Today’s immediate availability of metrics is resulting in a performance-based advertising environment that’s difficult to provide in the local linear environment. How that gets resolved is the question going forward, said a panel of executives speaking during a TVNewsCheck webinar, Attribution, AI and Optimizing Spot TV, last Thursday.
In digital environments, attribution is relatively easy to track. A person goes on a website, clicks on an ad or a link, goes to a purchasing environment and either buys or doesn’t. Digital cookies, IP addresses and other tracking methodology help marketers see what those customer journeys look like.
That is not the case in a linear broadcast environment, where a signal continues to be broadcast from a single point (or points in the case of repeaters) and received by a relatively large audience — also known as a one-to-many model. Because those signals don’t necessarily pass through a trackable device, it’s hard to deliver attribution information, other than ratings, to an advertiser. As TV becomes increasingly delivered over the internet to smart TVs and other smart devices, viewership patterns are becoming increasingly trackable, but it remains hard to drill down into demographic and targetable information at the local level. This lack of a reliable local measurement system is making it hard for broadcasters to operate as nimbly as digital marketers do, but doing so is necessary as advertisers and agencies demand more and more information about how their campaigns are performing.
“Advertisers want to see how their advertising is working, and if it is driving real world outcomes that move their business forward,” said Christopher Martinez, OTT director of sales, Hearst Television. “In the last decade, and even more so since the pandemic, performance around marketing has really accelerated. So being able to say I’m spending x number of dollars on my advertising, and I can point to x plus amount of revenue for my company, is a massively important thing to be able to do.”
Bridging this gap between digital and local linear advertisers is a key goal of local broadcast marketers. It starts with education, said Stacey Lynn Schulman, EVP, platform marketing and Intelligence, Nexstar.
“I think when we talk about the obstacles to getting this launched, a lot of it is education. A lot of it is just getting advertisers, whether they are local or regional, comfortable with using these tools and getting national advertisers to understand the differences between the capabilities that you can do at the national level versus a regional or local level,” Schulman said.
“[We want to make] sure that we don’t set up a system of haves and have nots, which typically we see in traditional measurement,” she added. “We don’t want local to be in the have-not bucket. We want our stations and our capabilities to be at the same level of sophistication as what you could get in any place in the marketplace, whether it’s national or local, or digital or analog. Local is digital as much as any other platform.”
While measurement of local markets remains lacking, it’s a problem that providers, such as Comscore and Nielsen, are working to resolve. There also are tools in the marketplace to help broadcasters better be able to do this.
“[Attribution] really empowers media sellers to have more data-driven conversations with their advertisers about what’s working, what might not be working and how to optimize their campaigns,” said Paul Cramer, managing director, media and broadcast, Veritone. “What we’ve really focused on [with our attribution product] is helping the linear side [correlate ratings performance] to the behavioral data, whether that’s Google Analytics, Adobe Analytics or what have you.
“It’s a really great tool — not only on the proof of performance and attribution side — but a lot of our clients are actually using it as a prospecting tool,” Cramer said. “One of the hardest things any salesperson has to do is get that meeting with the advertiser. When we overlay category data onto this, now salespeople can go out to an advertiser in a category and say, ‘have you thought about advertising with us? Because we typically see when you make this type of investment or buy this type of a campaign, we get this type of outcome.’”
While all of that sounds great in practice, the same problem exists around attribution in local markets that exists around demographic ratings data.
“A lot of these attribution models at a national level are dealing with scale and statistics that you just don’t get on a local level and that really impacts things,” said Michael Newman, director of transformation, Graham Media.
“This statistical significance in local markets when it comes to data collection around attribution cannot be ignored. It needs to be the first part of every conversation,” Martinez added. “Can we prove that an outcome happened? Probably. Can we prove that it happened at scale? Probably not. So, it’s really, really important that we have the education piece.”
Big Data will be part of the solution, but overlaying Big Data onto local markets remains an imprecise science. Big data is still too big to be able to accurately drill down into precisely what’s happening in local markets.
“When we talk about using any sort of models, it’s garbage in equals garbage out,” Schulman said. “What we’re all striving to get the best results for an advertiser. But if the way that we’re measured is not accurate, then you have to be concerned about the inputs going into that model that are spitting out answers for how media spend should be allocated across different media. The biggest concern, obviously, is that linear television is not getting its due because the measurement is not what it should be. So big data is a bridge that helps us get there.”
That’s where artificial intelligence and machine learning are coming into play. For the past year and a half, a lot of media attention has been paid to ChatGPT, Midjourney and other generative AI applications. But at TV stations, AI and machine learning are used much more to automate processes, creating more time for employees to do the parts of their jobs that require a human touch.
“A lot of AI is just automation, and AI, from that perspective, has been around for a really long time,” Schulman said. “It’s all about how we automate processes, and a lot of the reasons why we’re automating processes are because things are just moving that much faster, and we have to move at the same pace as the industry.”
And Newman said: “Any opportunities to bring some of the automation, AI and machine intelligence that agencies and buyers are already used to having on the digital side into the broadcast side, those efficiencies are crucial for us [as broadcasters] to continue to do what we’re doing, and to expose our sellers to more and more of our local clients who need this information.”
AI is currently helping TV stations to automate such tasks as information sharing, product management, rate card estimates and so forth.
“AI is really about workflow efficiency,” Cramer said. “The tsunami of content being created is just getting bigger and bigger. And so, whether it’s automating, clipping and recapping and things that used to take manual time, that’s what AI is really good at. The other thing that AI is really good at is distilling very large data sets into actual intelligence.”
Cramer noted that Veritone works with its clients across AI-powered platforms to create custom solutions. Some clients need automated translation because they work across markets where different languages are spoken, and other clients need to be able to scan transcripts quickly to provide advertisers with metrics on how many times their product or brand was mentioned.
When it comes to spot TV and political advertising, AI and automation are particularly helpful because inventory is moving fast with spots getting placed and pulled at a furious pace.
“Political gets really hot,” Newman said. “Automation is really where AI helps us on the political side of things. It can optimize campaigns in a way that otherwise would quite literally take days, if not weeks.”
While attribution and AI-powered automation are powered by high-end technology, what they really enable is something low-tech yet invaluable: better relationships between local broadcasters and their clients.
“At the end of the day, we have to weigh those automation AI capabilities with … personal relationships, because that’s what differentiates us from people who are just aggregating eyeballs,” Schulman said.
Digital Media is constantly trying to reframe the definition of “results”. Broadcast Television has lived in the world of “results” since its inception. Exposures, Click-throughs, time on site, conversions and every other digital analytic can be sky-high and still not result in any sales growth. Broadcast Television reaches both heavy and light web users. Broadcast TV stations are viewed via cable boxes in 100 million US homes. These cable boxes are sending data back to the cable company. Broadcasters will eventually rise up and retake their place at the top of the food chain. They can start by blocking advertising from companies who are trying to run them out of business.