Business models of news Web sites

This paper is included in the First Monday Special Issue: Commercial Applications of the Internet, published in July 2006. For author reflections on this paper, visit the Special Issue.

Business models of news Web sites: A survey of empirical trends and expert opinion by Frederick Schiff
The examination here yields eight business models that describe more or less comprehensively the commercial approaches to online news. Using the models, I review the secondary literature to see what has worked, what trends have emerged, and what the experts expect. Approximately 300 articles were reviewed, based on a search of those stories that appeared in the business and trade press in hardcopy and online in the past five years. Surprisingly, a consensus opinion exists among experts who predict that the interactive model offers the most promise of eventually capturing the majority of online news consumers. My intent is to develop the models to facilitate further empirical study of which news websites are most likely to survive and succeed in the dot-com shakeout beyond spring 2000.

Contents

Introduction
Business models
Empirical trends
Discussion and conclusion

Introduction

If the first decade of the 21st century will see the endgame for global media dominance as many analysts predict, what are the business models that will give one or a few media firms a definitive competitive advantage?

Many authors have evaluated and critiqued news Web sites and speculated about their future. A typology of eight business models is presented here as a basis to empirically assess news Web sites. The models are derived from a review of the secondary literature covering the online industry found in approximately 300 business press stories and scholarly articles from 1995 to date. A surprising consensus seems to have developed among analysts, consultants, experts and academics about the comparative advantages of the online medium and possible strategies that make for success.

The differences between the models can be summarized in terms of three cross-cutting characteristics: (A) features that differentiate the online medium from print, broadcast and cable media; (B) key variables or components that affect business operations; and, (C) the maximizing or optimizing behavior that guides management strategy and measures their performance. Business models are developed as "ideal types" (Weber, 1946) to represent the differences among management teams seeking to maximize or optimize selected factors that affect their online news operations, and to gain competitive advantage from their market position and from product features of the online medium.

Starting in the 1920s in the largest U.S. corporations, the central administration spun off a separate stratum of middle managers, staff professionals and technical personnel (McDermott, 1991). Policy and financial decision making remained under the exclusive and direct control of senior corporate executives while mid-level functionaries were given more autonomy, authority and responsibility for day-to-day operations. Executives and owners usually combine and prioritize features of one or more business models in a given firm or industrial sector. Firm-level analysis is useful in the empirical study of business behavior, but the examination of business models is typically a history of sector-wide management fads, generational investment waves and country-specific industry wisdom. Joseph Schumpeter (1992) described business cycles as 11-year periods of technological investment when the capitalist winds of "creative destruction" sweep away businesses in mature industries and allow businesses with new product substitutes to succeed.

In this article, the online news models are defined, and each model is described in terms of empirical trends cited in the secondary literature. The models are offered as a heuristic way to conceptualize the actual alternative strategies so far undertaken in the news sector of the online industry.

Business models

There are eight models that I will consider in this paper. They are:

Advertising revenue;

Online traffic;

Infant industry profits and stock values;

Digital content delivery;

Continuous breaking news;

Information retrieval and storage;

Portal conduit; and,

Interactive networking.

Table 1: Business model characteristics.

A. Performance outputs

1. Advertising revenue

(A) Medium Features: Deliver audiences to advertisers (B) Key Variables/Components: Advertising model: Price and demographic positioning. Category killers: Niche Web sites for classifieds, especially jobs, housing and automobiles (C) Optimize revenue side

2. Online traffic

(A) Medium Features: Unified medium of communication (B) Key Variables/Components: End user model: Uses & gratifications; wants & needs for cognitive orientation, need to know, societal early warning, normative watchdog, pattern maintenance; subscriptions, user fees, commissions, placements, registration. (C) Maximize general-interest readership, or optimize upscale readers through niche product consumption of news, entertainment, information and interaction

3. Infant industry profits and stock values

(A) Medium Features: Declining production/distribution costs with size; high risks; wide value fluctuations (B) Key Variables/Components: Financial model: Capital accumulation and reinvestment; volatility of revenue, profits, equity values and risks; monopoly profits; equity speculation through stock market price, market capitalization, IPO cashout, liquidation price (C) Optimize owners’ equity, return on investment, earnings growth, monopoly profits and stock market price appreciation (but not so far dividends)

B. Input content characteristics

4. Digital content delivery

(A) Medium Features: Digital platforms; origination push (B) Key Variables/Components: Translate content and offer unique content (C) Optimize expense side: Product synergy, cross-promotion and low-cost distribution

5. Continuous breaking news

(A) Medium Features: Immediate, ubiquitous, continuous coverage (B) Key Variables/Components: Traffic model: 24-7 ratings; miniaturized handheld access (C) Maximize updated content for the most repeat visits

6. Information retrieval and storage

(A) Medium Features: Low-cost storage; searchable databases; non-linear hypertext (B) Key Variables/Components: Story-specific links to media sources, research archives and libraries for, any date (not hot news) from in-house, chain/network, syndicated, partnered, competing sources (C) Optimize access and rewriting of texts and images from any historical period, culture or geographic region

7. Portal conduit

(A) Medium Features: Brandnamed oligopoly central site; end user pull via wired, wireless or satellite networks; broadband capacity (B) Key Variables/Components: Marketing model: Largest general audience (C) Optimize links to organizations and communities outside the media

8. Interactive networking

(A) Medium Features: Customizing links to live events, personalized services, banking, shopping, games, social movement activism, townhall meetings, instant messaging (B) Key Variables/Components: Transaction model: Free or user fees; micro commissions on product exchanges or time usage (C) Optimize communities of interest and tunnels/blinders of selective exposure

Essentially, the first three business models focus on "output" measures of success for online news sites and the next five models focus on "input" variables. Senior executives rely on one or another of the models to set the objectives and incentives for middle–level managers and to measure their own overall corporate success.

Empirical trends

1. Advertising revenue

(A) Medium features: Deliver audiences to advertisers
(B) Key variables/components: Advertising model: Price and demographic positioning. Category killers: Niche Web sites for classifieds, especially jobs, housing and automobiles
(C) Optimize revenue side.

Since the 1830s with the introduction of stream-powered presses in the United States and Britain, newspapers cut subscription and single-copy prices, and began to compete at below the cost of their production. Publishers who invested in the high cost, high technology of the period were able to radically reduce their per-unit production costs. Furthermore, the cost of producing each extra unit of the product decreased with the increase in the number of units produced. Like other "natural" monopoly industries, "marginal" costs decreased in inverse proportion to the "scale" of production.

After 170 years of below-cost competition, advertising allowed 98 percent of surviving hardcopy newspapers to become print monopolies in their local markets and to achieve the highest operating profits of any manufacturing industry. Newspapers survived because they were able to differentially attract local advertising. Picard (1989) pointed out that once the market share between the last two remaining dailies in a metropolitian circulation area reached a 60:40 split, the smaller advertising outlet faced an almost inevitable downward spiral.

By the mid-1990s, delivering digitized content through the Internet allowed online newspapers to radically reduce the operating costs of duplication and distribution. As for capital investments, a nationally representative sample of publishers gave me figures for front-end computer systems that were less than one tenth the cost of news presses of similar sized papers.

Meanwhile, hardcopy circulation revenue typically only covers distribution costs. Advertising pays for the cost of production (mostly newsprint and labor) and for any profits. Advertising constituted 82.2 percent of revenue on average for daily newspapers, according to the Newspaper Association of America (NAA) (2001a). Among the three broad types of advertising, 36, 33 and 13 percent of total newspaper revenues were accounted for by display, classified and national ads, respectively (NAA, 2001b). By comparison, broadcast television and radio were almost 100 percent dependent on advertising revenue.

Cable TV most closely resembles the Internet revenue profile, drawing 25 percent of their revenue from advertising and 75 percent from subscriptions according to Mackie-Mason and Metzler (2001). They reported that the Internet was 22 percent advertising supported while 78 percent comes from subscriptions. However, the revenue splits between advertising and subscription differ depending on service bundles that a firm offers. Mackie-Mason and Metzler distinguish four "service layers": Conduit services, ISP access, portal aggregation and content. At the extremes, conduit-only services were 100 percent subscriber-based while online content and aggregation services usually depend 100 percent on advertisers. Firms offering only access sevices differed widely; for example, the percentage split between advertising and subscriber revenue for Earthlink was 4:96. So, a company supplying ISP access and portal aggregation services seems to gain more subscription revenue while having content seems to open up advertising revenue.

Table 2 illustrates the effect of service bundles. Time Warner relied on being a cable conduit to generate 100 percent subscriber revenue, so its being a content provider seemed only to enhance consumer subscriptions. Prodigy, which added portal aggregation to ISP access, gained 13 percent from advertising revenue. American Online (before the merger with Time Warner) offered ISP access, portal aggregation and content, and ad revenue accounted for 31 percent of the total. The exception seems to be Juno, which only offered access and yet obtained 37 percent of its revenue from advertising.

Mackie-Mason and Metzler say advertising and subscriptions were separate and distinct consumer markets. In terms of bundling services and making decisions about vertical integration, they caution that layer boundaries only sometimes define markets.

Table 2: Revenue sources for different service layers
Source: Jeffrey Mackie-Mason and John Metzler, 2001. "Internet Media: What Are the Markets?" at http://www-personal.umich.edu/~jmm/presentations/industrialorganization/tpug-new-orleans-01.pdf, accessed 15 June 2002.

