Teng Yan

Teng Yan


Delphi Digital


As the head of NFT research, Teng Yan focuses his attention on NFTs and the metaverse, covering sectors such as NFT finance, digital collectibles, art, sports, fashion, entertainment among others

I'd say that Bittensor and Render focus on quite different aspects of the AI market.

Render: More infrastructure level by providing GPU compute for AI model training
Bittensor: AI inference, selecting and combining AI models to provide the best outputs.

So they aren't direct competitors. In fact, they can be complementary.

Yes rags, that's a great soft indicator. I think NFT markets are slowly cooking and strong interest could return again next year.

Report snippet


I believe the infrastructure necessary to build good apps is available today.

We are just lacking good consumer experiences right now. If brands can nail this, their customers will come.

Friend tech, Reddit NFTs are just a few examples to show how most of the crypto UX can be abstracted away from the customer until necessary.

Report snippet

"very low marginal costs (and nearly negligible if you use compressed NFTs on Solana), this can be done using free or cheap mints as a gateway into their brand community"

Cost profiles will be very contextual to the business. Every business will need to experiment and find which marketing channels works best for their target audience.

CAC will be a great metric if it can be measured accurately. Web3-to-Web2 attribution is still challenging today.

Report snippet

"but has managed to settle at a floor price of ~0.15 ETH"

they might not bother. but from an observer's perspective, price is a proxy for demand and gives us an indication of whether their initiatives are working.

Report snippet

"The Starbucks Odyssey Discord server"

Starbucks keeps up the excitement for the community with new drops and experiences for its members on a regular basis. Discord communities typically dissolve into the ether once teams stop making progress or engaging with their members.

Report snippet

"On-chain activity"

It's a "treasure trove" because you potentially have data on the customer's entire consumption habits (assuming more purchase transactions happen on-chain in the future)

You will be able to see what else they're buying on-chain, or what apps they're interacting with, and understand their behaviours better.

With a CRM, you're limited to your company's internal data and you won't know what they're doing outside your brand.

Report snippet

"And so this is the holy grail of Web3 and NFTs — giving brands the power to connect with their customers more closely and co-create with them."

They're pretty much on separate paths, with likely separate teams.

RTKFK is building its own universe and brand, with support by Nike. but Nike doesn't seem to be directly involved and want to let the team do their own thing.

Report snippet

"To conclude, we would like to provide a high-level framework for brands to consider when designing their NFT strategy:"

It would depend on the type of initiative.

In broad terms, there are two categories of metrics that should be tracked:

  1. Web3-native metrics: NFT floor price, holding time, transactions, discord engagement. These are proxies for customer demand and engagement.

  2. Brand business metrics:
    How does the initiative tie back to the brand's core business? Brand awareness, new customers, revenue

Report snippet

"And so this is the holy grail of Web3 and NFTs — giving brands the power to connect with their customers more closely and co-create with them."

A few interesting NFT distribution mechanisms I've seen that brands could consider include:

  1. Open editions + burning of NFTs -> brings an element of gamification

  2. Variable rate gradual dutch auctions (VRGDA), although so far no one has nailed the parameters correctly. Art Gobblers tried but failed.

  3. Slow reveal over time (similar to Opepens), this prolongs the anticipation and keeps interest on the collection

  4. Daily/Regular auctions, allows the supply to be dynamic and slowly increase over time

  5. Liquidity Bootstrapping Pools, although these haven't gotten much popularity.

I'd say that 5 - 6 ETH is a good level to watch and place bids. That was the support level before it went on its wild pump to 14 ETH. Disclaimer: Not financial advice.

Report snippet

"Will it succeed? No one knows yet, but it’s an interesting grand experiment by a dedicated, passionate community of developers and collectors."

  • Steep learning curve for people to access Ordinals, even more than for Ethereum NFTs
  • Success of the ecosystem will depend heavily on the creators, are they able to build things that will drive demand and users.

It depends on what the product is like by the time the airdrop comes. In its current state, it's going to be tough beyond the airdrop.

The FT team has a couple of months to get the product to a stage where people can feel that it has wider potential besides airdrop speculation.

That's the thing that makes me optimistic about FT: seeing people build tools around it. Still early days but could get huge in time.

They'll definitely need to adjust the bonding curve and taxes if they're looking to expand to a wider non-crypto native audience. It's too heavy right now

Outside of the game economy (agree that it's extremely difficult to balance and no one has succeeded so far), what do you think of the gameplay?

Is it a fun enough game to attract real gamers?

Teng Yan has not authored any research reports yet.