Jun 3, 2022·edited Jun 3, 2022

I know there was the whole conversation relating to Elon's acquisition attempt, around Twitter blue-checking every one of their users, to establish identity. If you could actually find a way to use your crypto identity to get paid back a fraction of the value back from the personal information you're producing as you navigate the web, that would certainly be interesting. Like, opt back into the world of detailed tracking (which Apple and Google have been up-ending by way of FLoCs and the like). Have your data attach to your BirdCoin wallet, and the system that discloses your detailed tracking also tosses you a few pennies each time somebody wants to use that information.

I still am skeptical you need to do this on a distributed blockchain for it to work, though; it seems to me like pretty much anything you can design on a distributed blockchain could be done at lower cost (in time, and cumulative processor cycles and storage bits) using a more traditional replicated/multi-sited database, administered by a trusted authority. That could be a for-profit company like Twitter or Google, or, maybe better, it could be some kind of co-op like entity, or a non-profit like the Mozilla Foundation. But fundamentally because proof-of-work costs orders of magnitude more energy, and proof-of-stake has such built-in rich-get-richer issues, the value proposition to the mass of users is going to look better with the model that uses a trusted central clearinghouse. The overall income less expenses of the business that does the central clearinghouse model is better, and they can distribute some of that back to the users, and pocket some of it, or reinvest it in improving their tech.

Expand full comment

What is the value proposition of a web3 attribution ledger over say a web2 MMP?

One obvious answer is decentralized trust between the participating parties of attribution. However in my observation trust between ad platforms, attribution providers and advertisers have never been a first order problem urgently in need of solving - in the web2 status quo ad platforms have no real way to verify MMP's attribution logic, MMPs have no real way to verify attribution claims from self attributing ad networks, advertisers have no real way to verify either, yet the system chugged along just fine with this notion of trust that is based much more on reputation than verification. One could argue that Apple ATT / SKAdNetwork (and soon Android privacy sandbox) took this to the extreme as they have become the ultimate, one and only attribution provider that other players have to trust unconditionally. SKAdNetwork has been (rightly) criticized from a thousand angles, but lack of trust in the system doing what it claims to do is not one that I have heard.

The primary driving force in the ads attribution game, IMO, is consumer attitude and choice towards what data is ok to be tracked vs not - this is what has been weaponized and manipulated so successfully by Apple in recent years to the tune of nuking the entire industry while still being perceived as the good buys in the eyes of the average consumer.

My hunch is web3 attribution has a future if and only if it can offer a sufficiently compelling value proposition to consumers such that they are willing to opt their data back into the ads machinery (or into the new web3 ads machinery if it looks very different from the current)

One obvious possibility for achieving so is paying users for their data - which is a flawed idea for reasons that you've written about.

Another is hoping that users will en mass come to appreciate the intrinsic value of a decentralized attribution architecture / ad ecosystem - BTC > fiat currency, by the same token web3 attribution > fiat attribution. It would seem that this requires web3 succeeding as a cultural movement first.

What are other possibilities?

P.S. I work on ads at one of the companies you mentioned above. Pretty fascinated by your explorations in this area. If you are interested, more than happy to bounce some ideas

Expand full comment

Good analysis.

I think the main problem is not how we store the user data though. This is relatively easy to solve , either centralized or decentralized.

What is difficult how to opt back in the users to the attribution system. We have gone from everyone in it to near zero on iOS in short time period. Real question is if its going to be sufficient to incentivize the users to get them back in , or the friction introduced by the platform gatekeepers will not make this approach viable?

Expand full comment

Very insightful - thanks Antonio. A friend of mine shared this post with me because funny enough, I'm building an impression attribution engine at a hackathon in NYC this weekend. For context, we're building an ad protocol for web3 social. Would love to tell you about it!


Expand full comment

Are you suggesting that the expensive Proof of Work and Proof of Storage could be processing attribution—with verification process though something like secret canari examples?

Expand full comment

Ah, the reason I signed up, Antonio. I wonder if you'd mind my assigning this to my NYU students in the fall?

Expand full comment

Attribution from Discord/Twitter to OpenSea NFT purchase would be insanely valuable for web3 brands and NFT creators. I guess OpenSea would need to develop an ad suite for creators, right?

Expand full comment