4. The Dot-com Bubble

The Marketing/Finance Feedback Loop

The dot-com bubble generated a kind of marketing/finance feedback loop in which the most important business competency was attracting investment capital. This was achieved to a significant degree through advertising and public relations, whereby companies sought to demonstrate their potential to become dominant in a given online market: to get big fast. Dot-coms with a strong market position and favorable media profile found it much easier to attract investors, while securing risk capital through IPOs and other means was deployed as a public relations event in its own right. At the same time, companies that had secured investment funding spent heavily on advertising to further build market share and enhance brand image, which in turn aided more fundraising. This feedback loop between marketing, finance, and the new internet medium directed significant investment toward transforming the web into an advertising channel, supporting and accelerating the broader efforts of the marketing complex.

What is often overlooked about Netscape’s success in the financial markets and the dot-com phenomenon more generally is the fundamental role that marketing communications played in all stages of the bubble’s investment processes. Standard investment analysis holds that a company’s valuation should be based on objective indicators of business performance and underlying market fundamentals. The field of behavioral finance adds a layer of complexity, showing that factors like brand awareness and public image—elements that are heavily influenced by marketing communications but that are difficult to account for on financial statements—can also be important determinants of a company’s valuations and fundraising outcomes. As Robert Shiller argues, “The role of the news media in the stock market is not, as commonly believed, simply as a convenient tool for investors who are reacting directly to economically significant news itself. The media actively shape public attention and categories of thought, and they create the environment within which the stock market events are played out.”13

During the dot-com bubble, advertising and public relations became important drivers of financial valuation, in many cases superseding more conventional metrics. A crucial detail of Netscape’s financial story is that on the day of its wildly successful IPO, the company had not recorded a single dollar of profit. This set in motion an investment rationality for the dot-com era in which profitability, a long-standing rule of thumb for companies filing IPOs, was suddenly seen as outmoded. In 1995, Netscape was an outlier. Almost two thirds of new stock issuers had profitable operations when they held initial public offerings. By the first quarter of 2000, fewer than one in five companies were profitable at the time of their IPO.14 The outlier had become the norm.

Netscape’s IPO triggered a retreat from the profitability standard. A new valuation model was articulated by professional investment analysts whose pronouncements were reproduced by uncritical and at times obsequious media coverage. Perhaps the highest-profile example is a series of reports from investment bank Morgan Stanley that downplayed established economic markers (such as cash flow) in favor of indicators related to growth potential. A proliferation of metrics such as “mind share” signaled the upswing of marketing-based asset valuation models that depended heavily on advertising and public relations.15

Howard Kurtz demonstrates the extent to which investment analysts were given media platforms to espouse marketing-based asset valuation models and the degree to which such practices were encouraged by media owners and professionals who often had vested interests in keeping the bubble going strong.16 The “rhetoric of the new economy was hot and glamorous.”17 It spawned a subgenre of internet investment market news (especially on cable television) and was featured more generally across mainstream media outlets. This developing media discourse had material consequences. As Nigel Thrift argues, “Telling the new economy story worked, and worked to the extent that it began to re-describe market fundamentals.”18 In practice, marketing-based valuation models created incentives for dot-com companies to pursue market share, measured perhaps most directly by web traffic, in order to attract investment and bump up their stock prices.19

Digging deeper into the VC/IPO risk investment process illustrates precisely how dot-coms utilized marketing communications to meet these objectives. Following Netscape’s pattern, the timeline for most dot-coms was to first secure funding from venture capital firms and/or wealthy angel investors, then hold an IPO as soon as possible. Demonstrating forward momentum was essential to fundraising. More important still was to show the potential to become a household name and win a given internet market. To this end, it became increasingly necessary to generate positive media publicity, or investment buzz, across all stages of the investment process, but especially when approaching an IPO.

