This is the seventh in AdExchanger’s series on the cross-device question, in which we examine what each data management platform (DMP) can provide in terms of connecting the identities or profiles of consumers across the digital, mobile and offline ecosystem.
AdExchanger: Last month, Lotame purchased cross-device tech provider AdMobius. What’s the status of that acquisition?
KALYAN LANKA: The deal has closed. We’re working on the backend processes and systems integration as we speak. That being said, we continue to offer some cross-device and connected device capabilities to some of our customers already using the AdMobius technology. Within the next few months, we’ll offer more scalable products using the combination of AdMobius and Lotame technology.
What can Lotame do with a fully-integrated AdMobius that it couldn’t do previously?
Where we are today is we are a cross-device DMP. We can collect in an intelligent and automated fashion consumer data, interests and behaviors across any device. Whether it’s a desktop, set-top box or CRM, we have the ability to collect that data.
We do not have the ability as a DMP to identify relationships between those devices. We don’t know if a consumer is using all of these different devices we’re collecting data from.
The answer, for DMPs, is doing direct match. ‘Hey, Client, can you pass a user ID or create login requirements for consumer to consume content?’ That’s not going to work.
If I start placing a login just to go, for example, to Yahoo.com, consumers won’t be happy with that. Very few companies can implement a firewall for content consumption.
So with AdMobius, we want to give scale to our customers to engage with consumers across channel.
What are the use cases you’re trying to address?
First, understanding the profile of consumers. Is there a specific reason why a consumer uses their mobile device or a tablet or a fixed web or desktop? That’s the first layer we want to offer to clients. Just a pure understanding of their consumer as they navigate through various devices.
The second use case is consumer engagement across these channels. Think of a scenario where my customer as a marketer can tell a story about their product as users go from one device to another device. I want to show a user a video ad. If that user interacts with that video ad, I want to show them either a coupon or some sort of [direct response] creative on their mobile device. Then I might want to tell a different story about the same product on their tablet. That engagement rate improves a lot as you tell that story across all of these devices.
That’s the capability we want to offer to our clients.
How can a technology provider enable this?
There are a couple of layers to this. The first layer is what everybody offers: basically a direct match. If you as a client have some sort of registration ID or user ID you can pass across devices, you can rely on that to say this [identity] is being used across devices. That’s a direct match and it’s the most accurate type of connection available.
The second layer is a statistical or probabilistic match. That looks at the underlying behavior of various devices, where are we seeing these devices pop up, what type of content do they consume, what sort of behaviors [they have], and we use statistical techniques to map the behavior together.
To what extent is this fingerprinting?
Let’s talk about fingerprinting. My definition, and I hope you agree with this, is placing something on the device so the consumer doesn’t have the opportunity to get rid of it. I’m trying to identify the same device again and again and again by putting some information on the device and the consumer doesn’t have the opportunity to get rid of that.
That’s not what we do. We’re not trying to place something on the device. And I understand where Aggregate Knowledge is coming from [Ed- see link above], but that’s not good for the market. That’s not fingerprinting. Fingerprinting is a completely different issue. We’re killing the growth of the market by calling something fingerprinting when it’s not fingerprinting. It’s completely wrong.
Not every DMP does probabilistic matching. What are the complications around it?
You need technologies and expertise to do this. Our chief scientist has been doing this for so long. He was with iAd, with Quattro before that. The AdMobius team, this is what they do. The second problem if you’re trying to do probabilistic matches, the ability to identify relationships between devices with a high degree of confidence depends on the amount of data you have across all these devices.
We see billions of cookies and devices every day. The same applies to AdMobius. They see a lot of traffic and we see so much information on every device in terms of behaviors, profiles and locations. The amount of data you can put in these technologies to build mapping between devices is what makes it powerful.
There are also DMPs that do both ID matching and probabilistic matching. How do the DMPs that do both differentiate?
It boils down to the scalability of your technology and the amount of data points you can push into the technology itself.
Are clients asking for any feature in particular, or do they ask just for general cross-device connection capabilities?
