New Progress for Multiscreen Advertising Data Framework

Suzanne Rainey, principal, Accord Media Group

Suzanne Rainey, principal, Accord Media Group

December 16, 2011 – In this discussion from ScreenPlays Media Innovations Summit executives leading the way in viewing measurement and data analytics explore strengths and weaknesses in various online and TV data environments and identify innovations that could take advertisers where they want to go with multiscreen campaigns.
Suzanne Rainey, principal, Accord Media – Just to put things in perspective before we get into our discussion, in 2010 the average person spent 2,000 hours using media on the Internet. This is a six percent increase from the previous year, and that was largely driven by social media and workers using software to access and manipulate information. Mobile was 77 hours, and that’s a 50 percent increase over the previous year. It’s expected to increase 35 percent this year to grow to 104 hours, and that’s thanks to the iPad and tablets. Astoundingly, the average person watches over 1,900 hours of television a year, and that’s a 0.2 percent increase over the previous year.

Let’s begin by having each of you briefly describe what you do.

William Merchan, VP, Business Development, MarketShare – I run business development for MarketShare to manage our partnerships with data and channel and agency partners. We’re a predictive analytics company using a lot of targeted media data.

Bill Feininger, SVP, Media Measurement, FourthWall Media – FourthWall Media provides a number of services in the cable space from EBIF platforms to applications. My specific area is in the data business.

Scott Brown, SVP, Digital Platforms, Nielsen – We’re obviously a company that provides measurement for television usage, product Internet and a variety of other products associated with media consumption. My role with Nielsen is largely to be an external speaker. Thought leadership is the more candid term that’s being bantered about these days. I’m in the communications and strategic relations areas as well.

Rainey – At Digital Hollywood last month one of the speakers said the current group of media buyers have to die in order to advance advertising. Is that what it’s going to take?

Brown – Hopefully there’s an alternative, because if it goes that far, then it’s gone too far. I think what the panel member at Digital Hollywood was probably trying to say – I was there – is that the media business as we know it today has been a slow grower – steady, sustained growth. It’s an industry that in effect doesn’t like a lot of change and in effect is very conservative for the most part.

The advertising community is very much the same. Programmer, buyer, seller, whatever, no one likes disruption. So what we’ve seen is an industry that every year grows by four, five or six percent. This past year in spite of a very rough economy we saw television usage increasing by 22 minutes more per month. We’ve seen that television is an amazing sustaining business. And the advertising community is obviously calling for improved measures of accountability, return on investment and harkening for an era when improved metrics can also follow. So I think every member of this panel in one form or another is interested in seeing the whole advertising paradigm move forward, interested in seeing improvements in the metrics and the insights that can be gained from those metrics.

Feininger – Not a lot to add to that. I do believe there’s a new wave of devices and a new wave of services, and somehow they’ve all got to get measured. Rather than relying on old techniques, and old services, there’s got to be new innovation in the research and on the collection side of that.

Merchan – We’re seeing that a lot of the media-buying decision making is being shifted to in-house CMO organizations, versus relying 100 percent on a media agency and a junior analyst at a spread sheet to make that decision, and thinking about it in terms of outcomes, revenue, profit, unit sales, instead of awareness and intent. I think those are fundamental shifts we’re seeing, and a lot of it is driven by the targeted media and the transparency of data.

Rainey – What was the biggest development in advanced advertising last year?

Merchan – I think the biggest development that we’ve seen is the larger brand advertisers putting more reliance and dependence on emerging media to drive impact. We’ve seen it with a lot of the clients we work with where, along the lines of what they were doing with search and display many years ago, where they were testing the waters before going all the way and making bigger buys in emerging media. Things like investing seven-digit media buys in Facebook and social media probably would have been unheard a couple of years ago. But last year you saw many large global advertisers do things like that.

