KERV.ai continues to mark progress for its AI-powered interactive and contextual connected TV advertising tech.
In June KERV tapped Joshua Koran as its new chief product and technology officer as the vendor works to grow its suite of AI-powered capabilities, including through agents.
And just this week, the vendor marked its latest customer win as it secured an expanded partnership with LG Ad Solutions that will see KERV’s full suite of interactive and commerce-enabled CTV advertising technology extend across LG’s streaming inventory in key markets globally.
Prior to the LG news we sat down virtually with Koran to give StreamTV Insiders his perspective on interactive CTV ads and how KERV’s detection tech helps align commerce with content across major media companies, OEMs and streaming services.
Read on for more as we dig into KERV’s capabilities, including the impact of pod-level targeting, unlocking the value of live news and finding new ad opportunities in existing content libraries.
KERV, LG pact expands globally
KERV’s expanded pact with LG follows earlier collaboration between the two in the U.S. and extends the vendor’s interactive CTV advertising tech to LG streaming inventory in Canada, Europe, Australia and New Zealand.
As we’ve covered in a deep diver before, new formats and interactive ads have made their way to CTV and are moving beyond the point of pilots and experimentation in the U.S.
Among many vendors in the interactive or shoppable ad tech space, KERV is a leading partner to content owners, streaming services and platforms in the U.S. and its tech is interoperable across top supply and demand sources.
KERV uses patented, AI-powered tech to analyze video content frame-by-frame, detecting and identifying objects (say a couch, for example) and products (say a specific couch made by IKEA) within scenes.
It not only recognizes products but can link them to specific product feeds and URLs (or find the closest match when an exact object correlation isn’t present), helping power dynamic shoppable experiences on a range of streaming services and platforms.
Interactive ad experiences from the vendor can include in-content shoppable formats or overlays like L-Bars and Pause ads with dynamic QR codes, show or brand-specific on-screen product carousels (as seen with Walmart and NBCUniversal on Bravo shows), and interactive ads that enable a second-screen mobile shoppable experience (as seen with Disney).
KERV’s tech also pulls out additional video metadata and contextual signals and analysis that goes beyond just genre, with separate and consistent taxonomies for scene-level elements like music, mood, location and more.
This granular contextual data, alongside an integrated dynamic ad creative and measurement platform, is meant to help advertisers and content owners uncover new and more relevant moments in content to align interactive ad experiences.
That alignment stands to make shoppable creatives or campaigns more relevant and resonate better with consumers. And, if successful, drive results or outcomes for brands at different stages of the marketing funnel while providing additional ad revenue opportunities for streamers.
“The future of premium video lies in an ecosystem where content, creativity, and commerce work seamlessly together,” said Brad Quinn, SVP of Strategy & Growth at KERV.ai. “Our expanded partnership with LG Ad Solutions is an important step toward that vision, extending interactive and commerce-enabled advertising across premium streaming markets globally. “
The broader LG pact also helps advance KERV’s wider interoperable ecosystem of leading publishers and platforms in additional markets.
Warner Bros. Discovery, among KERV’s bevy of partners, in June announced its own expansion of it KERV-powered “Moments” contextual advertising capability for HBO Max into select international territories by the end of 2026.
And cross-platform interoperability is one factor that already helps KERV as a vendor in the U.S., as a lack of it can be a hurdle to greater adoption of interactive ad experiences by advertisers since they often want to buy across services and require scale for TV investments.
KERV gets a new CPTO
Named CPTO of KERV this summer, Koran is tasked with leading the development and execution of the company’s next generation of panted AI-powered products.
That includes plans to accelerate development of its contextual enrichment and creative offerings. Under his leadership, the KERV product roadmap also involves investing in advanced AI agents for media planning, analysis and optimization – which are meant to offer greater signal transparency “to improve strategy recommendations and natural-language reporting across premium video environments.”
Koran started in digital marketing and advertising in the 1990s and brings more than 30 years of global ad tech leadership, as well as 26 patents related to digital marketing, audience intent and content optimization technologies.
Koran told us that part of what attracted him to KERV is the fact that, at a high level, it’s a company helping to bring capabilities that were previously available in the programmatic marketplace on desktop, laptops and mobile (in terms of transforming data into something meaningful to media owners and ad buyers) to the living room big screen on CTV.
