A Brief History of Web Monetization: From Banner Ads to Streaming Micropayments

In October 1994, the website HotWired ran what is generally considered the first paid banner advertisement on the World Wide Web. The advertiser was AT&T. The ad was a small graphic with the text "Have you ever clicked your mouse right HERE?" The reported click-through rate was around 44 per cent. The cost was approximately thirty thousand US dollars for a three-month placement.

It is hard to overstate how strange this number looks from the present. A 44 per cent click-through rate would be considered fraudulent today; the modern average for display advertising is between 0.05 and 0.5 per cent. But in 1994 the web was new, the population of users was small and technically curious, and the experience of seeing a clickable link to an external company's website was novel enough that people clicked on it just to find out what would happen.

That moment is the conventional starting point for the history of web monetisation. What follows is, in essence, thirty years of trying to find a way to keep that economy working as the web grew from a few thousand pages to several billion, from a curiosity to the world's primary medium of communication, and from a technical novelty to a saturated commercial environment. The story has several distinct chapters, each defined by an attempt to address the failure modes of the chapter before it.

The first era: banner ads and the dot-com era (1994–2000)

Through the late 1990s, web advertising was simple in form: graphic banners sold by the publisher to the advertiser, placed on the page, paid for by impression or by click. The economics were favourable for publishers, who could charge premium rates because the audience was technically engaged and the medium was new. The economics were also favourable for advertisers, because click-through rates were high and conversion was strong.

By the late 1990s the model was already showing strain. The novelty of clicking on banners had worn off; click-through rates had fallen by an order of magnitude. The supply of banner inventory had grown faster than the demand. Banner-blindness — the cognitive habit of ignoring rectangular regions of a page that look like advertisements — had become measurable in user studies.

The dot-com crash in 2000 ended the first era abruptly. Many publishers who had built businesses on banner-ad revenue could not survive the collapse in advertising rates. The survivors were generally those with strong direct audience relationships — search engines, email providers, content portals — who could substitute search advertising or email advertising for the failing banner economy.

The second era: AdSense and the long tail (2003–2010)

Google launched AdSense in 2003. The proposition was elegant: any website operator, regardless of size, could place a small piece of JavaScript on their page and Google would automatically fill the resulting ad slot with the most relevant text or image advertisement from its inventory. The publisher would be paid a share of the advertiser's bid. Google handled the auction, the targeting, the fraud detection, and the payment. The publisher had only to install the code.

This was a genuine technical and economic innovation. It made small-scale web publishing economically viable at a level of granularity that had not previously existed. A blogger writing about an obscure topic could now monetise their traffic without needing to negotiate with advertisers directly. The number of sites running AdSense grew quickly into the millions.

The second era ran for roughly the rest of the decade. By the late 2000s its limitations were becoming visible. The auction-based model rewarded publishers whose audiences were demographically valuable and penalised those whose audiences were not, even when the latter produced more or better content. The middlemen — Google, the ad exchanges, the demand-side platforms — captured a growing share of the value, and the proportion reaching the publisher fell. Ad placements grew more numerous and more aggressive on every page, eroding the user experience that had attracted the audience in the first place. Click-fraud became a significant problem at scale.

The third era: programmatic, data, and the surveillance economy (2010–2018)

The technology of digital advertising became substantially more sophisticated during the early 2010s. Programmatic buying — automated, real-time auctions for individual ad impressions — replaced bulk placements. Advertisers no longer bought "the front page of a website"; they bought "the next impression that will be shown to a user matching this profile". The data infrastructure to support this — cookies, fingerprinting, identity-resolution platforms — became central to the economy.

The third era is the era in which web advertising became, more than anything else, a data business. Whose data, what data, how that data could be combined with other data, how individuals could be identified across devices and contexts: these became the operational questions of the industry. The publisher's page was incidental. What mattered was the user.

The third era ended for two reasons, both originating outside the advertising industry itself. The first was regulatory: the European Union's General Data Protection Regulation, which came into force in 2018, made many of the data practices on which programmatic advertising depended legally fragile. The second was technical: Apple's changes to Safari's third-party cookie handling, which began in 2017, and the subsequent changes by Mozilla, Microsoft, and eventually Google itself, which removed the foundational identification primitive on which the era had been built.

