The Nielsen numbers have seemed a bit dire this year: the Emmy Awards were down 600,000 viewers, the American Music Awards saw a dropoff of nearly 3 million watchers and everyone’s been lamenting the loss of millions of viewers for NFL games in 2016.
For Campbell Foster, director of product marketing for Adobe, the numbers are indicative of today’s increasingly fragmented content and device landscape, where consumers have more choices of things to watch on more devices. And that means advertisers need to rethink how they engage specific audiences across both devices and platforms.
Enter Adobe’s new TV Media Management (TVMM) platform, which promises audience-based selling and more accurate forecasting. Instead of ad inventory being sold simply on the basis of the programming itself, TVMM aims to offer custom audience packages using analytics, taking into account existing audience data and viewer preferences to micro-target audiences on every platform and device.
Foster spoke with Next TV contributing editor Chris Tribbey about TVMM’s goal of better forecasting to avoid running ads against the wrong media; how the service can work within existing ad-tech infrastructures; and how media sellers can still make their delivery guarantees, without the guesswork of third-party measurement firms. Here’s an edited excerpt of their conversation.
NTV: How did Adobe’s new TV Media Management (TVMM) platform come about? What was the impetus?
Campbell Foster: No surprises, people are watching multi-streaming. They’re watching on their devices, they’re watching VOD, and more they’re streaming ads. There seems to be a secular transformation happening in the television industry right now, and even though it’s scaring a lot, it’s certainly an interesting time for TV. There’s been a lot of concern, as you know, about ratings declines. The sky is falling. All anybody can talk about right now is the NFL ratings decline, down 4 million viewers from 2015, according to Nielsen. Certainly with the Olympics, the reports were all universally negative about ratings and overall linear TV consumption.
But what we’re seeing for the glass-half-full view, and something that hasn’t been covered, is that the increases in multiscreening are more than offsetting the decline in linear. So while the coverage of the Summer Games was all about how people are tuning out the Olympics now, what we saw is that the authenticated consumption was up enough to offset the ratings decline, and nobody covered that. We were really surprised by it. That was a story that multiscreening is really picking up substantially enough that the media companies certainly took notice. They can get the reach they need by firing across channels. And that’s what our news here is, providing the tools to help them capitalize across all different screens.
NTV: What are the challenges facing advertisers in multiscreen and all these different platforms?
FC: In terms of multiscreening, advertisers aren’t dealing with fragmented platforms trying to get the consumers on apps, TV, Chromecast, desktop, Xbox, but dealing with fragmented testing. So, before you had consumers sitting down in their living rooms and watching one of three channels. Now they’re not just watching one of 150 channels: They’re checking their email, they’re doing Twitter and they’re doing a million other things. Just reaching consumers is harder.
Between Hulu going no ads, Netflix and Amazon, people are just watching less TV advertising. And for advertisers, it’s a pain in the butt to work at the availability of specific audiences across all of these different devices in a way that scales really effectively.
How do we look at a new season of a show? What’s the audience going to do with that, and how do I reach the consumer? And the media companies increasingly are dealing with these buyers, who comes up and say: “Look, yes, we want to buy audiences in these NFL games. We think that’s an attractive audience. We’re a beer company and we’re trying to reach 50 million people, but we have a very specific audience in mind that we are trying to reach.”
So, media companies are transitioning from pure content-based advertising sales, just showing in a NFL game, to audience-based selling. True audience-based selling, where the advertiser goes to them and says, “This is the audience that we want, these characteristics, and that’s only what we’re willing to pay for and, oh, by the way, we will pay you more. We will give you much more money to reach a specific audience, but we don’t want to buy anything that is not part of that audience.” So the media companies are scrambling to adapt to that.
NTV: What about the networks? How are they transitioning to audience-based sales?
CF: The largest cable network told us that they are going from 4% audience-based sales to 30% in 2017. That was a pretty shocking figure, and that’s across the upfronts and shadow markets, so that’s significant. Doing audience-based selling in the upfronts is pretty new and scary for these guys, but advertisers and agencies have gotten a lot more digitally savvy by borrowing techniques from display. They know that they can reach a specific audience, so they’re demanding more accountability from the media companies.
