Could the “C” in CTV also stand for contextual?
Companies with contextual ad products would have you believe so. Seedtag, IRIS.TV and GumGum are a few examples of players that insist they’re more than just a crutch for dealing with signal loss. They claim they can support effective targeting and measurement, including on connected TV.
On Wednesday, Seedtag announced a product called Contextual TV designed to help its clients apply contextual targeting tactics to their streaming ad buys. The product creates audience segments derived from online consumption patterns and viewing habits, which brands can buy through demand-side platforms, including The Trade Desk and Google’s DV360.
Seedtag started as a Spanish supply-side platform in 2014 and expanded into a contextual ad network for open web display in response to Europe’s GDPR. It launched in the US in 2022 after securing a private equity investment for global expansion.
As signal loss gets worse, clients in the US, now Seedtag’s largest market, have been asking for contextual products designed for streaming, said Mike Villalobos, VP of strategy and planning.
Planting the CTV seed
Planning CTV campaigns based on consumer viewing patterns allows advertisers to target viewers without focusing on an overly specific audience where there isn’t enough scale to justify the investment, said CTO Kartal Goksel. (There are only so many left-handed cat owners in market for a new sedan, for example.)
The challenge is that many marketers still perceive contextual advertising to be much less targeted than it really is because it’s not based on user IDs, said Victoria Gellatly, VP and media director at Mediahub, which is planning to test Seedtag’s new solution.
But contextual is still data-driven, she said, and can help marketers apply “a layer of personalization” by targeting audience cohorts based on interests and behavioral patterns.
Farming data
Seedtag’s contextual TV product is built on the company’s AI product, which is named Liz. (Yes, really, and there’s apparently no reason for it.)
Through its integration with publishers, the tool analyzes sites based on keywords and images to deduce the category and tone of online content in more than 15 different languages. Liz then packages related content that it (she?) deems to be brand safe. Brands can buy this inventory to reach certain audience segments, such as users who read business news or real estate content, for example.
Seedtag is extending this capability onto CTV via new data partnerships that help find connections between what people look at online and their streaming viewing. It does this by licensing automatic content recognition data from smart TV manufacturers, including aggregate data on viewing trends based on ZIP code, Goksel said. The company then compares online and streaming behavioral patterns at the ZIP code level.
Some would argue targeting based on location is not actually contextual because it involves data about a person. But Seedtag argues its product is, in fact, contextual because it uses location data in aggregate to detect patterns, not for one-to-one targeting.
In that sense, location can serve as the context for what someone is likely consuming, Villalobos said. Comedy lovers in New York, for example, may have different online behaviors than comedy lovers in Los Angeles, which can be a targeting signal.
I need context first
But measurement is also a priority for Seedtag, Villalobos said – in part because it helps improve targeting.
For example, Seedtag measures what types of content are trending online while CTV campaigns are in flight to determine whether the ads are resonating with consumers.
It also measures CTV ad impressions by unique household via partnerships with Nielsen and Samba TV. It works with TVision and Adelaide to measure attention metrics on CTV and with other third-party partners to measure the impact of CTV ads on brand and sales lift.
Based on how their original ads performed, advertisers using the contextual TV product can include or exclude certain groups of households through aggregated IP address data. By the second half of this year, the product will also allow advertisers to match their own data with that of Seedtag.
Adding IP addresses and first-party data into the targeting mix wouldn’t count as contextual, of course, but the option exists for advertisers that want to augment their contextual strategy with retargeting, Villalobos said.
Still, Villalobos said, it’s important to start with contextual targeting because it widens reach and helps buyers find new audiences they may not have otherwise discovered.
CTV campaigns pack a stronger punch, he said, when there’s “connective tissue” between streaming ads and what viewers are doing online.
