Kyle Roberts, CEO at AdImpact, joins Cross Screen Media CEO Michael Beach to share his thoughts on what advertisers across industries can learn from political advertising and how local advertising is reshaping the entire landscape. Watch our latest Screen Wars Thought Leader Interview here and read the full transcript below!
MB: Hi Kyle, welcome to Screen Wars.
KR:Good to be here. Good to see you, Michael.
MB: Tell us about your background and experience. In particular, the idea behind AdImpact and what problem you solve.
KR: Thank you for having me on. AdImpact started in 2014. My background is in politics. I noticed there was a huge need in the environment for competitive information and media intelligence. Spending, detecting creatives, finding out what messages are running in a marketplace, how often they’re running, how many different types, all those things. In 2016, we built out a detection system. After 2020, we upgraded it to include detection across local cable and CTV.
The genesis of all these things was politics. It’s a very fast-moving environment, and if you’re going to use data in a fast-moving environment, you need detection systems that collect this stuff as close to real-time as possible. That’s the only time you get value from it.
We develop tools to help campaigns do this stuff for spending, for creative, and for rates. We also scrape the FCC public file. We look at every single rate that all these campaigns pay, and put all that stuff together, and output a product trying to help campaigns. Outside of politics, we help commercial advertisers understand what’s happening in multiple markets, like local markets, or at the national level, if necessary, in as close to real-time as possible.
Things are constantly changing. You need to constantly innovate in the marketplace. To continue innovating, you have to keep asking your customers what are their needs, concerns, and problems.
”The political advertising ecosystem is easily a $9 billion marketplace. Now, as we look forward to 2024, are we going to eclipse 10 billion?
MB: I want to get more into the wider applications, what’s next, and a couple of questions on politics. Our companies collaborated on a video ad spend estimate for 2021-22, and it was a monster number, close to $9 billion as a projection. Where did that land?
KR: It landed at $8.9 billion ($8,987,000,000.)
MB: That’s incredible!
KR: We were 13 million off the prediction. It is incredible when you think about it because it was a midterm election. Three billion were spent for the Presidential election in 2020. And then you roll into 2022, and we’re saying, “Well, it’s going to be about the same, but there’s three billion that should come out because of the presidential.” Well, it didn’t really make a difference. That went back into gubernatorials. You saw increased house spending and increased spending on down-ballot. You also saw huge ballot initiatives with an enormous amount of spending in 2022.
The political advertising ecosystem is easily a $9 billion marketplace. Now, as we look forward to 2024, are we going to eclipse 10 billion? I’m not making the prediction now, but when you look at the trend, it seems likely that that will happen over the next 24 months. That’s a huge milestone.
MB: It’s interesting because I’m sure you’ve experienced the same thing before. For the last decade or more, politics led the way in a lot of innovations in advertising, targeting and speed, and test and learn. But now the market size is so big, and the pie is growing so fast, that it has attracted a lot of companies. Have you seen the same thing?
KR: Absolutely. For years, the commercial products that were available to help give insight to politics just were not good enough. A lot of agencies and political professionals, back in 2008-09, started to innovate and make investments in technology. Now here we are, 15 years later, and we’ve built really cool and interesting products. Because there was no one early on to fill the gap in terms of the technological needs in the space.
And over time the political space itself has grown. We were just talking about the $9 billion number for a midterm elections’ year. In 2018, that number was four billion.
So, there’s a need. Prior to that there was this culture of innovation which came out of it because there was no one doing anything. And today, you still see that in our space. Where there’s a lot of entrepreneurialism and innovation, because the needs are still not being met by the outside market.
MB: Looking back to 2018 and 2020, what do you think is driving that change?
KR: Well, certainly the change itself, for the marketplace. If you are talking about media intelligence, it’s because there’s so much more spending. Fundraising has increased a lot, and the internet has revolutionized it. You still have the high-dollar fundraiser, but you also have these millions and millions of dollars coming through low-dollar donations, which can power campaigns. That leads to a tremendous amount of advertising spent.
Because of all that advertising spend, AdImpact tracked around 6,000 campaigns in real-time in 2022. And around 80% of everything we see, we have it in someone’s inbox within two hours.
The need is because there is so much advertising activity. A single campaign, a single House race can have up to 10 or 15 different advertisers. It’s not just two candidates anymore or just two super PACs.
There are many outside groups that want to influence the outcome of an election. The balance of power, of course, is so close in politics right now. That also feeds this fundraising. It feeds this ability for each party to want an edge. Hence, so much advertising activity around that. Hence, so many dollars. Hence, the need for tools to help you understand what’s going on in the environment.
