How Wikipedia Can Save the Internet With Advertising

Robin Berjon is Principal at Supramundane Agency, Deputy Director of the IPFS Foundation, and a Senior Fellow with both the Future of Technology Institute and the Public AI network.

There’s a certain kind of irony in seeing smart people ignore their own insight, as if predicting a pitfall somehow excuses them for tumbling right in. Take, for example, Google’s founders, Sergey Brin and Larry Page, in their seminal discussion of Google, “a prototype of a large-scale search engine.” In an appendix discussing “Advertising and Mixed Motives,” the authors predict “that advertising funded search engines will be inherently biased towards the advertisers and away from the needs of the consumers.” It's what happened, and they'd seen it coming! It was foretold.

Clearly, it had to happen. Did it? There is ample evidence around the world that media organizations, when they choose to, have been able to develop imperfect, occasionally failing, but nevertheless viable institutional arrangements that have shielded newsrooms from advertising money across more decades than we've had digital computers.

That Google Search would be devoured by its ad business was a choice, and an outcome that could have readily been avoided by asking anyone with media experience and designing an institution to insulate search quality from short-term revenue considerations. And yet, it keeps being presented as the result of a mechanistic law of nature. Why? Because we want to believe that only the unsustainable can be pure, and that advertising must corrupt everything. This is not wisdom, but resignation disguised as common sense.

Which brings us to the thesis that likely brought you here, perhaps full of derision: Wikipedia can save the internet with advertising. I do mean the whole internet and not itself. Wikipedia itself is under threat and could always use more money to resist these attacks, but in any case, it will not stand alone long unless the rest of the internet is saved from authoritarianism as well. (And for the purists in the back: yes, I'm using ‘the internet’ to describe the entire digital sphere because that's how people understand it, and what I'm interested in is building for everyone.)

So, how could advertising possibly save the internet? Allow me to expand on the three reasons why this strategy could finally tip the balance in favor of an open, democratic digital sphere: because Wikipedia has the institutional strength to keep the corrupting influence of advertising at bay, because Wikipedia's scale is key to making a principled alternative to the current digital advertising system possible, and because any better future for tech requires money — a lot of money — while we still have a chance to make it happen.

In Garrett Hardin's "Tragedy of the Commons"—another phrase repeated ad nauseam despite being backed by no empirical work whatsoever and motivated by reasoning meant to support his deeply racist politics and fondness for eugenics, not to mention extensively debunked by Elinor Ostrom’s research—herdsmen just haplessly keep adding their own animals to the common pasture until it collapses from overgrazing because, as he puts it, "Freedom in a commons brings ruin to all." But in real-world commons, people aren't such politically convenient idiots. They see the problems they create, they talk to one another, and they build institutions to manage them together. That's exactly what happened with Wikipedia: it grew a complex set of internal institutions to manage the process of public editing of information and knowledge.

It is precisely this institutional spine, and the community that sustains it, that gives Wikipedia the ability to keep advertising from corrupting its content and its processes. In this sense, Wikipedia is far more socio-technically advanced than Google. Google is an OKR-controlled hierarchy that just happened to grow big on a handful of rent-extraction strategies. Wikipedia, in contrast, is a dense set of democratic and technical arrangements designed to efficiently organize information in the face of overwhelming incentives for corruption. It's no accident that Wikipedia, warts and all, remains a beacon of reliability on the internet while Google has become history's biggest slop machine. Democracy may not look sleek or come with hype, but it'll beat technocratic authoritarianism every time.

Second, having an actor with the scale and ethos of Wikipedia could be transformative for digital advertising as a whole. That may sound counterintuitive: why would anyone want to improve advertising? Because we could use it for good.

Depending on which sources you pick, global digital advertising currently grosses USD $500-600 billion annually and is on track to break past USD $1 trillion around 2030. That represents a massive share of the internet's overall budget. Refusing to engage with this system leaves money to flow to the same corporations and platforms that have been defunding public information and making our lives worse. A world in which we choose to have the worst technologies and content lavishly funded while democratic alternatives languish on crumbs isn't hard to imagine — it's the one we've built for ourselves.

But we can't just plug today's dominant adtech infrastructure into projects we value and hope for the best. Digital advertising is already infamous as a privacy train wreck, but that's just one of its many failings. Its technical architecture, built atop arbitrary code injection, is also a security nightmare. It functions as an espionage vector at scale, both for people and corporations. It undermines the revenue of quality publishers of every size, while funneling money toward disinformation and slop farms, just so that intermediaries can make more money.

A 2016 paper from the World Federation of Advertisers estimated that 10 to 30% of digital advertising is lost to ad fraud, and that figure doesn't even account for the vast sums that subsidize slop and disinformation. On top of that, we have good reasons to believe that a lot of ad spend isn't accomplishing anything, meaning advertisers themselves would benefit from systemic reform. Clearly, we shouldn't be incorporating such a system into any part of the internet we care about.

