Podcast: Standards for Binding IDs to Media Assets Using Audio Watermarking Technology

Aimée Ricca sits down with Steve Davis of Kantar Media, Chris Lennon of MediAnswers, and Paul Mears of Copperline Media at 2018 NAB Show to discuss a standardized approach — using audio watermarking technology — for streamlined cross-platform measurement by enabling faster and more accurate tracking and audience measurement of content and ads across media.

Transcript:

Aimée Ricca:                00:15               Hi and welcome to our series of informal chats with industry leaders so that you can get to know them and the technologies that you use better. Today I am with Chris Lennon, President of MediAnswers and Steve Davis, global product director of Kantar Media, and we also have Paul Mears, president of Copperline Media. 

Aimée Ricca:                00:35               Thank you for joining us today. 

Chris Lennon:               00:37               Thanks Aimée. 

Steve Davis:                  00:37               Thanks Aimée. 

Aimée Ricca:                00:39               So we are here to talk about something big that you guys are working on. It's the OBID suite Did I get that right?

Chris Lennon:               00:47               Good job. It's the OBID suite. We have another acronym which is the last thing we need as an industry. 

Aimée Ricca:                00:53               Another four letter acronym. 

Chris Lennon:               00:56               There was actually still one available. It's Open Binding of ID's is what it stands for. 

Aimée Ricca:                01:01               Okay. So can you give us a brief overview of what opening binding of ID's actually means?

Chris Lennon:               01:08               Sure. So what it means is it's a method for embedding content identifiers first of all in the content itself. So you can identify ads uniquely and unambiguously. You can identify content uniquely and unambiguously. And then also ... You want another acronym? We've got OBID TLC, isn't that a cute one?

Aimée Ricca:                01:30               That is a cute one. Yeah. 

Chris Lennon:               01:32               It's not tender love and care. It's our time labeling of content as well. So we can put in ... I like the way you put it, Paul, earlier, it's a breadcrumb trail. Right?

Steve Davis:                  01:44               It is.

Chris Lennon:               01:44               Of how the content got to the consumer. So we have OBID for the content, we have OBID TLC for how did it get to the consumer. 

Aimée Ricca:                01:53               And is this like a watermarking technology? Or how would you describe it?

Steve Davis:                  01:58               It is. And it's a watermarking standard now that really takes into one wrapper, two major and already well established industry ID's. It's the ad ID which is the identification that goes across all advertising. And it's the EIDR ID which is the identification that goes across all content that is produced. 

Aimée Ricca:                02:24               Okay. 

Paul Mears:                  02:25               Damn Steve. You got a good radio voice. Yeah, I think the important thing to know here is that in the past people insert data as a watermark but it was data that had to be mapped to something else. The important thing here is that the EIDR and the Ad-ID themselves are actually in the watermark. So there's no look ups or anything else. Once you detect it and recover it, that's exactly the data that you want and you can burn that into a program name or ad name or distribution path. So that's the important thing here that's novel and different than what's been done before. 

Aimée Ricca:                02:58               And so can you give me some examples of how this would be used in the industry?

Paul Mears:                  03:01               Sure, and one of the things that we're pretty excited about talking to the community this week is all the potential applications when it gets deployed across the ecosystem. One example we've been sharing is a concept around second screen synchronization. So where the advertising or the content is being run in the traditional channel on TV or on YouTube, and it can also have a different experience if the watermark picks it up on a mobile device. It makes for content to be a lot more interactive that the consumer can enjoy the content in different ways. 

Aimée Ricca:                03:41               Okay. 

Chris Lennon:               03:41               I think the mark to me of a good standard, something that is designed well enough that it ends up being used for things that the original creators of it never even envisioned. But it's architected well enough and flexible enough that getting something standardized is only the start. Then you let the innovators, and all the innovators that we got in our business, grab ahold of it and say, "Oh, I can use it for this. I can use it for this."

Chris Lennon:               04:10               So we've outlined probably a dozen kind of based use cases that we think it might be useful for. But I will put money on the fact that we haven't anticipated half of the things that it could be used for. 

Aimée Ricca:                04:22               And that's the case with the work horses like time code and the color bars. Who knew that that would be a thing on a t-shirt?

Chris Lennon:               04:30               Right.

Steve Davis:                  04:32               Or the Indian head that people used to use a long time ago?

