How AdTruth Adds Truth To Cross-Device Connections

adtruth martinWant to see a marketer rip out her hair? Ask her to begin identifying consumers across devices.

It’s not an easy thing to do, but vendors have been building tools to help. When Experian Marketing Services acquired 41st Parameter last October, it also got the latter’s subsidiary, AdTruth. AdTruth’s core product is the AdTruth ID, an identifier that enables its clients (both brands and tech players) to link consumers across devices.

“On mobile, there needs to be something that’s more consistent (than cookies),” said Martin Gilliard, who two months ago ascended to the role of AdTruth’s GM in North America. (Gilliard has spent the past 15 years in the data media tech space, working at Microsoft, 24/7 [which eventually became Xaxis] and building his own analytics company, DATAMYZE.)

AdTruth’s “technology looks at multiple variables to determine what device that is and if that device belongs to a particular person,” he explained. “Device recognition is so important because that’s where the most anonymous data exists, and everyone has the least amount of visibility into it. We let them create more clarity into that channel.”

To be clear, AdTruth doesn’t provide the analytics or linkage capabilities – at least, not yet. It provides the identifier and it’s up to its clients to make the connections. In other words, AdTruth provides the eyes and its clients need to provide the brain.

This might soon change, as the company’s future product plans includes a platform that manages multiple datasets (though Gilliard insists this isn’t a data-management platform).

Gilliard spoke with AdExchanger.

AdExchanger: How do you want to position AdTruth?

MARTIN GILLIARD: In general, what we do is we enable companies to have visibility into multiple devices in places where they haven’t had visibility before. We are a global company. We’ve made a lot of traction in a lot of markets. The challenge in the US market is that there are a lot of different tech players and one of the things I’m trying to establish is where we fit in the ecosystem with our tech partners and with our major brands.

What is AdTruth working on now?

Right now, we provide them with the ability to identify those devices and link them. That’s part of a much larger story.

What we’re trying to create is an ID platform that will let you connect users across all data sets. We allow you to be able to create more clarity in that channel and connect it to other data sets to create better experiences, better marketing, better analytics.

So the platform will link different data sets. What’s after that?

We’re looking to build a platform to manage all datasets in one place. Regardless of where your datasets are coming from, you have a platform that can find keys in those that cross every touchpoint you have with consumers, you can have access to data that can help you understand how you can engage with them in every experience.

Seems like the DMP guys like Acxiom and [x+1] are moving or have moved into this area.

We’re in a space where it’s not as much competitive as it is complimentary. We have relationships with companies that might externally seem like competitors to us. But there are two things we do very well. One is device recognition. A lot of companies can create that.

The second is you have to cross a number of businesses and an entire ecosystem, which we’ve been able to do. The global nature of our product in general gives us larger reach that lets us work with these guys as partners versus competitors.

What’s the status of this platform?

We think over the next two to three business quarters, we’ll have a version of this available to some of our strategic clients. Right now, what we have today is the ability to recognize these devices, that companies building out data-management platforms can use.

Who are your competitors, if not the DMP guys?

There are a couple of ways we should look at this. There are companies that create device IDs, those that try to solve cross-device across streams and those that provide it as a service versus a licensed product.

Those that try to solve the device ID recognition part are data-management platforms that exist in the space. Those that try to solve cross-screen or cross-device are Tapad or Drawbridge. They provide a service that actually does that, but do you want to do this yourself and own the data? If this is strategic to your company, AdTruth provides it as a licensed software. That distinguishes our organization: We enable our customers to do that themselves. We are not trying to manage that process for them.

Will you ever branch into services?

We are not looking to branch into services.

How many companies use the AdTruth ID? Are they vendors, agencies or direct-to-brand?

We have large direct clients. We have tech players. And we have data-management platforms. Some are internal to companies and others are licensed through DMP players in the space. We have roughly 80 to 100 clients globally. Most are pretty large clients. We started working with large brands and large tech players and we continue to work with them. In our space, most of the big tech players are enablers for these brands.

How do you build an AdTruth ID?

AdTruth came from 41st Parameter, which is the platform used to identify fraud. It gives you a number of variables to identify a single user and how many applications are coming from a device. Most of the financial services companies are very familiar with 41st Parameter. We have been able to leverage the knowledge of that to create an ID – what we call a recipe – that will read things from a device. Operating system, how many calls it’s making, etc. It goes into an algorithm that helps predict if this is a single person or not. We use multiple variables that are available, that’s all public information, to determine probabilistically if we know this is an individual user or not.

Those match rates fall between 80-90% for us, which is much higher than the 0% people have right now.

That’s pretty high. I thought the industry average is around 60-70% for probabilistic. How do you pull 80-90%?

That’s a question our data scientists can answer. I can’t answer that question. From my own experience, there are a number of companies can create a device ID. Most of the information is already there. But the science that goes behind that is the part where it gets tricky. Right now, we can match up 50-60% through Web analytics, but the math isn’t effective enough, it’s not pulling enough variables. What they’ve been able to do here is look at that information in context of everything.

How do you verify that 80-90%?

There are people much smarter than me that can answer that. It doesn’t mean that every single user you look at may be accurate, it means there’s an 80-plus percent chance that this is that person.

How have you integrated into Experian’s technology or datasets, since the acquisition?

It’s been a very good direction for Experian. It filled in a gap they had in the digital space. It’s easy to see how this product fits in with the clients they currently work with. Having been here for only two months, it’s amazing how quickly they integrated the backend parts of this. We expect that to continue over the next year, so current Experian clients will have access to the software we have either through tech providers or us directly.

Where is AdTruth integrated into Experian?

That part I can’t communicate just yet.

What are the biggest challenges for brands going forward around using their data?

Facebook and Google don’t share information with the brands that use them. They share information in aggregate. If you want to identify a user across Facebook, you can’t do that. You’re blocked from anything happening in their application. If I’m a brand and want to understand what’s happening in all my digital interactions, right now, I can only do that outside of Facebook. We’re enabling them to do that across all the other digital channels. The challenge will be as those platforms grow, how do I become less dependent on them as my data provider and create my own data and understand it myself?

Brands increasingly are trying to own their own data, which is why so many are bringing their ad-buying apparatuses in-house. How has that impacted AdTruth?

It’s a great thing for the industry and for AdTruth. Companies are starting to see this data as a strategic asset more than they have before and that they want to harness this internally is important for their business and very valuable for us. We enable those companies to have visibility into other data assets they didn’t have before.

Are you having more conversations with brands to integrate AdTruth directly into their stacks?

Absolutely. Weekly, we’re having those conversations with big brands. Some are further along than others. For some, it’s a discovery process. Others are full speed ahead. Their challenge is figuring out where the data fits. Is this a platform they build internally or do they use someone else’s platform?

Enjoying this content?

Sign up to be an AdExchanger Member today and get unlimited access to articles like this, plus proprietary data and research, conference discounts, on-demand access to event content, and more!

Join Today!

1 Comment

  1. When you consider the match rates of a solution like this, it’s important to also take into account accuracy rates. It’s easy to confuse match rate and accuracy rate, but when you shop for a cross channel solution you should take into account both. An 80 percent match rate looks pretty good on paper, but without incorporating the proper techniques to achieve a high accuracy rate as well, that number is less significant. This means developing a solution that goes beyond basic statistical inference – which only leads to an accuracy rate of about 50 percent – and instead working to achieve true matches between data points, such as across devices and consumers. For example, our algorithm has a 97 percent matching accuracy.

    Jeff Zwelling, co-founder of Convertro