Potential business models

The world's largest data commons won't build itself — incentivizing data provisioning is at the core of the protocol. For this economy to be sustainable, it shouldn't rely only on inflationary token incentives — we need to secure external revenue for data providers (see Potential partners for some examples of who these could be).

Plain data selling

Below we show a few examples of data and value flows for some protocol use cases. All assume a mechanism by which:

  • we identify and compensate data providers;

  • we ensure user preferences are respected.

Color codes for diagrams below

Selling data in bulk

DSPs and SSPs use data from DMPs to conduct ad auctions. (see Targeted Advertising). DMPs collect data from multiple sources (see Data Collection) and, when they buy it, they usually buy it in bulk.

One way to tap into this chain is for Fractal to get data from the protocol, package it, and sell it to DMPs.

Additionally, we could sell to business already providing data to DMPs. This broadens our business development target.

Building our own DMP

Another possibility is for us to build our own DMP. Depending on how downstream integrations with DSPs and SSPs look like, this might enable us to control the value flow more tightly, better coupling payments to usage.

Selling directly to publishers

Integrators of the protocol, such as browser extensions, could provide data to publishers when the user visits their page. Publishers can then use these data to enrich their ad requests, customize content, or feed their own DMP.

Selling directly to advertisers

Integrators of the protocol, such as browser extensions, could provide data to advertisers when the user visits their page. Advertisers can then use these data to enable retargeting, customize content, or feed their own DMP.

Enabling browser-based ad infrastructure

We could build our own alternative to Google's Privacy Sandbox, or offer an infrastructure that helps mitigate the downsides of the loss of tracking to businesses working on alternatives (see Enabling privacy-preserving tracking in Crumbs for an example, and Needs of Advertisers, Publishers and Users for an exhaustive list of ad tech use cases).

For example, instead of requesting context data to determine brand safety for a bid, the advertiser could send this algorithm to be ran on local data without exfiltration (compute-to-data). Computation results could be verified using zero-knowledge proofs and/or written to the blockchain for non-repudiation. This would help with reputation management, and would lower latency during the RTB process by allowing for post-hoc verification.

Data provided to the protocol, if used to enable targeting, would stand to share in whichever revenues we'd get to collect from this operation.

A clear downside to this idea is that Google's Privacy Sandbox and friends don't currently have a revenue angle.