Reducing advertising fraud with adChain’s token-curated registry

adChain (token: ADT) is a community-driven list of trusted publishers (i.e. websites where advertisers can place ads). The community curates this list through a so-called token-curated registry. A token-curated registry (TCR) is, as the name suggests, a curated registry of items that satisfy some criteria. As such, token-curated registries are one of many Cryptoeconomic Primitives.

More about Cryptoeconomic Primitives here.

More about token-curated registries here.

In the case of adChain, the list is the sum of all publishers (also called members), and the criterium is “trusted”. Through voting, token holders (i.e. those that ADTs) decide which publishers are added to the list.

Screenshot of adChain's token-curated registry (Source: Website)
Screenshot of adChain’s token-curated registry (Source: Website)

As adChain’s token-curated registry is designed as such that token holders will only profit if they add trusted publishers, this voting-based determination is supposed to remove “views fraud”. Publishers charge advertisers for their ad space based on impressions. Because distinguishing between real and fake impressions (e.g. bots) is difficult, publishers tend to artificially increase ad impressions to charge higher fees. Because incentives are aligned neutrally on adChain’s token-curated registry, publishers are naturally incentivized to act correctly. If advertisers regard adChain’s publisher list positively, publishers will be highly motivated to be on it. And because token holders only profit if their selected publishers are trusted, they will naturally only vote for such. Thus, because publishers want to be on the list, they must act legitimately because token holders, which decided whether publishers are added, only add legitimate publishers.

More about token-curated registries here.

adChain’s application process

adChain’s application process is the archetypical application process for token-curated registries consisting of the initial application and the associated Challenge Period.

Application

Publishers apply to the list by depositing ADTs. They are now in the Application stage

An application in Application stage (Source: Website)
An application in Application stage (Source: Website)

Challenge Period

After the application, the publisher must wait in the so-called Challenge Period. During this Challenge Period, token holders can evaluate the applicant and raise a so-called challenge if they mistrust the applicant. A challenge is an election where token holders actively vote for or against the publisher, whereas “for” means “not fraudulent” and “against” means “yes fraudulent”. Before raising a challenge, the token holder must make deposits equaling the publisher’s deposit. The token holder who initiated the challenge is now called the “challenger”.

An application in Application stage (Source: Website)
An application in Application stage (Source: Website)

If nobody considers the publisher fraudulent, she is added to the list and gets her deposit back.

An applicant that was added without a challenge (Source: Website).
An applicant that was added without a challenge (Source: Website).

If a token holder considers the publisher fraudulent, she can raise a challenge.

If the majority of token holders vote for the publisher, the publisher wins, if the majority votes against her, the challenger wins. The election outcome determines the publisher’s status and the distribution of deposits.

  • Challenger winning: If the challenger wins, the publisher is not listed. The publisher’s deposit is distributed among the challenger and the winning voters. Unsuccessful applications can reapply as often as they want
A rejected application (Source: Website)
A rejected application (Source: Website)
  • Publisher winning: If the publisher wins, she is added to the list. The challenger’s deposit is distributed among the publisher and the winning voters. However, publishers can be delisted (through a challenge). This is relevant if a publisher goes rogue.

Long-term effects resulting from adChain’s token-curated registry

This voting model brings with it a range long-term effects

Application deposits prevent spam

Because publishers lose money if their application is unsuccessful, only legitimate ones will apply.

Increasing token value

The more useful adChain’s list becomes, the more demand it will see from all participants (publishers, advertisers, token holders). As a consequence, adChain’s token grows in value. Thisembodiess one of the essential attributes of cryptocurrencies: only if they are inextricably linked to a platform, they get real-life value.

Token-curated registries are not necessarily a product, but rather a feature or paradigm

Although this post might give the impression that the token-curated registry is adChain’s main product, it is actually more a feature as adChain sees itself as an „advertising supply chain“. If we consider that lists exist everywhere, one can go so far and question whether token-curated registries will become a paradigm (e.g. like object-oriented programming) in the future. Thus, in the long-run, token-curated registries might be part of all kinds of projects without anybody even noticing it.

Open questions

Although adChain was one of the first (if not the first) to propose a token-curated registry, there are a few open questions:

Assessing legitimacy and on-chain as preconditions for token-curated registries

Token holders must asses publishers’ legitimacy (i.e. fraudulent or not). However, as long as publishers do not share their actual advertising data, voters cannot make fully informed decisions based on facts but rather only their own opinions. This rises the question whether adChain’s use case can actually be fully implemented via token-curated registries and what preconditions for token-curated registries are. One such pre-condition is that voting only happens based on on-chain data. By having a more transparent access to the data that is used for voting decisions, more confident votes can be issued.

As vote following is impossible, the likelihood of attacks increase

adChain’s commit-reveal voting makes „vote following“ impossible. Vote following, like in the case of DFINITY’s “Automatic voting by following other neurons“, is a setting where token holders automatically follow the decisions of other holders. The absence of vote following could lead to “non-voters” who hold tokens but do not vote (if vote following were possible every token holder would have a vote even if that vote was based on somebody else). Non-voters are problematic because they make 51% attack less costly.

More on DFINITY and its voting process (as part of their Blockchain-Governance) here.

Competing interests among members

adChain’s token-curated registry contains competing publishers. These publishers have two conflicting goals. On the one side, they want to be part of a respected list. By design, such lists contain multiple publishers. On the other side, they do not want competition. Whereas competition is a problem that exists everywhere, it is especially problematic in the context of token-curated registries because publishers can be voters as well. Through that, they can influence the list’s members. As such they are likely to try preventing competitors from joining.

Member by default-model decreases list quality

Applicants, after the challenge period has passed, are added automatically to the list without anybody questioning them. This can be problematic if applicants outweigh token holders; due to the sheer amount of applicants, token holders won’t be able to review all applications and fraudulent ones would slip through.

Sign up for Newsletter: