The Ad-Based Internet Is About to Collapse. What Comes Next?
The web as we know it relies on advertising, but that model is headed for a crash. Fortunately, we can build something better from the wreckage.
Online, ads are everywhere.
They bombard you when you try to read the news. They pop up between your friends’ Facebook updates. They’re disguised to look like regular results on Google. And one, maybe two play before every video you watch on YouTube — with more peppered throughout.
From the perspective of an internet user who is desperately trying to ignore, avoid, or block this constant deluge of ads — ads that have to get more and more intrusive in order to force us to pay attention to them — the power of the online advertising industry might appear unstoppable. Yet the digital ad market is a lot more vulnerable than it seems.
Advertising has played an integral role in shaping the internet as we experience it today, as it provided an easy and obvious way for services and websites to generate revenue. Web developers, website editors, and online business owners simply had to pop a few ad boxes onto their sites, and they were making money. But behind those ads is a vast infrastructure designed to track, target, and serve them, all paired with increasingly complex markets whose lack of transparency makes them ripe for speculation. A growing number of experts and analysts believe that these inherent flaws are creating a financial bubble that, when it bursts, could fundamentally reshape the internet as we know it.
The big question is: What kind of web would emerge on the other side? What would an ad-free internet look and feel like?
An internet designed to serve ads
“You can almost take any piece of the web, from the smallest thing to the biggest thing, and basically say, ‘What does the role of advertising play in shaping this?’ It’s sort of a gravitational force,” says Tim Hwang, author of Subprime Attention Crisis: Advertising and the Time Bomb at the Heart of the Internet. “It acts on everything you see online.”
The platforms that comprise the bulk of the internet as experienced by the average user — Facebook, Google, Twitter, Instagram — have been built from the ground up to profile you for advertising purposes. To target their ads, the companies need details about you and what you are likely to be thinking about as you use the internet, so they created ways to quantify your attention. “The creation of a ‘like’ button is driven by the financial motivation of making attention legible online,” Hwang says.
Since the beginning, Facebook has had a “real name” policy that required users to create their accounts using their actual names. The company said the policy was intended to make it harder for people to anonymously harass others. But if you know who everyone is, it’s also much easier to target them with ads. Forcing people to register for accounts with verified information helps the company connect the data it collects over a long period of time with that user, and when people add friends, input their interests, and indicate where they live and work, these things all help serve the platform’s targeting capabilities.
Advertising’s influence goes beyond the design of the platforms and also affects the content that gets created. When I spoke to Hwang, he pointed to the listicle as emblematic of content crafted for the ad model. “You need a form of media to keep people on the page and maximize time on site,” he says. “One way of doing that is to string an enormous number of images together that they literally have to scroll down to.” Hwang also believes content made to maximize ad revenue has affected broader problems like polarization. Yet this economic power can also mask fundamental problems within ad markets.
There’s a common narrative that the algorithmic targeting abilities of platforms like Facebook and Google are changing how people think, but a closer look reveals these claims to be dubious. In his book How to Destroy Surveillance Capitalism, which was published here on OneZero in September, author and digital rights advocate Cory Doctorow argues that new technologies have “greatly boosted the efficacy of Big Tech’s sales pitch as marketers have exploited potential customers’ lack of technical sophistication to get away with breathtaking acts of overpromising and under delivering.” As Hwang puts it, “The whole edifice of online advertising, in short, is bunk.”
How financialization created an ad bubble
Every time you see an ad online, “your attention has been sold by the platform and bought by the advertiser,” Hwang writes. Ads are not selected by the platform or website you’re visiting. The ads you see are determined by algorithmic auctions that take place in the milliseconds between when you click a link and the webpage is displayed for you.
“In the process of buying an ad, there can be tens or dozens or hundreds of intermediaries and no one would ever know,” says Shoshana Wodinsky, a journalist who covers the ad tech industry for Gizmodo. “In fact, the advertiser might not know — and the consumer on the other end definitely doesn’t.” That’s all facilitated by marketplaces that have been highly automated and were explicitly designed to work that way by dominant internet companies.
“Online advertising is more Wall Street than local book fair,” Hwang writes in his book. As the internet began to be commercialized, it was not a given that advertising would emerge as such a central business model. It was the product of decisions made by corporate actors, and Google played a key role.
Hwang explains that the development of Google’s advertising platforms — AdWords for advertisers and AdSense for publishers — were led not by executives with business backgrounds, but rather by Eric Veach, a Stanford computer science graduate. Co-founder Larry Page wanted to replace traditional sales teams entirely, but that couldn’t be done by computer scientists alone. To fulfill its vision, Google hired economist Hal Varian in 2002, who would go on to become its chief economist. It was Varian who modeled Google’s digital ad markets on financial markets, replacing salespeople with automated algorithms. But adopting this model also came with trade-offs.