Service layers Conduit Access Aggregation Content

Media firms Telco Earthlink Lycos, Yahoo! CNET, MSNBC

Advertising 0% 4% 100% 100%

Subscriptions 100% 96% 0% 0%

Table 3: Revenue sources for different service layer bundles

Service layers Conduit & content Access & aggregation Access, aggregation, & content Access

Media firms Time Warner Prodigy AOL Juno

Advertising 0% 13% 31% 37%

Subscriptions 100% 87% 69% 63%

The NAA's Market & Business Analysis Department (2001b) finds that total expenditures on newspaper classified advertising grew 35.7 percent from 1995 to 2000 or almost six percent per year. In a representative national sample of 122 newspapers before the downturn of 2000, most publishers told me they saw online advertising as a supplement to hardcopy classifieds and were reasonably confident about the competitive position of newspapers to continue to dominate advertising.

Ad revenue, trends and market share

Total Internet advertising revenues grew to $2.9 billion in 2000 from $1.9 billion in 1999, a 52.6 percent increase (Digital Edge, 2001). Nevertheless, the basic ad revenue model — selling advertising Web space or Web features on the basis of the size of the Web audience or its characteristics — has been difficult.

A Forrester Research (2000) report predicts that ad revenue won't generate more than half of total revenue, and so it proposes more direct affiliation with e-commerce. The non-news authors don't seem to anticipate how that might compromise the editorial product.

Despite the growth of online advertising, Paul Farhi (2000) argued that revenue has been minimal. He blamed the weakness of advertising buttons, the glut of banners and e-mailed spam, and low "clickthrough" rates. Perhaps the reason was that computer users were sophisticated media consumers who were innoculated with the saturation advertising in broadcast and print media. Computer users may find it easy to avoid the "white noise" of Web sites cluttered with ads. Like remote TV controls and pre-set radio buttons, which consumers routinely use to block out advertising, no-spam filters have become customer service features offered by ISPs or purchased and installed on a PC to counteract e-mail spam, intrusive pop-up ads and no-exit sites with multiple screens triggered when an offensive ad site was closed.

The average duration of an online ad is three weeks or less, according to a study by AdRelevance (Magid in Presstime, 2000a), suggesting that online stickyness and shelf life may be lower than in other media. High exposure and pass-along rates seem to be driven by unique content or functionality of Web sites.

Besides user resistence to a new medium, Quinn (2000) offered two generalizations that together pose an even starker online advertising dilemma: Because online audiences are fragmented and dispersed, local and niche advertisers are not interested, and at the same time because online users are still disproportionately upper-middle income, national advertisers of common-place products are not interested.

Some evidence seems to contradict Quinn's premises.

Regarding income and Internet access: Internet access increases with increased household income, according to a U.S. Department of Commerce survey in 2001. In all major ethnic groups, households with more than $35,000 in income have more than 50 percent Internet access. Ethnic differences in access decrease among higher income families though Asian Americans consistently have the highest Internet access. Differences between rural, urban and central city households were minor in all income groups.

Approximately 54 percent of the total U.S. population have Internet access (Nua Internet Surveys, 2002). Out of 145 million American people with Internet access, 12 million were new in 2001, according to the story posted on Nua's Web site by owner Scope Communications Group of Dublin, Ireland.

"The digital divide seems to be narrowing in the U.S. Internet use among households earning less than USD $15,000 per annum increased by 25 percent a year between December 1999 and September 2001, while the rate of growth among households earning USD $75,000 or more was just 11 percent."

The division of the advertising market of the past 45 years may be changing. In that period, newspapers captured the largest share of local advertising and broadcast television receiving the largest share of national advertising. Since the mid-1990s, the NAA has sought industry-wide standardization through its Classified Advertising Standards Task Force initiative. Newspapers' share of national advertising grew 80 percent to $7.65 billion in 2000 from $4.25 billion in 1995. So, as TV viewership was falling off, newspapers were gaining a share of the national advertising base that TV traditionally dominated. Online newspapers have enhanced the ability of hardcopy newspapers to increase their share of national advertising as readers increasingly come from outside the local circulation area. Far up-market, the New York Times has attracted national advertisers and grown total national circulation even though it lost 13 percent of New York City subscribers between 1990 and 1999 (Shaw, 2001a).

As for local advertising, the portal site run by Cox Interactive Media fits the online growth pattern. "[D]ot-com advertising constituted less than a third of CIM's revenue by the end of 2000. Most of CIM advertisers are conventional local accounts, including automotive dealers and realtors," said President Peter Winter (Focus, 2001).

Ad packages and partnerships

To boost online advertising, newspapers were promoting themselves on radio and television and creating advertising package buys (Sullivan, 1999). Advertising consultants were often behind the move to become a local portal. Newspapers are selling buttons and advertising text as well as banner ads. The new development was bundling ads on portal sites that are jointly branded or even owned by competing media companies. In the Los Angeles metro area, Dean Singleton's Media News local newspaper group has partnered with Donrey and other newpapers companies. In south Florida, competing papers owned by Cox and Knight Ridder have paired up with dominant local television stations. To become the major local portal, media firms are joining together to offer content and share advertising revenue.

Meanwhile, the media themselves are advertising increasingly online to drive traffic, budgetting $156 million in 2002 (Myers Reports, Inc., 2002). Television (combining local, syndicated and network as well as broadcast and cable TV) spent 35.9 percent of total online media advertising compared to 27.1 percent by newspapers.

Ad placements and customized services

Some marketers have switched to ad placements, pop-up ads and Web sites dedicated to niche classifieds (Farhi, 2000). For example, GeoCities and TheGlobe.com, two of the early leaders among online portal sites, hosted personal home pages on their sites and arranged to feed headlines and media content for free in exchange for placing hyperlinked advertisements (Outing, 1999). GeoCities placed a banner ad on the top of every branded content page and keep the revenue, and the syndicator iSyndicate sold a side panel ad slot. According to Outing, content providers expected the feeds will act as gateways or entry points to drive traffic back to the media source.

Niche category killers: Help wanted, real estate

According to a survey done by American Opinion Research (Casale, 2000), nearly three fourths of newspapers were losing classified revenue to online niche competitors in 2000. In response to these niche advertising Web sites like monster.com, name.biggest.autos.com and name.biggest.apt/home guide.com, NAA newspapers have created separate classified advertising sites for help wanteds, autos and real estate. Newspapers are invested in or linked to national or regional niche advertising sites like CareerPath.com, PowerAdz.com, AdOne and Classified Ventures (Boynton, 2000). Visits to real estate Web sites, for instance, increased 75 percent from summer 1999 to summer 2000, according to NPD Group Inc. and Media Metrix (Rabasca, 2000).

Free-and-fee model

Partially free, partially for a fee, the online news Web site may yet succeed as a customized information delivery vehicle in a consumer pull market. The mixed choice model might offer both loss-leading, free online content for a general-news mass audience and profitable fee-based niche content for a few customers. So far, marginal fees have not contributed enough to total revenue to sustain general-news content production.

2. Online traffic

(A) Medium features: Unified medium of communication
(B) Key variables/components: End user model: Uses & gratifications; wants & needs for cognitive orientation, need to know, societal early warning, normative watchdog, pattern maintenance; subscriptions, user fees, commissions, placements, registration
(C) Maximize general-interest readership, or optimize upscale readers through niche product consumption of news, entertainment, information and interaction.

Subscription revenue

Online news sites initially charged end users and visitors for access to information and entertainment. For example, Knight Ridder's leading edge San Jose (Calif.) Mercury News and Hearst's San Antonio Express-News both dropped charges in 1998 after about two years. In three years, the Mercury Center made 25 to 30 percent of its revenue from charges (Buel, 1998). Many newspapers discovered that if they charged or if they held back a large part of their print edition, few users visited their Web sites. Most newspapers have stopped charging for online subscriptions though many still require online registration and only make archives available at a fee.

Even the most prestigious papers have discovered that general news was too widely available to impose subscription fees. Eric Meyer, associate professor of journalism at the University of Illinois at Urbana-Champaign, told Editor & Publisher's Noack (1999c) that research by his consulting firm, Newslink Inc., showed that only a fraction of online readers are willing to pay about half the hardcopy subscription price for access and that, even with the reduced cost of distribution, e-copy subscriptions are insufficient to cover the cost of production.

Online circulation revenue might be considered a form of business-to-customer sales like product sales by brick-and-mortar retail stores. Online advertising revenue is equivalent to business-to-business sales. The newspaper model of a dual market is an attractive option, but then a subscription online news site would have to double as an online advertising site and sell its audience to the advertisers. Online retail revenues were expected to reach $24 billion in 2000, based on a Jupiter Media Metrix survey in September of U.S. Internet shopping. Although the transactional model seems to work for e-commerce products, the exchange model of fee-for-services for news has met resistance. An April 2001 survey of Internet users by Lyra Research found that, out of an estimated Internet population of 104 million, about 19 percent of users pay for content (only 1.9 percent for specialized or premium news sites compared to a high of 8.3 percent for adult content). James Fallows (1999) writes, "The Net's [free] news model won't last forever either. The current 'business model' was for hard-working organizations to spend good money providing insights, analyses, news flashes, occasional enlightenment and then give all away for free."

Subscription revenue for business and financial news has found a niche online, for example, for specialty sites with exclusive proprietary information, like WSJ.com (Wall Street Journal) and ESPN SportZone. Brooke Shelby Biggs, fellow at the Berkman Center for Internet & Society at Harvard, told Noack (1999c) that more information can be delivered more quickly online compared to print or even broadcast. A June 2000 survey by the Pew Research Center for the People & the Press confirms that active investors — those trading stocks within the past six months — use the Internet as their main source for stock market updates and advice. Meanwhile, the general public still relies on newspapers and television and the bulk of all investors use the three media about equally.