MB: One thing I want to highlight is that you talked about the 6,000 advertisers. That’s a unique challenge for the local market. At the national level, about 200 advertisers buy about 90% of national TV. So most of the tools developed are for 200 people that spend a minimum of tens of millions of dollars to get into that group. But when you start to talk about the local market, there are tens of thousands, or hundreds of thousands of advertisers. There have been 6,000 just in one year in political alone. It’s a unique challenge.
KR: Yes. 6,000 all local. And you’re right, it’s a concentration of national advertisers. A lot of tools are getting built for those organizations. With our detection system now, we define a local advertiser as anyone that advertises in five markets or fewer. We’re tracking 60,000 advertisers across the entire country in real-time, across all 210 DMAs. Within that, tens of thousands are just local. I’m talking plumbers, car dealerships, electricians. There’s an enormous amount of activity going on underneath the national advertising level right now. We see it.
The genesis of our tools is politics. And because everything is local, we had to build an ad library that can find ads in all 210 markets. Why? Because in Rapid City or Sioux Falls, South Dakota, there could be a huge campaign going on. No one cares about those markets, but because there could be a lot of money getting spent there, we have to have detection systems, and we have to have the ability to go and find ads in those markets. Since we do that, we can find all these other local advertisers easily. We’ve built all that out now, and it works perfectly.
MB: Looking back to the last two years, what surprised you the most on the 2022 cycle?
KR: I would say the one thing that surprised us the most was the decrease in digital spend. We define digital as in Facebook and Google. We expected it would be a $1.7 billion market, and it was a $900 million market. It was about $800 million less than what we thought it was going to be. That was a big surprise.Cable/Linear, was about where it has been. It’s approximately 75% of the overall spend if you put broadcast and cable together.
Broadcast dropped a bit in terms of their overall share, and then cable actually went up a couple of points, so they saw growth. But digital was the biggest decline. And then, of course, the largest new player was CTV. It’s not surprise. We had the ability to track it. And the first cycle we were able to track it, it clocked in at a billion dollars. That’s just huge. We know that, as that impression base continues to grow, more dollars are going to flow there.
MB: It’s interesting. When you start to put the targeting and measurement restrictions on the big players like Facebook and YouTube, looking at it in a vacuum, you don’t think it has an impact. But all of a sudden, the buyer is looking at all their options they have. That’s probably one of the biggest use cases.
And our product is to say, “Now that I no longer can target a group of individuals, I’ve got to target a huge bucket, which 70% of my audience isn’t the target.” Then you’re taking an ad that has basically the same targeting as broadcast, but far less attention and a lower impact creative.
We’ve argued for years that that’s going to have an impact on how people buy. You saw that in politics this year, where the overall digital video market really grew, powered by streaming, but Facebook and YouTube were flat or declined. That’s because buyers finally are like, “We have other ways to reach these people. You can’t just drive up our costs and not have any impact.”
KR: I think that’s right. I mentioned earlier that the presidential election was north of nine billion. And plus, in 2020, Bloomberg himself represented $1 billion in the overall spend. One person. That’s just crazy. If you take that out of 2020, 2022 was larger. The midterm was actually larger than the presidential if you take Bloomberg out of the presidential primary. Pretty incredible.
Talking about digital, a good portion of their dollars in 2020 came from the presidential. For linear and traditional media, we saw that it didn’t make a difference if it was presidential or non-presidential. They still had the same amount of dollars flow. Five billion for broadcast, 1.7 billion for cable.
Now, if you look at Facebook and Google, when that presidential money came out, it didn’t return to them. Not at all. That was another big factor. There are hundreds of millions of dollars in there that just didn’t show up. Then you talk about the ability to evaluate the efficiency of the different platforms, the targeting versus non-targeting. That had an impact as well.
”As streaming takes over and people watch whatever they want to watch, content becomes less important and ad delivery becomes much more important.
MB: Yeah. If you can put a 30-second ad up and the person watches it, versus three to five seconds of attention, that’s a big deal. And people’s attention spans are going down the drain a bit. But if you’ve got a choice between how to reach people– We just saw that universally, where the buyer or the planner started to allocate differently than they did two years ago.
Pivoting into the wider market, after every election, we spend a lot of time meeting with general market advertisers. And they’re really interested in learnings from politics and what they can apply, especially in the period we’re in now. What trends do you see in politics that can carryover to the wider market?
KR: One thing in politics that we are solving for now is refining the detection. Most of the detection systems are only tracking linear broadcast, so we’ve enhanced it to detect local cable and CTV.
Because of that, we can now see what everyone else is doing across local cable networks and connected devices. But we can also see how many impressions these ads deliver.
There was so much focus on content and where you were running within content, but as streaming takes over and people watch whatever they want to watch, content becomes less important and ad delivery becomes much more important. We are moving towards that system.