Alternatives that avoid these downsides are, however, possible. They are more sophisticated than the dismissive "just use contextual" ( though well-designed contextual systems do exist), and exploring them in detail would make this piece far too long. But what matters is that they are feasible, and they would benefit people, publishers, and advertisers alike. So what's holding them back?

Part of the problem is seed funding, but by far the greater challenge is scale. Scale is key because marketers must invest effort in each channel that they advertise on. This means that even if a new adtech platform promises better conversions, greater spend transparency, and safer, faster, and more sustainable campaigns, it's still not worth the effort unless it can deliver enough reach, i.e., put ads in front of enough people, to make it worth the time it takes to set up and run a campaign. Worse, achieving scale is particularly hard because advertising is a two-sided market: not only can you not get marketers interested if you don't have a sufficient volume of impressions, but you also cannot get publishers interested if you don't have enough marketers buying on your platform. In other words, neither the egg nor the chicken wants to go first.

Wikipedia, of course, solves that problem. It consistently ranks among the top ten most visited sites globally, well ahead of major advertising properties like Amazon, LinkedIn, or Pornhub. A global advertising infrastructure built and governed to operate in the public interest, pioneered by Wikipedia and adopted by other services over time, could quickly emerge as a major player. It would be able to capture funds that currently fuel our broligarchic dystopia and redirect them toward projects that serve the public good. And in the process, it would introduce an alternative advertising model with the potential to reshape privacy norms and incentive structures across the entire internet.

Which brings us to the third argument: we need the money. The internet needs regime change, and regime change takes more than hope. Much of the ideological underpinnings of today's digital world come from a place of privilege in which we can pretend that money doesn't matter. To say that it hasn't served us well is an understatement.

The alternative to ignoring money isn't necessarily crass commercialism. A public interest internet still needs money, and advertising, however much we may dislike it, is a more plausible, scalable, and sustainable option than donations, tipping, or the perennially vague promises of micropayments. In this sense, advertising can be seen as a tax on screen real estate, one we can choose to govern in the public interest.

Pragmatically, many of the people maintaining the open web, open source, or open content are burning out for lack of financial support. Promising services built on open protocols, designed to empower users, remain stuck in semi-usable limbo because their creators cannot find a business model that they are willing to subject their users to. Meanwhile, newsrooms around the world are facing extinction by dint of an advertising ecosystem actively working against them.

Wikipedia is uniquely positioned to change this. It could build and support the foundational technologies and processes for the privacy-first, publisher-first advertising online system that is desperately needed.

We can ballpark how much money is at stake using a conservative set of assumptions:

• Page views. Wikipedia receives about 300 billion page views per year (296 billion in 2024).
• Ad slots. Assume just one unobtrusive ad slot per page. In fact, given how heavily traffic skews toward the most popular articles, it may be possible to limit ads to those pages and still generate substantial revenue.
• CPM. Numbers that I've seen reported for purely contextual advertising are around USD $7-8 CPM (cost per mille). Using a conservative figure of USD $5 CPM accounts for uncertainty for safety, and in practice, the number may be higher, since we are assuming a single ad per page.
• Viewability. Factoring in ad blockers and viewability issues, we can assume that about a third of page views won't lead to an ad impression.

This works out to: (300 × 10^9) × 0.005 × ⅔ = USD $1bn.

At the scale of the internet, one billion US dollars a year isn't that much. It won't save the internet on its own, but it's a toehold. It is enough to begin turning the tide. It makes it possible to fund shared digital infrastructure in the public interest, which can help roll back the spreading authoritarianism and open up new avenues for sustainable support.

I'm keenly aware that this is not—by far—the first time that advertising on Wikipedia has been raised, a debate that has, in true Wikipedia fashion, been helpfully documented on Wikipedia itself. But I believe that the times have changed. We know what the internet looks like when we aren't realistic about funding. We see how urgently democracy needs a funding champion with both deep pockets and the institutional capacity to direct resources where they are most needed.

My proposal isn't just about funding Wikipedia: it's about mobilizing Wikipedia's earning potential to also seed public interest alternatives to today's digital dystopia. Wikimedia could dedicate much of this new revenue stream to a Public Interest Internet Fund with which to launch projects, and the advertising infrastructure could additionally be used to produce a real revenue source for news, public service information, and other sites that utilize it. We know that such an approach couldn't possibly use an adtech provider like Google (as had been previously proposed) but would instead be in a position to develop an alternative that would help the rest of the internet work better.

Importantly, I don't have numbers to back this, but I am convinced that those of us in the internet community who still believe in democracy now understand that we must be far more ambitious in our goals and far more realistic in the means we deploy to achieve them.

Strategy is the art of creating power. The Wikipedia community is in a position to create substantive power for the public interest internet at a time when it is being pushed to the edge of extinction. We will not survive without the funds to take a stand. We will not survive without a principled advertising infrastructure that can sustain usable, viable public interest digital services. Perhaps, if we can stop retreating to self-defeating principles with no grounding in empirical reality, we can try something different: We can choose to win.