ALAN WOLK (AW): What exactly was the problem you were trying to solve here? SERGE MATTA (SM): From the LG side, the ask from clients was very consistent. We had great national coverage and then the top 32 DMAs. That doesn’t get the job done if you care about what’s happening in a specific congressional district in Iowa. For verticals like tier two auto or political, that gap meant we were basically shut out of certain budgets. Even if we had plenty of LG TVs in those markets, if we couldn’t report and measure what was happening locally, we weren’t going to make the plan. So the idea was: take our ACR scale, then plug in partners that can make that data truly usable at the local level across every DMA. That’s where CCR and AdImpact come in — they complete the picture around what’s airing and what’s advertised in every market, not just the big ones. AW: Can you explain how that all works? DON NORTON (DN): AdImpact’s role is to capture the ad side of the world. We monitor ad streams throughout the country — national network and every DMA — using a mix of hardware, software partnerships and proprietary tech. Every time we identify an ad, we attach metadata: category, subcategory, where it originated, the geography, and then a deeper layer for key verticals. In politics we tag the campaign, candidate, district, message and tone. In auto we go down to vehicle type, model, tiering, even the URL on the end frame. We then feed all those ads and their metadata to LG multiple times a day. LG can instantly match that catalog against their ACR data to see which households saw which ads, and build local media products off that. AW: What about CCR? MARK BONHAM (MB): CCR historically provided what I jokingly call “dumb feeds” — just getting the local channel signals where they need to go. What this partnership is about is evolving those into “smart feeds.” To do that you need someone like AdImpact with a detailed ad catalog and you need an ACR team like LG has that can actually stitch it all together. Our job is to make sure coverage across local channels is complete and economically viable so LG can scale this, first across the US and eventually both internationally and into streaming. Taken together, it’s content plus ads plus distribution, all lined up at the local level. AW: How do you actually maintain accuracy in these smaller markets? MB: That’s exactly why you can’t just throw a raw feed over the fence anymore and call it a day. In local markets, there are still a lot of “unmeasurables” if you only have a basic feed. You get into nuances like the same auto spot running in Pocatello and Portland with identical creative except for a tag and phone number at the end. If you’re not bringing in more granular data and working closely with the ACR team, you’ll miss that. So we license more detailed local inputs, we work with partners like AdImpact who distinguish those variants, and then Serge’s team does the hard work of reconciling it all at the ACR level. That’s why I describe it as a real partnership. It’s not just CCR shipping a dumb feed and walking away. It’s the three of us aligning so you can truly say you’ve got coverage in every market, not just the big ones. AW: How granular are you able to get in terms of what’s being advertised or the tone of the ad? DN: For every ad we pick up, we collect the audio transcript and analyze the video all the way down to the last frame. On the political side, that means we can classify party, race, district, issues and the tone of the message — positive, negative, contrast. We use keyword analysis and we’ve brought in large language model technology to evaluate the transcript at a deeper level, so we’re not just hunting for one or two trigger words. On the auto side, we distinguish creative even when the difference is just the last second or two — a different call to action, a different URL, a different dealer tag. That lets us identify the dealership, the tiering (tier one, two or three), the type of vehicle being pushed and so on. Once that’s all in the catalog, LG can decide: “I want to target households that have seen a lot of SUV spots,” or “I want to build an audience off exposures to specific tier two dealers,” or “I want to find people who’ve seen negative ads about a particular issue.” The metadata makes that possible, and the ACR makes it actionable. AW: For someone who is already buying a lot of linear, but wants to extend their reach on streaming, how does this help? SM: We know which IPs and households are being targeted in CTV through LG, and we can see which households are only exposed through linear, only through streaming or to a combination of both. We can also classify them as light, medium or heavy viewers if that’s what the advertiser wants. Before this partnership, we could do that kind of analysis, but only in the top 32 markets where we had the full classification layer. We knew we had viewing in the rest of the country, but without a way to label what those households were watching and which ads they were seeing, we couldn’t turn it into usable segments or clean incremental reach stories. Now, because CCR gives us the local channel footprint and AdImpact tells us which ads are running where, we can create those segments across all 210 DMAs and report on them. So yes, you can get to an incremental reach number, you can see how much of the opportunity you delivered versus other advertisers in a market, and you can do it with a level of geographic precision we just didn’t have before. |