A shift that’s been well on its way but has more recently advanced.
Interactive CTV shifts from rudimentary to advanced
In the digital space interactive capabilities for OLV are more mature than the CTV environment, Koran explained, mainly because the interoperability that has existed in mobile and desktop web is just now making its way to CTV.
Still, interactivity on the TV screen isn’t entirely new.
Koran’s prior experience includes the advertising division of AT&T’s U-Verse called AdWorks, which he helped grow. He described how roughly 15 years ago they had channels where users could click an ad and go to a custom channel – an early form of interactive TV.
However, looking at it from the user experience the industry has well advanced and is moving away “from just a clickable TV experiment.”
“We’re now moving beyond that rudimentary interactive type of creative, where with the second-screen of a mobile device you can scan and actually [have] a shopping experience, or even on the big screen itself interact with what you’re viewing,” he said.
Others in the space like BrightLine are serving up remote-control enabled interactive ads that feature things like quizzes, trivia or custom advergames.
Interactive and shoppable ad experiences are also prominent on Amazon’s Prime Video, for example.
The e-commerce giant has a variety of interactive video ad formats, such as click-to-send more information to a mobile device and the ability to add specific products directly to a linked Prime Shopping account cart by clicking an ad on the TV with the remote, as well as its Shop the Show effort.
Smart TV maker Samsung this year represented the first external platform partner to integrate the Amazon DSP’s remote control-enabled interactive shoppable video ad tech.
“In my mind, this is really an improvement of another version of what is already on screen, already the media that people are interacting with, uncovering these additional signals that they get to explore, interact with, and sometimes shop from,” Koran commented.
Amazon might be a top example of what closed-loop commerce from the TV screen can look like, as it has its own video service, ad tech and servers, and massive digital online retail business.
But in terms of interactivity and shoppable experiences, KERV’s product and tech chief suggested he’s excited to bring similar capabilities on the interactive and shoppable front at scale to all other players in the ecosystem.
KERV tech, signals allow for pod-level targeting
Helping KERV to get the right shoppable or interactive CTV ad experience in front of viewers at the right time is pixel-level scene detection. This helps produce signals that allow interactive ads to be targeted at the pod-level or based on and aligned with the scene that viewers saw just before the commercial break.
Alongside frame-by-frame analysis and object detection, KERV’s consistent and separate taxonomies that group content signals play an important role in its product offering.
Separate object and product catalog taxonomies that can be matched help underpin the vendor’s ability to pair contextual content and commerce signals.
It also has additional taxonomies, such as location, and signals related to emotion, such as music, tone, whether people are laughing or arguing on screen and so on.
Not only does KERV identify and uncover the products and contextual signals in content but uses them to help build the dynamic, interactive creatives.
Together these allow for both interactive shoppable TV experiences directly tied to exact or similar products seen in programming.
It also means ad creatives that align not only with broader content themes, genres or moods as some contextual targeting does, but specific scenes, frames or objects that viewers just saw.
Content alignment boosts consumer interaction, willingness to shop
Per Koran, interactive ads that more precisely align with content are more relevant to the audience at that moment in time and have been shown to help drive interactivity by the viewer.
“What we found is when you show the creatives in the next ad slot relative to what was most recently seen on screen, it increases the interaction rates by double digits.”
For example, Wayfair employed a KERV-powered mobile QR-code enabled interactive campaign on WBD’s HBO Max. Using KERV’s pixel-level contextual signals, Wayfair saw a 12x lift in their scan rate, as well as in the remote-control enabled second-screen click to shop format with a 15x boost in people clicking an ad to shop.
“That click-through rate increased 15x when the ads were relevant to what [viewers] had just seen on TV,” he said.
Consumer survey results also support the notion that viewers are more willing to shop via TV ads when the creatives or products are contextually relevant.
A study from KERV found that 48% of consumers say they’re more likely to scan a QR code when a product is relevant to the content and 43% would shop a product directly tied to the scene. Additionally, 35% of consumers say they have interacted with a shoppable ad when it feels relevant and seamless on TV.
