TV buying is entering the realm that digital marketers have long operated in. Advertisers can target tens of millions of US households using third- and first-party data rather than buying in broad demographic strokes. TV buys are also becoming programmatic , results no longer take months to get in, and campaigns can be optimized on the fly. That’s the promise of data-driven TV targeting.
It sounds great — especially to a digital marketer’s ear — but how much of the talk around advanced TV, an umbrella term (perhaps first coined by IPG’s Cadreon) for data-based TV buying on linear addressable TV, over-the-top (OTT), programmatic TV and more is reality-based, and how much is, well, still hype? (For a primer on some of the terminology and capabilities discussed here, see our FAQ: Addressable TV & the convergence of digital video and TV ad buying .)
We talked to media buyers, strategists and others in the space to find out what it’s really like on the buying side of advanced TV these days, not just what’s promised.
“Everybody in this ecosystem — platforms and supply partners — boast they have stuff they don’t,” said Erica Schmidt, EVP, managing director of Cadreon, IPG Mediabrand’s ad tech unit. Schmidt, who has a digital and search marketing background, says part of her job is to figure out what’s reality.
“We wanted to take the principles of data and automation and apply them to TV to find networks, programming and dayparts above the point of diminishing return that a linear TV plan will have and find ‘spots and dots’ that can extend reach,” says Schmidt of Cadreon’s entry into TV. “Anything that can be delivered programmatically, we believe should be. That’s where TV came into play — let’s take the principles of data and automation to TV.”
One big shift that is happening is that brands are thinking much more strategically about customer data.
“We are increasingly finding clients are taking data seriously and have gotten their ‘data house’ in order, including CRM and other digital programmatic signals that are now in place,” said Schmidt. “That gives them the opportunity to tie that data throughout the planning process.”
As data enters the realm of TV buying, it is taking on sensibilities already native to digital practitioners and strategists. One would assume digital marketers are well positioned to inform the trajectory of advanced TV in their organizations with the convergence of data-informed audience targeting with TV.
In terms of data execution, addressable TV is quite similar. “You’re using data sources to identify and triangulate audiences,” says Tobias Wolf, executive director at GroupM’s Mindshare, who got his start with addressable TV 10 years ago in his previous role running the American Express US account. “The delivery is slightly different, but the planning is essentially the same as digital in identifying audience segments and delivering personalized messages to them. You can be running five targeted ads [in addressable] versus one in traditional TV.”
There are similarities in terms of accountability, too. “You can get direct match-backs to the household (anonymous) and provide ROI or ROAS, which has proved very effective,” says Wolf. “Particularly for [GroupM client] Volvo in terms of ROAS. If I’m in the market for a sedan and you’re in the market for an SUV, I’ll see a Volvo ad for an S60 and you’ll see an ad for an XC90. In programs we are watching, not what Volvo thinks we’re watching.”
In terms of the actual buying and activation processes, though, the level of targeting and automation depends on the type of delivery.
A linear addressable buy via an MVPD [multichannel video programming distributors, including cable operators and direct broadcast satelite] is currently the most common method targeting households with some sort of audience segment, says Anthony Laurenzo, SVP of non-linear strategy and investment at Dentsu Aegis Network (DAN). Laurenzo provides input across internal agency teams about opportunities for TV and video planning.
OTT buying on platforms like Sling, PlayStation Vue or Roku is slightly more complicated, says Laurenzo. Those buys can leverage subscriber data and can target households via connected TVs. Roku, for example, announced a licensing deal just this month to include its streaming services on Philips-branded smart televisions in the US. But says Schmidt, “OTT is more like digital video and has more flexibility in terms of on-the-fly optimizations.”
“With MVPDs, Schmidt says, “we can activate and send over insertion orders on an automated basis, but in reality they are manually engaging. And for national broadcast spots, automation isn’t at all possible.”
Addressability to households on providers like Dish Network, Cablevision and Comcast is more in the spirit of programmatic with the ability to do some optimizations in-flight. For example, Schmidt explains, if they’re doing a zone-addressable buy targeting in-market auto buyers and are able to see sales match-back rates quickly, they can make adjustments. But only by going to the addressable providers, and that still involves some back and forth.
For brands that already have a traditional TV strategy, addressable is seen as an add-on, with a look to extending reach and efficiency of a traditional linear TV plan.
“Eventually there’s a point of diminishing returns with broad demography targets,” says Schmidt. “In order to be more effective, we need to identify high-value audiences and use data sets to identify them cross-stream, including digital video, OTT, addressable linear, etc. It’s a holistic planning process using traditional TV planning and audience planning underpinned by data.”
IPG has created an in-house data stack, called AMP, that’s designed to be a single platform to plan, buy, measure, and manage all advertising campaigns “whether your target audiences are viewing ads from set-top boxes or smartphones and complement traditional TV,” says Schmidt.
Still, traditional TV measurement sources play a role. “You can debate the reliability of Nielsen and Rentrak data,” says Schmidt, “but it’s still the common currency, so we’re going to start with that and then layer on additional data depending on the client target. That may include the client’s first-party data as well as data from an OTT platform like Samba [TV] and from IPG partnerships to augment Nielsen and Rentrak.”
DAN uses in-house tools to identify the right media mix to reach the right audiences. “In some cases, it might be a large linear buy. In others it might be a combined buy with Dish and 2 million+ incremental SlingTV households, for example,” he says.
“You’re not trying to reach everyone if you’re doing good targeting.”
