How the Media Lost the Traffic War to Facebook and Google

If the first half of 2023 was a low point in the news business’s free fall, the second half is shaping up to be the era of the autopsy. Like earthquake victims searching for a lost handbag, the question on every media person’s tongue (apart from “Are you hiring?”), is “What happened?” Less than five years ago, companies like BuzzFeed and Vice enjoyed multibillion-dollar valuations. Now The Huffington Post is little more than an afterthought, BuzzFeed News has folded after its parent company’s brief foray into life as a public company, Gawker was reborn only to die again, and Vice has filed for bankruptcy, along with a growing list of other beloved magazines. What happened?

Ben Smith—BuzzFeed’s former news chief, who recently ventured from his perch at The New York Times to help launch Semafor, a new new media site best known for courting (and then returning) a large investment from disgraced crypto billionaire Sam Bankman-Fried—offers one answer. In May, Smith published a reported memoir called Traffic: Genius, Rivalry and Delusion in the Billion-Dollar Race to Go Viral, in which he traces the history of the media landscape from the bygone days of the early social web.

In Smith’s telling, the story of this period is framed in the shadow of an intellectual argument—a clash of internet philosophy, playing out in the lives of two eccentric geniuses. One was Nick Denton, the hardened British founder of Gawker, whose vast agglomeration of toxic blogs became the go-to spot for dick pics, scene reports, and celebrity gossip. The other was Jonah Peretti, Smith’s onetime boss at BuzzFeed, who built his empire of quizzes, cat memes, and hard news by blurring the lines between advertising and editorial content, mashing together an endless stream of advertorial clickbait into a slurry of libido and identity-inflected kitsch. Whether viral traffic is an art (Denton’s view) or a science (Peretti’s theory) is treated as the central tension of the book.

Both men steered their organizations with a nearly religious fixation on traffic and data, Smith explains, a mentality that would help reshape media as a whole. And both believed that traffic was a commodity: a sort of “digital gold” that could be exchanged for real money, status, and power. Their ultimate failure resulted from the shortfall of their respective hypotheses: Denton believed that publishing unvarnished stories about celebrities’ genitalia would bust open the curtain of power (think: Anthony Weiner), and this earned him the enmity of an angry tech billionaire obsessed with privacy (at least his own, that is). And Peretti’s insatiable hunger for growth found him drinking from the fire hose of venture capital only to sacrifice his autonomy to impatient investors. The story of new media, as Smith tells it in this “first draft” of history, is also one of hubris—tenacity, cleverness, and megalomania, writ large.

But seeing this period as a study of character (and characters, with Denton and Peretti just two players in a cast ranging from Arianna Huffington to Andrew Breitbart to Smith himself) ignores critical economic facts that help explain why the BuzzFeeds of the world were destined to remain flashes in the media pan. Because while Peretti and Denton had a hand in the media era that brought us the listicle, the sponsored content post, and the now-ubiquitous style of writing known as internet snark, the story of their rise and fall has as much to do with the society-scale transition from television to internet advertising as anything else. The story of digital media throughout the 2000s, in other words, is really the story of a resource that was seemingly plentiful and inexpensive to mine—until it wasn’t.

In this sense, a gold rush is an apt metaphor for those heady days of the early social web. The California gold rush, which started in 1849, proceeded in three phases. In the beginning, early diggers were given cheap, direct access to a resource that was flowing in abundance. But those early adapters quickly set up systems to box out newcomers, such as machinery to extract the gold faster and taxes on newly arriving diggers. The once-astronomical rate of return available to those who just showed up started to dwindle. Eventually, the unit economics no longer made sense for the average digger. There was still gold—but it was a lot more expensive to get it.

The history of internet traffic—and, by extension, internet advertising—in the early 2000s followed a similar pattern. With relatively few people on the web, and even fewer creating quality content, it was possible to show up and quickly amass an audience with the right niche and some technical savvy. The so-called “blog era,” in which Smith got his start, was a product of this environment.