If you look at really specific use cases that customers are asking for, on the top of the list is sequential messaging. Being able to tell a story as users navigate from one device to another device.
The second one is content personalization and creative optimization, depending on what device the user is using and the data you have on that consumer from some other device. How can I rely on behavior or interest of a consumer that I know from a fixed-web device, and use that information to improve my campaign performance on mobile ads or a mobile web environment?
Finally, it all boils down to your cross-channel, cross-device consumer engagement strategy. Today our customers still look at these different channels as silos. But their goal is being able to engage with their customer. So using technology like ours to understand those relationships helps improve cross-channel engagement strategies as well.
How much data is available through mobile channels?
The amount of data you can collect from mobile is limited today. We’re seeing a shift in the opposite direction. We’re seeing lots of information coming in from mobile ads and mobile web devices. But compared to your desktop environment where there’s so much data about a consumer, we’re not there yet with mobile.
What data can you get on desktop that you can’t get on mobile?
On fixed Web devices, we collect interactions, such as: this user engaged on a campaign, clicked on rich media creative. This consumer just commented on something or shared something. We collect search terms. We collect triangular interactions about the consumers as they navigate through various pages. Metadata on the page, that information. We collect creative level information like whether someone clicked on a video, paused or skipped.
We do that in an automated fashion. We have a proprietary technology we use to literally write rules to collect all of these interactions as consumers navigate through that content. We’re not killing the user experience by asking them to provide that information, and we’re not adding a lot of burden on our clients by asking them to pixel each and every interaction and send it to us when it happens.
Why is this difficult to do in a mobile environment?
When you go to mobile devices, there's not a consistent format yet on applications, for example, and devices are so different. Operating systems are so different. It’s tough to come up with an automated way [of collecting information]. Now, we’re working on it, but we’re not there yet. So technology limitations are a big issue.
The second limitation is consumer intent and behavior. Today, most of the consumption that happens on mobile devices is driven by a very specific goal. You want to read something or check your flight status? Go to your mobile app. Want to send an email? Open your email app.
You have 30-45 minutes and want to consume news? You typically open your desktop, unless you’re remote. Unless you’re playing some games, you won’t spend a lot of time just consuming content on your mobile device. Now with bigger devices, that’s changing a bit. We hope as it improves, as more content is consumed on mobile devices, we’ll be in a position to collect more data from them.
Looking into your crystal ball, when will those technological limitations no longer be the hurdle they are today?
It’s a demand and investment question. The technology limitations are going to go down. There will be consistent formats. We’re seeing privacy and regulatory bodies trying to implement certain regulations on mobile. I’m so glad Google is moving to Advertiser ID [instead of] using an Android ID. While it’s not the same as IDFA on iOS, it’s a consistent ID and is very similar to IDFA. We are seeing those changes. And I believe, unless you’re talking about a fundamental shift in devices like when Apple introduced iPhone, if the current trend continues, we should get to a higher consumption rate and automated data collection across mobile applications in the next six to 12 months.
What percentage of your customers are publishers versus marketers and to what extent is that split changing?
We deal with different types of consumers. We don’t look at them as publisher or marketer. We’re vertical focused. The way they’re thinking about data, it’s not publisher or marketer-focused any more. Every marketer is a publisher and every publisher is thinking like a marketer. The end goal is engaging with your consumer.
Any big trends you’re seeing?
The shift across our customer base is moving away from that ad tech or pure advertising philosophy, taking a step back, and saying “My goal is not to either buy or sell an ad, it’s to engage with my consumer.” That’s where DMPs come in. We help them with that fundamental shift in their philosophy. That’s the major trend we’re seeing across our customer base, whether they’re marketers or publishers. They don’t care about selling or buying anymore. They want as much information from their consumers as possible and advertising is just one of several use cases we support with this information. We do a lot of content personalization and help with A/B testing and we help with consumer engagement surveys.
These are not publisher or marketer use cases. They apply to everybody.
Email This Post