Feininger – I think there’s been a lot of experimentation that has gone on. And unfortunately for the television side a lot of the dollars have left to go somewhere else to do some of those experiments. As an EBIF and technology provider for the cable industry, we think the opportunities are going to be coming back to television. That’s where the audience is. We think that as the platforms get out there, there’s much opportunity for that to take place and for more innovation and more reach to those consumers to happen in the television space.

Brown – I guess I would add to it from the Nielsen perspective, we’ve seen interactive television go through lots of starts and stops. We think back to the period when the Aurora trials were going on in 2001, and a lot was learned with AT&T Broadband and with Navic in that period of time. And most of what was learned was that the technology was not ready. Today it would appear from a Nielsen standpoint the technology is ready, beginning to deploy. Monetizing digital media is a huge challenge, because most people don’t understand what they have at this point in time.

Rainey – What do you think they have that they don’t understand they have?

Brown – There’s a gentleman by the name of Bob Liodice, who’s the CEO of the ANA [Association of National Advertisers], who got up and made a comment at a recent show that the advertising community expected that digital media would be much easier to measure, because of the nature of the devices supplying the content, the signals, the information and the entertainment. But, in effect, what they found is that it’s just the opposite, that the digital media has been harder to monetize, harder to estimate, harder to get their hands around and understand, and particularly to understand the flow of media dollars, whether it goes from traditional media to digital media, whether digital media is purely an additive.

The great issue that all companies face, and in particular Nielsen Co. faces, because we are first and foremost a measurement company, is to be able to attach measurement to each of the various small kernels, slivers of usage that are beginning to emerge. Fragmentation is multiplying. So every client is dealing with smaller slivers of usage.

The question is what is effective within those slivers. We can talk a lot about what has really changed. From my perspective the thing that’s most important for Nielsen is that we’ve been able to put in place partnerships with third-party companies that were talked about for many years but finally have reached fruition.

We have a partnership with Google where we actually provide them very deep information from our measurement panel to help them monetize their end of the business, which is separate from ours. We have an arrangement with Facebook for our new online campaign ratings, which basically dovetails panel-based viewing with subscription information and demographics that they’re enabled to collect. And by merging those two sources of data we now have a new currency that we can offer the industry for the online world and begin to allow advertisers and programmers to understand the monetization model that must exist there.

So a lot of change has occurred in the last three years, much of which gets lost in the shuffle of media coverage. And there are other companies that are going to put in place relationships, and ours is certainly seeking relationships with any provider of data that can help Nielsen provide more insight to our clients.

Rainey –It’s been said that people using analytics are overwhelmed by data that’s meaningless, and that they need data that’s actionable. What’s the reality of using the available data? And please comment on the quality of the data.

Feininger – I agree. There’s a lot of data emerging. There’s a lot of it that’s not very well done. It’s not very consistent.

We collect a lot of set-top data in our services. Today we’re in over two million set-top boxes across the United States in 14 MSOs. We understand the problem. We can collect from every single one of the devices we’re inside of.

We’ve beat the problem of how do you handle the back channel and how do you get consistent and clean, concise information back. We feel like we’ve created a new area for the industry. It’s a new way to get set-top data nice and actionable and useful into all kinds of other services.

Merchan – The thing we’ve seen is that the big data problem is obviously out there where marketers are swimming in terabytes of data. it’s only useful if you put context around all that data. We’ve used social media data, digital data and set-top box data and lots of different sources of that data to give insights to clients.

But I think just because the data is available, doesn’t mean it’s useful. How do you put context around that to help inform media decision or a marketing decision? With social media it’s great just to hear what’s happening out there is great, but if you don’t understand how to get more of that sentiment through traditional media buys it’s kind of useless. So we think providing context and insights to clients on how to drive more activity through the data that’s available now is really where there’s going to be a lot of value created.

Rainey – How do you do that?

Merchan – We do it with statistical models using these big data sets and build predictive models for clients on investments and marketing decisions for them.