That’s very helpful in the commercial space because, “I’m running all these ads, but how many impressions am I actually delivering? Am I losing any market share to my competitors?”
We can show you all that in real-time. We can show you exactly where your competitors are, and you can compare, almost like for like, who has the more efficient media play. You get much more depth now. You get a lot more data, and a lot more information. You see a larger part of the media space that we haven’t seen before.
MB: It’s incredible. One area that is really helpful to our team is the speed of your ad tracking and notifications. It powers a lot of our products. Data ingredients like that power the test and learn culture in politics. Do you see the general marketers, or advertisers, you talk to utilizing that more?
KR: Yes. Bringing this to the general market, it’s a different way of thinking. You’re not thinking about, “I want to be across these platforms,” you’re thinking more about, “What’s going to deliver the most bang for me?” Less than 24 hours later, we can tell you how your ads delivered, how they performed across all devices. You’re getting much quicker feedback.
When you go to the general market, this is a new concept that we’re bringing to advertisers. This initial phase of it is crafting just broad enough metrics so that people can begin to understand exactly what they’re getting. We’re delivering just two plus impressions across a specific ad. We’ll tell you, “Here’s a specific creative. Here’s what ran. Here’s where your opponents ran,” or, “Here’s where your competition is running.” We have brought that to the marketplace. And it’s new, it’s different. There’s no one doing that right now.
When you have that approach, there’s a lot of conversation that needs to happen between a customer and us, so that we make sure that we are crafting this the right way. So that we are delivering something that they need. But it’s certainly a problem that exists in the space, because people don’t have a good sense of how well they’re performing.
Especially when you’re tracking linear, and the reports update days or even weeks later.
We’re having great success outside of politics, with audience companies and data companies. Because now we can deliver so much more insight. The detection system now works at the ZIP code level.
We’re also breaking this idea of a DMA. If the population of your DMA market is only within three or four counties, maybe you only need to measure ad delivery there, because you have these other lower populated rural counties that don’t have any of your customers. Or if you’re a local plumber, and you only perform service within 20, 35 ZIP codes within a DMA, we can tell you ad delivery within those ZIP codes.
These are all new ideas. You’re crafting new UIs to help deliver this information. You’re crafting APIs for data companies who just want to ingest the raw data. But because the detection happens at a very granular level now, we can build out whatever people need.
MB: Back to the fast ad detection at the market level, one way our team uses your data is to show TV in a more digital-like report, displaying daily impressions, reach, frequency. Looking at test and learn culture, you can change things mid-flight, as opposed to waiting until you get your logs two or three weeks later. And put on even footing with one buying and planning team, and one analytics team. That really powers that information.
Looking at the local video ad market, if you could bring out a magic wand and make one innovation happen, what would that be?
KR: Right now, it’s all impressions, but that’s not that good. It would be nice to have a GRP across all devices, cable, CTV, and broadcast, and say, “Within this ZIP, here’s exactly how many GRPs.”
Right now, we can only do impressions. It’s good, but it’d be great if we could do GRPs. Even though we are migrating everything to an impression-based system, because it’s a neutral form of measurement across multiple media type, it would be nice to have. It seems like a lot of people are very comfortable with GRPs. Since we’re moving away from that, to say, “You have to convert it, and is it necessarily right?” It’s just much easier to say, “We’re now at 800 points, we’re now at 200 points.” Of course, the impressions vary by market because they’re all based upon population.
Having a neutral GRP system is helpful. It’s not going to go there. I sound like I’m stuck in the past in a way, since I don’t want to change from GRPs. But I still think it is a very good measurement system, and I wish we could somehow figure it out, where everybody’s comfortable with the number, but incorporate it across multiscreen. As you move to audience-based, it becomes much more complicated. With 200 GRPs across any market in the country, the impressions are going to vary, but you really know what you’re going to get in terms of reach and frequency based upon that number.
MB: I’m interested in that. I was on a measurement panel at CES last week, and Warner Brothers’ Andrea Zapata, from the Comcast days, just announced a deal with VideoAmp to offer measurements. It’s a great deal with two great companies. Talking to people there, they were like, “Well, the buyers aren’t integrated with this.”
With my previous life, putting a buyer hat on, when we were buying linear we felt that we didn’t care what the negotiation was on. We were going to negotiate. I think this was a unique/incorrect view of the market. If the broadcast group wanted to negotiate on 35 plus, or two plus, or whatever it was, we had our own metrics for what it was worth, and we were going to value that inventory in that own way. I’ve always felt like that’s a digital way of looking at it.
Back to your GRP question, do you feel like it does it matter? You’re the most sophisticated, buying shops out there. Do you think that if you were able to look at the GRPs in a certain way, would you care what the seller was selling it to you as?