Steve Davis:                  04:38               The other important aspect here I the fact that with EIDR and Ad-ID, they're open registries. So in the past people have had proprietary numbers, if you will, to identify content and they thought that was an [inaudible 00:04:50] property. But it's not. So having EIDR and Ad-ID that are available to everybody, that is open to anyone to sign up for and register, that's really important. 

Aimée Ricca:                05:00               So this is ... I guess this would go back, probably not as far as I think it would, but there used to be the days where it was literally a spreadsheet or a fax-

Steve Davis:                  05:09               It still is in some cases. 

Aimée Ricca:                05:12               That's why I say maybe not so long ago right?

Chris Lennon:               05:14               It went from spreadsheets to pivot tables because media has fragmented so much. It's just becoming so much more complicated for the clients, the marketers, the agencies, the media owners, to really get a sense for where all of their content is being distributed. It's great that the consumer is more connected to their content but it's really frustrating that it's hard to aggregate that back. And that's what this standard is trying to solve.

Aimée Ricca:                05:39               Because there's so many channels now. I mean multi-channel distribution, we're talking not just the hundred channels you have on television, this is also the channels that you have on your phone and digital and-

Paul Mears:                  05:52               Screens everywhere. 

Chris Lennon:               05:53               Yeah, that's kind of a key point. And this was architected from the beginning to be multi-platform enabled. Because if we didn't solve this for all the different platforms we're not looking forward to even where we are today and where this industry is going in the future. So yeah, very important to know this works for however the content gets to the consumer, this mark will stay on there and will be able to be detected. 

Paul Mears:                  06:16               And we got this point by testing pretty much all the different watermarks that are available in the market today. We had an independent lab run all these tests and the Kantar Media technology survived everything. So any form of distribution, the watermark survives. So that's like compression, everything else. Also it can be picked up acoustically through a microphone, like on a phone or a tablet or any other device as a mic. The watermark can be recovered.

Aimée Ricca:                06:43               Interesting. So yeah, I could see how that would be extremely useful. Is it already being used now?

Steve Davis:                  06:50               So parts of the watermark is being used by Kantar Media, right? So I represent that business. In the United States we're an ad intelligence business primarily. We track the advertising and we really don't use watermark for that purpose. But outside of the US where the TV currency business and about 42 markets, where the watermarking technology is a critical component to the audience measurement piece. 

Steve Davis:                  07:16               We've tried different applications of the watermark for some of the things that we're trying to explore here now with OBID. But it just takes so hard to deploy. It takes time. And that's what everybody just doesn't have. You know, technically it can be done and solvable. We've done it a bunch of times. But what's so great about this initiative is you embed once and then you let all the applications-

Aimée Ricca:                07:43               So the master-

Steve Davis:                  07:45               Totally.

Aimée Ricca:                07:45               Gets embedded in any version there of the master automatically has it too. 

Chris Lennon:               07:50               The other thing that's worth talking about, Paul, is that the POC that we've done and then the upcoming kind of field trials that we're looking at doing with the [inaudible 00:07:58]. 

Paul Mears:                  07:58               Yeah good point. We did a POC with Fox Labs because it's great to talk about and you know, have something as a standard. But to actually make sure it really works. So Fox Labs-

Aimée Ricca:                08:10               It's all about the adoption of the application.

Paul Mears:                  08:11               Exactly. Until you actually put it to use you don't really know. So we ran a whole bunch of tests to Fox Labs, through various distribution means, normal [inaudible 00:08:21] TV, playing off of DVR, internet distribution, and we also picked up the content on a tablet and we showed that the watermark survived pretty much everything we could throw at it. So we were able to pick up the EIDR and the Ad-ID codes unscathed which was great. 

Paul Mears:                  08:39               Coming up, I'm going be involved in the ATSC 3.0 test in Phoenix. So we can actually put those into another real world scenario and that means that this watermark would be kind of a template on how it would be embedded into broadcast workflows. 

Aimée Ricca:                08:54               Okay. 

Steve Davis:                  08:56               And Chris, maybe you want to peak a little bit about kind of CIMM, and they were the sponsor of this initiative, but so many of their members were behind this. I think it's so critical to know who really was involved.