Brown – I agree with your comment. Data by itself is an interesting commodity, but when it becomes information it becomes actionable. One point I guess I would like to echo is that, as all of you know, Nielsen has provided measurement data to national and local client advertisers and agencies for years and years and years. We move about as fast as the industry allows us to move, because half of our clients unwittingly will tell us we’re moving too fast and the other half say we’re too slow. So where do you meet?

The offerings of data we’ve been developing over the past couple of years have really grown to rather change our business model. Instead of being a purely panel-based company upon which we project audiences, we’re increasingly becoming more of a hybrid-based company, meaning we see the virtues of panel-based measurement, because it’s an opt-in, it’s an in-depth sort of analysis of who’s viewing, what they’re viewing and all of the kind of market characteristics that are important to clients, but also it’s one of taking advantage of third-party data sets, set-top data, data from any number of companies that actually can be integrated to create more depth of measurement, more breadth of measurement by virtue of the large volumes of data that can be secured.

So we’re really moving down that path of pushing with our client base this notion of going hybrid and trying to destroy some of the myths around set-top data, the myths being that set-top data doesn’t give you enough actionable information. We think it does. We’ve compared it to panel-based measurement, and set-top data is pretty good stuff for the most part.

There is no perfect data set. And anybody that believes there is is mistaken. Even our data is not perfect. There are weaknesses in every data set. But we think by virtue of combining the two and developing a number of models which we have developed for set-on, set-off demographics and so forth, we can offer to our clients at an affordable price a much more expansive business and a much more expansive set of metrics. And also having the ability to do the deep analytical dive into the data to understand the nuances, which so many of our clients require.

Rainey – Do we need standard demographics such as adult 18-49 in an advanced advertising world?

Brown – Gee, I thought Nielsen kind of standardized demographic reporting in one context. The answer is probably sure. I wouldn’t disagree with that. But I’d probably pass the mike to my friend Bill.

Feininger – I think the answer is, yes. But I think it’s a qualified yes. The reason I say that is for doing predictive modeling or for any kind of measurements you can place your products and services in front of consumers, you want to make sure you get it to the right place. And those services, advertisers and products still are trying to reach specific audiences. Many of them are built on age, gender and other kind of characteristics that are out there today. So I think the answer is yes.

Merchan – I’ll go ahead and be the no guy on the panel. I think the idea of demographic targeting or reach-based targeting was a means to an end historically. The reason you did that was you had no way to directly figure out whether this user was going to end up purchasing my product.

In a lot of ways those barriers are gone. You look at data providers that let you know attributes that are more relevant than a demographic – is this person in the market for a new car, do they have kids in the household – things that historically you wouldn’t even have a possibility of targeting against that now you can. As more and more devices light up more and more of that data will be available across platforms. I’d say historically it was a proxy for getting to that point, but now that you can do targeting in a more meaningful way to get better predictions as to how it might impact outcomes, you don’t need it any more.

Rainey – Analytics is driving the development of new metrics. What’s taking hold and what new metrics are being developed?

Merchan – A lot of our clients are using marketing ROI and doing it across all their portfolio of investmnets. What we’re doing is building resource allocation models for them. And so it’s not about what was my reach in a particular niche. It’s more about, if I’m going to dial up my investment by five or 50 million over the next quarter, how does that translate into sales? Ultimately their CEOs or CFOs are asking them to be accountable now whereas in the past they could have said, well, look what this says to our awareness or brand tracker. That’s not good enough anymore. The measurement we’re seeing many of the marketers we work with being accountable to now is all about marketing ROI, regardless of what channel they’re doing that marketing investment in.

Feininger – For new metrics there are two things we see over and over from our clients. They want to provide some way to determine engagement with the programming or the advertisement. Having set-top data and very granular information allows them to get to that, whatever particular algorithms they’re using to drive to that. That’s one.

The other I agree is with ROI. We see a lot of taking set-top data and merging it with other third-party data sets so they can try to find those segments, those groups or those consumers who are looking for something else. And so they can get that entire feedback loop now so that they can understand that.