So, for general-interest online news sites, subscription revenue rarely constitutes a major income stream. Nevertheless, in June 2000 after the launch of the subscription-fee business Web sites netimperative.com, insider.com and the451.com, Ian Hargreaves wrote, "At the moment, the pendulum is swinging back towards payment."

Audience size and share

Building a mass audience of tens of millions of end users is valuable even without circulation revenue. At the level of corporate strategy, consolidating a general-interest online audience is a primary management objective for the leading competitors. For instance, the merger of Tribune Co. and Times Mirror Inc. in March 2000 meant that Tribune Interactive had access to 34 million unique visitors (Williams, 2000). Reportedly, Harry Chandler, executive vice president of GoTo.com and son of Los Angeles Times ex-publisher Otis Chandler, was behind the merger, the removal of then-publisher Mark Willes and the strategic move to online media. Tribune Interactive's president, Jeff Scherb, said that, as broadband becomes more accessible, his group will be able to deliver a wider range of content from newspapers, radio and TV stations, and niched Web sites than more specialized media competitors. Tribune Co. newspapers will have the dominant market share position in the top three U.S. urban markets.

Among major media in the top 50 media markets, daily newspapers reached 52.6 percent of adults compared to 37.8 percent for prime-time television, 12.0 percent for prime-time cable and 23.4 percent for morning drive radio, according to the Newspaper Association of America (2001a). Earlier cross-media studies by Scarborough Research Corp. (NAA, 1999; Philadelphia News Interactive, 1998) reported similar findings of audience retention using a Competitive Media Index that measures total readership rather than net paid circulation. The trend in the size of online audience from 1999 to 2000 to 2001 shows a 50 percent year-over-year growth rate, according proprietary data for the September home and work panels released by MediaMatrix to the author.

Newspaper vs. broadcast Web sites

The NAA (2001a) claimed that online newspapers attract more visitors than any other local Web sites in 60 percent of the top 81 U.S. markets (based on its Media Audit survey). Newspaper sites received six times the number of monthly page views on average than local TV sites (Magid in NAA, 2001b) and local online news sites have twice the percentage of loyal customers as national online news sites (using cPulse Internet Satisfaction Monitor). In 67 of those markets, the newspaper Web sites attracted more than 10 percent of the market's adult population (Digital Edge, 2001). Online newspaper portals also out-performed the leading online city guides with 6.8 million visitors per month compared to 1.4 million for Digital City and 1.1 million for CitySearch.com, according to the NAA (2001b) citing NFO Communications for third and fourth quarter 2000, using population projections in 17 combined markets. In 1999, two online newspapers ranked in the top 10 of online sites (Stone, 1999a, referring to Media Metrix data). WashingtonPost.com was No. 8 and Cox's AccessAtlanta.com was No. 10 behind Yahoo!, AOL, MSN, GeoCities and Lycos.

In comparing media usage, online traffic growth coincided most closely with the rapid decline in television viewership. A Paine Webber index (Presstime, 2000b) showed that traffic to U.S. newspaper portals increased 30 percent from June 1999 to June 2000. Meanwhile, from 1996 to 2000, television viewership fell 7.5 percent versus 2.6 percent for newspaper reach. A Pew Research Center survey (1998b) of 3,002 adults was even more negative for broadcast news: Prime-time nightly network and local television news fell to 15 percent in 1998 (from 30 percent in 1993). By comparison, the long-term decline in newspaper circulation remains gradual. Average weekday newspaper readership fell to 55.1 percent in 2000 (from 80.8 percent of adult population in 1964). The decline represents a 25.7 percentage point drop in 37 years or about 0.7 percent per year. The Lyra Research survey reported that some Internet users say they have decreased use of other media. Of Internet users who reported both substantial decreases or increases in traditional media usage, 12 percent said they watched less television compared to five percent who read the newspaper less (while 17 percent reported less use of magazines and three percent less use of radio).

Online readers may even expand the reach of newspapers to people who do not read hardcopy news. News Web sites seem to be differentially taking away TV viewers and to a lesser extent attracting visitors who would otherwise not pay attention to current events (Pew, 2000). Online "sticking" time seems to have cut more heavily into the TV audience than into newspaper readership. Of heavy news consumers, online news readers are more interested in hardcopy news than broadcast news (Pew, 1998a). Heavy Web users are also heavy newspaper and news magazine readers. However, because the heaviest traffic to online news sites is from work during the weekdays, the net effect of online news may be to heighten interest in evening news broadcasts and the weekend Sunday paper.

Online demographics and uses

The online audience has grown increasingly younger, more educated and more affluent (Pew, 1998c). The more experience users have with the Internet, the more often they go online. The 50 percent of users who have four years or more of online experience go online five or more times a day, which is more often than net newcomers. A study by Media Life Magazine (Digital Edge, 2001) found that more than 23 percent of the U.S. population logs onto the Internet more than once a day. The average time spent per person online varied from 14.9 hours to 17.5 hours per month in the fourth quarter of 2000. Significantly, regular Internet news users constitute 53 percent of all U.S. Internet users, according to a 1998 Internet survey conducted by Market Facts Inc., headquartered in Arlington Heights, Ill. (MSNBC in Noack, 1999c).

As a reason for going online, personal and special interests (like weather, stock market prices and sports scores) took first place with 79 percent, according to a panel of 1,112 Internet users over 18 years old by Lyra Research (2001). News tied with entertainment and fun for second place with 63 percent, and work-related information motivated 42 percent of visitors. However, users spent only about 14 percent of their time online for news compared to 14 percent for work-related information, 30 percent for entertainment and 42 percent for personal needs.

Database marketing

In a "dual market" (Picard, 1989) — where media outlets sell or offer content to users and, in turn, sell the audience to advertisers — digital news sites have a distinct advantage over other media in accumulating detailed, household-addressed data on each audience member. By April 2000, the New York Times had created a direct marketing database from its 11.4 million registered non-paying readers. Big-name newspapers, like the New York Times on the Web and the Chicago Tribune, require online registration for access to unique news content and premium features, such as comics and crossword puzzles (Stone, 1999b). Owen Youngman, director of the Tribune's interactive media, told Stone that the enhanced content is a fair trade for private registration information. Registration adds valuable demographic details to the media parent's ability to use databased marketing to become a regional portal, according to Stone. Unlike unsolicited junk mail, mandatory registration and captured e-commerce records offer advertisers more targeted use of e-mail spam.

3. Infant industry profits and stock values

(A) Medium features: Declining production/distribution costs with size; high risks; wide value fluctuations
(B) Key variables/components: Financial model: Capital accumulation and reinvestment; volatility of revenue, profits, equity values and risks; monopoly profits; equity speculation through stock market price, market capitalization, IPO cashout, liquidation price
(C) Optimize owners’ equity, return on investment, earnings growth, monopoly profits and stock market price appreciation (but not so far dividends).

Typically in an infant industry, a few competitors are the only suppliers and consumer acceptance is uncertain. As a result, profits and risks are usually high, and equity values are volatile. First-in firms tend to capture the technology patents and know-how, use their market power to erect barriers to entry, and seize the dominant market share. According to value-based management, stock price appreciation supposedly depends on profit in the longer term. But according to newspaper publishers I interviewed, investors and parent firms vary widely in their tolerance for startup costs, level of endebtedness, cost allocation methods and minimal threshhold profit expectations. Early profitability may not be a predictor of long-term profitability, and yet, a long startup period (beyond three to five years from launch to the break-even point) is a barrier that threatens the survival of online news sites.

Bichler and Loebbecke (2000) suggest that industries engaged in online delivered content are inherently monopoly markets because the incremental costs of production (and distribution) are near zero. The largest firms can expand their market almost indefinitely at no cost for each extra unit produced and impose enormous promotional startup costs for smaller competitors to gain an audience.

Bichler and Loebbecke favor market self-regulation with firms imposing preferential pricing based on the market power of business and residential customers. They also recommend classic oligopoly pricing mechanisms based on product differentiation, which means content providers would offer personalized Web products, up/down market versions of Web products and features, and administered pricing using filters and consumer surveillance technology to set optimal maximum prices for each demographic group. If Bichler and Loebbecke are correct, the early stage of "perfect market" competition among many startup news content providers may already be over.

In the long term, as the market increasingly approximates a global oligopoly of cross-media conglomerates, the same firms may extract monopoly profits by restricting widespread access to the cheap or free substitutes that are currently available. The problem in the short term may be that monopoly pricing may only apply when and where audience segments have limited access (e.g., in remote locations) to the close product substitutes offered by print and broadcast news firms.

In mid-December 1999, newspaper industry stocks seemed to peak. The NASDAQ had risen to over 5,000 from under 1,000 when the Internet began to be widely used in 1995. By March and April 2000, the dot-coms in the technology sector had lost more than a third of their market capitalization (Business Week Online, 2000).

In 1999, Tribune Interactive Group, Belo Online Inc., Times Co. Digital and Washingtonpost.Newsweek Interactive were preparing to cash in on initial public offerings (Stone, 1999c). CBS MarketWatch and TheStreet.com launched earlier in 1999. But by late 2000, three leading online media groups (Tribune Interactive, AOL Time Warner's CNN and N.Y. Times Digital) abandoned plans to spin off their online properties as tracking stocks. The economic downturn had affected advertising lineage. In early 2001, Gannett, Knight Ridder and New York Times Co. acknowledged the sudden fall in advertising and revenue.

