Web3 Marketplace Mechanisms
I did research on what exist in the web3 ecosystem on the topic of marketplaces. Specifically marketplaces similar to the Waku service marketplace idea. Many projects want to build decentralized markets for specific purpose. AI is very popular but I focused on other niches to avoid having to wade through too much marketing B.S. Time is a great filter and so I tried to look at older projects too.
The projects that are interesting to us have the following properties;
- Services cannot be proven.
- egg. Messages sent, Search are done correctly, bandwidth has been provided etc…
- Payment amount are close to the lower limit of what is economically viable.
- Must use L2s, micropayment, state channel, etc…
- Sufficiently decentralized, or aim to be.
The goal was to find useful mechanisms that could be used by Waku Service Marketplace.
Huddle
What started as a web3-powered content creator toolkit pivoted to DePin for webRTC. I think they wanted to replace Twitch and Youtube.
Overview
As I understand it, webRTC connections and bandwidth are provided by nodes. The goal is to optimize latency and lower cost versus centralized providers to gain an edge.
Multiple nodes roles have been designed to handle load balancing, tokenomics and verification.
Tokenomics
- Token Burn & Mint → Mechanism Institute
- Data Credits pegged to USD
- Data Credits <> HUDL token variable exchange rate
Verification
Pick random peers and have them verify each other by testing bandwidth capacity.
The Huddle whitepaper doesn’t describe in detail how peers are chosen but one can imagine an algorithm based on the hash of the latest block in a blockchain serving as source of randomness. Although who the challenger is should probably be obfuscated so that the prover can’t distinguish the challenge from normal “work”.
A trusted third party might be the only solution if a smart contract can’t be use as the registry of challenges.
Rewards are calculated in a smart contract based on the information from challenges.
LivePeer
A marketplace for live video transcoding and other video effects. Better latency and cost advantages.
White Paper → wiki/WHITEPAPER.md at master · livepeer/wiki · GitHub
Overview
Nodes use GPUs to transcode video and get payed via tokenomics. Stake act as signal for load balancing and fee sharing. Multiple node roles have been designed to fulfill different needs same as Huddle. There’s a mechanism to handle dispute but the broadcaster (buyer) has to initiate it.
Tokenomics
- Staked Token → Mechanism Institute
Verification
The claim is that TrueBit is used to verify segments of video in case of a dispute but I did not found sufficient explanation as to how this verification is done exactly. Transcoding AFAIK is not deterministic, what exactly is verified and how?
Others
Many other projects exist but the mechanisms are all very similar. Some more notable ones.
- Orchid → https://www.orchid.com/whitepaper/english.pdf
- Simple staking mechanisms but with probabilistic micropayment.
- Nym → https://nym.com/nym-whitepaper.pdf
- Staking mechanism and quality of service measurements to calculate rewards.
Conclusion
Most project use tokens for value capture and do not use decentralized reputation systems (Havelaar, EigenTrust, etc…) instead stake is used as reputation proxy and a load balancing signal. Verification methods are varied, some even dubious, but nothing we could not have came up on our own. Each project has a tech stack tailored to their use case but the problem they solve are common (load balancing, payments, reputation, verification, value capture, etc…).
My take on this is that what we want to do is ambitious since we want not only to create a market but several. The good news is that I believe most of the problems can be solved generically. Verification is the only doubt in my head. It seams very specific to the market and what already exist might not be that good (too expensive, easy to fake, etc…).
This research had me learn about some good mechanisms but much less than I’d hope sadly…
Drop your favorite projects in the comments if you think they use some interesting mechanism!
DeCoM
Decentralized commerce is also interesting to look at but further away from our use case. In crypto-time this is pre-historic. This below taken from a very good post on Twitter
OpenBazaar 2014
Promises of free exchange of goods using crypto payment. Very old project born out of a Bitcoin hackathon, now dead due to lack of funding. It seams like the infra was too costly to maintain long term. Not very interesting technically, no reputation mechanism, smart contract, identity system or data storage mechanism. Fortunately we now have most of that tech. more or less figured out.
Swarm.City 2018
Older project as well. Lack of funds killed it. Weird project management style. On the technical side, I could not find much beside the fact they used smart contracts. I have a feeling they didn’t know what they were doing but to be fair it was so early, can’t really blame them. They were part of some incubator program the IFT had internally. They also intended to use Status for the chat capability.