AL: When we acquired Red Aril there was a lot to love. A lot of pretty solid technical people had built a pretty cool product that had not yet been truly applied to marketing challenges.
The whole thing is absolutely built for real-time. We gather data and create audience profiles in real time. We're appending to a record in real time as users are moving through various touch point. We can absolutely power an RTB buy, and our position is that that audience profile should also power bought earned and owned touchpoints, not just RTB.
When iCrossing and Hearst purchased Red Aril, one of the things we brought to the table was a lot more resources. Peter [Randazzo] over the years has built a very strong agile development team. We had more access to infrastructure [than Red Aril had previously]. A lot of what we've done is beef up Red Aril's overall capabilities. The whole thing is now running on Netezza, running on a 40-node cluster, gathering data in the cloud with Amazon Web Services. A lot of the work we did over the last six months has been heavy lifting technical work to make sure the platform is ready for industrial grade use.
The reason for the rebrand is we felt we could apply that amazing platform to specific use cases in the market.
eMarketer says there will be $40 billion on digital spend this year. Of that spend only $15 billion is display and of that only $2 billion is RTB.
So from our point of view, the idea of a DMP being built primarily just to provide audience profiles to inform an RTB buy is a limited use case. Brands should be using that audience intelligence to inform site-side personalization, email customization, all the touch points they have for clients.
We know that if a brand is going to buy our audience management platform to build audience intelligence, it's going to want that audience intelligence to inform its content management systems, its email platforms, its ad serving, etc.
A number of the DMPs are coming from a world purely of media and exchanges. The idea of "what is an audience" has been unfortunately reduced to people saying, "An audience is just a BlueKai segment." An audience should be defined by a blend of first party, third party and first party offline CRM data. If you're a CMO, all your marketing dollars are not spent buying media. You also spend an awful lot of money on owned platforms and on earned attention. We're trying to power all of that.
You mention Core Audience is now integrated with iCrossing's connected marketing platform. How does that work from a client's functional perspective?
iCrossing has taken the Core Audience platform and done an integration of it with the Connected Marketing Platform, which is mainly iCrossing's bid management platform, similar to a Kenshoo or a Marin. That platform buys search media, display media, and has APIs into Facebook.
iCrossing doesn't SaaS that platform, but it uses it for its clients. We've been integrating our DMP with that platform. We're now able to start providing audience oriented data into that platform to inform buying. It will start to optimize all the display buying that's done by iCrossing, and it will also be powering the search buying as well.
How is the DMP roll-out going across Hearst properties?
Core Audience has been actively deployed across every Hearst property. The number of uniques is north of 80 million. The platform is at scale, managing all of that. Hearst is using that data to create its own audience segments.
With regard to first-party data security, are there specific steps you've taken or restrictions you've placed on where the media will run to guarantee there's no risk?
The construct of an audience profile should be a broader definition than just one piece of third party data. Our value proposition to brands is that brands have a lot of first party data that is extremely relevant, it's powerful, and it's probably leaking out the back side with some partner that's taking it and getting smarter themselves. We believe brands should own their intelligence.
Our fundamental business model is we don't sell data. Our platform is built from the ground up to have record-level data ownership – assigned at as atomic a level as possible.
One scenario that would be fascinating to us, that we're actively talking to people about, is to get two companies to do an ad hoc private data network. Each brand would determine what it chooses to share. It will share a cookie pool, but it can decide what data about that cookie pool it wants to share.
Have you done any of those yet?
We are definitely in conversation with several clients about doing an exchange between them and Hearst. Because Hearst is a publisher that may take the form of a buy. But there also should be brand-to-brand.
Will you directly broker brand-to-brand deals?
We do not have an active brand-to-brand, but going into next year I could see that happening.
I believe Red Aril had about 30 employees when iCrossing bought it. Has that number changed, and what are the development priorities?
We have just over 35 employees now. I've stepped in to be the CEO of the company, which I'm all in on. I will be very busy. We've been bolstering an engineering team under Peter and we're also bolstering up our data sciences group, led by Doug Bryan.
He taught consumer science at Stanford for 10 years, he ran the team at Amazon that wrote the recommendations algorithm. I don't even say numbers aloud around him because he just looks at me like I'm silly.
We've got some real horsepower on the data and tech side. I bring a lot of marketing prowess to the practice. The rebrand has been a big part of my effort to clarify our position. This category has been very confusing. I'm on a mission to simplify the value proposition, and get brands using it so they can see the value from it.