Learnings from MetaX on Launching The First Token-Curated Registry (TCR)
Our white paper brings to light a vision for a token based platform built on Ethereum called adChain. The idea behind it was to give people from around the world the ability to curate a single list of digital publishers so that advertisers could use it as a public utility.
In short, advertisers have a hard time figuring out what websites are fraudulent or not and this list would help them in determining whether a digital publisher was legitimate. The elegance of this list, however, is as much in the substance of the list itself, as it is in the form of how curation is accomplished.
We designed a clever incentive game that leveraged a cryptocurrency called adToken (ADT). Using adToken, a community could curate a list of digital publishers based on their own set of quality standards. If they did a good job curating that list then advertisers and publishers would adopt it as part of their digital media buying campaign strategy.
Adoption has always been the main goal for adChain, with token holders standing to gain the most if it were achieved.
The crypto mechanism we created to facilitate curation became more generally known as Token-Curated Registries, or TCRs, for short.
If you are unfamiliar with TCRs and would like to know more about how they work then we encourage you to read this before continuing.
Soon after describing our TCR concept to the world it became clear that other projects in the Ethereum space had a need for something similar. It turns out, leveraging a community of token holders to curate a list could be applied to a number of verticals. TCRs began springing to life across multiple teams in variant ways.
Take Civil, for example, they are using a TCR for determining what Newsrooms should be on their platform. FOAM is using a TCR for Crypto-Spacial Coordinates (CGC) and Points of Interest (POI). Even Vitalik got in on the hype when he tweeted out his VCR (Vitalik Curated Registry) of his social media accounts.
The TCR codebase is open-sourced on Github under the Apache 2.0 license and has been forked a total of 96 times. Seeing the Ethereum community embrace the TCR concept with such open arms has been extremely validating. Of course, there have been detractors too. But most of them focus their critique on design considerations and/or tweaks to the mechanism rather than outright condemnation.
Our team consists of people from all walks of life. And so it should come as no surprise to hear that we have deliberated internally about the nature of TCRs and the mechanisms employed in adChain extensively. The current adChain Publisher Registry is the final output of countless discussions and debates around mechanism design, game theory, UX, token voting, attack vectors, governance, and many other topics.
Aside from TCR design, we were also looking to create a new future for the digital advertising industry, which was valued at $209 billion worldwide in 2017. MetaX and ConsenSys built the adChain Registry together to address issues within the digital advertising industry such as:
- Combatting the Duopoly
- Domain spoofing
- Ad Fraud
- Brand Safety
- Quality Assurance for Media Buyers
The adChain Publisher Registry officially went live on the Ethereum Mainnet in April 2018 marking a historic accomplishment.
But being the first also has its disadvantages. There are no signposts that lead to the way or lessons passed down from those who have gone before. This is something many in the Ethereum community can relate to. In a way, we are all in the same boat together plotting a course into the unknown.
Leaving the safe harbor of the Rinkeby Testnet for the unchartered waters of the Mainnet has provided us with a number of valuable lessons. We feel that it is our duty to share those lessons with all of you.
What We Have Learned
TLDR: TCRs are really great but they still need a lot of attention.
The following are insights and learnings from our entire team. We felt this post deserved all of our attention because TCRs are forming the backbone of many Ethereum dapps and it’s important that we pass along our findings so that others can benefit from the information we provide.
So without further delay here is our 4 main lessons from launching the first TCR on Ethereum:
TCRs Need Guidance
In the adChain white paper, assumptions were made as to how people would interact with the adChain Registry. Initially, we believed that users would form out of band communication channels to standardize guidelines and best practices for voting. We assumed users would come to an agreement on what is a subjectively good or bad digital publisher by using these standards. It was important for us (MetaX) to take a hands-off approach to offer guidance in the registry. We did not want to be a centralized entity telling participants how to vote. However, without proper guidance, many users did not know what criteria they were voting on and the token holders haven’t yet taken it upon themselves to form any out of band communication channels to form guidelines.
In the absence of collective opinion, token holders have turned to personal opinions when making voting decisions. Facebook was kicked out of the registry due to data privacy concerns during the Cambridge Analytica scandal. Spankchain was voted into the registry and this sparked controversy over brand safety concerns. In both of those cases, there were token holders arguing on both sides without any clear indication as to who was right or wrong and most importantly, why?
All of this led to us (MetaX) drafting a Google doc of Publisher Guidelines that we hoped would be a starting point for the adToken community to come to a consensus about what should qualify (or disqualify) a publisher from being in the registry. But, much to our dismay, the Publisher Guidelines have largely been ignored by the community.
Lack of user participation in Ethereum seems to be a symptom of a larger condition. In a recent blog post by Loom Network they called attention to the fact that,
“If you add up ALL the 100+ active DApps on dapp.review combined, it’s still less than 10,000 users in total who are interacting with an Ethereum DApp on a given day.”
There are many reasons for this, but the main takeaway from that article is that given the complexity and bad UX of dapps right now, only a tiny minority of highly-motivated users can be expected to use dapps. Most ordinary people (non-crypto enthusiasts) don’t have the time or energy to devote to using Ethereum dapps because of transaction wait times and the high learning curve associated with using them.