To sell attention on a large scale, ad inventory had to be standardized and abstracted. “The amorphous, shapeless concept of attention has been transformed into discrete, comparable pieces that can be captured, priced, and sold,” Hwang writes. But losing the context of where that attention comes from can lead to “irrational levels of market confidence, a regular feature of financial crises going back hundreds of years,” because the worsening foundations of those markets become harder to see or easier to ignore. For example, even as the value of the digital ad industry was continuing to rise, the average clickthrough rate on Google’s display ads fell to 0.46% in 2018, ad fraud was expected to jump 21% to $42 billion in 2019, and a Google study found 56% of its display ads may not even be seen by a human. These stats suggest the product being sold is not nearly as effective or valuable as many purchasers of digital ads believe it to be.
Standardization is just one piece of a structure that makes digital ad markets increasingly opaque. The automation of auctions for digital ad space, the growing number of trades that happen on private exchanges, and the effective oligopoly of Facebook and Google add their own layers of secrecy to these markets. As Wodinsky puts it, “Unless these major players actually own up about what they’re doing, you’re playing by their rules and you just have to hope they abide by them.” Needless to say, it’s not a recipe for a well-functioning market.
Hwang argues that this lack of transparency has created an ad bubble resembling the housing bubble that came crashing down in 2008 and dragged the economy down with it. If the ad bubble were to burst, causing the value of ad inventory to plummet, the business model of major platforms like Facebook and Google would collapse, and they would need to find another way to commercialize social media, search, and all of their free, ad-supported services. But beyond that, a major source of revenue for news media would decline, potentially imperiling even more journalism jobs, and websites offering everything from dictionaries to recipes would be in trouble.
The impact of this crash could be lessened, but industry and lawmakers seem unlikely to take the necessary action. “We’re in this situation where, long term, it looks very bad,” Hwang explains, “but short term, we have incentives to keep playing the game.” He suggests setting up a commission similar to the Pecora Commission, which drafted new regulations for the financial industry after the Great Crash of 1929, and establishing an activist research body to deflate the claims of the digital ad industry. These initiatives would inject more transparency into the system to wind down the bubble, but Hwang thinks it’s far more likely simply to burst — and that could start a battle over the future of the web.
An internet beyond advertising
The ad-dominated internet is the only internet many of us have ever known, and it can be difficult to seriously consider what an alternative might look like. The collapse of an ad bubble could usher in many possible futures that could deepen existing problems or make new opportunities available to us — its users. But what future ultimately comes of such a collapse will depend on who is prepared with an alternative and who has the power to push their agenda forward. This could go in many different directions, and this section will consider three very different scenarios.
A further monopolized internet
Let’s begin with the most obvious possible future. If the ad bubble bursts and the market is allowed to respond with little regulatory response, the outcome could intensify the monopolization that has already gripped much of the web we currently experience.
Doctorow, an advocate of antitrust enforcement, believes a major disruption in the ad market could lead to “multiple, competing firms being bought out by the same deep-pocketed entities, who will zero out any equity stake held by founders or workers, wring concessions out of suppliers and creditors… and a bid to buoy up ad prices by restricting inventory and using cartel-like behavior to put the advertisers’ backs against the wall.”
The ad industry itself is “sprawling,” Wodinsky told me, with 8,000 adtech companies in the United States alone. Those companies would see their values plummet, which Hwang argues would “generate an intense phase of consolidation beyond the rarefied confines of the Google-Facebook duopoly.” And while the ad duopoly may face its own challenges, he told me, “Some of the companies may actually be able to get bigger.”
Unlike smaller companies in the advertising space, Facebook and Google would have the cash reserves to survive an ad crash, but that doesn’t mean a big hit to their primary revenue source wouldn’t make them vulnerable. They would have to swiftly find a new business model that would likely see their previously free services supplemented with subscriptions, if not placed behind a paywall entirely. Imagine a Facebook that limits the number of friends you can add unless you buy a subscription, offers an ad-free tier, or requires payment to create groups and pages. Alternatively, Google might still let you search for free, but the Google Docs software suite might require a monthly subscription.
The collapse of Facebook and Google’s business models would also leave an opening for other monopolies to eat into their market share. Amazon’s business strategy has revolved around using the profits it extracts from its existing businesses, especially cloud infrastructure, to expand into everything from film production to grocery stores. Meanwhile, Apple has been using profits from its hardware products to expand its suite of digital services. The company already has email and office software it could make more attractive, and maybe it would even take another shot at social media.
Facebook and Google’s services may become less attractive with new restrictions, while the other monopolies could create alternatives for Prime members or iPhone users — if they don’t simply try to buy their struggling competitors outright. Yet such a shock as the one we’re describing could also spur a serious effort to break up the monopolies, which would in theory create an opening for more experimentation with business models that reorient how we use the internet.