Netscape’s media acumen was critical to its fundraising success. As cofounder Jim Clark put it, “Anyone starting a company that doesn’t try to influence the press’s impression surrenders the future to fate, a tremendous mistake.”20 Shrewd deployment of advertising and public relations was critical for securing venture capital and holding a successful IPO. In other words, the emerging characteristics of the speculative financial bubble encouraged dot-coms to allocate inordinate resources to marketing communications. On this point, the Morgan Stanley analysts were clear: “For now it’s important for companies to nab customers and keep improving product offerings: mind share and market share will be crucial.”21

Netscape’s IPO was consequential because it demonstrated that financial success could be achieved through advertising and public relations in the absence of profitability. Subsequent dot-coms and investors took the lesson to heart as branding became “essential for web companies” seeking capital.22 After starting a dot-com, journalist and would-be internet entrepreneur Michael Wolff quipped that his “primary job was now to get the company’s name in the paper.”23 “Publicity,” Wolff noted, “is the currency of our time.”24

In this context, dot-com start-ups routinely hired advertising agencies and public relations firms to launch strategic communication campaigns in order to attract attention from potential investors. As one internet executive declared, “You’ve got to get an [ad] agency to show VCs that you are making progress with your business plan; having an agency is a comfort factor for VCs.” Some observers noted that dot-com ad campaigns seemed to be “disproportionately skewed” to investor publics in order to “generate confidence.”25 During one eight-month period in 1999, TBWA/Chiat/Day, an agency known for its edgy creative work, reported meeting with no fewer than 174 dot-coms as potential clients.26

Dot-coms raised billions of dollars from IPOs during the bubble, a significant portion of which went directly to fund marketing communications. Retailer E-Stamp committed nearly two thirds of its $110 million IPO earnings to “ads, marketing and brand-building” efforts.27 Likewise, online insurance peddler HealthExtra used its IPO to finance a $25 million ad campaign.28 Goldberg Moser O’Neill, an ad agency with Interpublic, claimed that its dot-com clients planned to spend in excess of $1 billion in the fourth quarter of 1999, roughly equivalent to the annual U.S. ad spending of McDonald’s and Burger King combined.29

Advertising spending data are more meaningful when compared across business sectors. One method for gauging a given company or industry’s relative emphasis on marketing activities is to compare sales and marketing expenses as a percentage of revenue, known as the SME ratio. A study commissioned by Advertising Age found that the dot-com sector had an average SME ratio of 94 percent in the fourth quarter of 1999, meaning that the typical dot-com spent 94 cents on sales and marketing for every incoming dollar of revenue.30 Although high SME ratios are not uncommon among new businesses, dot-coms allocated disproportionate resources to sales and marketing efforts compared to off-line retailers, which averaged SME ratios of 25 to 40 percent.31 These data show that the dot-com sector funded marketing communications, and advertising in particular, at rates that far outpaced comparable off-line businesses. This trend can be attributed in part to certain large-scale branding efforts via traditional media channels. This was most pronounced in television, where dot-coms bought ad inventory at premium rates, including collectively purchasing seventeen ad spots during the 2000 Super Bowl.32 Yet the majority of traditional media spending came from an unrepresentative cluster of the most highly capitalized internet companies.33

Although Super Bowl commercials were noteworthy expenditures, most dot-coms spent the bulk of their ad dollars on a medium much closer to home. In 1996, six of the top ten online advertising spending leaders were dot-coms.34 The list included Excite, Netscape, Infoseek, Yahoo, Lycos, and CNET, all of which had recently held IPOs. Infoseek spent 60 percent of its advertising budget online, while Yahoo spent nearly all of its ad dollars on the same.35 In 1997, all but CNET again ranked among the top ten online spenders. More generally, dot-coms and companies in the computing and technology sectors, like Microsoft and IBM, accounted for more than half of all online ad spending in 1996 and 1997 and about 40 percent in 1998.36 In 1999 and early 2000, a group of about eighty well-capitalized dot-coms paid for over three quarters of all web advertising.37

