AI and Contextual Advertising: A Q&A With Comcast Advertising’s Kristin Shumaker and FreeWheel’s Larry Allen
Comcast Advertising has released new research on the effects of contextual alignment in advertising.
FreeWheel's Viewer Experience Lab released the new report, Maximizing Brand Recall Through Contextual Alignment. It examined the value and effects of contextual advertising by analyzing the video, audio, text, and metadata of more than 5,000 advertisements and more than 200 programs using Comcast's VideoAI tool.
In this Q&A, Kristin Shumaker, Comcast Advertising Analyst, and Larry Allen, VP of Global Strategy for Addressable, Data, and Measurement Partnerships for FreeWheel, expand on the details of the report’s findings and how it shows the ways that advertisers and publishers can best harness the power of contextual alignment.
Comcast Advertising Analyst Kristin Shumaker, Ph.D. (left) and Freewheel VP, Global Strategy Larry Allen
What are some of the primary benefits of contextual alignment in advertising?
Larry Allen: Almost all of us as viewers have experienced the benefits of contextual alignment – like watching your favorite cooking show and seeing an ad for a new kitchen appliance. That connection and relevance of the program and ad content is contextual alignment, and it offers benefits to both the buy and sell-side of the industry.
For buyers or advertisers, it means that they can align ad creative to more relevant content environments, which increases consumer enjoyment, brand recall, and purchase intent. In fact, our latest research from the FreeWheel Viewer Experience Lab found that brand recall increases by +20% when one of the top three keywords are contextually aligned.
Additionally, with contextual alignment buyers can build campaigns based on specific keywords or emotions that are important to their brand or more closely aligned to values or performance drivers. On the publisher or seller side, it allows them to recommend inventory based on the context, which is incredibly valuable when audience signals aren’t available or don’t match to a narrow audience of a targeted campaign.
Contextual alignment not only benefits advertisers and publishers, but it also helps improve the overall viewer experience. For example, when contextual advertising is used it allows for more personalization of ads for the viewer, as well as ensures the ad content is aligned with the program content, creating a more complementary viewing experience.
Previous research from the Viewer Experience Lab found that relevant ads capture the attention of the viewers and help increase overall engagement – with audiences reporting 2X more liking and 2X more engagement. This relevancy goes a long way with audiences: research from Comcast Advertising found that 3 in 4 millennial consumers pay attention to ads for products and services that they are in the market for.
Ultimately, brands benefit from this increased attention, which has been found to lead to greater recall and a lift in purchase intent.
Could you describe the importance of keywords in contextual advertising and how they are identified and ranked?
Larry Allen: Keywords are extracted from video scenes, objects or sentiment of the content that represent details beyond the genre or general theme of the programming, enabling advertisers to more closely embed their advertising messages with content that matches either thematically or to their category of interest.
AI tools like Comcast Technology Solutions’ VideoAI use computer vision to watch the content, evaluate it frame by frame, determine the various attributes or meaning of a scene (or asset) and then determine relevancy and intensity of the keyword in the content providing a ranking. We then use this ranking (or threshold) to determine keywords that best represent the scene or asset.
Kristin Shumaker: The VideoAI tool from Comcast Technology Solutions uses a combination of methods, including computer vision for scene and object recognition, as well as audio and text (closed captioning) analysis, to tag frames for the presence of a wide variety of objects, scenes, actions, and concepts. These tags are then aggregated over time, whether the time is the whole length of the program or just a segment between ad breaks.
VideoAI then maps individual tags onto keywords, effectively aggregating labels for objects, actions, and more, into the kind of unified descriptor words people would use to explain the content of a show or ad. In this project, we chose the Interactive Advertising Bureau (IAB) Content taxonomy as our source for keywords because it broadly represents video content.
VideoAI ranks keywords by the strength of their association to the tags, so we can differentiate the most relevant descriptors from less relevant ones. Having a validated taxonomy of keywords as a shared language to describe video content is vital to making apples-to-apples comparisons between programs and ads.
Please discuss some of the ways that Comcast Technology Solutions' VideoAI measured brand recall due to contextual alignment. How did AI also help refine this process?