According to Stone (1999c), the New York Times estimated 1999 earnings at between $24 million and $26 million on almost 50 Web sites while the Tribune with 200 Web sites projected earnings at $25 million to $30 million. In retrospect, site traffic has not grown with 87.8 million pages views recorded in September 2001 compared to 86 million in April 1999, and the average time on site per person per month declined to 24.2 minutes from 41.7 minutes (Nielsen/NetRatings, 2002).

Prior to March 2000, the stock market value of dot-com companies was based on demonstrated rapid growth in user traffic, not in revenue or profitability. Even without profits, the expectation of profits allowed many startups to raise investment capital or sell their niche brands and know-how. The end of the market cycle and the fallout from corporate scandals seem to have altered the infant industry assumptions about the dot.coms.

Costs of production

Analysts are beginning to question the assumption that online is a low-cost medium. Farhi (2000) argued that the March 2000 debacle in hi-tech stocks demonstrated, "[D]elivering information via the Internet is expensive." Farhi pointed to high upfront capital expenditures for hardware and software development and to advertising and marketing expenses needed to drive traffic. Farhi said, "Creating content requires manpower, which means labor costs." He said firms that expected low-cost distribution and took a shotgun approach got hurt. His example was APBnews.com, a crime news niche site which supported 140 professional staffers and 130 stringers, invested $27 million and spent $11 million in promotions before its closure. Farhi argued that the slowdown confirmed that the print-media model won't work online.

Staff size and costs

Online advertising revenue appears to be significant enough to support a larger, more experienced Internet staff, according to Brooke Shelby Biggs, fellow at Berkman Center for Internet & Society at Harvard (Noack, 1999b), but most publishers told me that they have closely limited the growth of online staff. Sizes vary widely, but typically, the online news staff consists of only three or four people.

Singer, Tharp, and Haruta (1999) found "mixed support" for trade press claims that online newspaper journalists make significantly higher salaries than their print colleagues or are drawing away experienced personnel. In fact, online and print staffers received comparable benefits. At more than half the papers, online editors shared staff with print although larger papers were more likely to have separate print and online staffs. The survey found that top online editors tended to be less experienced, and to be lower in the paper's hierarchy or removed from the direct line of editorial command. The researchers concluded, "[S]urvey results indicate staff sizes are too small to adequately support a quality online product."

Profits

American Opinion Research (Casale, 2000) reported that while online classifieds generated revenue at 77 percent of papers, only 24 percent of newspaper executives said they were making a profit, and 17 percent said online services produced no revenue. Singer, Tharp, and Haruta (1999) said senior editors at 184 newspapers with online editions claimed their Web products were making money in 1998 but were significantly below double-digit profit margins for hardcopy editions.

Shaw and Hofmeister (2000) reported that Tribune Co. has repeatedly produced profits of 25 percent to 30 percent above the industry average. They said that Times Mirror Inc. was consistently below average with profit margins falling from a 1986 high of almost 22 percent and circulation falling almost 20 percent from its 1991 high. TMI stock prices oscillated widely under chairman and CEO Mark Willes (1995 to 2000) and his hand-picked publisher-successor Kathryn M. Downing from $23.25 when he took over in June 1995 to a high $72.63 in fall 1999 to $47.94 when the buyout was announced in March 2000.

The Wall Street Journal Interactive Edition posted its first monthly profit in September 1999 (Nicholson, 1999). Tom Baker, vice president and general manager of Wall Street Journal Interactive Edition, said advertising was slightly more than 50 percent of revenue with 330,000 subscribers (Nicholson, 1999). One-third are hybrid readers who get the hardcopy and electronic versions of the Journal or its sister magazine Barron’s. The 50:50 split in advertising and subscription revenue is closer to the magazine norm than to the traditional 80:20 newspaper split. New daily subscriptions doubled in the second half of 1999 to more than 1,000 on many weekdays as a result of a seven-fold increase in direct mail and e-mail discount promotions. He projected 500,000 subscribers by 2001 and eventually matching the 1.8 million hardcopy circulation. In fact by August 2000, the main site had grown to 460,000 subscribers and had produced $31 million in gross revenue (Fine, 2000).

Pre-tax profits for 11 publicly held newspaper companies declined to 17.6 percent in the first quarter of 2001 down from 22.6 percent for the same period in 2000, according to Morton Research Inc. estimates reported in the Los Angeles Times (Shaw, 2001b). The companies that were hurt the least were Gannett (ranked No. 1 in total revenue) and Media General (ranked No. 11). Dow Jones Print Publishing, the second most profitable newspaper group with 30.4 percent in the first quarter of 2001, suffered the most severe drop to 6.8 percent.

Selloffs

Infant industries are characterized by the rapid and repeated mergers and acquisitions as entrepreneurs cash out on innovations, patents and early brandname dominance. Major Canadian media corporations, like Hollinger International and Thomson Corp., sold off newspaper properties to achieve liquidity to develop online companies (Wilson-Smith, 2000). Wilson-Smith reported that Thomson Corp., which was the second largest Canadian chain with 54 dailies and 75 non-dailies in 1999, was selling all its papers except Toronto's The Globe and Mail and starting up a live online business news service. Meanwhile, Conrad Black, who owned the largest Canadian newspaper chain, sold when the multiples for newspapers were high in 2000 and bought equity in fledgling online firms after tech stocks cratered.

Thomson Corp. formed a $2.7 billion multi-media company linking Canada's leading private broadcaster (CTV), the leading interactive portal (The Globe and Mail's globeinteractive.com), the telecommunications company BCE Inc., and Woodbridge Co. Ltd. (Rabasca, 2000).

Firms in mature industries may also go through a wave of mergers and acquisitions at the end of a technogical investment cycle. Product substitution and regulatory regime changes affect the pace of consolidation, but the infant industry model of capital accumulation remains the same: To create monopoly markets, to cash in on accumulated value or retained earnings, to breakup and selloff assets, to slash costs, to pay off debt and to write down "stranded" investments. In March 2000, the Tribune arranged to purchase the Times Mirror Inc. for $5.8 billion. The Tribune Co. offered the highest premium ever paid for a newspaper company, almost double the TMI stock price (Shaw and Hofmeister, 2000). The deal was a stock swap for the Chandler family that owned 66 percent of TMI voting shares. In addition, the Tribune paid out $2.66 billion in cash and assumed $1.4 billion in debt. The newly combined firm became the third largest newspaper company with 11 dailies — three in the three largest cities. Its 22 television stations, four radio stations, cable operations and satellite coverage enable it to reach more than 80 percent of U.S. television households.

In short, the so-called new economy and online dot-coms are more than magazine-cover trendy. The remaining consolidated content news providers are creating a phase change to an oligopoly market with a quantum leap in the size of the mass audience to a global scale. The double dip in the world’s two biggest economies in the ’90s and early ’00s (even without depression) has helped the acquisitive eventual survivors by lowering stock values. The accumulation of windfall profits in worldwide near-monopoly markets are sure to follow.

So far, I have outlined three "output" business models for online news sites. The next five models focus on "input" business models.

4. Digital content delivery

(A) Medium features: Digital platforms; origination push
(B) Key variables/components: Translate content and offer unique content
(C) Optimize expense side: Product synergy, cross-promotion and low-cost distribution.

Online communication is still primarily text-based and reader oriented. Single medium, text-based providers, with relatively low distribution costs, possess an advantage with regard to delivery of content to a worldwide mass audience. Digital display and bandwidth have not yet caught up with the capacity for automated, customized narrowcasting, so long-form content is not as valuable as on-demand, brief news-you-can-use. This phase delay is especially true in the areas of entertainment listings (82 percent), science and health (64 percent), chatrooms-forums-bulletin boards (61 percent), technology (60 percent), and personal finance (48 percent) (Pew, 1998b).

However, as broadband reaches a price point where it becomes economical for the upper middle class, multi-media content providers will begin to compete head-on with text-based sites for audience attention. Synergy between cross-media products will become more important, and those providers that concentrate on core content products that can be translated across media subsidiaries will be favored. Cost savings are anticipated in information gathering and editing as professionals supply content and find synergies across all partnered media outlets and are increasingly freed to focus on production within their own medium. Savings from the elimination of redundant same-city reporters may be minimal, but the costs of printing, broadcasting and/or disseminating the news are expected to fall significantly as the marginal cost of reaching an ever wider Internet audience is expected to be nearly zero.
In fact, local partnerships increasingly pair dominant newspapers and television stations on shared Web sites. For example, Philadelphia Online added real-time video to its breaking coverage from "Inquirer News Tonight," a program it launched in 1994 on the Tribune Co.-owned WPHL-TV station on the Warner Brothers network (New Media Index, 1997). Newspaper companies like Tribune Co. and Media General Inc. are ahead of television networks in creating hybrid multi-media newsroom operations.

Tribune executives explained that the strategy behind convergent news delivery was to reaggregate the audience. "Multiple distribution channels in the same market enable media companies to do at least three things better: share and enrich content, cross-promote brands and cross-sell advertising," said John W. Madigan, chairman, CEO and president of Tribune Co (PRNewswire, 2001). Three Tribune publishers told the author that the company's primary strategy was content translation across its news producing properties.

Robert S. Boynton (2000), a journalist and sometime professor at New York University, argued that the synergy between old and new media is on balance positive, in that, it extends the brand and reach of old media, especially for newspapers. He mentions three reasons: Newspapers have moved aggressively into online classifieds and gained market share in national advertising. Newspaper-driven city guides, which could substitute for in-print entertainment tabs, have done better than those produced by local television stations and have driven Microsoft's Sidewalk out of business. Online banner and button ads have been so ineffective as advertising vehicles that so far they have not hurt display advertising in newspapers.