Online advertising represented a small fraction of total U.S. ad spending at the start of the bubble. By 2001, it had surpassed outdoor media and trade publications.38 Annual outlays more than tripled from 1996 to 1997, more than doubled from 1997 to 1998, doubled again from 1998 to 1999, and grew by 75 percent from 1999 to 2000.39 Dot-coms drove this growth, offsetting the more tentative outlays of traditional marketers. Some traditional advertisers experimented with banner advertising and corporate websites as early as 1994, but others were hesitant to move ad dollars online in any systematic manner until the end of the decade. Although only intermittently available, ad impressions data confirm this. In the second and third quarters of 1999, the ten largest dot-com marketers collectively purchased over 7.5 billion banner ad impressions, more than double the amount purchased by the top ten traditional marketers over the same period.40

In addition to providing material support for ad spending, the bubble helped to foreclose alternative media funding models. Publishers considering user subscriptions ran into the powerful headwind of “get big fast.” “Our investment bankers were urging us not to charge [for content],” said the cofounder of financial news site TheStreet.com.41 Although start-ups did not need to be profitable to get funded, they were required to sketch out plans for future revenue, particularly in the bubble’s later stages. As Ethan Zuckerman notes, dot-coms were not universally excited about adopting the advertising business model. It just so happened that advertising was the “easiest to market to investors.”42 Advertising became a kind of default business plan for dot-coms who could not risk losing market share by attempting to charge consumers for content or services.

Propped up by risk capital, dot-coms maintained robust marketing budgets even as burn rates—industry jargon for negative cash flow—climbed steadily. These expenditures were legitimized through a new economy discourse, or what Patrice Flichy calls a “frame of representation,” that helped to coordinate the activities of the period.43 The new economy discourse was steeped in “mythologies of entrepreneurial risk taking” and promises of monopoly profits for those who could dominate yet untapped internet markets.44 Vincent Mosco characterizes the period as being overtaken by a “myth of the digital sublime” whereby the dot-com designation “conferred a mythical power that allowed firm[s] to transcend accepted marketplace conventions.”45

The new economy was a compelling narrative that not only justified spending large sums on advertising but also made it effectively mandatory. As technology writer and free market evangelist Kevin Kelly wrote, the new economy meant a new set of rules for conducting business: “Those who play by the new rules will prosper, while those who ignore them will not.”46 The new rules maintained that while dot-coms would eventually need to attract actual customers, their immediate priority was to generate excitement among investors and stockholders. A portfolio manager with the investment firm Neptune Capital Management spelled out what this meant: “In the internet world, you have to look for the dominant player. We’re not looking for profitability. Now, we’re only looking for growth.”47 For dot-coms pursuing get big fast, advertising was the gateway to the next round of essential investment capital. “If budget-busting advertising campaigns or product giveaways are what it takes to propel your company into the ranks of web giants, well, that’s okay,” wrote Fortune magazine. “Profligacy pays.”48 For the venture capital firms and investment banks that exploited the financial mania to collect substantial investment payouts and service fees, profligacy paid quite well, at least during the bubble’s upswing. As John Cassidy notes, “Instead of using the stock market to build companies, venture capitalists and entrepreneurs use[d] companies to create stocks”—a strategy that was remarkably lucrative.49 In 1999, U.S. venture capital firms collectively realized a return of nearly two and half dollars for every dollar they invested.50

Venture capitalists used advertising and public relations to build valuation before exiting investments via IPO markets. Execution of this strategy was dependent on VCs’ exerting managerial influence within the boardrooms of their portfolio companies.51 As the Wall Street Journal reported, “When it comes to the marketing craze among web-based start-ups, the most powerful advertising executives aren’t in the advertising business at all. They are the people of Sand Hill Road, Silicon Valley’s venture-capitalist enclave.”52 In late 1999, an estimated 80 percent of venture funding provided to internet companies was spent on advertising.53 This is indicative of the instrumental power of an investor class that deployed marketing communications as a means to deliver returns on speculative investment. Inflating demand for internet advertising was in many ways an externality of this moment of financial capitalism.

