and measurability of other media will rise.
New creative approaches
The timeliness and direct tone of advertising on the Internet will increasingly inspire marketers operating in other media. Seeing the daily updates of information that the Web makes possible and the lengths to which online advertisers must go in order to keep users’ interest (for instance, renewing banners weekly) may sharpen their appetite for replicating Internet practices on TV and in print.
The notion that creative approaches pioneered on the Web will spill over to more traditional media should surprise few. Historically, the emergence of new media has always prompted content changes in existing media. Consider how print changed after radio, and later television, arrived on the scene.
Fidelity Investments recently attempted to mimic the immediacy of the Internet in its television advertising. It refreshed its ads on a daily basis by incorporating current news headlines. However, the campaign met with mixed success, perhaps because it lacked a distinctive point of view.
Marketers’ adoption of creative techniques pioneered on the Internet will grow as technologies like broadband, WebTV, and virtual reality begin to influence traditional media. Wink and Worldgate are developing technologies that allow viewers to "save" a commercial to watch later, or to obtain more detailed information. These technologies are in their early test stages on television.
The enormous creative flexibility offered by the Internet will increase pressure for more choices of delivery in traditional media. The (probably apocryphal) story of Helena Rubenstein asking to buy an extra three seconds for a 30-second spot to realize her creative vision suggests how we may start to question accepted standards and constraints in traditional media.
Marketers may also need to reexamine the theories that underpin their advertising practices. As we noted, online advertisers have found that banners must be renewed frequently if consumers are to keep clicking. Their experience defies the conventional wisdom in advertising that any ad must be seen at least four times to make an impression. On the Internet, greater impact can be achieved by showing a wider range of ads that are repeated less often. Insights like this cast doubt on the effectiveness of current television campaigns, most of which are still based on old ideas of frequency.
Reevaluating media investments
Everyone has heard the advertiser’s lament: "I know 50 percent of my advertising is working; I just don’t know which 50 percent." The greater measurability of Internet advertising will prompt marketers to reevaluate all their investments in media, especially in the addressable categories of print and direct marketing. Not only are response rates often higher in Internet advertising, but the cost of reaching target customers can be lower, with better information received in return. As a result, we may well see a migration of targeted marketing spending from direct mail and other traditional media to the Internet.
Consider a recent example. AT&T used the Internet to generate awareness of and shape attitudes toward its toll-free collect-call service, which is mainly targeted at 16- to 24-year-olds. The company had previously found this audience difficult to reach cost-effectively through print or broadcast media. The results of the online effort were excellent. Top-of-mind awareness increased by over 30 percent, and AT&T opted to replace its print advertising with an Internet campaign.
The traditional approach to customer response and lead generation has been to use ads in trade magazines and customer response or "bingo" cards. However, findings announced by one large publisher of trade titles indicate that more than two-thirds of bingo cards either go unanswered or are not responded to promptly because of the time it takes to qualify and manage leads. The study suggests that the Web is an excellent tool for generating quality leads and may even supersede bingo cards in time.
Migration of this kind will reallocate the slices of the advertising pie. Interviews we conducted with marketers reveal that most believe their initial spending on the Internet did not come at the expense of other media (in other words, their overall advertising budget grew). But many expect that future increases in their Internet expenditure will be taken from other areas, probably print and/or direct marketing. They also see their Internet advertising budgets growing much faster than their traditional media budgets.
Migration may also take place in non-addressable media spending. Striking levels of media displacement are already evident among Internet users. Most notably, TV viewing has declined among a third of adult Internet users (Exhibit 3). Similarly, in a recent Wall Street Journal poll, 21 percent of respondents cited spending more time on their computer or in using online services as a reason for watching the major TV networks less than they did five years earlier. When marketers accept the idea that brand building can be accomplished on line, some spending on TV, radio, billboards, and other non-addressable media may migrate to the Internet.
Getting closer to the consumer
We believe marketers will soon start to use the Internet as a kind of testbed for campaigns planned for print, TV, or radio. One leading-edge marketer, London International, the maker of Durex condoms, is already trying out advertising concepts on its Web site before transferring them to other media where their effectiveness is harder to track. It is testing three concepts ultimately destined for conventional media: "On-line Lovers," "Dr Dilemma," and "The Nurse." By monitoring pages selected, click-throughs, responses generated, and other indicators, the company is able to discover which parts of a prospective campaign work and which don’t, thereby reducing the risk of launching the equivalent of a box-office flop.
Conducting market research and obtaining feedback from consumers can be expensive and difficult. The Internet offers cost-effective alternatives to conventional methods, and may yield more revealing information. Several of the marketers we interviewed said that their presence on the Web had taught them a tremendous amount about their customers’ views of their products and services. They maintain that the Web offers a non-judgmental way of providing feedback and ideas, and is less intimidating for consumers to use than standard toll-free numbers.
