The skeptic might point out here that it is easy to imagine a doomsday scenario but quite a different thing for that speculation to become reality. Would a sustained, global collapse in programmatic advertising actually be possible? How might it come about?
Even as the world grapples with a global recession in the wake of the COVID-19 pandemic, programmatic advertising appears poised to grow, grow, grow over the long term. The dominant, ad-driven platforms of the web continue to be some of the largest and most profitable businesses in the world. Traditional forms of advertising—the dollars that put ads in newspapers and magazines, for instance—are not yet fully integrated into the ecosystem of online programmatic advertising. This means that there is room for growth: these dollars may yet become part of the ad exchange system.
But such a rosy situation may mask deeper systemic risks lurking within the system. Like the financial markets prior to the crisis of 2008, the modern infrastructure of the ad economy might have produced explosive growth while simultaneously introducing a set of vulnerabilities that could make this money machine less stable over time. Indeed, the parallels between the financial industry and the evolving economy of online advertising give us clues as to where things might be going and the less evident cracks that might lie beneath the surface.
2Market Convergence
It is easy to take the presence of advertising on the internet as a given. The infrastructure of programmatic advertising has so robustly supported the rise of the modern giants of the technology sector and operates so efficiently that it often seems as if the internet was always meant to work this way.
That the internet was predestined to be one way or another is an attractive but ultimately wrong idea. There is nothing essential about the internet as we know it. The internet is not inherently open or democratizing, nor is it inherently closed or authoritarian. Digging into the history of the internet quickly reveals that the present-day internet is just one of many different information networks that could have come to be.1
In that sense, the internet has been actively designed at each stage in its history. An ever-compounding set of choices has resulted in the particular experience that we have of it today. This is true of the selection of advertising as a primary money machine for funding the web. It is also true of the precise way that we set these advertising marketplaces up.
To that end, we need to say why the programmatic advertising ecosystem has ended up bearing such a close resemblance to the structure of the financial markets. What forces encouraged the online programmatic advertising marketplace to take the shape that it did? How much was coincidence, and how much was an active effort to structure the market for attention online in this precise way? How much does the rise of programmatic advertising owe to the structural similarities with finance?
This chapter argues that the resemblances between the programmatic advertising markets and the financial markets are more than skin-deep. While software engineers might have laid down the code that enabled advertising to explode as the primary moneymaker for the web, its design has been inspired not by the culture of nerds in garages but instead by the more buttoned-up world of the capital markets.From Wall Street to Silicon Valley
Online advertising predates the rise of the internet as we know it today. Companies like America Online (AOL), Prodigy, and CompuServe were precursors to the modern web in the 1980s and 1990s, allowing customers to use their computers to access a limited “walled garden” of chat rooms, media, and shopping over their phone lines.
These services received monthly fees paid by their subscribers, but they also relied on advertising as a major source of revenue. Prodigy was pitched as “a new advertising medium … that would assemble audiences for marketers much as niche TV channels do.”2 AOL became a major channel through which other emerging dot-com companies of the era could advertise to prospective users.3 AOL also experimented with allowing businesses to pay to sponsor content on the platform. CBS would pay to be the exclusive provider of sports reporting, 1-800-Flowers would pay to be the official flower delivery service, and so on.4
But the advertising taking place on these early platforms was still a far cry from the large-scale automation and split-second bidding of modern programmatic advertising. Myer Berlow, who ran AOL’s advertising operation, hailed from the advertising world of New York and had “no experience with anything computer-related.”5 In the 1990s, the early internet was increasingly commercialized, but it was still organized in ways that relied on teams of human salespeople working to sell advertising space on these platforms.
Search engines, which were made necessary by the ever-increasing sprawl of the internet and the lack of effective tools for exploring it, would alter this model entirely. During their early years in the 2000s, search engines like Google were able to acquire huge numbers of users who relied on them to find things on the web, but these search engines faced the tricky challenge of figuring out how to make such a tool pay off as a business.6
Advertising was by no means the most obvious way to make operating search engines a solvent business. In its early years, Google anticipated that only about 10 to 15 percent of its revenue would be derived from advertising.7 The rest of the revenue was assumed to come from the licensing of its “search technology to other websites … [and] a hardware product that would allow companies to search their own operations very quickly, called ‘Google Quick Search Box.’”8 These initial projections, of course, would soon be blown away by the torrent of money generated by advertising.
Getting advertising to