For years, simply learning how to game the mechanics of Google rankings, while getting on the good side of aggregator sites like The Drudge Report was enough to build a meaningful audience—an approach known as search engine optimization, or SEO. By 2005, however, as Google began to launch new ad products, such as Google Analytics, it became easier for advertisers to track the performance of their own campaigns, and with better products Google became more assertive in charging people for its digital real estate. As this process accelerated, Google made it harder for brands to appear in search rankings without a substantial investment in SEO or search engine marketing (SEM). It triggered an internet arms race, and things became more expensive.

Social media marketing on Facebook followed a similar trajectory. When Facebook launched its ad platform in 2006, brands and publications mainly used their Facebook pages to reach their audiences. The platform’s algorithm optimized for such interactions, showing a brand’s “organic” posts to nearly everyone who liked a page, so brands went on a buying spree to acquire page likes using Facebook’s ad platform. These were cheap at first—often costing no more than a couple of cents, and then they became more expensive, coming with the added risk of buying bots and fake followers. Eventually, Facebook realized that instead of charging advertisers for page likes—a transaction that, once completed, would provide lifetime access to their intended audience nearly for free—by reducing the reach of an individual Facebook page it could charge advertisers on a post-by-post basis, forcing them to pay each time they wanted to reach their audience.

As a 2014 Gawker article put it: “Facebook Is Ending the Free Ride.” In 2012, the average organic reach of a branded Facebook page was 16 percent (meaning that for every hundred page likes, 16 people would see your post); by 2014 it was less than 6 percent, and eventually it fell nearly to zero. By 2015, Facebook effectively killed the reach of the average corporate or branded Facebook page, which made it a requirement for brands to advertise each individual post.

But even then, the cost of reaching an audience on Facebook was so inexpensive that it could produce multibillion-dollar companies in a matter of years. This dynamic produced an e-commerce boom, giving rise to internet-darling firms like Harry’s Razors, Warby Parker, and Allbirds. These firms’ standard operating procedure was to spend 60 to 80 percent of the millions they raised from venture capitalists on marketing, with the vast majority of those dollars spent on Facebook ads. From 2013 to 2017 the average cost for a click on Facebook was between 21 cents and 27 cents—a pittance compared to the vast sums required to advertise on television or elsewhere. Facebook ads had been around for nearly eight years, but large advertisers like Proctor and Gamble or Nike were slow to migrate to the platform. Partly, this was because of an industry-wide perception among corporations and their advertising firms that running Facebook ads was highly technical. This allowed savvy upstarts—such as BuzzFeed—to take advantage of the cheap media and serve as middlemen. As a 2013 story in The Atlantic put it: “One Secret to BuzzFeed’s Viral Success: Buying Ads.”

This began to change around 2018. Whereas “once, BuzzFeed’s clever advertisements had gone viral all on their own,” Smith writes in Traffic, “gradually, Facebook had taken a larger and larger cut.” It’s a somewhat clumsy way of explaining that, by the end of that year, BuzzFeed found itself spending “millions” to distribute the branded posts and videos that had long been its main source of revenue: It spent $386 million to make $307 million in revenue—a $79 million loss.

By 2020, as the onset of the Covid-19 pandemic prompted a mad dash online for large brands, the cost of advertising on Facebook continued to rise. Large advertisers drove up the cost of Facebook’s “attention marketplace,” leaving the average bidder behind in a marketplace where the unit economics stopped making sense. In 2022, the average cost for a click on Facebook was $1.86.

While the fact that companies like BuzzFeed were spending millions on Facebook is no secret, Smith ignores this part of the story; presumably because it contradicts BuzzFeed’s core marketing claim. Because from the moment that BuzzFeed began selling advertising—or, as it called it, “native advertising,” where links paid for by brands like Verizon would be folded into posts that appeared editorial—the premise of its business model was its supposedly unique ability to make things go viral. Where a JetBlue post on the company’s Facebook page might only reach a few thousand people, BuzzFeed’s promised virality could supposedly bring that same post more bang for the buck.