Ad Networks Get Big Fast

Beyond creating the primary source of demand for internet ads, speculative investment also played a more direct role in launching the internet advertising industry. Advertising start-ups raised billions of dollars in risk capital in the dot-com period. The ad network DoubleClick and its main competitor, CMGI, epitomized this trend. Bolstered by public and private investment, these two companies grew to become the new industry’s top dogs. The balance of this chapter offers a case study of DoubleClick and CMGI, showing how these companies used risk capital to support the development of surveillance advertising on an increasing scale.

As Michael Indergaard observes in his study of New York City’s “Silicon Alley” district, DoubleClick’s “prowess for developing technology was matched by a knack at raising capital.”54 Founded during the bubble’s early stages, DoubleClick was led by an executive team that counted fundraising and internet advertising evangelism among its primary competencies. CEO Kevin O’Connor was an outspoken figure who saw public relations as a pillar of the company’s growth strategy.55 Kevin Ryan, hired in 1996 as chief financial officer and later succeeding O’Connor as CEO, was a former investment banker and senior vice president at United Media, a syndication service owned by the newspaper conglomerate E. W. Scripps.56 Under their leadership, DoubleClick repeatedly used risk capital to fund aggressive business expansion. In June 1997, DoubleClick secured its first private financing in the form of a $40 million venture capital investment.57 This was the largest private financing for a dot-com start-up outside of Silicon Valley to date, valuing DoubleClick at over $100 million and making it “far and away” the most exciting internet company in the New York area in the eyes of the business press.58 It was a seal of approval from the investor class, bolstering DoubleClick’s brand and the still unproven internet advertising industry more generally.

DoubleClick emerged as a figurehead of New York’s dot-com start-up scene, “setting benchmarks for risk-taking” and catalyzing a wave of internet investment from traditionally more conservative East Coast venture capital firms.59 Capitalizing on the stock market’s simmering infatuation with dot-coms, DoubleClick held an IPO just eight months after its first private venture round.60 The IPO provided another $62.5 million and generated further publicity. In what was described as “one of the hottest IPOs of the year,” DoubleClick’s stock rose 57 percent on the first day of trading, resulting in a company valuation of more than $400 million.61

As CEO Kevin O’Connor put it in his signature bombast: “Our goal is to deliver every ad in the world to every consumer.”62 The timing seemed right for such audacity. Capital and publicity were available in spades. Over the next several years, DoubleClick raised money whenever it could, netting over $1 billion by the end of the decade.63 As more and more capital poured into the dot-com sector, the value of individual stocks rose sharply. This was especially true for top-tier companies like DoubleClick, whose market capitalization—the total value of its outstanding shares—increased from $424 million at the end of 1997, to $1.9 billion at the end of 1998, to $10.7 billion at the end of 1999.64

“Flush with cash and possessing a highly touted stock,” DoubleClick set out to “dominate the internet advertising business.”65 In its annual shareholder reports between 1997 and 2000, DoubleClick outlined its plan “to significantly increase its operating expenses in order to expand its sales and marketing operations, to continue to expand internationally, to upgrade and enhance its [ad-serving] technology, and to market and support its solutions.”66 DoubleClick was going all in on the notion of internet advertising as a winner-take-all proposition. “It’s clear what the market’s telling us,” said O’Connor. “They want to give us a lot of money, so we take it and we invest. Why? Because this is the biggest thing that’s ever hit. Market share is everything.”67