The fourth era: the unraveling and the search for alternatives (2018–present)

The current era is in some ways a search for what comes next. Several distinct strands have developed in parallel.

The first strand is the consolidation of programmatic advertising into a smaller number of "walled garden" platforms — primarily Google, Meta, Amazon, and TikTok — that own enough first-party data to operate independently of the third-party-cookie ecosystem. For publishers outside these walled gardens, this consolidation has continued the long-running trend of rising intermediation fees and falling per-impression revenue.

The second strand is the rise of subscriptions and direct support. The New York Times, the Financial Times, Substack, Patreon, Bandcamp, and many smaller efforts have built businesses on the proposition that some readers will pay directly for content they value. These models have produced sustainable economics for a small number of high-quality publishers. They have not produced a generally applicable solution: most web pages are not the New York Times, and most readers are not willing to subscribe to most pages they read.

The third strand is the development of alternative payment primitives — micropayments, streaming payments, and tokens — that try to bypass the advertising economy entirely. Coil, founded in 2018, attempted a streaming-micropayments model based on the Web Monetization standard developed at the W3C; the company sunset its consumer service in 2023, though the underlying standard continues to be developed by the Interledger Foundation. Brave Browser, launched in 2016, paired its own browser with the Basic Attention Token (BAT) to pay users for watching opt-in ads. These efforts have produced working systems with real users, though none has yet reached a scale that materially shifts the economics of the wider web.

The fourth strand is the use of cryptocurrency-based reward systems applied to existing advertising flows rather than replacing them. Several projects have proposed mechanisms in which the value generated by advertising is made transparent on a public ledger and distributed to multiple parties: viewers, publishers, recruiters, network operators. This is the strand we work in directly, and the one that is the subject of our own pending UK patent application (GB2606991.4, filed March 2026), which describes a method for distributing tokens via verified content exposure events across a multi-level attribution network.

What the patterns reveal

Looking at the thirty-year arc, certain patterns are visible. Each era began with a structural innovation that made monetisation economically viable for a wider range of publishers. Each era ended when the structure of intermediation that supported the innovation captured enough of the economic surplus to make participation marginal for anyone outside the largest platforms. The transitions between eras were not driven primarily by the failures of the existing system; they were driven by exogenous changes — regulation, browser-vendor decisions, audience behaviour shifts — that broke the existing system's assumptions.

The current era seems likely to follow the same pattern. The walled gardens will continue to consolidate as long as the data-and-identity advantage they hold remains compounding. Subscriptions will continue to work for a small number of premium publishers. Alternative-payment systems will continue to grow at a slow but real rate. None of these strands is on its own a successor to the AdSense era of broad-based small-publisher monetisation. The question is whether some combination of them will produce a sustainable web economy for the long tail, or whether the long tail will simply recede into the walled gardens and disappear.

Our own bet is that transparency of value flow is the missing element. The successful eras of the past worked because the relationship between the publisher's value contribution and the publisher's revenue was legible: a banner sold for a known amount; an AdSense impression paid a known share. The current era has lost that legibility. A publisher running a page with five different programmatic networks does not know, in any detail, how the per-impression revenue split was calculated, what data was used to calculate it, or what fraction of the advertiser's spend reached them. Restoring legibility — making the value flow visible to every party in the chain — is, we think, the missing technical primitive of web monetisation.

What that primitive looks like in practice, and how it interacts with the existing advertising stack, is the subject of much of our other writing on this site. The history above is the context in which the question arises. The web has been trying to pay its creators for thirty years. The next chapter is being written now.

Sources and further reading

The history above draws on a number of well-established sources. The HotWired story has been documented by the Internet Archive and in the trade press at the time. The AdSense launch is recorded in Google's own historical filings and contemporary reporting. The dot-com crash and its effects on advertising are described in many standard accounts of the period. The development of programmatic advertising is documented extensively in the trade publications AdExchanger and Digiday. The privacy regulation history is in the public record of the European Commission, the United Kingdom Information Commissioner's Office, and the corresponding bodies of other jurisdictions. The Coil and BAT projects have published their own retrospectives. Anyone wishing to go deeper than this short summary can find substantial primary material on each of the topics covered.