EMarketer estimates the US linear addressable TV ad spend will reach $1.26 billion in 2017, up 66 percent from the prior year. Still, that’s just a fraction, 1.7 percent in fact, of total TV ad spend. By 2019, that percentage is expected to increase to 4 percent of TV ad expenditures in the US.
There are somewhere between 75-85 million addressable households on linear cable and video on demand services. OTT has been growing rapidly over the past year to 51 million households, according to comScore. (There is some overlap with households that have both set-top box and a streaming device).
Still, most I spoke with see scale as beside the point.
“Scale is not a problem now,” says Wolf. “Early on, 10 years ago, we started with Cablevision with 2.6 million households.”
“You’re not trying to reach everyone if you’re doing good targeting,” he adds.
Schmidt echoes that sentiment. “Scale should be less of a concern when you have precision. And it will be a lot more accountable than if you hit your GRPs [gross rating points, a metric TV buyers use to determine how many people within a target demographic might have seen their ads].”
With addressable linear TV in about half of US households, the addition of that plus OTT and digital video can “provide greater reach than with traditional linear only,” adds Schmidt.
“You can use addressable to find incremental reach, increase frequency against a high-value audience, negatively target to exclude certain households (say, loyal competitor buyers), use multicultural household targeting — and that’s probably just half of the use cases,” says Laurenzo.
In terms of volume, however, the reality is it’s still quite limited.
For one, rather than opening up data targeting to their programming network affiliates for the full 14 minutes of TV ad time typically available per hour, TV operators only make it available on their own share of local ad time, which is roughly two minutes per hour.
Dave Morgan, CEO of Simulmedia, which runs a SaaS platform for data-based linear TV advertising targeting, says the real scale of addressable should be put in context.
“While linear is being challenged, it’s still massive and it’s not going away anytime soon. People seven or eight years ago said that by today you’d be seeing all or most of TV fully addressable.” But that hasn’t happened yet.
To illustrate his point, Morgan says, “A ‘Judge Judy’ episode will deliver more advertising time than on all of the videos on YouTube all day. One 30-minute show will deliver 12 to 15 minutes of ads. That scale doesn’t match linear.” (I haven’t fully fact-checked this, but note that most of YouTube’s ads are skippable, and according to a recent VAB study, music accounts for more than 30 percent of all time spent on YouTube.)
However, the fragmentation of watch time across a huge number of channels is a challenge for media buyers, says Morgan, because “the standard methods of measurement that inform advertisers where to buy aren’t very effective anymore. Vast amounts [of watch time] happens on shows rated 0.5. You have to buy a lot more spots … and the error rate of Nielsen panels break at that 0.5 point, so it’s as likely to be wrong as it is right, which is why people buy big buckets of GRP averages.”
“Fundamentally, change needs to come from the sell side.”
The opportunity, of course, is “if you could get really granular data on thousands of ad spots, it’s going to be more like digital with the ability to use purchase behavior or CRM data or other digital behaviors. If you have the data for targeting and software to analyze all of the spots to ensure you have the maximum reach and penetration, instead of a $3 million buy for a month for 100 spots on 10 networks, a computer system might suggest you buy three thousand spots on 85 networks,” says Morgan. “That’s where we are starting to see this data-optimized advertising show innovation. The reason it’s starting to happen faster is there is a win-win-win realization.”
That equation works out like this: Buyers can use data to optimize their ads and get better results for less. And they can track exposures back to see impact on ROI. Media sellers benefit because they’re not getting credited for wasted frequency, and they can optimize their sales at a premium based on data rather than the time of day the ads serve. Consumers win because they’re less likely to get redundant and irrelevant ads – that also means they’ll be more likely to stay tuned in.
Morgan says that at most 1 to 2 percent of TV ads are being managed this way. “But there are a lot of incentives that are going to force it … Fundamentally, change needs to come from the sell side and they’ll need to invest in systems to do this.”
Morgan points to AT&T’s Brian Lesser and Comcast’s David Clark as evidence that the sell-side is moving in this direction. “Some are going to aggressively invest in this and they will grow faster,” predicts Morgan.
“Legacy TV owners like NBC are starting to change the game and build addressable into their offerings. How the large media owners start to change and bring the ability to target ads to households will be a trend to watch,” says Laurenzo.
NBC Universal has launched an API to enable brands like Target to programmatically buy ads using first-party data across NBC programming.
Another example is Open AP, a joint initiative by FOX, Turner and Viacom to support audience targeting that launched in October. Advertisers can create audience segments and share them across sellers on the platform.
“TV is becoming more accountable like digital,” says Laurenzo. “It’s not going to replace linear TV in the near-term, but leveraging data we can now find out faster if a campaign was effective, and just as important, you can use those learnings across your media plan and roll that back into linear TV buys, such as insights into which networks are driving performance locally and back that in nationally.”
For advertisers, says Wolf, “If you want a competitive advantage for core audiences, you can do that in addressable TV at scale.”
Wolf speculates that we’ll start to see digital and addressable TV buying teams coming together more in the future.
Laurenzo agrees, “I can see the skill sets coming together. If you think about something like addressable, you’re thinking about television the same way as digital with targeting and things like frequency capping. The underpinning of data across everything fundamentally changes the business and puts everyone on the same playing field. There are still going to be specialists in TV and digital, and those roles are really important, but they’ll share a skill set in data.”
Cross-stream targeting that syncs up addressable linear with digital video campaigns, so you’re looking at unique reach and frequency across channels is also a trend we’ll see develop in the near term, says Laurenzo.