One of DoubleClick’s primary competitors during this period was the internet holding company CMGI. If DoubleClick was the “godfather of the ad services game,” then CMGI was the “web giant nobody knows.”68 With roots in direct marketing and software development, CMGI was reconfigured in the early 1990s as a kind of hybrid venture capital/holding company for internet businesses.69 One element of CMGI’s strategy was to build a portfolio of interlocking subsidiaries whose operations were meant to enhance each other. Seeing DoubleClick’s success in the targeted advertising sphere, CMGI’s chief executive David Wetherell was drawn to the idea that consumer data could function as grist for his “virtuous circle” business plan.70

CMGI maintained investments in audience aggregators such as Lycos, a leading search portal, and GeoCities, a large network of user-created web pages, as well a number of internet advertising start-ups.71 In an attempt to create synergies among these holdings, CMGI created Engage Technologies to develop technology for consumer profiling and targeted advertising. Wetherell believed his company could win the internet ad market by deploying Engage’s targeting technologies across CMGI’s portfolio of high-traffic websites. With his sights set on DoubleClick, Wetherell also bought a controlling stake in the web portal AltaVista, one of DoubleClick’s most important clients. The goal was not only to reach AltaVista’s large user base with CMGI’s ad systems but also to deprive DoubleClick of the opportunity to do the same.72 In 1998, the Wall Street Journal highlighted the success of CMGI’s investment portfolio, which had grown to include full or partial ownership of twenty-two dot-com businesses. CMGI also attracted investment from established tech companies, including Microsoft and Intel.73 As the financial market reached a fever pitch in 1999, the value of CMGI’s shares shot up 700 percent on the year, shattering the relative gains of the likes of Amazon (342 percent) and Yahoo (166 percent).74

Like the rest of their dot-com counterparts, DoubleClick and CMGI used public relations and advertising to keep the marketing/finance feedback loop churning. As DoubleClick’s O’Connor told Advertising Age, “It’s important to make sure [investors] know what your company is about and what you do.”75 O’Connor became a fixture in the business press and cable news networks, while CMGI was featured on the covers of BusinessWeek and Fortune.76 CMGI’s Wetherell was said to promote his companies “with the passion of a true zealot,” even purchasing the naming rights to the newly constructed New England Patriots football stadium.77 By late 1999, the two companies were the “toast of both the Street and the Valley,” and each boasted market capitalizations that hovered around $10 billion—about the same as Omnicom, one of the world’s largest advertising conglomerates.78

The advertising dot-coms were remarkably successful in raising capital, but what did they do with it? The answer mirrors the plot device of Brewster’s Millions, the 1985 comedy where Richard Pryor’s character has to spend a ridiculous sum of money as fast as possible on the condition that if he succeeds, a vastly bigger fortune awaits. For DoubleClick, CMGI, and a few other contenders, the promised fortune was the elimination of meaningful competition in the internet advertising market. The goal was to compete for the market, not in the market. So like Brewster, DoubleClick and CMGI went on a shopping spree of epic proportions to pursue acquisitions and other kinds of strategic partnerships designed to limit competition. The strategy was certainly in vogue, as concurrent waves of mergers in the media and telecommunications sectors were ongoing in the wake of the Telecommunications Act of 1996.79 The free market orthodoxy discussed in earlier chapters permeated the internet space and provided political cover for all kinds of dot-com deal making. One key distinction of the internet sector mergers is that very little money actually changed hands in the execution of these transactions. Instead, stock was the primary currency. In the context of the bubble, market capitalization readily translated into purchasing power as inflated share prices made it viable to conduct stock-based acquisitions. As the influential venture capitalist John Doerr noted, the upshot of going public is that it provides an immediate boost to liquidity in order to buy out competitors and branch into new markets.80

In October 1999, DoubleClick acquired NetGravity, a major competitor, fully financing the transaction with $530 million in stock.81 The takeover increased DoubleClick’s customer base of web publishers by 50 percent, adding 350 new clients including high-profile sites such as CNN.82 It also made DoubleClick the ad-services provider for more than half of the web’s top fifty publishers, solidifying its status as the market leader. The deal was about “operating system dominance,” said one observer. “They don’t want to give [publishers] . . . another alternative.”83