Original Content

Original content was still rare in 1999 (Noack, 1999a, quoting Eric Meyer, associate professor of journalism, University of Illinois. at Urbana-Champaign). Most online news content came from print editions (Noack, 1999a, cites Adam Clayton Powell, vice president of technology and programs at The Freedom Forum in Arlington, Va.). Companies like Knight Ridder created little original content, concentrating instead on the accurate transfer of information from hardcopy products. Reportedly, however, the least used newspaper Web sites were those that simply posted content online (Dotinga, 2000).

In its critical review of online strategies used by major European newspapers, Forrester Research (2000) is an example how media consultants tend to create management fads. The report dismissed content translation from the core news product and demanded that content be refreshed throughout the day.

Lomangino (2000) offers yet another critique of content translation in his survey of the most popular Web portals. He criticizes the sites for offering headline briefs, for repackaging syndicated stories and for posting few in-depth stories. Noack (1999a, quoting Eric Meyer again) claims automated shovelware menus of inserted story headlines are commonplace, as are early "bulldog" versions of spot news stories and syndicated boilerplate from niche business, sports or weather sources.

Analysts like McHugh (2000) disagree. He said online publishing was complicated and online reading was fundamentally different from scanning a newspaper or magazine. In fact, consumers may not want long-form in-depth stories, magazine articles, investigative reports or political-philosophical commentary delivered online. They may prefer headlines, breaking news flashes, weather bulletins, stock market alerts, lists of to-do and to-buy items, and difficult to access reports about science, health, technolog and business. This possibility has led other analysts (Quinn, 2000) to suggest that moving from general online news to multiple niche online news would generate bigger audiences for news sites. In the future, online content may consist mostly of breaking news and database access.

Brandname extension

Brand extensions across media have proven successful. Successful online-only, stand-alone Web sites like American Online, Yahoo! and ebay.com are gigantic but rare (Fahri, 2000). Old media brandnames and "free" in-house promotions have driven traffic to news Web sites affiliated with www.CNN.com, www.MSNBC.com, www.Yahoo.com, www.nytimes.com and www.usatoday.com. In September 2001, the unique home audience for these five were among top 20 most popular Web sites with 14.1, 12.0, 9.6, 4.6 and 2.9 million vistors, respectively (Nielsen/NetRatings, 2002).

Brandname extension to capture an infant industry or niche market is a venerable strategy. Clearly, in 1919, Westinghouse, AT&T and General Electric were behind the Radio Corporation of America consortium. Clearly, the three radio networks extended their control of broadcasting into the new medium of television for another 40 years. Clearly, established film formulas, sequels and imitations work. The biggest single obstacle is that, without promotion, Web sites can remain unknown in random-access cyberspace. But commercial promotion is not the only model. Napster demonstrated how quickly a Web site can grow at a geometric rate through word of mouth. In a competitive multi-channel cyber-market of news, ideas and cultural products, intrusive promotional "push" communication may work much less effectively. Brandnamed news content providers have been less able to marginalize free high-quality alternative news sources by controlling the distribution channels and by gaining privileged home-page placements. Certainly, targeted promotions and shot-gun advertising work, but in the decentered "pull" of cyberspace, the primary filtering network may be "personal influence" (Katz and Lazarsfeld, 1955).

5. Continuous breaking news

(A) Medium features: Immediate, ubiquitous, continuous coverage
(B) Key variables/components: Traffic model: 24-7 ratings; miniaturized handheld access
(C) Maximize updated content for the most repeat visits.

Another unique medium feature of online news is its potential to offer immediate, ubiquitous, saturation coverage. The key variable in the model is generating high traffic 24-7 ratings by continuously updating content for a maximum of repeat visits.

The medium itself eliminates a fixed deadline schedule because news can be digitally assembled, edited and, most importantly, delivered at any time. The latest continuous updated information is available on the wire-agency model of 24-7 coverage. The difference is that digital news providers are retail wire services — direct to end users. Ironically, online news sites become even more dependent on news agency wholesalers and features syndicates to fill the expanded news hole. The continuous production cycle eliminates a key advantage of broadcast over print. Print and broadcast outlets both operate on fixed schedules, but broadcast outlets break into regular entertainment programming to offer saturation coverage. However, they do so only for exceptional events. So, online news is potentially more immediate than news delivered by radio, television, newspapers, magazines or documentary film. Wireless hand-held computers will probably become almost as ubiquitous as portable radios. Permanent saturation coverage could be available on any subject or topic through niche Web sites. Digital news means the loss of the editorial prioritization of important news. For long-form investigative exposes, the multiplicity and anonymity of digital sources means the loss of an "authoritative" editorial filter. Do-it-yourself video vérité and participant documentaries may improve on the televised quality of news as reality-based non-fiction. Yet self–serving claims and secondhand gossip may degrade and crowd out news as a product based on professional standards — for example — on norms about verifying facts, finding primary sources, seeking out expert interpretation, and allowing for a variety and balance of points of view.

As to the increased competition between newspaper and television news, William S. Morris III, chairman and chief executive officer of Morris Communications Corp. told publishers at the Texas Daily Newspaper Association meeting in San Antonio in March 2000, "The Internet gives us two things we've always wanted but never had: unlimited space and immediacy. ... It can give us a better, more timely, more flexible attchment to our customers and our readers than we could ever achieve in print alone."

The number of online visitors from work surpassed the number of online visitors from home, according to national panel data for September 2000 and September 2001 supplied to the author by Nielsen/NetRatings. Partly, employers with high-speed online access are unintentionally subsidizing higher news consumption. Partly, the medium may not yet be available at a practical price for fast SDL or cable connections for most consumers in their homes. Partly, the digitally divided audience may still be rather narrowly limited to professional-managerial-technical personnel who have computers in their offices. Mid-week and mid-day is "high surf" prime-time for online news consumers.

The Washington Post started a PM Extra online edition (available at 1 p.m. EST) in 1999 aimed at office workers who surf the Internet through fast company connections on company time. The all-day updates and continuous deadlines may not be as crucial as having at least one afternoon edition.

News outlets may increasingly syndicate and distribute news content to businesses Web sites, acting more like on-demand wire agencies for non-news sites rather than bringing readers to a news-intensive home site. Just before the March 2001 merger with Tribune Co, the Los Angeles Times Syndicate New Media reported that 80 percent of its 200 Web site clients were non-media, non-newspapers (Astor, 2000).

Charles Piller (2000), a business staff writer for the Los Angeles Times, argues that major news events recurrently bump up online readership and that the primary beneficiaries are online newspaper sites. He said traditional media with print and broadcast news brand recognition benefit from online spikes more than Internet-only news sites. After each spike, traffic to online news sites falls back but at a higher level, augmenting the upward trend. His examples are the release of the Starr reported on the Monica Lewinsk scandal in 1998 and the 2000 post-election controversy. His analogy was the boost that CNN got from the Gulf War crisis in 1991.

Piller contends that Web sites backed by a local newspaper have advantages over other media firms in covering complex controversial events. Their staffs have the depth to offer complex, data-intensive analysis and graphic subtlety. They can offer a range of local angles on a national story and a more comprehensive view. Piller reflects the industry consensus about what online news users want with the qualitification that to keep them "sites must add or update stories and graphics continuously."

If Piller was correct about inter-media competition, the workday terrorist attacks on Tuesday, September 11, should have shown the advantages of Web site accessibility from work compared to television. In the first four hours of after the 8:46 EDT event, student monitors in my classes found four of 15 major wired Web sites were overwhelmed and inaccessible and only three were regularly updating the story from their own reporting staff, whereas broadcast and cablecast news outlets offered unlimited and simultaneous access. The performance of most of the leading online news organizations did not match their promise of continuous breaking coverage.

6. Information retrieval and storage

(A) Medium features: Low-cost storage; searchable databases; non-linear hypertext
(B) Key variables/components: Story-specific links to media sources, research archives and libraries for, any date (but not hot news) from in-house, chain/network, syndicated, partnered and competing sources
(C) Optimize access and rewriting of texts and images from any historical period, culture or geographic region.

Advances in computer technology have allowed digital storage costs to continue to diminish geometrically. Online sites can offer more extensive and elaborate retrieval capability of archived information. Web sites are quickly moving toward a future in which searchable access will be available to databases and archives of every cultural product ever produced, as well as to databases of in-depth historical and comparative knowledge. Hyperlinked access to information from any reference is driving a consumer "pull" technology, leading to more consumer–driven, customized, ratings–based content production and programming. According to Jim Schulte, editor of USAToday.com in 1995, the Web has evolved from information delivery to information retrieval (Noack, 1998).

The ease with which information can be retrieved from a site ranks highly among criteria by which users rate a Web site. In fact, 20 percent of users leave a Web site because they are dissatisfied with the retrieval–related capabilities of the site such as poor content organization and slowness in searching for articles (Presstime, 2000b, reported a Digital Edge pop-up survey of 123 Web sites by cPulse for first and second quarters of 2000). Presumably, other inhibiting factors such as archival access fees or registration requirements might act to deter users from visiting Web sites. Most sites do charge for access to archived stories although some, like the Boston Globe at Boston.com/freearchives, have experimented with offering such access for free. Active web users are looking for one–stop aggregated information, such as at Yahoo!, with access in–house to credible news sources (ranked No. 3 nationally and No. 1 internationally). However, in the digital environment, the capacity to share information often conflicts with corporate interests in maintaining exclusive content control. Since 1998, newspapers that dominate the cooperative that owns the Associated Press have pressured it to drop its agreements to supply news content directly to competing online–only Web sites and city portals.