A more diverse, pluralistic internet
“The ubiquity and the monoculture of ads has prevented other types of business models from forming,” Hwang says. But if we reached a point where ads didn’t produce the same kind of returns, “people would finally be able to experiment.” Those new models could also bring significant benefits to a web that’s currently dominated by a few massive platforms.
“We should have a LOT of firms, with a LOT of publications, paying a LOT of creators, to reach a very wide diversity of audiences,” Doctorow told me. “For that to happen, we need both a restructuring of ad markets, but also a restructuring of search, social media and app stores.” At present, the monopolies use their power to make it hard to compete with their dominant platforms — but that can be changed.
In the United States, federal and state authorities are already conducting antitrust investigations into Amazon, Facebook, and Google. On October 20, the Department of Justice sued Google over its monopoly on search and search advertising. But given the right political circumstances, the rupture caused by an ad collapse could accelerate antitrust action to not only break up tech monopolies but also enforce rules that would limit mergers, acquisitions, and even the predatory pricing that has allowed companies to use venture capital to undercut their competition. Changing the rules that govern the market could then lead to very different outcomes.
Doctorow argues that enforcing interoperability — when your various services work together regardless of who owns them — would eradicate the walled gardens erected by the major platforms. That would allow people to access posts, messages, and other data on whatever app they wanted to, instead of being beholden to the platform giants. Your iOS Messages would be portable to Signal, Facebook posts to Mastodon, and so on. Platforms would have to compete on features and user experience, not by locking your friends and family into one platform by default.
Then, if you were fed up with Facebook and there was another app that promised not to track you or sell your data, with interoperability you could switch to that new app and still communicate with your Facebook friends. In essence, the platform loses the ability to trap you and instead has to win you over.
For example, maybe you want to keep using Facebook, but it requires you give up even more of your data to remain profitable. Some people might be okay with that, but maybe you really care about privacy. So, thanks to interoperability, you subscribe to a new app called Securebook, where you can still communicate with all your Facebook friends, but there are no ads or tracking. Or maybe you go to Lookbook, where you get similar features, but it’s a barebones operation and the owners display only enough ads to keep it running, not turn big profits. But that’s fine because Facebook has to let you access your data from anywhere.
Both Doctorow and Hwang are excited about the different prospects that could emerge in such a scenario, but it could also present challenges to maintain equity online. Hwang explains that without financial speculation to inflate the value of advertising, the web may not generate as much revenue. For some, that would be “a painful transition, because it’s unclear if the aggregate size of the pie is just smaller,” he says, but for others, it would create an opportunity that was previously stifled by the dominance of the ad model. Hwang also warns that this future could create a more fragmented web where “some people will be able to afford the internet they knew, while other people will be pushed to a less featureful internet.” Advertising has funded all these free services that are the same whether you’re Mark Zuckerberg or one of Facebook’s poorly paid content moderators, but a proliferation of business models could also correspond with a greater inequity in our online experience.
We already see that in news, where declining ad revenue has not only led to the loss of thousands of journalism jobs but also incentivized major publications like the New York Times and the Washington Post, among many others, to add paywalls to their websites. While this brings in a new revenue stream for the publications, it arguably makes high-quality reporting harder to access, while the misleading, if not outright false, stories tend to be free and spread like wildfire on social media. This is a phenomenon that will need to be addressed in the ad-free web, and to locate possible alternatives, perhaps we should look to the past, when a more public-run internet reigned.
An internet that serves the public good
The digital economy emerged during a peak neoliberal period — the early 1990s to the 2000s — but it was possible only because of certain public policies. Historian Margaret O’Mara explains that for decades, the U.S. government poured public research money into key hubs around the country — Silicon Valley being one, Boston another — to ensure it didn’t fall behind the Soviet Union. That public funding produced many of the technologies that have been key to the digital revolution, including the ARPANET and NSFNET, the forerunners to the internet. But in the early 1990s, as the Soviet Union collapsed and traffic on the NSFNET was picking up, the privatization of the internet became an explicit policy of Bill Clinton’s administration. Under his watch, the public backbone of the internet was retired, the web was explicitly opened to commercial activity, and few rules were put in place to ensure what happened online was in the public good.
Those decisions, paired with a lack of antitrust enforcement, led to the monopolized internet we have today, where most activity takes place on a handful of commercial platforms. While this produced those free, ad-funded services that were available to everyone, it has also served to place the profit motive before the public good. A less monopolized and more experimental web may revive the question of whether there should be more public ownership or worker control.