Kristin Shumaker: For this research, particularly, our brand recall measure comes from Mediaprobe’s database of audience responses to in-home content viewing. In post-session surveys, Mediaprobe asks viewers to name brands shown in the ads they have just seen in the session (e.g., unaided free recall). The measure of brand recall is independent of the keywords VideoAI generates to describe program and ad content. The independence of the brand recall measure from the keyword-based contextual alignment measure is part of what gives us confidence in the validity of the relationship between the two measures.
How has the IAB Content Taxonomy 3.0 helped to refine these measurements better?
Larry Allen: The IAB taxonomy provides a standard by which we can evaluate and categorize the content attributes. It provides a set of well understood keywords for advertisers to choose from when targeting or evaluating content. It also allows us to scale the contextual capability broadly across many publishers and content genres, mitigating the use of custom models and approaches for advertisers one by one, limiting scale and operational efficiency.
Ultimately, the widespread adoption of the IAB taxonomy across publishers and advertisers is what will allow the activation of contextual alignment at scale.
What are some of the best approaches that advertisers and publishers can take to best harness the power of contextual alignment?
Larry Allen: As the research shows, contextual alignment leads to greater brand recall, meaning more impact for advertisers when they align ad and program content. Advertisers can harness the power of contextual alignment in three key areas: performance, creative optimization, and targeting.
Advertisers should consider choosing contextual keywords that best drive a desired outcome to enhance performance. Additionally, matching ad creatives with content can improve brand relevancy and engagement for likability for consumers. Advertisers should look to contextual alignment as an effective targeting alternative when identity of audience signals are not available or allowed due to privacy or other technical or regulatory restrictions.
Publishers should look to a standard taxonomy when tagging program content and tag keywords that best describe the content while avoiding sending too many keywords. This will help showcase the value of content for advertisers and maximize its impact.
What are some of the unique features of FreeWheel’s Contextual Marketplace that assist buyers in increasing the value of their advertising?
Larry Allen: FreeWheel’s first-of-its-kind Contextual Marketplace was built to provide optionality for both buyers and sellers to leverage a contextual vendor of choice. We currently partner with leading contextual providers such as KERV.ai and Proximic by Comscore, which further expands targeting solutions across premium video at scale for buyers and unlocks more value for sellers.
By embedding contextual capabilities within the ad server, we help sellers create new monetization opportunities at scale across their premium, multiscreen inventory through a privacy-safe mechanism, with performance advertisers in mind.
This approach gives buyers a sophisticated way to reach the right audiences with more personal and relevant ad experiences without identity signals in a more granular way than ever before, helping to drive outcomes such as increased engagement, purchase intent and brand sentiment. An example of this is a buyer, such as an airline, delivering a commercial within holiday travel content while avoiding any news programming about travel delays.
In what ways does FreeWheel help to enhance the worth of buyers’ advertising even in the absence of identity signals?
Larry Allen: Contextual advertising is a key way that FreeWheel is enabling buyers to direct advertising toward audiences in more relevant content environments. There are many programming environments (like FAST) where there are limited identity signals and context is a great way to provide more specific segmentation of an audience beyond program genre or channel to increase performance and brand suitability.
Related Articles
The buzz around contextual targeting is ramping up in the online advertising world, with the promise that it will solve issues surrounding privacy, transparency, and more. But where is its impact registering in the ad-supported streaming ecosystem? TVREV's Alan Wolk, Fremantle's Laura Florence, Philo's Aulden Kaye Yi, Roku's Charlie Goodman, and Vevo's Melissa Sofo offer their perspectives in a clip from the latest Streaming Media Connect 2025.
05 May 2025
"Contextual advertising is a really hot topic in TV advertising right now," declared TVREV's Alan Wolk at Streaming Media Connect 2024. But what exactly is it, and how is its growing presence in CTV and streaming changing the TV experience and particularly the way advertisers buy media? Wolk, Estrella MediaCo's Christina Chung, and Mad Leo Consulting's C.J. Leonard explore these issues in this clip from their panel at Streaming Media Connect.
04 Oct 2024
A recent study by BrightLine, the leading platform for interactive streaming ads, revealed fresh data about the incremental impact of interactive advertising for CTV, including major new findings showing how effective interactive ads are with encouraging unaided recall.
09 Sep 2024