The Club Good Protocol

Bringing a Balanced Economic Model to the Web

The economics of the open web have been distorted by the rise of AI agents and crawlers. Content, which was once monetized by turning limited space and reader attention into advertising value, is now treated as if it were infinite and free. The surface where attention gathers has shifted away from the pages publishers control to the interfaces of platforms and agents, leaving the underlying content disconnected from the presentation that captures value. As a result, publishers are stripped of pricing power. They risk becoming mere data vendors rather than experience creators, losing their hold on customer relationships in the process. To rebalance the system, they need mechanisms to define and sell inventory that agents recognize as scarce and valuable. The key is to realize that scarcity does not reside in the bits themselves, but in time, context, and proven utility.

To restore a fair marketplace to the open Web we can apply fundamental principles of economics to select a model that works for these new realities. The model that works best is to treat Web resources as a Club Good (https://en.wikipedia.org/wiki/Club_good). This model supports a “non-rivalrous” environment - consuming the goods and services doesn’t prevent others from consuming the same resource (as opposed to advertising where only one ad can chosen for an impression). A Club Good is also characterized by being excludable - the ability to selectively choose who participates. The Web is naturally non-rivalrous however, excludability on the Web still requires some innovation.

In order for this Club Good approach to be as open and scalable as the Web itself, we propose a small extension to the HTTP that we call The Club Good Protocol. This protocol extension is the minimum work necessary to establish the Web as a Club Good and is independent of business model, pricing model or payment method.

This article describes one approach at using The Club Good Protocol to help publishers rebuild their content based revenue streams by using another fundamental economic tool - auctions. This technical white paper provides specific details on how The Club Good Protocol is implemented as an extension to the HTTP and how it leverages the 402 Payment Required response header to make it all work.

There are three primary forms of inventory that publishers can control: base access, quality and first look (or freshness).

The first is base access. This is the everyday availability of baseline content: the ski report that says “40-inch base,” the game score that reads “3–1 in the second period,” the election result that tallies the first precincts. Publishers know this material is valuable, even if it feels like a commodity. By setting a floor price, publishers prevent their work from being treated as worthless. An AI agent can still request as much of this information as it likes, but every call carries a minimum cost. At scale, this changes the calculus: ten thousand low-value requests a day no longer round to zero but accumulate into a predictable revenue stream. Base access does not attempt to extract monopoly rents; it simply establishes that the floor is not zero.

The second form of inventory is quality, which captures the distinctive, nuanced, and contextual content that agents find improves user satisfaction. Not all snow reports are equal. One may simply note the base depth, while another observes that the morning will be powder, the afternoon will crust over, and the east slopes will be best after lunch. For some users, the first fact is enough. For others, the richness of the second report leads to better decisions, higher trust, and more engagement. Agents know this, because they measure their own performance in terms of accuracy, retention, dwell time, and user satisfaction. The challenge is that publishers cannot always know in advance which pieces of content agents will rate as high-utility. The solution is to allow broad sampling at a baseline price, but then introduce floors that rise as demand repeats - essentially dynamic reserve pricing. As a simplified example, a publisher might start by charging $1 CPM. After a few hundred hits, if demand is sustained, the floor automatically rises to $2 CPM, and so on, with a ceiling determined by optimal yield. Each response can signal the next minimum bid, so agents are on notice that continued use requires paying more. This mechanism lets the market discover the true value of quality in real time. It is not a static two-tier representation of the content; it is a sliding equilibrium where the floor ratchets upward as agents prove their willingness to pay.

The third form of inventory is first look, which monetizes freshness. Timeliness is one of the few real scarcities in digital content: there is only one “first five minutes” after the Iowa caucus polls close, the first report of morning snow conditions, or the initial posting of a new house for sale. In these moments, being first carries disproportionate value for agents whose users demand immediacy and who wish to outcompete their rivals.

One way to capture this scarcity is for publishers to sell options through daily auctions, organized by subject domain. Each day, agents bid for tomorrow’s coverage in categories like politics, sports, real estate, or entertainment. The auction allocates exclusivity windows in ranked order: the top bidder wins the first slot, the runner-up takes the next window, and so on. When the event occurs, those winning bidders exercise their rights to the time slots they secured. After the exclusivity periods expire, the content returns to baseline access for all.

For example, on caucus eve, a political news publisher might run an auction for “tomorrow’s results feed.” The top bidder secures the first five-minute window, the second bidder the following five minutes, and the third bidder the five-minute slot after that. Similarly, a real estate listing site could hold a daily auction for “tomorrow’s new property postings,” assigning early-access windows in sequence before listings enter the general feed for AI consumption.

This mechanism blends predictability with competition. Publishers earn guaranteed revenue each day through option sales, while still capturing dynamic upside when multiple agents value the same coverage. AI operators gain confidence that their customers will have the first and best access to quality information—whether that’s a quick take in the moment or a thoughtful summary soon after—which strengthens user trust and loyalty.

Taken together, these three controls form a new set of publisher inventory. Base access ensures that no request is ever priced at zero. Quality allows initial discovery but then captures upside as demand and utility prove themselves. First look transforms freshness into a tradable commodity through options and windowed auctions. In combination, they offer publishers a way to reassert pricing power in an economy that too often assumes their supply is infinite. The scarcity is there—it lives in time, attention, and measurable utility. These inventory controls make it explicit, enforceable, and monetizable.

Join the party. There will be dancing. It will be fun.