“Advertising has allowed us to punt the question of whether or not some of these services are fundamental services that you have a right to,” Hwang says. “But in a world where everybody needs to pay, the question of whether or not it’s a utility becomes a lot more clear.” He believes “we’re likely to see more nonprofit, cooperative models emerge.” Could they be the key to a better internet and, by extension, a better society?
Logic editor Ben Tarnoff suggests that the proper organizational structure would depend on the scale of the service. In some cases, cooperatives would be ideal. Instead of a global Uber that mistreats and underpays drivers, there could be a federation of local ride-hailing cooperatives sharing technology and best practices so their communities benefit while drivers can still make a good living. They could even work with local transit agencies to build applications that encourage people to use sustainable modes of transportation instead of just what’s best for a company’s business goals.
In Europe, CoopCycle serves as a federation of bicycle delivery cooperatives, while Fairbnb is trying to create a nonextractive vacation-rental platform. But these models will always struggle when they have to compete with private competitors backed by venture capitalists who can spend billions of dollars undercutting them with business models that shred workers’ rights and can go years losing money in a bid to monopolize their sectors.
Different approaches may also be necessary for the key infrastructures and major platforms of the internet. When you watch a Netflix video or check CNN, data is being transferred from massive data centers through a vast network of cables and other infrastructures to reach your device. Nick Srnicek, author of Platform Capitalism, argues that all those servers, often called “the cloud,” should be a 21st-century public utility, while Tarnoff believes it should also include fiber optic cables and physical pieces of the network, reflecting the public ownership that was common before the early 1990s. If people are going to be experimenting with new business models or cooperative forms, public ownership of this infrastructure would allow the government to keep prices low to encourage innovation.
Srnicek also recommends regulating dominant platforms as utilities, as railways were in the past, to ensure they’re serving the public good. But we could also imagine a different way of organizing them altogether. Instead of simply leaving that experimentation to private actors, Dan Hind proposes the creation of a publicly owned digital cooperative that would be governed not by bureaucrats, but by assemblies composed of workers and randomly selected members of the public who engage in democratic deliberation to ensure the cooperative and its activities are serving the public good.
This cooperative would build on the achievements of public broadcasting, while updating that model for the digital age. It would be tasked with creating a series of new services that foster cooperative enterprise, including e-commerce and payment platforms that don’t abuse their position as intermediary to extract high fees. Hind also thinks it should be responsible for building a new social network that brings together content from public institutions like museums, universities, and theaters to promote education and community instead of engagement and profit.
Even more exciting, it could reverse the trend of economic centralization we’ve seen in recent decades. Monopolized tech companies have hoarded jobs in a few key cities, but the digital cooperative would distribute innovation around the country by establishing offices and storefronts in cities big and small — not only to help educate people about the use of technology, but also to develop digital solutions to meet local needs, like an app for a cooperative ride-hailing service, tools to promote local governance, or solutions that help farms without taking away their rights. We could even make it part of a revamped postal service.
When we start thinking beyond the logic of a privatized web, different ways of organizing the internet can emerge. Promoting a greater role for public companies and worker cooperatives also doesn’t remove the incentive for innovation and experimentation, but rather alters the incentives so that profit is no longer the driving force and technology can be developed with more humane goals in mind.
Building a better digital public sphere
Back in 1998, when Google co-founders Larry Page and Sergey Brin were still students at Stanford University, they wrote a paper about the perverse incentives of advertising and how it “causes enough mixed incentives that it is crucial to have a competitive search engine that is transparent and in the academic realm.” More than two decades later, we can see the consequences that came with ignoring their own research, and we must recognize that the ad model is more fragile than it seems.
Having less advertising on the internet, and more relevant advertising where and when it’s still used in limited cases, stands to improve online life in multiple ways. Hwang believes this shift would open the door to “creating more diverse culture online again,” while Doctorow argues that “it will be MUCH harder for states to run mass surveillance programs,” and “our political discourse may refocus on the material conditions that make conspiracism thrive, rather than blaming high-tech manipulation.”
If we are in the midst of an internet advertising bubble, then this is the moment to think beyond the constraints of the present-day web to imagine the kind of internet we really want — and how we might build and support it. Are we content having our online lives limited to a few massive platforms while venture capitalists treat the digital economy as their personal casino? Do we want a digital sphere that is more diverse, not sanitized to conform to commercial interests, and a place where the public good comes before the profit motive? Then we should start organizing to bring it to fruition now — so we can be prepared if the digital ad economy does indeed implode.
The importance of the web to our daily lives and interactions will not lessen in the coming years and decades. The pandemic has shown both how essential it is and how the inequities that arise from its current form have material consequences. Massive corporations have been allowed to structure the internet to serve their interests, but it’s time to develop a new, more equitable vision for its future. One that isn’t dependent on advertising to operate.