One month later, after overcoming opposition from privacy advocacy groups, DoubleClick finalized a merger with data broker Abacus Direct in another stock deal worth $1.7 billion.84 DoubleClick intended to combine Abacus’ consumer purchasing information with its own profile database in order to improve its ad-targeting capabilities.85 Stock-based transactions of this nature were almost self-perpetuating in the sense that they sent share prices upward, providing even greater purchasing power to the combined entity. As one dot-com CEO told Fortune: “Valuation is a sign that investors are actually rewarding us for being aggressive.”86

In early 2000, DoubleClick again exchanged stock for a 30 percent stake in the discount ad network ValueClick. The move was significant because it established “a beachhead for DoubleClick in the emerging cost-per-click advertising model,” a market segment in which it had previously not maintained a significant presence.87 Other DoubleClick acquisitions in this period included Opt-In Email and FloNetwork, providers of email marketing services; Flashbase, an online sweepstakes servicer; and @plan, which offered research services for media buyers. Through this series of stock-based transactions, the biggest fish in the online advertising sea grew significantly bigger.

Like DoubleClick, CMGI threw the weight of its financial valuation behind a push to become a “powerhouse in the rapidly emerging market for targeted advertising.”88 The company’s primary tactic: using its highly valued stock to make strategic acquisitions. In 1999 alone, CMGI completed seven acquisitions of online advertising companies, all of which were financed through stock trades valued in total at nearly $2 billion.89 As the CEO of one ad tech company told BusinessWeek upon its acquisition: “CMGI wants to become one of the cornerstone players of internet advertising, and they are investing in the backbone pieces to do that.”90 Between 1999 and 2001, CMGI spent “a staggering $13 billion on acquisitions, nearly all paid for with its own stock,” on an array of internet business in the ad sector and beyond.91

In addition to mergers and acquisitions, DoubleClick and other leading companies used risk capital to invest in technology development, real estate, data centers, and a significant expansion of sales labor. DoubleClick invested heavily in its core technology systems for ad serving and consumer profiling, including developing redundant capacity for emergency systems failure scenarios and investing in server architecture upgrades.92 By late 1999, the company maintained about 650 ad servers across twenty data centers housed in the United States and abroad. In 1996, DoubleClick employed thirteen people at a single location.93 Shortly after receiving venture financing, the staff was expanded to over a hundred people, and the company relocated to a high-rise at Madison Avenue and 26th Street, the heart of what became known as Silicon Alley.94 After its IPO, DoubleClick began opening sales offices domestically and internationally, establishing some thirty locations throughout Europe, Asia, and Latin America.95 By the end of 2000, DoubleClick’s workforce had grown to nearly 2,000 with a sales staff of 1,040 people, including 380 working internationally.96 CMGI had a similar trajectory, employing 6,000 people in fiscal year 2000.97

The outcome of this investment was a dramatic scaling up of ad-serving capacity among internet advertising’s biggest companies. In its first thirteen months, DoubleClick delivered a billion and a half ads to more than 26 million unique users.98 This was during the bubble’s early stages and before the company received any risk financing. By 1998, the year of DoubleClick’s IPO, the company served 34 billion ads reaching 46 million web users in a single month.99 Although this initial upward trajectory was strong, it pales in comparison to the growth that occurred during the final two years of the bubble period, when DoubleClick’s risk capital–fueled expansion enabled it to deliver a remarkable 621 billion advertisements in 2000.100 In other words, the company served about 200,000 more ads on the average day in 2000 than it delivered in its entire first year of existence, just four years before. Although not operating at the same massive scale, CMGI expanded its ad-serving capacity as well. A significant portion of its growth came from within as CMGI invested heavily in various online publishers and retailers throughout 1998 and 1999, bringing new companies into the fold. By mid-1999, CMGI’s AdSmart network comprised 300 websites and delivered over 2 billion monthly impressions.101 A year later, CMGI’s ad properties combined served some 8.6 billion ads a month.102