Brown –Nielsen is also looking at how to cement data bases that we already have at our disposal to create more value for our customers and clients. Be it Catalina marketing, be it our Home Scan, all of that purchase data and linking in with the viewing data and then looking to see if we can basically create a cross-platform offering is being met with our clients with enthusiasm. Although the growth of mobile year over year seems very large, in the context of just regular video usage in the home, it’s a small fragment. The same is true for Internet. Yet we still launched the TV Everywhere extended screen measurement so that today homes that are part of our national panel that have PCs, we’re measuring those homes.

We’re measuring the PC streaming consumption, and the estimates, if you will, are additive for the reported C3 rating. That is a big step forward, to be able to measure the PC usage. It is cross platform in nature. So we’re really trying to tie it all together.

We’ve been issuing these quarterly cross-platform reports to our clients. It’s giving an interesting picture of what viewers and viewership is doing. We’ve been able to tease out some kind of interesting things from the data as well. We’ve learned, for instance, traditional light viewers who don’t tune into television a great deal are actually showing evidence of becoming heavier Web video users.

We’re seeing that over-the-top is becoming a bit more of a force. I don’t want to mislead. It’s not a huge segment of the viewing pie, but it’s a consistently growing segment. In concert with over-the-top content, people are increasingly making use of DVR and non-linear sorts of viewing arrangements in their homes. It skews younger, to the 18-34 year-old audience, which is more and more watching TV in a non-live environment.

All of those things have tremendous implications for an industry that measures progress sometimes in terms of single percentage points. A single percentage point change for an NBC or CBS is a big deal. So we try to provide a lot of the granularity behind that change and are trying to predict some of the shifts with the analytics we produce so that clients are not taken unaware. The worse thing a researcher wants to do is go to his boss and say, we missed this, we didn’t see this coming, because usually that costs somebody their employment.

Rainey – How does social media fit in and impact advanced advertising?

Brown – In many ways. The data has supported very strongly the contention that social
media is a very strong influencer. It’s a very strong influencer for video applications. We’re seeing more and more video moving into the social network world.

There’s a direct correlation. If you think about it, some of the major social media suppliers – Facebook, LinkedIn and others – you’re talking 300 million, 400 million subscribers. That’s a lot of people you’re impacting and influencing. We’ve seen correlations that directly indicate to us that more and more people will go to social media for viewing if the media is made available to them.

As a result Nielsen is really going down this path of cross-platform measurement but also trying to put out a kind of newer metric, if you will, a metric that we would call “total video.” Total video is unconstrained. It’s not video from your television, from your DVR, from your mobile phone. It’s video from all available sources that we need to aggregate and put together as one fixed number.

Feininger – I agree. The ecosystems may be around programs or profiles of people or what have you. That takes us in our business at FourthWall to developing technologies that allow you to connect these other devices back through to the television itself so that you can do some level of synchronization or some level of having your iPad work in conjunction with your EBIF-enabled set-top boxes. We see that as a way to help drive into that social environment and to be able to collect some metrics there and to be able to see the things that are going on between those two companion devices.

Merchan – We’ve seen that traditional media is a strong driver of social media. There’s this idea that you need to pick one or another, and it’s really not the case. You need to figure out how to orchestrate all of your media efforts together. If you launch the best Facebook page in the world, no one is going to go there unless you’re running traditional media against it in volume that actually matters. The more innovative brands are building a cross-platform approach, building a presence through their TV campaign that then drives people to social. And that’s where you’re going to see the best success. You can’t just have social be your only strategy, because you’re going to fall flat and lose.

Brown – I’d like to bounce it back to you as our moderator and ask, what are your thoughts?

Rainey – I think social is a whole category. I was talking to a researcher at USC a couple of weeks ago who’s an expert in gaming. He has a theory that social may impact 40 percent of decision making for a person. You see the growth of Facebook, 100 percent year over year. I think it’s going to continue to grow bigger and bigger. I wouldn’t think anyone would just use social as a strategy. It’s going to be a complement to other forms of advertising.