Second-screens are also important as 68% say they’d prefer to complete a purchase on their mobile device.
Adding to KERV’s tech is an integrated closed-loop performance measurement functionality to see how well signals are driving performance – be it interaction, QR code scan or purchase rates.
That in turn allows the vendor to further enhance the data signals it produces and improve the interactive creatives and experiences it designs. And since it has purchase and other information from shoppable creatives, KERV can see the feedback loops and optimize campaigns in-flight to the outcomes brands aim to achieve.
Precise contextual data on live streaming increases the ad value of news
KERV’s capabilities now extend to live streaming as well.
It’s one avenue where the tech can help content owners and media buyers find ad opportunities and value in places or content that previously haven’t been as attractive or well monetized.
For example, Koran described how the contextual signals in live streaming can be particularly useful for media owners offering news programming.
The category historically has been a somewhat harder sell or not as high of value as advertisers need to contend with both happy and less feel-good news stories.
“If, however, you can target just the happy stories, you’re probably more likely to pay a higher amount, especially when they’re contextually related to the brand you’re advertising.”
Such as an automaker that aligns with a news segment featuring a new car development or feature.
“That’s probably going to be highly valuable relative to the car crash on the highway, where [carmakers] probably don’t want to be advertising next to that.”
With KERV’s ability to distinguish between the two and target accordingly on live streaming, he believes it unlocks a “win-win” for both the streamers selling advertising and the brand buying it because, as seen with the news example, it unlocks additional opportunities for each.
Existing content libraries, new ad opportunities
While advertisers and brands are likely keen to get their messages around fresh and tentpole content, there’s also the challenge for streamers of effectively monetizing large content library catalogs with TV shows and films that have been around for a while and could benefit from more relevant ad targeting.
It’s another area where KERV can help find and identify high-attention and relevant moments within content in ways that owners and ad buyers haven’t been able to buy or sell previously.
Koran used the classic fan favorite Friends series as an example, where traditionally the long-running yet older show would be grouped by a more generic genre like “sitcom”.
However, there are many scenes in the series that take place in the kitchen of the apartment. Koran pointed to a Cuisinart mixer, for example, as a product that might appear and a brand could easily already see Friends as a safe environment.
“But if you look at only the kitchen scenes, all of sudden this is more relevant to you [as a brand],” he noted.
By identifying all of the objects and products that are actually in a scene and KERV’s ability to do product-level ads that match those frames or items, it’s finding “new opportunities for the buyers and for the sellers to now advertise in existing libraries of content”
Part of the appeal is also identifying less obvious matches within large libraries.
A kitchen appliance brand, for example, might already be well aware and know it wants to advertise against or align with top shows or channels that, at a high-level, are relevant to their product, like the Great British Bake Off or the Food Network. But KERV could identify not only food-related programs, but existing scenes with kitchens or even specific product matches vastly beyond the first few to mind.
“By uncovering these additional episode scenes and moments where the product is already being featured in the native content, all of a sudden, it opens up a lot more buying opportunities for the brand to get in front of the audience when it’s going to resonate with them most,” Koran explained.
And there can be a lot more moments.
WBD in June said its contextual advertising capability powered by KERV has identified more than 25,000 “Moments” across content on HBO Max. And touted the value of contextually relevant ads, noting that in the U.S. the matching capability had already delivered a 19% lift in viewer engagement and a 13% lift in purchase intent.
Matching existing ad creative to frame-level moments
In addition to building interactive ad creatives that contextually align with content, KERV is also doing a little reverse engineering.
It has a new AI-powered recommendation interface capability that was demoed during Cannes and aims to reduce friction for advertisers.
Brands can simply upload a brief or an existing creative asset they’ve already built, enter in brand goals, and KERV’s recommendation interface will match the creative to relevant moments in content that the vendor has already scanned and identified.
For example, if an automaker wants to be associated with an “adrenaline” moment and has their own ad of a car driving through a mountain pass or city downtown. The carmaker could upload that creative, enter their aim and get the ad matched to car-focused adrenaline moments within content – be it a car chase or other scene.
“It really does help bring together and connect the buyer and seller in a new way with this new dimension of information that previously they have not had,” Koran said.
Read the full article on StreamTV Insider.