Although DoubleClick and CMGI were market leaders, they were not the only ones in the field. DoubleClick functioned as a kind of proof of concept for dot-com ad companies. By securing its initial investment from top-tier investment firms Greylock Partners and Bain Capital, DoubleClick earned a seal of approval from the finance capital community, helping to establish New York’s Silicon Alley as a hub of dot-com activity.103 More importantly, DoubleClick’s early successes helped pave the way for broader investment in the internet advertising sector. Between 1998 and 2000, at least nineteen internet advertising companies held IPOs, raising a combined total of over $1.3 billion. Notable among these were ValueClick, AdForce, FlyCast, 24/7 Media, MatchLogic, and Real Media, several of which were later incorporated, to varying degrees, into DoubleClick and CMGI.104 Others such as 24/7 Media and Real Media became significant industry players by completing big mergers of their own.105

By 1999, several second-tier companies including AdForce, ValueClick, and FlyCast had reached the milestone of serving over 1 billion ads per month, while 24/7 Media was delivering 3 billion and Real Media, 10 billion.106 No data on aggregate impressions are available, but some observers put the figure well into the hundreds of billions per month by late 1999.107 Such estimates appear to be credible considering that DoubleClick alone was delivering approximately 50 billion monthly impressions. In any case, the sheer volume of advertising on the web grew by an order of magnitude from 1998 to 1999 and again from 1999 to 2000, at which point it was estimated that three different companies—DoubleClick, CMGI’s Engage, and 24/7 Media—had the capacity to reach over half of global internet users.108

As described in chapter 3, DoubleClick’s foremost technological breakthrough was to unify targeted ad serving and consumer profiling into a reciprocal process of data transmission and collection. In this regard, increasing the scale of ad serving was about more than simply delivering as many ads as possible across as many sites as possible. It was also about expanding data collection, consumer profiling, and other data-driven advertising practices to reach increasingly segmented groups of consumers wherever they happened to be on the internet. As DoubleClick’s Kevin O’Connor put it: “The great paradox with targeting ads is that the more you are micro-targeting, the more reach you have to have.”109 “Critical mass is important,” added DoubleClick’s Kevin Ryan. “The bigger your [profile] database, the more targeted you can be.”110

Consumer data became a central factor in “dictating merger and acquisition strategy for the industry’s leading players,” noted Adweek.111 DoubleClick’s acquisitions of competitors such as NetGravity not only reduced competition but also provided significant additional surveillance capacity.112 Likewise, CMGI’s controlling investment in AltaVista “vaulted [it] into the number three spot among advertising networks” in terms of reach, but it also provided an abundant new source of consumer data.113 The portal’s 10 million monthly visitors fed a steady stream of consumer information into the databases of CMGI’s Engage advertising subsidiary.

At CMGI, extensive consumer profiling became the lynchpin of its virtuous circle investment strategy of vertical integration. Our “vision is to have the largest reach on the web and monetize that reach better than anyone else,” noted a CMGI executive.114 The goal was to “build interactive marketing services and infrastructure to generate revenue across that reach.”115 Portals like AltaVista and Lycos were positioned as hubs to funnel consumers through CMGI’s roster of internet properties such as the financial information site Raging Bull, retailer Furniture.com, and genealogy site Ancestry.com.116 Undergirding these connections were CMGI’s advertising operations, Engage and AdSmart, which delivered ads for and collected consumer information from all CMGI-affiliated sites. As one journalist observed, “If CMGI has a core technology that weaves through its patchwork portfolio, it’s the ability to track computer users through their every browser click.”117