Brown – One of the interesting things we’ve delved into recently is that one of the fastest growing categories is in the 55-plus age group. What’s happening is, particularly on mobile devices, there’s a way of interfacing with family and friends, friends going back perhaps generations that you’ve lost touch with. It’s an interesting kind of experiment, if you will.

There are other companies looking at social media sites who are doing innovative things with trying to license and permission traditional video content and populate it within social media applications. Who knows if social media becomes another outlet for video consumption, but from a technology standpoint I don’t think anything prevents that. And it probably will be a viable video outlet for the future.

Rainey – How well are the cable, satellite, telephone companies and others using set-top box data, and are there significant differences in how well they’re using the data?

Feininger – All three have various amounts of data. They’re using it for various different reasons. On the cable side, we’re kind of early. We’re one of the first there that are really able to get a lot of cable data. We’ve heard lots about Canoe over the years, but frankly there hasn’t been a lot of data flowing there yet. In our case we’re doing what we can to help our clients. There’s a good amount of cable data that’s beginning to flow back both as raw data and as research reports, so they’re getting something they haven’t had before.

On the telco side there are two very different systems available today in the U.S. There are different thoughts as to how good it is or how much of it is available. There’s not a lot that’s available commercially today, but there is some. It seems to be pretty decent data.

And on the satellite side it’s kind of hit or miss. Satellite, because of the back channel not being consistent, there are some issues of reliability with it. Although there are large volumes, it just doesn’t seem to be as complete and as consistent. They do have some interesting interactive data in some cases. As we look at new ad models and new ad units that’s something that’s very important.

Our own system, our own ad widget system and TV widget systems which allow you to enhance programs or provide applications do provide a significant amount of data. So there are new areas on the cable side that are really jumping to the forefront.

I think it’s an interesting mix of data among the three. Cable in my opinion is ahead of the other two. It’s not quite as broad and wide, but I think it’s a little better data with the other two following behind fairly closely.

Brown –The process has led to incremental improvements over the years so that the data in our estimation and the data and we’ve seen – we have data from Charter and from Direct View today – actually it looks pretty good, clearly usable. And we will use the metered samples we collect for both our local People Meter markets and for the meter markets and for the national People Meter market as a control, if you will. Again, building the model so that we can fuse the data elements together and eventually get to a stage where we’ll be able to offer expanded sample sizes to our clients.

But everything is an iterative process of improvement. Nothing will look perfect out of the gate. For the Direct View data we’ve taken license to it will probably be sometime next year before we actually begin to share and release data. But we’ve already done three market trials with some of the data that we do have in house.

Rainey – Electronically traded bidding markets are developing alongside the upfront and conventional media markets resulting in decisions that are more data driven than relationship driven. Will the two markets come together?

Merchan – I think we’re seeing a lot of that happening already. You look at what Google is doing with Google TV; you look at the announced merger between MediaBank and Donovan. I think you are going to see a convergence there. There will always be some upfront, premium version of the TV marketplace, and you’ll still see it in digital media as well. Yahoo!’s Front Page is a valuable property. There are other places that you’re still going to see there’s need to guarantee delivery, guarantee that premium content.

But I think more and more you’re going to see media move into this world of bidded marketplaces and automation and linking the data sets together. The agencies are building up their tools and teams to be ready for that. Every major holding company has a group that’s focused squarely on that area. And most of it’s in display media today, but you can see how that can morph over time as more media comes online.

Feininger – I think all decisions will be more data driven. Whether it’s through the traditional up fronts or whether it’s through some bartering or trading system, more and more will be done by data-driven decisions. There’s more data out there. You’re certainly going to see more people use it. I think that as far as a trading system goes, it’s more likely it’s not going to happen in the up fronts. That kind of thing will be in the scatter market.

Brown – I agree.