KR: Yes, I always care what the seller is selling to, because as a buyer, you want to negotiate the best deal.
MB: You want to get the best deal. But what if they come back and say, “Our ad was worth $10,” and your data says it was worth 11?
We always wanted to find the asymmetric information, where we could get something that was more valuable. We would look at two spots that had an eight rating on the same demo, but to our data, one had an 11, and one had a seven against our audience. We wanted to cherry-pick the 11 because we knew the rating against the target, and the broadcaster didn’t know that, or the cable provider. Obviously, they didn’t like this.
But the conversation, even last week, surrounding arguments for why Nielsen is still in this incredible and strong position, were, “If the buyers and sellers aren’t using the same currency…” But, “Why does it matter?” If someone does a deal with VideoAmp, or Samba TV, or iSpot or somebody, even if they have asymmetric information, why does it matter? It seems the wider market still thinks we both need to use the exact same data set.
KR: That is right. It matters because even moving to impressions now, going back to a customer and saying, “You delivered 50,000 or 100,000 impressions on this particular ad across these different markets” they don’t necessarily understand what that means. And if you did a good job for the customer, and they don’t understand what it means, you, as the agency, are not really succeeding.
People still understand GRPs. So if you say, “We’re supposed to purchase 200, but we actually got 210, because our research showed that we could deliver 5% more value. Got it at this cost,” etc., that computes. That’s why you do need the standard currency, and you do need both sides understanding what the currency means as we move into these different platforms, with each evaluating things differently.
Nielsen came out and said that the Thursday Night Football games on Amazon average an audience of eight million. Amazon said it was 11 million. That’s a pretty big gap. So, who’s right?
I’m sure they’re both right, but they have different methodologies. Nielsen has just been that trusted methodology for so long, and in the marketplace, that GRP measurement is so burned in that it’s difficult to just shake it.
Agencies have to be very conscious of translating these things, providing meaning. But this issue of multiple currencies now is going to be a big issue because, which one is right? Which methodology is accepted? It really is going to be a big deal. That’s why I think Nielsen still has the advantage. I haven’t seen the Nielsen One system yet, but I think this is why they still do it, and why it’s so difficult to dislodge them in the marketplace, even though a lot of people complain about Nielsen.
”The need for local is enormous. It's an enormous market. It's an $80-90 billion marketplace, and no one is paying a lot of attention to it.
MB: Local could create a huge challenge for us because you’re going to take the big data sample, or a panel, and you’re going to break it into 210 pieces. Maybe in New York and LA, you’re going to get enough of a sample, but outside the top markets, that’s one of our biggest challenges. How do you bring in more and more viewing sources to allow people to do audience-based planning in Dayton, Ohio or Baton Rouge, Louisiana, where our advertisers are. And that’s the disconnect also at the national level. We’re going to have a whole other host of issues locally.
KR: Definitely. And the need for local is enormous. It’s an enormous market. It’s an $80-90 billion marketplace, and no one is paying a lot of attention to it.
The measurement tools right now are in need of help. There’s no question about it. Even some platforms themselves. You and I reach in frequency at local levels, and we had to build our own tools to do that. Because there’s nothing out there that’s actually any good that’s believable. We had to build our own.
Going back to the point I was making earlier. In politics, these things didn’t exist, and we had to innovate. Now we’ve built the right tools that are not only used specifically in the political space anymore. They’re very powerful outside of that.
MB: Excellent. Well, Kyle, I’m grateful for your time and I know our audience is going to love this conversation.
KR: Great to see you, and thank you for having me on. I really appreciate it.
See the rest of the Screen Wars Thought Leader Interviews here!
Cross Screen Media is a marketing analytics and software company empowering marketers to plan, activate, and measure Connected TV and audience-driven Linear TV advertising at the local level. Our closed-loop solutions help brands, agencies, and networks succeed in the Convergent TV space. For more information, visit CrossScreenMedia.com.
Kyle Roberts is a New Jersey native and has been in the media buying industry for over 20 years. He founded and led one of the top media buying shops in DC for 9 years before starting AdImpact. Originally Advertising Analytics LLC, Kyle started the company in order fill holes he saw in advertising intelligence offerings. Kyle led the company from a small four-person outfit to a team of 30+ people. AdImpact is the leading provider of political advertising intelligence thanks to his vision and dedication. But Kyle did not stop there. He knew that AdImpact’s success could translate into commercial advertising. In 2020, under his leadership, AdImpact expanded from political ad tracking to commercial advertising analytics.
Kyle graduated with his Bachelor of Arts in History from Manhattan College and has been the President of Smart Media Group for 16 years and has been at the head of AdImpact from its inception in 2014.
