Four Foundational Pillars to Usher in the Self-sovereign Data Monetization Era in Web3
Via DataDAOs and DataNFTs
In 2021, Google & Meta/Facebook made $210B & $115B in advertising revenue, and here’s how they did it in a nutshell
collect our data
centrally analyze & decide what to do with them (e.g. matching them with parameters/demographics that advertisers look for)
monetize off of our data by selling our attention and eyeballs to advertisers and keep the entirety of the profits
I’m simplifying the complexity of data analytics & the algorithmic excellence in the process — and having spent time at Apple as a Data Science manager, I truly appreciate the innovation & value-add there.
But the parallel between this status quo and a dictatorial government regime is still pronounced: in the Data Nation, we are utterly powerless with
No control over who has access
No idea the whereabouts
No say over the use of, and
No profit share over the monetization of our data.
My ears would perk for #4 — as it is material vs. #1#2#3 being more philosophical. Imagine you can make real money by contributing your personal data, and believe it or not, it’s a win-win-win situation for you, Google/Meta, and the advertisers.
Here’s a future state picture of how a self-sovereign data monetization model would work
Let’s first define DataDAO vs. DataNFT:
DataNFT is to turn a dataset (whatever format it takes - e.g. CVS, Excel, PDF, etc. into a digital asset in the form of an NFT (non-fungible token)
DataDAO is defined as a community of relevant contributors or stakeholders to manage the use, access/permission, and monetization of one or many datasets, which ideally will be tokenized to a DataNFT and stored on IPFS.
Scenarios & Steps:
I’m planning my travel schedule for 2023. I want to make money off data about my travel plan (where I want to go, when, and what I need to buy to prepare for the trip).
I elect to tokenize my travel schedule data, along with the contact information I’m willing to share, as a dataNFT and upload to a dataDAO platform like Kamu as a data publisher. Examples of data points that I’m comfortable sharing are
approx date/month of travel
destinations,
categories of activities (such as shopping, sightseeing, and museums) she intends to do.
Either Google/Meta or Direct Advertisers (e.g. airline, hotel, tour agencies — basically any service providers who want to win my business before I start booking) can get on the platform, bid on access to my dataNFT, and sell the best offerings to me
It’s a win-win-win situation because
For data publisher (me): I get to make extra cash off of data that I’m comfortable with disclosing + get suitable offerings/promotions from advertisers and vendors
For Google/Meta: the ability to measure conversion & target customers directly + access to precise, preemptive activity data (no data science can model our travel plan for the entire year)
For Advertisers: direct access to target customers to sell products and offerings.
To take a step further into the DataDAO domain, I can join a DataDAO of fellow contributors who also want to monetize their travel data. We will collectively decide on
Who we will share/sell our dataNFTs to
Granularity & format of the data that we each contribute so dataNFTs can be composable with one another
Access & privacy rules: duration & permission of access
Computation policy & requests
Who we want vs. not want to sell to
Distribution rights
etc.
However, to turn DataDAOs and NFTs from a rosy narrative to mass adoption, we are still missing 4 critical pieces
Venue
Usability
Operability
Conversion
1. Venue
A marketplace & discovery platform to connect data publishers to data consumers/purchasers.
Key features & use cases:
- UI/UX to guide data publishers to tokenize their dataset -> set privacy, distribution, and computation rules -> upload to the platform
- Enable browsing, filtering, and discovery by data consumers (for example, United Airline may use the API of this platform to filter/AI/ML to find data publishers/consumers who disclosure their upcoming travel plan to market airline ticket promotions to them)
- Record dataNFT transactions on-chain and make sure it's traceable & verifiable
- Allow data publishers to create a database (DataNFT) on the platform and grant various levels of access to advertisers based on their willingness + monetary incentive
Existing players
Ocean Protocol, Kamu, DataDAO.io
2. Usability
A data structure/format/model standardization layer: the ETL process in the traditional big data transformation process to make data useable and computable. As a data consumer, I don’t want to pay for 20 DataNFTs to get genomic data and end up with 20 different files in csv, excel, and worse, PDF (which is completely unreadable).
There needs to be some sort of standards to make DataNFTs composable, compilable, readable, and computable - like how ERC 721/1155 came to be community recognized on Ethereum. Hence, an ETL layer will be foundational to make these DataNFTs useable at all.
Key features & use cases:
- Provide some data format standard so that data in the same category (e.g. education history data from Amy, Bob, and Catrina) can be easily aggregated for analysis
- Provide data model & ETL abstraction as a service to remove technical burden from data publishers and consumers
- Preferably provide "drag & drop" style non-technical SDK for users to manipulate data
Existing players
Kamu, Mind Network
3. Operability
Ability to compute directly over data stored over IPFS: once a data consumer buys a dataNFT and gains access, will he/she be able to download the entire dataset?
If yes, the privacy rule will basically be wiped out, as once a data consumer makes a local copy, there’s no way to track and trace its use.
If no, how to make this dataNFT computable & readable without allowing download to a local copy? This is a key topic for the Compute over Data working group at Filecoin/IPFS and a very key topic to be further explored, but below is one way of working around.
“Data publishers can choose to share data only for privacy-preserving computations (e.g. allow advertiser's ML model to run on their data to classify them as “interested in traveling to Portugal / for business / in November", but never to actually see the raw data)” - Sergii, CEO of the DataNFT Marketplace Platform, Kamu.
4. Conversion
Innovations to bridge web2 digital ad tools (cookies, targeting, tracking, conversion) with dataNFTs to drive conversion.
Ultimately, demand for data & analytics is driven by its effectiveness in driving sales through conversions. The web2 track-and-trace model historically implemented in the form of cookies & ad tracker will persist into web3, with the difference being now data suppliers (consumers) can also profit from their data contribution, hence more willing to participate in the tracking program. The purchasers of dataNFTs need a mechanism to reach their target customers with precision to present their sales offers/promotions, ultimately to convert impressions/clicks into purchases. One example of such design is to let data contributors opt-in for targeted sales offers based on their interests in exchange for extra rewards from their data contribution.