Aimée Ricca:                09:08               And CIMM is just-

Chris Lennon:               09:11               CIMM is the Coalition for Innovative Media Measurement which is a coalition composed of media companies, advertising agency organizations, that sort of thing. All of the stakeholders who would benefit from this. So CIMM has existed for quite a while and they're trying to enable this multi-platform measurement. So OBID is a key part of that. You know, it's a technology that's going to enable that. That's an important point, Steve, is that people shouldn't think that this is just a science experiment and a cool technology, this is something that came out of real world business needs. Which, I think, every good standard, every good technology initiative starts with some real business needs that the industry said, "We desperately need this." So they formed this coalition, they supported this across, about five years now, of developing this. And getting to the point now where we're getting ready to deploy it.

Chris Lennon:               10:07               So I think those you might say, "Hey, you've developed a nice technology. Is it ever going to go anywhere?" I think you just need to kind of look at the people who have supported it to this point and say, "Hey, some very big organizations, the big media companies-" and I'm not going to list them all. But the people who you would expect to have behind this, are behind it. And the agency community is behind this. The advertising community is behind this. So I think that really bolsters the chance of success of this in the field. 

Aimée Ricca:                10:36               Well it sounds like it would make everybody's workflow easier and the tracking much easier. I mean you used to think you need to track so that you know if there's a makeup or something like that. But now you just need to track because there's so many different channels. 

Paul Mears:                  10:52               Yeah, and especially for cross platform, to your earlier point, that was really the whole initiative, was to measure cross platform. 

Chris Lennon:               10:59               I always say, you know, I must be a really boring person at parties because when I talk to somebody at a party about something like this, if it's somebody outside of this business, assumes that we figured this out decades ago. You can't identify what you're sending to people? This is like fundamental. Every other business that you can think of identifies what they're selling. Right? And tracks what they're selling. It's amazing that we haven't really nailed this yet. But I think this shows we're finally on our way to that. 

Steve Davis:                  11:30               So it is fundamental but I don't want it to be thought of as just an efficiency initiative. There is that, but I think the fun part ... I mean I think the thing that a lot of our constituents get excited about is it allows us to also make great advertising in the use cases that we're seeing around the marketer need for this. When you can focus on using the platforms in different ways for your creative experience, that's game changing in so many ways. 

Steve Davis:                  12:03               Consumers aren't necessarily just looking to avoid all advertising. You know, that might be a myth out there that people don't want to see any advertising. Yet we all tune into the super bowl here in the US for the advertising. 

Aimée Ricca:                12:17               Yeah. 

Steve Davis:                  12:17               Consumers love good advertising. This allows for better advertising and you know, when we think about all the efficiencies, yeah it's that. But there's also some really good positive stories I think that's going to come because we're enabling them.

Aimée Ricca:                12:31               Well there was even this past Super bowl you reminded me that there were some ads that asked you to view the full ad on YouTube. 

Steve Davis:                  12:38               Sure. And another great example and use case that we're seeing, you can only do that when everything talks to one another. Right? Where that might help for the ad component, the analysis of how that ad performed is also obviously really important. Connecting the ad with the content or where the ad ran, whether it was the Super bowl or ran against the Grammys, being able to see how the ad performed relative to the content is also critical. And also really hard to track right now. So again, this solves a lot of use cases. 

Chris Lennon:               13:21               And like you were saying Aimée, today kind of state of the art is, "Oh, go into the general motors app and click this." What OBID enables is take that whole manual staff out of there so somebody doesn't have to do that. It can come up automatically. Your phone can hear the fact that that general motors spot is airing and take you to an ad content automatically that's targeted to you. 

Aimée Ricca:                13:45               That's really cool because everybody wants to see things that they actually need and be able to find them. If it's tailored to the behavior of the person I could see that being very useful. 

Paul Mears:                  14:00               And that's just one use case. The nice thing about having an open standard is there could be two kids in a garage and doorway that come up with a great application based on this technology that we don't even know about yet. 

Aimée Ricca:                14:10               I'm sure he's thinking about it right now. 

Steve Davis:                  14:13               Well we already know we're going make t-shirts. So we already have one application. 

Aimée Ricca:                14:16               Well there you go. Now we've talked about it being a standard, but what does that really mean? And is it a standard that's available now? So what does it mean to be a standard?

Chris Lennon:               14:31               So the base standard that defines how to encode and decode in OBID is fully published and available on the SMPTE store right now. It's known as SMPTE ST 2112-10 for the propeller heads out there. It's all known as the 2112-x suite. That's the first document to be published. It's the most important one. Then there's a bunch of other documents that define the TLC. And some recommended practices for how to implement this and how this kind of fits into the ecosystem that are following right on the heels of that standard. 