Marketers at Fidelity, London International, and Coors found that users of toll-free numbers mainly called to ask questions about products. On the other hand, Internet users, even when given answers to the most frequently asked questions, would often provide feedback about the quality of a product, new variations on it, and ways that it might be changed. To be sure, some of the additional interaction may be down to the different demographic profile of Internet users, but gathering information of this kind is becoming an increasingly important way to use the Web.
To gather deeper feedback, marketers are experimenting with Internet focus groups. LiveWorld has already hosted several sessions for NFO, a company specializing in this area. The advantage of conducting a focus group on line is that participants are anonymous and can speak their mind without worrying what others in the group think. In addition, geographically dis-persed participants can be assembled at a fraction of the usual cost. London International is planning to conduct an online focus group to assess the effectiveness of its Web efforts in the near future.
Finally, the opportunities for testing new product ideas on the Internet are legion, particularly for electronic or intangible items such as magazine covers, entertainment concepts, and personal financial services. The possibilities are just beginning to be exploited.
Rising expectations
Two features of Internet advertising—the measurability of its impact and the probability of some form of results-based pricing emerging—are likely to raise marketers’ expectations of traditional media. If they do, pressure may build for a more accurate measurement system or a shorter measurement cycle. The demand for greater accuracy in measurement is already coming from the broadcast networks in any case. The coding technology tests being carried out by SMART (the emerging competitor to Nielsen), by Nielsen itself, and by its joint effort with Lucent to develop Media TraX indicate that improvements are technically feasible.
In fact, it would not be surprising if new measurement tools and tech-niques originally designed for the Internet were to spill over and be applied to traditional media in the not so distant future. Moreover, in those traditional media that are already more measurable, such as print, we foresee increasing pressure from advertisers for results-based or tiered pricing like that offered on the Internet.
The developments we have described are necessarily speculative, and may not materialize as broadly or as quickly as we suggest. All the same, they are worth watching out for because of their implications. Most of the media industry is affected by the billions of dollars spent every year on consumer marketing. If key advertisers were to reallocate their media budgets, the impact on traditional media could be profound.
As the aspirations, techniques, and expectations associated with Internet advertising spill over into traditional media, both marketers and advertising agencies will have to rethink the capabilities they bring to bear on selling products and services.
Implications for marketers
The growing importance of Internet advertising and its effect on conventional marketing will have profound implications for practitioners. First, the Internet model will set new standards for building relationships in the physical world, challenging many current practices and expectations. Second, a new concept, value exchange, will emerge as a core marketing capability. Finally, the move toward organizational structures and processes designed around consumers’ experiences with specific products or services will accelerate further.
New standards in relationship management
The Internet will set new standards for total relationship management in both breadth and depth. "Breadth" means that a relationship will increasingly last for the entire ownership experience, including the time before and after the purchase of the product or service. Consider Coors, which used consumer feedback received via the Web during both the development and promotion of its beverage Zima—thus involving customers at all stages in the product life cycle.
"Depth" reflects the degree of interaction with consumers at any given point in their experience of a product. The book retailer Amazon.com, for instance, is beginning to use the information it gleans from customers to create value-added services such as suggestions about books that a particular reader might enjoy. This raises the bar for competitors on the Internet and in the physical world, posing a challenge that other players must meet if they are to retain customers’ loyalty.
The Internet’s role in consumer relationship management has important consequences for marketers. Network-based interactions must be integrated into the rest of a business, with all that this entails. If car purchasers make fewer trips to the showroom, say, doing their own online research into different models instead of talking to salespeople, dealers will need to rethink the way they manage the whole consumer relationship. Eventually, customers may go to them only to place an order; at this point, the role dealerships play may no longer justify their cost, and they will have to find new ways to offer buyers value if they are not to disappear. Moreover, as consumers’ behavior changes, so will the skills that salespeople need. And how are those salespeople going to be compensated when consumers make their purchases through channels other than dealerships?
Design and funding is another key area. If the Internet’s role is to grow beyond advertising, the design of online activities should probably not be constrained by the priorities of a single functional area such as marketing, or by the limitations of the marketing communications budget.
Value exchange as a core capability
Much of the Internet’s potential relies on the creation of a dialogue between consumer and marketer in which information is exchanged for value. Marketers need to develop the new skill of rewarding consumers for giving them access to personal information such as who they are, what they like, and what they buy. This reward may take the form of discounts toward future purchases, or benefits such as valuable information or a personalized product or service.