How did it make this work? At the peak of BuzzFeed’s reign, Peretti told New York magazine that internet virality was determined by a formula, expressed as R = ßz (where z represents the number of people who come in contact with something, and ß represents the probability of transmission). Peretti claimed that BuzzFeed’s content style (its listicles, quizzes, eye-catchy headlines) juiced ß, while its transmission tactics (which included things like SEO but chiefly meant buying Facebook ads) were the z. But as the aforementioned Atlantic piece concluded: “Peretti’s formula for virality really adds up to a more mundane sales pitch: Buy lots of ad impressions and realize a modest, if unpredictable, viral bonus.”

That approach was well suited to the early days of the social web, when it was relatively inexpensive to get something in front of a large number of people. But as the cost of advertising went up, that window closed. BuzzFeed may have been among the best purveyors of ß, and this strategy may not have accounted for the majority of its traffic overall, but its ability to buy ads inexpensively and then turn around and sell the same traffic, on its website, at a higher rate was also a critical part of the equation. Advertisers call this process “media arbitrage”—buying traffic and engagement on social media platforms, or other inexpensive traffic sources, at one rate, then selling that traffic from their site to large legacy brands (like Virgin or JetBlue) at a much higher one. In its heyday, according to Smith, BuzzFeed and Gawker were charging about $9 for a thousand impressions. The average cost to acquire the same views on Facebook was around $1. They built their business model on the difference.

Indeed, this approach is what the much-decried “advertising-driven business model” of journalism really means. For years, BuzzFeed, Vice, and others essentially acted like advertising agencies on behalf of big brands. A 2019 case study published by BuzzFeed and Nielsen illustrates this clearly. In it, marketers at BuzzFeed explain how they spent $139 million on behalf of consumer packaged goods companies to promote their products. BuzzFeed was paid to create sponsored content posts and other ad formats, while running ad campaigns that directed traffic to its own website, an approach BuzzFeed called “social discovery.” The point of the study was to convince advertisers that buying ads through BuzzFeed was superior to buying them on television. But it was also selling the idea that giving BuzzFeed money to buy ads and direct that traffic to its own website was a superior strategy to just buying the traffic directly. This meant that BuzzFeed was directly competing with the social media companies from which it was generating traffic in the first place. At its peak, over 75 percent of its traffic came from Facebook.

So rather than build their platforms on things like audience integrity or cultural esteem, companies like BuzzFeed were built on predicting, and then responding to, the fleeting “content” whims of increasingly assertive tech platforms. Their focus was mastering the internet’s game of technical dark arts in a glut of venture capital—and then they were slowly priced out of the game as larger firms with deeper pockets came to learn the same tricks.

In this sense, the story of media throughout the 2000s is also a story of digital enclosure—a process in which a shared, open resource (attention, control of the internet’s roadways) was gradually restricted to those who could pay large sums for it. It’s the result of an internet in which two companies came to control 70 percent of web traffic and 90 percent of digital advertising. BuzzFeed is merely the most dramatic example of the risks of hitching yourself to the whims of an ascendant tech platform. For a few years, cultural savvy and media awareness beat out legacy capital by taking advantage of the society-scale transition from television to digital media—a transition that saw traditional advertising decline year after year while digital advertising rose by roughly 3,400 percent in less than 20 years. Then, as large companies caught on to the shift, the cost of media rose, and buying attention on the internet became a lot like buying attention on television: expensive and competitive.

BuzzFeed didn’t collapse because the internet moved on from its love affair with listicles and cat memes (as a cursory glance at Instagram proves), and it didn’t collapse because of its genius founder’s hubris. It collapsed because the unit economics of online attention changed and the arbitrage opportunity went away. Big brands caught up.

Smith is right to think of this period as the end of an era. But rather than remember these companies as cultural protagonists, run by wily geniuses who discovered why we click on salacious headlines or endlessly share heretofore personal details, we should see them as the casualties of a brief gold rush—the forty-niners who showed up early and made it big, only to be booted off the land by the technologists who never really needed them that much in the first place.