Fallows (1999) notes the ever increasing supply of links to historical background, descriptive details and primary documents that give readers supplemental material. Noting that long-form stories online (more than 2,000 words) are rare, Fallows speculates that neither the Web nor television will replace print in "the role of developing news ideas about public life." He argues that serious discussions that explain, rebut, criticize and react will remain confined to print.

7. Portal conduit

(A) Medium features: Brandnamed oligopoly central site; end user pull via wired, wireless or satellite networks; broadband capacity
(B) Key variables/components: Marketing model: Largest general audience
(C) Optimize links to organizations and communities outside the media.

A news site that seeks to become a portal conduit for a city does so to attract the largest possible audience by organizing general–interest information and making it accessible. In a sense, the Yellow Pages have offered a similar type of general–interest local information and taken a major share of locally generated advertising revenue for years. Essentially, the portal site is a strategy to extend the brandname of the news organization and make it the dominant local information and advertising source in the area. A news organization also builds on its identity with the local community as an interactive means to answer useful–information and how–to questions through lists, databases, links and hypertext.

Paradoxically, providing links to sources of information offered by competitors and rivals may be as important as providing on–site access to syndicators, database archives, content providers and primary sources. Traffic depends not on unique content but on repeat visits, and winning sites are those that are more convenient, user friendly, better organized, and more visually appealing. Fallows (1999) said that the key online advantage is the frequency of updates throughout the day that draws repeat visits, especially during the workday from fast office connections.

Unfortunately, while open access adds to user value, corporations are often only interested in bundling one–stop in–house brands. Filters and tunnels can be placed imperceptibly on information to drive flows and restrict networks (Lawrence and Giles, 1999; Willey, 2000). E–businesses and special interests pay commissions to a browser to be part of an exclusive virtual mall or to flood an on–site search engine with first–listed keyword copies. Organizations and points of view may be excluded by secret contracts or exclusionary site affiliation fees. Software manufacturers, Internet service and access providers, database wholesalers, search engines, national backbone officials, Homeland Security police, corporate and household hardware owners can all censor and selectively control access to information. Central "push" technology, bundled products, secretly imbedded ID codes, required registration, cookies, automatic formating, decision–making by artificial intelligence, all limit portal access to perceptual tunnels.

As dial-up wired connections give way to broadband capacity, dedicated DSL cables, microwave wireless telephony and direct satellite networks, user demand is likely to increase. As the scale of the mass market drops the price point for cell phones, Palm Pilots and e–books, miniature computers are likely to make portal access as well as interactivity and continuous saturation news coverage ubiquitous features of daily life. End users might prefer universal access to "pull" from all potential information or product sources, but a portal site by definition monopolizes generic information access. Willey (2000) said, "[S]earch engines can scan approximately three million servers and 800 million indexable Web pages." In a disorganized cyberspace crowded with unverified information and product sources, portals compete with search engines and ISPs as the go–to sites whose success depends on anticipating and marketing what appear to be the widest range of alternatives. The highest payoff may go to sites claiming the widest possible access but actually offering a more manageable subset of the entire Internet.

Portal sites optimize links to organizations and communities outside the news site itself. Many news organizations have begun to partner themselves with competing organizations, sharing content in order to spread costs and newsgathering ability, as well as to cross-promote. The (Norfolk, Va.) Virginian-Pilot in a deal with an all-news cable channel owned by Cox Communications was the first daily to partner with a rival local TV station (New Media Index, 1997). Since then, New York Times Digital agreed to supply content to Yahoo!; the Internet Broadcasting system partnered with CBS affiliates in Los Angeles and Minneapolis to create local content; and CNN.com agreed to feed video clips to newspapers such as Hearst-owned Houston Chronicle and the N.Y. Times-owned Boston Globe (New Media Index, 1997). In 2000, there were deals between the Washington Post and MSNBC, and between the New York Times and ABC News.

Within six months of the Tribune's acquisition of Times Mirror Inc., the number three newspaper company arranged a partnership deal with No. 2 Knight Ridder to deliver syndicated features and photographic content online (Associate Press, 2000). TMI's NewsCom and Knight Ridder's PressLink formed the joint venture to draw from more than 50 providers to serve 8,000 domestic and international media customers. Forrester Research (2000) recommends extending the brandname to online portals as another way to drive traffic. In terms of comparative advantage, Forrester says syndicating unique content and selling access to proprietary databases will be increasingly more important.

Jon Katz (Fahri, 2000) called for "open" sites rather than for a "closed" media model. He favored the portal model of Yahoo! and CNET.com over the top-down news-you-need-to-know William Safire model. Web sites should point users to information and services they tell you they want, like shopping, bill paying, stock tracking and search functions. By allowing users to go elsewhere, the open site encourages them to make it their "home base" and to come back repeatedly for searching.

Yahoo! Inc. reached 51 percent of the worldwide total Internet population and was the leading Web destination for the 16th consecutive month in September 2001, according a survey by Nielsen/NetRatings (Yahoo! Media Relations, 2001). A survey of 225,000 users in 29 countries showed that average home users spent 1 hour and 12 minutes per month on Yahoo! However, within the United States, Yahoo! news Web sites ranked No. 3 behind CNN and MSNBC (Nielsen/NetRatings, 2002).

Newspaper sites are ideally suited for Web portal hosting because of their size, local expertise and relationships to the community. Compared to Web sites owned by local television stations, newspaper Web sites and city guide sites have more users, raise more money and have more employees, according to a study by Frank N. Magid Associates Inc. (Presstime, 2000a). At the same time, larger newspapers (over 20,000 circulation) often have news coverage partnerships with competing media organizations, thus increasing their breadth and depth of reach. Pew Research Center said 69 percent of such newspapers had news partnerships with local cable TV, 34 percent with radio and 21 percent with public TV stations, according to a national survey in early 2001 (Campaign Study Group, 2001). Conversely, non-newspaper media organizations often have trouble filling out community sites, especially at the local level. In October 1999, Microsoft discontinued Sidewalk.com because it did not have enough local content to compete with entertainment sections in hardcopy newspapers. In its subsequent partnership, MSNBC has relied on NBC to provide local content.

In September 1999, Knight Ridder New Media launched a nationwide network of regional portals (Phipps, 1999). RealCities.com is a national homepage that overlays regional online sites. Dan Flinnigan, president of Knight Ridder New Media, told Phipps, "National reach is a great thing to have, but the game is going to be won locally." Flinnigan reportedly said that being first on the scene and marketing cross-media to become the "go-local button" were the keys to market dominance. Knight Ridder's partners include Belo and Central Newspapers Inc. In April of 2000, KnightRidder.com agreed to a partnership with E.W. Scripps Co. to add four news Web sites to Real Cities, giving its network 36 regional hubs (San Jose Mercury News, 2000), and by July 2002, it operated in 54 cities.

Burl Osborne, who recently retired as publisher of the Dallas Morning News and president of Belo's publishing division, told Phipps, "We joined Real Cities in Dallas (www.DFW.com) because we think it offers the kind of national support that local media companies need to succeed in their own markets. National power means to us national ad sales, [the] portal functionality we need, stronger positioning in the interactive industry, and the traffic benefits of the network effect." Kight-Ridder's Flinnigan argued that without a full array of services, newspaper Web sites were not enough. He described the deal as a turnkey solution with a menu of products, including national and local news, local directories, search functions, free e-mail, community publishing, e-commerce and one-time registration.

A number of other media groups have also jumped into the city guide arena. In March 2001, after three years of online presence, Cox Interactive Media decided to close or sell all but one of its national Web sites and to focus on city portals (Focus, 2001). Peter Winter, CEO of Cox Interactive Media, emphasized the importance of useful and local information that would differentiate the online newspaper product and generate repeat visits to the site (Stone, 2000). In 2000, the Record in Hackensack, N.J., linked 2,000 community-group Web pages to its NJCommunity.com site (Boynton, 2000). NJ.com was the l5th largest local guide and the 58th largest general news site with 10.7 million page views and 476,000 unique visitors in September 2001, according to a survey by Nielsen/NetRatings Inc.

The Tribune Co. has had mixed success in its attempts to dominate city sites. Tribune Co. was an initial investor in America Online Inc., contributing $10 million for a 10 percent share in the Dulles, Va.,-based company and a 20 percent share of Digital Cities, a network of sites with entertainment guides in 200 U.S. cities. Tribune Co. was unable to hold on, however, selling its stake in Digital Cities to AOL in June 2000. The Tribune Co. was more successful when it teamed up with Nielsen and Gist Communications Inc. to offer local listings and specific channel numbers for each of more than 11,000 cable franchises and 220 broadcast markets nationwide (Vonder Haar, 1997). The service competes with TV Guide Online, Microsoft Network and NetChannel Inc.

8. Interactive networking

(A) Medium features: Customizing links to live events, personalized services, banking, shopping, games, social movement activism, townhall meetings, instant messaging
(B) Key variables/components: Transaction model: Free or user fees; micro commissions on product exchanges or time usage
(C) Optimize communities of interest and tunnels/blinders of selective exposure.

