Simple Summary: We claim the most valuable NFT data is created after its mint. However, blockchains are optimized for consensus of truth, permissioned writes, and historically sticky data. This creates friction for creators trying to update NFT metadata or crowdsource external contributions. 

We propose a new way of approaching these challenges. Rather than trying to control and update the single metadata file referenced from an NFT’s tokenURI, why not mint additional blocks of metadata that backlinks to the original? This crowdsourcing-style approach would allow the ability to search, sort, and filter by the public keys of each contributor. If done correctly, NFTs with a large, quality set of contributions could grow in value because of their richer stories and datasets.

Demo The Metadata Minter UI

Seeing the tool in action may answer some questions and spark ideas before we go into details. Here is a video.

What did you just see? Let’s break it down.

1. Select NFT: Artist loads token with frozen metadata (i.e. it can’t be changed).

2. Select Template: They pick the copyright template.

3. Fill Out Form: They input copyright license & owner; create a draft metadata block.

4. Sign: They cryptographically sign this new metadata block with their private key.

5. View Results: All metadata contributions chunks are displayed noting public key.

What didn’t you see in the background? Let’s break it down.

• Conversion: The metadata and signature were converted into IPLD blocks.

• Storage: Blocks are stored to various web2 (S3) or web3 (IPFS) endpoints.

• Privacy: If enabled, blocks would only be stored in private, secure endpoints.

• Index: Searchable index in Postgres and served over a GraphQL API endpoint.

When we release this demo to the public, you’ll be able to see for yourself. There will be an option to click “raw JSON” button to get a copy of the consolidated metadata w/contributions, and IPFS CIDs.

Why This Matters

Imagine 27 pieces of art selling at an auction for $1.6 million dollars. Months later, an NFT collector stumbles upon one of the pieces on Opensea, but sees no transaction or sale history. Next, a financial analyst fires up Dune Analytics to look at monthly NFT sales volume, but is missing all off-chain sales at galleries all over the world. 

This isn’t a fictitious example. This recently happened in June 2022 at the Christie’s event Cartography of the Mind: A Curated NFT Sale to Benefit MAPS. While Dustin Yellin’s piece Psychoprogeny - Mother Tree was sold for $63,000, there are only transfers listed on its Opensea page. 

Now consider all the other scenarios where art is changing hands without any onchain record. What about private trades between two collectors? What about private sales that happen just to skirt around NFT marketplace royalties?

Then there are the non-financial aspects. Up and coming artists rely on an up to date CV to showcase their work and land their next opportunities. Not only does it matter what their work sold for, but where it was on display and who purchased it also matters. It adds to the overall story, The Lore of a piece of art.

So how could we fix this?

Minting Sales Data

Dustin was fortunate enough that he placed the Christie’s event data in the metadata.json before the auction. In doing so, any NFT marketplace (that knows how to ingest that information) could benefit themselves and the artist by displaying it on the NFT page.

Now what about the sales information? For that, we need to go back to the metadata minting UI.

Before we used this Copyright template. This time, we will use the Off-chain Sale template. Here we enter:

• Date

• Sale Amount

• Sale Currency

• Event Name

• Other

While this information can be added by anyone with an Ethereum account, the veracity of this information is strengthened by the signature. If two people submit different entries for this sale, we can decide which one is more trustworthy. Since the public keys from Christie’s have been verified in the public domain, we can extend more trust to those signatures versus unrecognizable keys.

We could further amplify the veracity of these submissions by adding counter-signatures. Let’s imagine an artist submitted the sales information themselves. A person seeing this data might question whether an artist is lying to inflate sales figures. However, if a gallery or auction house signed that same block, we effectively have a handshake of two verified accounts submitting the same information. This provides further strength on the data claims.

And Beyond

In a previous article, we identified 28 potential metadata fields that are not commonly present in most NFTs minted to date. Similarly to how websites went from being simple business cards in the 90s to full-blown applications in the 2010s, we predict NFTs will become more advanced in terms of the types and quantity of data connected to each one. Given that not all of this information will exist onchain and at the time of mint, we need a strategy for properly appending data to these NFTs over time.

This Metadata Minter UI is our attempt to create a web3-friendly mechanism of creating, storing, and serving crowdsourced contributions to an NFTs metadata. In doing so, we hope that it’s a valuable approach to add value to the NFT ecosystem.