Chris Lennon:               15:05               We expect a knocking on something that appears to be wood here, we expect the full suite to be fully published within the next couple of months within SMPTE. 

Aimée Ricca:                15:15               Great. I know that there's a couple documents I think that are that are about ready to go. 

Chris Lennon:               15:23               But the most important one is available right now and I know that even though it was just posted to the store a couple weeks ago, I've already got inquiries from people who are downloading it who have noticed there was something not quite right. That's good because we know that it's being downloaded. 

Aimée Ricca:                15:40               Well and part of the standards process, like I believe part of the standards process is to revise them overtime to examine them and go back and revise them if that's necessary. 

Chris Lennon:               15:54               Yeah absolutely. I think any standard that does not get constantly reviewed and updated is either obsolete or completely irrelevant because it means people aren't paying attention. So many examples in SMPTE of standards that have been out there for ... One I've been involved with for 15 years now, and we update it pretty much every year. But that validates that it's being used and the industry today is totally different than it was 15 years ago. So I think again that speaks to good standards within SMPTE of the ones that stand the test of time but also are constantly updated. 

Steve Davis:                  16:29               Yeah, and once people start implementation they're going to find things that are maybe not totally clear that will be refined and added into the standard. So that's a good point. 

Aimée Ricca:                16:40               It's good to know that these things are looked at, reviewed, and it's not just once a standard it's a standard it doesn't change. 

Chris Lennon:               16:47               No and there's a formal process in SMPTE that first of all one year after every standard is published, there's a mandatory review of that. So whether or not you get any input in it, there's a flag that comes up. It says, " Just take a look at this now that we're one year down the road. Let's just make sure." And it can be as easy as just saying, "Yep, rubber stamp it. It's good." Or you know, it could prompt some discussion. So I think it's a good procedure. There's one in five year reviews that are mandatory. 

Aimée Ricca:                17:13               Well either way it's a good procedure to make sure that people are looking at it and reviewing it. So what do people that are working in your industry now, what should they do if they really want to be able to leverage the technology?

Steve Davis:                  17:26               Yeah, we've been calling out a couple of next steps. You know the first one is depending on who you are, if you're a content producer and you're not a member buyer, we're suggesting that you look at their organization and register and get your content EIDR coded. Similarly if you're a marketer and you're not already using the great work that Ad-ID provides, you ought to. And then we want you to start to embed OBID. 

Steve Davis:                  17:59               One of the things, you know again, we've talked about trying to now unleash the community. CIMM, Kantar Media, a lot of our partners with Ad-ID and EIDR and others, are looking to open up some workshops. We want to bring the community to have these kind of hack sessions to think about ways that we can improve upon what we've kind of unleashed. So there'll be that kind of informal process going on. Then Paul mentioned ATSE 3.0 in Phoenix and maybe want to say what our plans are there. 

Paul Mears:                  18:31               Yeah so the plan there is to have the content actually watermark, file based watermarking, so the content itself, the ads, the program content will be watermarked and it can sit on a play out server. And then you have the TLC embedder which actually operates in real time that's a separate watermark that's ... So you have these two watermarks that coexist. That separate watermark basically will allow within ATSE for us to measure the distribution as well. Because as part of this pilot people are very interested in understanding how the MBPD's actually ingest the ATSE 3.0 content and build up those workflows. So that's really the goal. 

Aimée Ricca:                19:14               So is there anywhere people can go right now to ... What website should we send them? It sounds like multiple websites maybe?

Paul Mears:                  19:23               I don't have the URL in front of me but www.kantarmedia.com is a great place to start for us. 

Chris Lennon:               19:27               Yeah and if you want to learn more about CIMM and its mission and a little bit about the OBID thing it's sim-us.org is a good place to go. And of course SMPTE-

Aimée Ricca:                19:38               S-M-P-T-E.

Chris Lennon:               19:39               smpte.org we have some resources available there in addition to of course the published standards. 

Aimée Ricca:                19:45               Great. Well thank you for taking time out of the very busy show to sit down with us today at NAB Live. We look forward to hearing more about the technology in the future. 

Chris Lennon:               19:55               Thanks Aimée. 

Paul Mears:                  19:56               thank you. 

Aimée Ricca:                19:57               Thank you.