The cryptoeconomic model baked into TCRs places control in the hands of the token holders because they are economically incentivized to self-regulate the TCR they have a stake in.
However, that assumption seems to have been made in error, at least with respect to the adChain Publisher Registry up to this point. Token holders are largely content to sit on their laurels and wait for somebody else to do the work. This is known in game theory as the free rider problem.
So how do we solve for that? Hint, hint.. it involves rethinking the role token holders play in curation. We have been working on some new ideas for that and we will be revealing more about that in the coming months. But rest assured we are working hard on the problem.
Token-Weighted Voting is a Two-Sided Coin
Token-weighted voting is the principle of one token having the same value as one vote. When token-weighted voting is introduced to a TCR that hasn’t yet defined standards or guidelines, whales can wield a lot of arbitrary influence. If a whale uses the power of their token-weight to vote something into the TCR that falls under the umbrella of “questionable” for the quality of the overall TCR, and there is no court for the community to rely on for a ruling or standards to reference, then the community is at the mercy of the whales broadcasting signal being louder and more powerful than theirs.
Again, other projects like Civil are looking to mitigate this by having a constitution and a governance council. That approach may work. But it still doesn’t solve the issue of tiny rewards for voters. If it takes 2 weeks to commit and reveal a vote for one specific candidate, pay transaction fees in ETH to the network, and only win $3.00 (because you are not a whale), it is not very enticing to participate as a voter in a TCR. If your reward is less than the MetaMask transaction expense paid to submit the vote that is also not very ideal.
Problems with token voting is not an issue unique to TCRs, it seems nearly all projects that have tried to implement some form of token voting on a blockchain fall victim to lack of voter participation. But nonetheless, it does beg the question,
Is token-weighted voting the best we can do in the TCR community?
Vitalik’s great blog post about token voting and blockchain governance addresses many of these concerns and we highly recommend reading if you haven’t done so already.
The concept of Futarchy is an interesting way of addressing voting issues. In futarchy, votes are made based on prediction markets rather than personal opinion, providing some validity to those who have no background in the digital advertising industry. But it is still untested.
Another idea that we find intriguing is quadratic voting “where participants cast their preference and intensity of preference for each decision (as opposed to a simple for or against decision).”
Any way you slice it, token voting on the blockchain remains unsolved. Token-weighted voting is Sybil resistant, and for now, it is the best solution we have, but we need to find better alternatives. As part of the TCR working committee, it is one of the things we plan to address.
The token distribution audience for a TCR is critical. Currently, we cannot say with confidence that the main adToken holders are our “target market”. Our ideal target market would include advertising operations professionals, content creators, publishers, and ideally any stakeholder in the digital advertising industry.
A consideration for would-be TCR pioneers is to think about how to get your tokens into the hands of the people who have an additional incentive to contribute to your TCR. Either because they are professionally qualified, or otherwise, to curate effectively. The reason being is that if the people holding tokens are merely crypto enthusiasts, and aren’t all that interested in your specific TCR or the problem it’s trying to solve, that may present an issue down the line when it comes time for them to actually do real curation work.
If we had to do it all over again we would redistribute the token in a way that aligns incentives with the ideal token holders. That could include a process of screening potential token holders before they purchase the token through a test about digital advertising or by only selling to players at premium advertisers, publishers, or agencies.
Of course, at the time we did our token generation event on June 26, 2017, no project had successfully implemented any type of alternative mechanism to mitigate whales accumulating a lot of tokens. Hindsight, as we know, is always 20/20.
Our goal to address this issue now is to build a complementary platform where advertising industry professionals can audit domains in the Registry.
This will ensure that the domains have been vetted in accordance with current industry standards and guidelines. Furthermore, a scoring mechanism will also be provided to parse domains in the Registry into well-defined categories. This will take the binary element out of the TCR. It will no longer just be about whether a publisher is in the registry or not. If all domains in the registry have a quality score then even domains in the registry have a ranking. This is much more valuable to advertisers and agencies in the digital advertising industry.
TCRs (and Ethereum dapps as a whole) need to focus on user experience. When presenting the adChain Registry to advertising professionals, we have often encountered the complaint that there is a high barrier to entry, such as acquiring ETH and adToken, as well as, installing basic tools like MetaMask. There are also a high number of transactions and wait times. These issues can be addressed through second layer solutions or perhaps even delegated voting. But one thing remains certain, TCRs need more user-friendly interfaces.
TCRs may also need some bootstrapping mechanism to get started such as incentivizing early adopters and users of the TCR. Without those first movers, there is no way to test TCRs and understand human behavior surrounding them. We have learned that people make emotional decisions in lieu of rational ones. Perhaps this is the new way to look at TCRs, not as logical decision makers but rather as examples of human behavior.
We are excited to introduce new mechanisms into the adChain Publisher Registry based on what we have learned and to provide better signaling to the adToken community. Through continued research, we will work to remain leaders in the TCR space and to improve the user experience for the benefit of the entire TCR community!