Interactive networking depends on offering the variety of entertaining features, such as chat rooms, bulletin boards, free e-mail, free Web site hosting, professional consultation, technical feedback, file sharing, customized searching, personal finance alerts, driving maps, weather aids, bill paying, banking, shopping, gaming, auctions, prizes, polling, romance ads, cyber-sex, and a range of stunts and gimmicks reminiscent of the peak of the press wars and jazz journalism of the early twentieth century. Media firms can extend their advertising brandnames to become marketplaces to host business-to-customer (B2C) e-commerce. A Web site can also extend the news media’s role as a watchdog and societal monitor to become a forum to host community townhall meetings, social movement activities and interpersonal communication. To be more relevant, for-profit news organizations are already taking on such roles — nowadays called "public journalism" or "civic journalism." However, the interactive model can also compete with and supercede for-profit news arrangements through free file sharing, indy media centers and online educational outreach projects.

Newspaper editors are clearly dissatisfied with the current level of newsroom-reader interaction and 90 percent agree the health of their industry depends on more interactivity, said a study funded by APME and the Pew Research Center (Campaign Study Group, 2001). The survey found that editors have adopted the outreach techniques of civic journalism. Cox Interactive Media CEO Winter told an the E&P Interactive Newspapers Conference in Atlanta that interactive applications like polls, chatrooms, personal pages and group pages are one of the keys to driving page views (Stone, 2000). Sites such as Washingtonpost.com have responded with customized offerings like free e-mail newsletters. Washingtonpost.com includes News Headlines, News Alerts, Sports Updates and Entertainment Best Bets through Bigfoot Interactive, based in New York City (Digital Edge, 2001).

Fallows (1999) argued that interaction online is a nightmare akin to one-way talk radio filled with a majority of cranks while relatively few get to speak up. Certainly, interactivity may fail because of uninformed trash talk, message crowding, information overload, a learned incapacity to engage in authentic public debate or a distaste for inauthentic public spectacles. Even so, instant messaging has become one of the fastest growing Internet features since the late 1990s, especially among teenagers who often use it as a conference call alternative to the telephone. Still, a difference exists between user experience and news Web site acceptance of the widespread opinion among experts who promise the rewards of providing interactivity.

Discussion and conclusion

So far in the first decade of the twenty-first century, observers and analysts seem to have arrived at a few consensual conclusions about competing business models.

Advertising revenue does not and probably will not become a primary online revenue source.

Online traffic is still one measure of success and user fees may become increasingly important for news Web sites that offer unique, especially business-related, content.

Infant industry profits remain relatively rare but are even more important since investors now recall that speculative appreciation of stock values does not necessarily coincide with "inevitable" technological progress.

Digital content delivery is a primary product that draws advertising, but revenue to content providers may depend on the intermediary mix of service layers. Access owning firms can offer or channel content; service firms that aggregate may highlight or exclude content. The next major expansion of the Internet and a phase shift to entertainment programming will probably occur as broadband access drops to a more affordable price point.

Continuous breaking news operates more in mid-day, mid-week, work environments but has not yet given online news outlets a niche advantage at home.

Information retrieval and storage (one form of aggregation) continues to pull in news seeking audiences. The audience seems divided between a segment who want headlines and useful information bits and a fraction who want in-depth data documents and in-context reports.

Portal conduits (another form of aggregation) successfully draw audiences, and the competition is over its potential to steer consumers to points of view and products.

Interactive networking offers perhaps the most potential for consolidating and mobilizing audiences, but so far it is the least explored unique medium feature.

The media firms that survive to the endgame for global oligopoly domination cannot afford to opt out of the market for news in a convergent media environment. Nor can they ignore the fact that daily, non-fiction, news content has an impact on the legal and regulatory environment in which they operate. The media firms that seek to dominant will need either to develop or acquire the most successful news Web sites. The empirical and near-term historical question-to-be-answered remains: Which news Web sites with which business models will succeed?

About the Author

Dr. Frederick Schiff received his doctorate in political sociology at UCLA. For 10 years, he worked as a reporter and foreign correspondent, covering Latin America, Europe, North Africa, the Middle East and the United States. He was a staff reporter for United Press International in Lebanon and later became the resident correspondent for USA Today in Beirut. He has written more than 2,000 news and feature stories for print and broadcast media. In the year 2000, he won a three-year grant from the National Science Foundation to study newspaper management and news content, based on a stratified random sample of papers representing eight percent to all U.S. dailies. He is currently writing a book about the U.S. daily newspaper industry. He is a tenured associate professor in the School of Communication at the University of Houston. He teaches journalism, media studies and media economics. Schiff is also the associate director of the International Telecommunications Research Institute.
E-mail: fschiff@uh.edu

Acknowledgments

Research for this article was supported by a grant from the National Science Foundation and by an undergraduate student research grant from the University of Houston.

Bibliography

Dave Astor, 2000. "Not Enough Net From Net for Syndicate," Editor & Publisher, volumn 133, number 5 (31 January), pp. 33-34.

Martin Bichler and Claudia Loebbecke, 2000. "Pricing Strategies and Technologies for Online Delivered Content." Journal of End User Computing, volume 12, number 2, pp. 4-10. http://dx.doi.org/10.4018/joeuc.2000040101

Robert S. Boynton, 2000. "New Media May be Old Media's Savior." Columbia Journalism Review, volume 39, number 2, pp. 29,34.

Stephen Buel, 1998. "Mercury News Drops Internet Edition Fees," View Source, Pennsylvania Newspaper Publishers' Association, (16 June), at http://www.pnpa.com/newmedia/viewsource/061698/mercurynews.htm, accessed 8 July 2002.

Business Week Online, 2000. "A Little Less Irrationally Exuberant?" (10 May), at http://www.businessweek.com/investor/content/mvpw/mpw0510a.html, accessed 8 July 2002.

Campaign Study Group, Pew Research Center for the People & The Press, 2001. "Journalism Interactive: New Attitudes, Tools and Techniques Change Journalism's Landscape," a paper presented at a conference of Associated Press Managing Editors, Pew Center for Civic Journalism, and National Conference of Editorial Writers, Springfield, Va.

Anthony M. Casale, 2000. "Papers Need to Recast for Net," Editor & Publisher, volume 133, number 1 (3 January), pp. 22-24.

Randy Dotinga, 2000. "Only a Handful of Papers Out of Site," Editor & Publisher, volume 133, number 21 (22 May), p. 15.

Digital Edge, 2001. "New Media Index — Weekly Edition," (26 February), at http://www.naa.org/edge/webnews/, accessed 31 March 2001.

James Fallows, 1999. "But Is It Journalism?" American Prospect, volume 11, number 1 (23 November), pp. 58-60.

Paul Farhi, 2000. "Surviving in Cyberspace," American Journalism Review, volume 22, number 7 (September), pp. 22-27.

Jon Fine, 2000. "New Web Site in WSJ's Plans for 'Weekend'," Advertising Age, volume 71, number 34 (14 August), pp. 1, 42.

Focus, 2001. "Been There, Done That: Cox Companies Use Time-Tested Strategies for Dealing with a Soft Economy," newsletter of Cox Enterprises, Inc., (July/August), pp. 1-2.

Forrester Research, Inc., 2000. "Forrester Predicts Career Networks Will Capture Majority of Online Recruitment Market in 2005," (14 February), at http://www.forrester.com/ER/Press/Release/0,1769,246,FF.html, accessed 3 October 2001.

Elihu Katz and Paul F. Lazarsfeld, 1955. Personal Influence. New York: Free Press.

Steve Lawrence and C. Lee Giles, 1999. "Accessibility of Information on the Web," Nature, volume 400, number 6740 (8 July), pp. 107-109.

Kevin Lomangino, 2000. "Go Beyond 'News Lite' on These Sites," Internet Medicine: A Critical Guide, volume 5, number 3 (1 March), pp. 11-12.

Lyra Research, Inc., 2001. "The Emerging Web-Driven Society. Content Intelligence," at http://www.contentintelligence.com/current/Frontpage.asp, accessed 29 June 2001.

John McDermott, 1991. Corporate Society: Class, Property, and Contemporary Capitalism. Boulder: Westview Press.

Jeffrey Mackie-Mason and John Metzler, 2001. "Internet Media: What Are the Markets?" at hhttp://www-personal.umich.edu/~jmm/presentations/industrialorganization/tpug-new-orleans-01.pdf, accessed 15 June 2002.

Josh McHugh, 2000. "Will Online Publishing Ever Fly?" eCompany Now, volume 44, number 3 (1 April), pp. 140-146.

Myers Reports, Inc., 2002. "Media Spending, 2002," at http://www.jackmyers.com/research/reports/advertisingspending.html, accessed 23 July 2002.

New Media Index, 1997. "A Monthly Wrap-Up of Key Develoments on the Interactive Frontier," at http://www.naa.org/presstime/96/PTIMEp197nmi.html, accessed 31 March 2001.

Newspaper Association of America, 2001a. "The Competitive Media Index," NAA's Market & Business Analysis Department, at http://www.naa.org/info/facts/, accessed 15 August 2001.

Newspaper Association of America, 2001b. "Facts About Newspapers: A Statistical Summary of the Newspaper Industry," at http://www.naa.org/info/facts/, includes additional findings from their Market & Business Analysis Department, accessed 15 August 2002.

Newspaper Association of America (NAA), 1999. "Newspaper Readership Remains Competitive in Top 50 Markets; NAA Analysis of Scarborough Data Also Shows Online Usage Not Hurting Print Product," PR Newswire (26 October), Financial News Section.

Nielsen/NetRatings, Inc., 2002. "Audience Measurement Below the Radar Report for September 2000 and September 2001," personal communication with Mary Gutierrez, Custom Research Manager (2 January).