By the end of the 1990s, CMGI’s Engage had stockpiled a profile database containing over 70 million entries, which it used to refine and develop new ad-targeting methods.118 In the third quarter of 2000, nearly 50 percent of CMGI revenues came from online advertising.119 Yet even this dramatic growth could not match DoubleClick, which through its much larger network and own spate of acquisitions had amassed a collection of 120 million user profiles—twelve times what it had at the start of 1997.120

An Industry without Income

The big asterisk to this story is that DoubleClick and CMGI were consistently unprofitable throughout this period. Like many dot-coms, they maintained balance sheets that contrasted sharply with their stock valuations. Although DoubleClick posted steady revenue growth throughout the bubble, its losses grew at a much faster rate. Revenue increases stemmed from the absorption of acquired companies and greatly expanded sales efforts, but these income sources were not nearly enough to make the company profitable. Warnings regarding its “history of losses and anticipated continued losses” appeared consistently among the mandatory disclosures of risk factors contained in annual SEC filings. Digging deeper, financial documents reveal that losses in 1999 more than doubled those from the previous year, while revenue grew at a significantly slower pace.121 CMGI too was consistently unprofitable and by 2000 maintained a “burn rate” of around $50 million a month.122 It was, in Michael Wolff’s memorable, if inexact, phrase, an “industry without income.”123

Risk capital was the critical enabling factor that allowed DoubleClick, CMGI, and their contemporaries to aggressively expand despite sustained losses. Whereas DoubleClick raised funds by repeatedly offering pieces of itself via the public stock market, CMGI primarily took the approach of engaging directly in speculative investment. Regardless of these tactical differences, the goal was the same: get big fast. As Candice Carpenter, CEO of the web publisher iVillage, explained, because investors “will accept losses at this juncture, we are able to rapidly acquire other companies and really build market share. This is a land grab.”124

DoubleClick’s SEC filings plainly state the connection between risk capital funding and growth. DoubleClick applied the capital raised via its initial public offering “toward the expansion of international operations and sales and marketing capabilities” in addition to financing general operating costs.125 It is reasonable to conclude that subsequent capital infusions beyond the IPO were applied to similar purposes. How else could the company “significantly increase its operating expenses” year after year while continually losing money?126 Likewise, CMGI’s business model was “built on the stock market’s enormous expectations for the internet.”127 As BusinessWeek put it: “As long as investors keep paying high prices for shares in his companies, [CEO] Wetherell will have the currency he needs to keep doing deals.”128 Reciting a refrain of the new economy discourse, Wetherell shrugged off concerns regarding profitability, maintaining, “It would be sinful to be making money on the internet right now, when it’s growing this fast.”129

CMGI directly applied funds from the sales of its appreciated investments in Lycos and GeoCities to finance its money-losing advertising operations such as Engage.130 Likewise, DoubleClick used proceeds from its VC investments and public offerings to supply working capital to maintain business operations.131 Amazon founder Jeff Bezos clarified what was really going on here: If “ecommerce had been subject to the regular discipline of the market, early setbacks would have been fatal. But consumers were not driving online commerce, Wall Street was.”132 Although Bezos was talking about online retailing, his observations hold true for the internet advertising market and the dot-com sector at large, which, not incidentally, produced the greatest portion of early demand for online advertising.

Temporary as it turned out to be, the dot-com bubble shielded surveillance advertising companies from the pressures of profitability, enabled their rapid expansion, and funded a significant buildup of consumer monitoring capacity. Eyeing a future beyond the bubble and seeking to conquer the online advertising market, DoubleClick kicked off a commercial surveillance arms race, bankrolled in large part by speculative capital. At the same time, ad network executives understood that they would need to attract major off-line marketers in order to prosper in the long term. In their efforts to woo the more recalcitrant elements of the marketing complex’s mainstream, ad networks sought to reposition themselves as intermediaries for a broader range of targeted marketing communications, a process that involved bringing marketers themselves into the business of online data collection and exchange. This trajectory, which I call protoplatformization, is the subject of the next chapter.