And the media companies and the agencies, the buyers and the sellers, are both saying: “Don’t make us rip out any of this legacy ad-serving infrastructure. It’s too expensive to do.” They’re saying the costs of replacing that are not enough to justify what you guys are offering, so let me leave the technology in place so we can further milk those layers of the infrastructure investment. Those are some of the problems we’re trying to solve with Adobe Primetime TV Media Management (TVMM).
NTV: What makes TVMM unique? What does it offer that other solutions don’t?
CF: TVMM is ad planning, forecasting and it deals in optimization solutions. We go to the sales teams at media companies, and say, “We are going to help you better forecast the availability of specific audiences in specific time periods, or new and existing programs.” That’s the core value problem: How much of a given audience? The advertiser comes to them, they go to Fox, they go to CBS, they go to Turner and say, “This is the audience I want to reach, I want a guarantee of this. Can you do that?” Today, [networks] go into Excel, they hold their fingers up to the wind, and they talk to their research person and say, “Guys, Procter & Gamble wants to do this $200 million buy, can we do this?” And there’s a lot of winging it and guessing.
With TV Media Management, including Adobe analytics data and Adobe audience manager segments, they can forecast the future availability of specific audiences at any given time with roughly 86% accuracy versus about 40% accuracy of what they’re doing today.
Today, they’re using Excel or they’re using their ad server forecasting tools, which don’t work very well. I used to run telemarketing for DoubleClick, and we really didn’t invest in the forecasting because it wasn’t a money-maker for us. It wasn’t a huge pinpoint for the customer; it was always a secondary problem. It’s becoming a much bigger problem for them, because there is a lot more money at stake, and audiences are moving to all these screens, so it’s becoming more of a pinpoint.
We’re helping the media companies make more money, because they can charge more to reach these specific audiences, and the tool takes into account viewership on all of these different devices. Historically, advertisers have relied on cookie lists to try to identify the audiences and reach them, but in these cookie list environments, like Roku, McAfee, and others, it’s harder to identify segments and reach those audiences, certainly in a planned and forecastable way. That’s the capability we’re bringing to the media sellers: help me forecast, help me understand, help me optimize yield against the future availability of these advertising opportunities six months from now, so they can sell it right now. So we’re only talking about direct sales. We’re only talking about negotiated deals. This is only premium inventory. This really has nothing to do with programmatic, or remnant or short-form. We’re only talking about episodic TV sales, and live broadcasts. Those are the use cases that we’re addressing.
NTV: What’s the overriding problem with today’s ad-planning systems?
CF: The current systems for ad planning are heavily reliant on technology from display days, and we’re really built to handle TV programming, which has things like seasonality, which has programming characteristics — things that aren’t relevant or didn’t exist in this play world. TV is very specific. It’s a highly codified way of buying and selling advertising that doesn’t translate well today; they’re very different technologies. So the TV Media Management platforms looks at signals specific to television, taking into consideration of seasonality, audience characteristics, and programming characteristics, to better forecast and help plan those ad sales for the media companies.
Ultimately, the goal is to eliminate media labels, so it’s either overselling or underselling. So the media company can look at all of the deals they have in the hopper, and they can say, “OK, we’re negotiating with these eight different advertisers; there’s four different agencies. What am I going to be able to deliver on, and how do I make sure I’m not overdelivering or underdelivering,” in which case they have to issue audience efficiency units or AEUs, things that sales guys at media companies hate. They hate underdelivering and they certainly hate overdelivering, because they’re losing money in both cases.
NTV: Who is Adobe looking at for TVMM?
CF: We’re going after tier ones, the big distributors like Comcast and DirectTV. We’re doing a lot with AT&T right now, and the big broadcasters as well. So Turner, NBC, and so on. Internationally, we’re doing a lot with Sky in the U.K. Time Warner is a customer. Rogers in Canada. NHK in Japan. So the product continues to do well.