David Noack, 1999a. "Online News is Growing Up," Editor & Publisher volume 132, number 4 (23 January), p. 26.

David Noack, 1999b. "Profitability Will Dominate Interactive Newspapers '99," Editor & Publisher volume 132, number 7 (13 February), pp. 32-33.

David Noack, 1999c. "From Novelty to Mainstream," Editor & Publisher volume 132, number 7 (13 February), pp. 36-43.

David Noack, 1998. "Crossfire: Print vs. Online Newsrooms," Editor & Publisher (14 February), pp. 38-41.

Joe Nicholson, 1999. "WSJ Interactive Turns Its First Profit," Editor & Publisher volume 132, number 49 (4 December), pp. 4-6.

Steve Outing, 1999. "Your Headlines On Thousands of Personal Web Pages," Editor & Publisher volume 132, number 4 (23 January), p. 33.

Pew Research Center for the People & the Press, 2000. "Investors Now Go Online for Quotes, Advice: Internet Sapping Broadcast News Audience," (11 June), survey conducted under Andrew Kohut, Carroll Doherty, and Kimberly Parker.

Pew Research Center for the People & the Press, 1998a. "Online Newcomers More Middle-Brow, Less Work-Oriented: The Internet News Audience Goes Ordinary."

Pew Research Center for the People & the Press, 1998b. "Media Consumption: American News Habits," at http://www.peoplepress.org/medsec3.htm, accessed 15 April 2002.

Pew Research Center for the People & the Press, 1998c. "Event Driven News Audiences: Internet News Takes Off: Pew Research Center Biennial News Consumption," (June).

Philadelphia News Interactive, 1998. "Newspaper Readership Holds Steady While TV, Cable Viewership Erodes," PNI Between Editions, (21 April).

Jennie L. Phipps, 1999. "Knight-Ridder: Local Battles Will Win the Web War." Editor & Publisher, volume 132, number 39 (25 September), p. 37.

Robert G. Picard, 1989. Media Economics: Concepts and Issues. Newbury Park, Calif.: Sage.

Charles Piller, 2000. "The Cutting Edge: Focus on Technology; Vote Recount Yields Clear Winners: Online Newspapers; Media: Complex Election Story Plays to Major Dailies' Strengths. Trick Will Be to Retain Cyber Readers After Controversy Wanes," Los Angeles Times (4 December), Business Section, p. C1.

Presstime, 2000a. Cited Frank N. Magid Associates Inc. study (9 October).

Presstime, 2000b, reported a Digital Edge pop-up survey of 123 Web sites by cPulse for first and second quarters of 2000 (9 October).

PRNewswire via COMTEX, 2001. "Tribune Executives Outline 'Multimedia Company of the Future'," (6 March).

Kathleen Quinn, 2000, "Journalism on the Web Has No Future," Interactive Week, volume 7, number 34 (28 August), pp. 34-28.
Lisa Rabasca, 2000. "Extended New Media Index," Presstime (9 October), at http://www.naa.org/presstime/0009/mmi.html, accessed 31 March 2001.

San Jose (Calif.) Mercury News, 2000. "San Jose, Calif.-Based Newspaper Publisher Adds Sites to Web News Network," (20 April).

Joseph A. Schumpeter, 1992. Capitalism, Socialism, and Democracy. New York: Routledge.

David Shaw, 2001a. "N.Y. Times Unfolds Paper Across Nation and Thrives," Los Angeles Times (24 June), pp. C1, C6.

David Shaw, 2001b. "Papers' Cuts Put Readership at Risk," Los Angeles Times (3 July), pp. A1, A10-11.

David Shaw and Sallie Hofmeister, 2000. "Times Mirror Agrees to Merger with Tribune Co.," Los Angeles Times (13 March).

Jane B. Singer, Martha P. Tharp, and Amon Haruta., 1999. "Online Staffers: Superstars or Second-Class Citizens?" Newspaper Research Journal, volume 20, number 3 (Summer), pp. 29-47.

Martha L. Stone, 2000. "Getting 'Real' in Redesigns," Editor & Publisher, volume 133, number 23 (12 June), pp. I24-I28.

Martha L. Stone, 1999a. "Newspapers Still Playing Catch Up on the Web," Editor & Publisher MediaInfo.com, volume 132, number 9 (27 February), p. 30.

Martha L. Stone, 1999b. "Chicago Tribune Web Site Moving to Registration," Editor & Publisher, volume 132, number 10 (6 March), p. 24.

Martha L. Stone, 1999c. "Spinning on the Web, Old Gives Birth to New," Editor & Publisher, volume 132, number 23 (5 June), pp. 24-25.

Carl Sullivan, 1999. "Marketing the Web in Other Media," MediaInfo.com, Editor & Publisher, volume 132, number 9 (27 February), p. 30.

U.S. Department of Commerce, 2001. "Percent of U.S. Households with Internet Access, By Income, By U.S., Rural, Urban, and Central Cities, 2001," NTIA and ESA, using U.S. Bureau of the Census, Current Population Survey supplements, at http://www.ntia.doc.gov/ntiahome/dn/hhs/ChartH9.htm, accessed 9 July 2002.
Steven Vonder Haar, 1997. "Gist Teams with Nielsen, Tribune Co. on TV Listings," Inter@active Week (28 July), http://www.zdnet.com/intweek/daily/970728b.html, accessed 1 April 2001.

Max Weber, 1946. From Max Weber: Essays in Sociology. Translated, edited, and with an introduction, by H. H. Gerth and C. Wright Mills. New York: Oxford University Press.

Susan Willey, 2000. "The Pitfalls of Cyberspace and Electronic Database Research," Journalism & Mass Communication Educator, volume 55, number 2 (Summer), pp. 78-85. http://dx.doi.org/10.1177/107769580005500209

Jason Williams, 2000. "The New Spider in the Web," Editor & Publisher, volume 133, number 12 (20 March), p. 9.

Anthony Wilson-Smith, 2000. "When Old Becomes New," Maclean's, volume 113, number 19 (8 May), p. 17.

Yahoo! Media Relations, 2001. "Yahoo! Ranked No. 1 Global Internet Destination," (23 October), at http://docs.yahoo.com/docs/pr/release847.html, accessed 9 July 2002.

Editorial history

Paper received 18 November 2002; accepted 16 May 2003.

Copyright ©2003, First Monday

Copyright ©2003, Frederick Schiff

Business models of news Web sites: A survey of empirical trends and expert opinion by Frederick Schiff
First Monday, volume 8, number 6 (June 2003),
URL: http://firstmonday.org/issues/issue8_6/schiff/index.html

A. Performance outputs
1. Advertising revenue
(A) Medium Features: Deliver audiences to advertisers	(B) Key Variables/Components: Advertising model: Price and demographic positioning. Category killers: Niche Web sites for classifieds, especially jobs, housing and automobiles	(C) Optimize revenue side
2. Online traffic
(A) Medium Features: Unified medium of communication	(B) Key Variables/Components: End user model: Uses & gratifications; wants & needs for cognitive orientation, need to know, societal early warning, normative watchdog, pattern maintenance; subscriptions, user fees, commissions, placements, registration.	(C) Maximize general-interest readership, or optimize upscale readers through niche product consumption of news, entertainment, information and interaction
3. Infant industry profits and stock values
(A) Medium Features: Declining production/distribution costs with size; high risks; wide value fluctuations	(B) Key Variables/Components: Financial model: Capital accumulation and reinvestment; volatility of revenue, profits, equity values and risks; monopoly profits; equity speculation through stock market price, market capitalization, IPO cashout, liquidation price	(C) Optimize owners’ equity, return on investment, earnings growth, monopoly profits and stock market price appreciation (but not so far dividends)
B. Input content characteristics
4. Digital content delivery
(A) Medium Features: Digital platforms; origination push	(B) Key Variables/Components: Translate content and offer unique content	(C) Optimize expense side: Product synergy, cross-promotion and low-cost distribution
5. Continuous breaking news
(A) Medium Features: Immediate, ubiquitous, continuous coverage	(B) Key Variables/Components: Traffic model: 24-7 ratings; miniaturized handheld access	(C) Maximize updated content for the most repeat visits
6. Information retrieval and storage
(A) Medium Features: Low-cost storage; searchable databases; non-linear hypertext	(B) Key Variables/Components: Story-specific links to media sources, research archives and libraries for, any date (not hot news) from in-house, chain/network, syndicated, partnered, competing sources	(C) Optimize access and rewriting of texts and images from any historical period, culture or geographic region
7. Portal conduit
(A) Medium Features: Brandnamed oligopoly central site; end user pull via wired, wireless or satellite networks; broadband capacity	(B) Key Variables/Components: Marketing model: Largest general audience	(C) Optimize links to organizations and communities outside the media
8. Interactive networking
(A) Medium Features: Customizing links to live events, personalized services, banking, shopping, games, social movement activism, townhall meetings, instant messaging	(B) Key Variables/Components: Transaction model: Free or user fees; micro commissions on product exchanges or time usage	(C) Optimize communities of interest and tunnels/blinders of selective exposure

Service layers	Conduit	Access	Aggregation	Content
Media firms	Telco	Earthlink	Lycos, Yahoo!	CNET, MSNBC
Advertising	0%	4%	100%	100%
Subscriptions	100%	96%	0%	0%

Service layers	Conduit & content	Access & aggregation	Access, aggregation, & content	Access
Media firms	Time Warner	Prodigy	AOL	Juno
Advertising	0%	13%	31%	37%
Subscriptions	100%	87%	69%	63%