A new San Francisco startup wants to use the blockchain to create more trustworthy and collaborative databases for consumers and businesses. Dirt Protocol raised $3 million in new investment from 26 investors, including venture capital heavyweights like Greylock and General Catalyst and crypto investors Digital Currency Group, Pantera Capital and Linda Xie.
Dirt cofounder and CEO Yin Wu, 29, is a serial entrepreneur who dropped out of Stanford in 2011 to found her first software company. Her second venture, Double Labs, was a mobile notifications startup that Microsoft bought in 2015.
The concept for Wu’s platform centers on information registries, also known as token-curated registries, that allow people to vote to verify their accuracy. For example, in the wacky world of cryptocurrency initial coin offerings (ICOs), some teams make false claims about who their advisors are to make their project look more prestigious. In response, the advisor can tweet about the lie and contact the ICO team to request her name be removed, but she has little other recourse.
In Wu’s vision, there will be a registry on the Dirt platform for cryptocurrency information, where different stakeholders (token-holders) can vote to verify accuracy. For an ICO team to add itself to that registry and list its advisors, it would need to “stake” tokens, thus making a bet that the information is correct. To dispute a claim, people must wager their own tokens, and whoever wins the challenge, decided by number of votes, wins the opposing party’s staked tokens.
The goal is to create a widely used, reputable list of verified ICO projects, in the same way the Better Business Bureau badge has served as a stamp of approval for small businesses. And like Wikipedia, the list can be edited by anyone, but it will be moderated by stakeholders. For Wu’s project to work, many people need to use and trust the list, to the point where ICO projects who aren’t on it would have less credibility.
Dirt doesn’t plan to build registries as its main business—it wants to be the platform on which registries are built. But it will break that rule at the outset to build traction. “I have a strong belief that you need to be your first customer,” Wu says. “Amazon Web Services is such a good product because Amazon was the first customer.”
She plans to launch the first registry in a month or so, which will focus on cryptocurrency markets information. Wu says other registries that could be built on Dirt are an Etsy-based list to verify whether sellers make authentic goods and a Yelp-like listing of restaurants, where users could veto reviews that were deemed fake.
The name Dirt stands for Decentralized Incentive Registry Token, but Wu also chose it for another reason. “There are a lot of projects, especially in the cryptocurrency space, that have names like Galaxy or Multiverse. We need to grow into that name if we’re going to call ourselves something as lofty as that. Dirt is very foundational. We see it as a data layer,” she says.
To make money, the Dirt team will hold a portion of the tokens, which could increase in value if the service delivers on its promises. Dirt could also sell additional services to help companies build applications on top of the platform, in a similar way that open-source database company MongoDB makes money.
But getting usage and participation won’t be easy. Usage of Ethereum-based applications like CryptoKitties has dropped steeply. Another registry-related crypto project, AdChain, launched in April 2018, but participation has been low so far.
AdChain aims to create a list of high-quality websites with user-friendly advertising practices, where token-holders vote on the quality of the websites to maintain the list’s integrity. Although 60 publishers have been added to it so far, and there are nearly 1,300 wallets, or accounts, holding adTokens, “We currently have under 1% of tokens participating,” says Mike Goldin, the lead engineer at crypto consulting company ConsenSys who’s working on AdChain. “That’s formally bad.” Goldin wrote the paper that formalized the term token-curated registry.
Originally, Goldin’s design relied heavily on financial incentives. He thought people would be motivated to vote because, if they didn’t participate, the quality of the list would suffer, and the value of their tokens would decline. But those incentives turned out to be insufficient.
For incentives to work in Dirt, “The early voters need to have extrinsic motivations for the data being correct beyond financial—especially before the token has value,” Wu says. “For example, the stakeholders for whether a token project lists the correct set of investors would be the investors who want to protect their reputation.”
In crypto today, perhaps one of the biggest mysteries is how economic incentives will play out after projects go live, since many of them raise money and build hype before launch. At ConsenSys, Goldin is now trying to create a simulation where he might spend $10,000 on research that could test how these incentives would play out before spending much more time and money building a real product. He recommends that Dirt try to do something similar and run a “credible experiment” before launch.
“We are lucky to be able to learn from the work that Mike and AdChain have done,” Wu says. “However, the only way to know whether any of our ideas will work is to launch to product and test our assumptions.” She and her three cofounders plan to hire four more engineers in the coming months, and they’re aiming to launch the product this fall.