Comcast’s FreeWheel ad server business has been developing a blockchain-based data onboarding service called Blockgraph since December 2018.
JASON MANNINGHAM: We’ve been focused in the past year on moving out of the proof-of-concept phase and into the early stages of platform adoption. We’ve run a series of successful pilots.
A priority right now is establishing a significant deterministic identity layer, which becomes the base of the platform and the necessary activity. That’s really when you start being able to demonstrate how you can address the right audiences in a privacy-considerate way. It’s a two-sided system where we’re adding identity and distribution partners, and having that allows us to build advanced ad features.
Some of the basics are like targeting and frequency capping.
The first pilots were for data matching on a peer-to-peer basis. So, an advertiser with its customer data matches with an MVPD [cable or satellite service providers such as AT&T, Comcast and Cox] to size out their target audience. It also allows them to optimize on linear or in addressable video environments.
This year, our pilots are focused on cross-platform features, now that there are a few distributors involved. That means taking a given audience segment, doing a match across distributors and passing that data to ad management platforms like FreeWheel to optimize campaigns across platforms, instead of necessarily with one-to-one advertising deals between a buyer and a broadcaster.
Who are the kinds of partners you need right now?
I can’t name names right now, but the types of companies we’re talking about have distribution in TV. Not surprisingly, that means the MVPDs. [Charter, ViacomCBS and Comcast sibling company NBCUniversal are Blockgraph’s public partners.]
And in the broader video ecosystem there are other distributors, like OTT services and even smart TV manufacturers themselves; perhaps publishers launching their own direct-to-consumer video apps.
The key is to add partners who can authenticate the user, or viewer, and the household.
Are “identity” and “distribution” practically the same for you, or could certain partners provide one and not the other?
Some companies in the ecosystem may provide audience segments or inventory. Those would be trusted third-party services. But in terms of being able to establish our identity layer base, we are focused on TV industry distributors.
Distributors control enablement, where and when the ad is going to be served. But they also have the appropriate basis for using audience data. So in that sense, identity and distribution are closely connected, in the form of cable deals and other relationships directly with households.
What’s the revenue model?
We are a subscription-based business, with software licensing costs to join our network. Those costs would vary depending on the type and size of the company’s data.
The idea is for it to be very economical. And as we add value and more use cases, we want to keep that price low and at a fixed rate. That means as it scales it becomes cheaper and more effective, compared to providers today that charge a percentage of media or a fee per impression.
So, it’s similar to data onboarding companies?
It’s a peer-to-peer version of data onboarding. So with a company like LiveRamp, both parties would onboard data and audience files to the third-party company, which sends back synthetic IDs that work amongst themselves.
Blockgraph replicates that, but instead of the data going to a trusted third party, they’d run our software in their systems to anonymize audience records. From there, the data can be matched and shared across our common identity layer.
Another important difference is that we provide household-based identity. Most digital identifiers are focused on being people-based, whereas in TV and streaming video media is still transacted at the household level.
Is “blockchain” still part of the pitch?
It’s a fair question.
I will say that the name “Blockgraph” was always about building the identity block by block. We did get some confusion around that. But a lot of technology we use is inspired by distributed, blockchain-based systems. We use a distributed data base and cryptography so that our identity layer can span the whole system, but the data isn’t visible to any partiers without its owner’s permission.
One of the reasons we couldn’t rely on blockchain is we need to be able to identify individuals and the data and impressions attached to them because there are “right to be forgotten” laws if people want to be removed from the system.
This is a common issue in collaborative TV industry initiatives, but why should non-Comcast companies join your network?
There are two answers. For one, our system is decentralized and can be hosted by any distributor. We have no visibility into the data: They run the software on their systems with full transparency. That isn’t something you get in the world of black-box algorithms and platforms.
Comcast also has a strong vested interest in making sure the TV industry grows and thrives based on data-driven advertising. Comcast has been a leader with FreeWheel and addressability initiatives.
Television is being redefined at this moment, and what we want is to apply the benefits of traditional TV reach and frequency management with the best of digital, which means ROI metrics and optimization.