This process of value exchange will become critical as new standards are created to protect consumers’ privacy. The proposal announced by Netscape in May 1997 to capture information on consumers’ hard drives rather than on marketers’ computers marks a step in a new direction with its implicit acknowledgement that consumers will "own" information about themselves and control the release of that information to marketers. The demand for value among consumers is likely to grow as they become aware of how highly marketers prize their demographic profiles, product preferences, and transaction histories.
A few marketers are beginning to manage this process effectively. In exchange for basic information such as name, address, age, and income, Vogue provides readers with discounts, special offers, and previews of forthcoming articles. Saturn’s approach is to offer convenient access to information. Consumers who reveal a small amount of information about themselves are able to use Saturn’s interactive pricing center to research new cars, saving them trips to a showroom.
Organizations centered on consumers
As the Web merges marketing with other business processes such as customer service, it will put more pressure on the organization of most marketers. The coming of age of interactive networks will accelerate the move toward new organizational models in which marketers will structure their various functional capabilities around an integrated customer front end.
For a real-life example, take the insurance company USAA. Its customer center receives and manages all communications with consumers, whether direct via telephone, mail, and the Internet, or indirect via intermediaries. The rest of the organization revolves around the customer center. Sophisticated information systems help the company to process interactions and maximize their value.
The benefits are many. Customers feel that USAA knows them better, and the company is quick to respond to a complaint or learn about important market changes such as a cut in a competitor’s price in a particular territory.
As more and more companies reorganize themselves around their customers, intranets linked to the Internet will become crucial. They will make it economically feasible for managers within an organization to have more information about consumers—and more interactions with them—than ever before.
Implications for agencies
The rise of Internet advertising, with its unique economics, may well call the validity of current business models and processes into question. It will also compel agencies to rethink the way they create and develop campaigns, and the skills and capabilities they need to survive.
New business model
So different are the revenues generated by conventional and Internet advertisements that traditional agencies will have to think carefully about their approach to online advertising if they are to pursue it profitably. At present, most agencies incur high fixed costs in developing campaigns. Big creative teams and the like were fine in the days when agencies could rely on the commissions they earned from large media buys associated with a small number of creative executions. On the Internet, however, this cost structure is inverted: the creative element of the total advertising cost is much larger in relation to the media element. The resulting commissions will no longer be sufficient to cover agencies’ high operating costs.
We believe that traditional agency business models simply will not work for Internet advertising. A trend toward retainer compensation is already emerging. Agencies may well seek to enhance their revenue streams by taking a cut of the results of their efforts in the shape of a commission on leads or sales generated. In future, agencies will increasingly share in the risk of their advertising instead of—as they do today—leaving all of it to be borne by marketers.
Compensation models will be transformed. The measurability of Internet advertising makes results-based pricing more feasible than in any other media, as we have seen. Some examples are already in evidence. Site Specific is using performance-based contracts for clients including Duracell, CUC International, and Intuit’s TurboTax division. Though these arrangements are not yet making it any money, they are expected to do so as advertising effectiveness increases. In time, results-based compensation will probably spill over into traditional media as the measurement of advertising impact improves. It will then have its most profound impact, affecting agencies’ core business and revenue source.
New capabilities
This vision of the future calls agencies’ current capabilities into question. Many have seen themselves as the guardian angel of the brands they represent. But agencies have a patchy record of orchestrating brand-building activities across the full range of marketing disciplines: media advertising, direct mail, promotions, and so on. The emergence of interactive media means that agencies must not only manage a broader and more complex mix of marketing tools, but also master radically different skills. Three main gaps will need to be filled:
In summary, the future holds many challenges for agencies. The emergence of new business models and the need for new capabilities are likely to shake up an industry that has been under pressure for some time. Some agencies have shown that they can customize their processes and economics to specific industry needs like those of grocery retailers or auto dealers. Now they must learn to institutionalize these capabilities within their organizations or spin off a cluster of flexible, technology-savvy boutiques with low fixed costs. Viewed another way, the emergence of Internet advertising may represent an opportunity for renewal—a chance for agencies to reclaim the high ground of brand stewardship that some marketers argue they have let slip away in the past two decades.
The emergence of Internet advertising is likely to have wider implications for business than many imagine. Its effects will not be confined to the online world, but will extend to traditional marketing activities and processes. For those who look closely, Internet advertising holds many more opportunities and risks than is commonly assumed. And the payoff waiting for those who rise to the challenge will more than justify the efforts required. 
About the Authors
Caroline Cartellieri and Varsha Rao are consultants, Andrew Parsons is a director, and Michael Zeisser is a principal in McKinsey’s New York office.
We would like to acknowledge the contributions of Eric Simonson, Inder Soni, and John Leibovitz to this article.
Notes