The landscape of contextual targeting is evolving, thanks to advancements in generative AI technology. Unlike traditional methods where media planners relied on basic audience segments, the new tools now allow marketers to input creative briefs directly into prompt fields, generating extensive lists of relevant URLs for ad placements. Recently launched platforms for editorial ads like Medialister aim to deliver more value from paid content while enhancing targeting precision.
Companies like Cognitiv and RTB House are leading this charge by utilizing generative pre-trained transformers (GPTs) to refine contextual segments. This innovative approach aims to resolve some longstanding challenges associated with contextual targeting.
Traditionally, contextual targeting can be either too broad, where options are generalized, or too narrow, focusing solely on specific keywords. For instance, Cognitiv’s co-founder and CEO, Jeremy Fain, highlights the limitations of keyword blockers, which can unintentionally eliminate relevant content. Cognitiv’s sentiment model addresses this by allowing brands to filter words based on context. For example, it can distinguish between negative and positive uses of the term “cheat,” thereby enhancing targeting accuracy.
RTB House has also embraced generative AI with its PrimeAudience platform, launched last June. This platform analyzes web content to create contextual audience segments. Marketers seeking to target tourists planning a trip to Paris can leverage PrimeAudience’s algorithm, which identifies summer-related content and local attractions, ensuring targeted outreach.
How Generative AI Transforms Contextual Targeting
Both Cognitiv and PrimeAudience operate by starting with a marketer’s prompt. For instance, if targeting consumers interested in Yeti cups, marketers can paste relevant text into Cognitiv’s tool, which then generates and ranks URLs based on relevance. Users can refine results by approving or rejecting individual sites, allowing for tailored targeting.
Cognitiv’s tool also includes custom models that assess sentiment and inclusivity. Marketers can specify their desired tone, while the inclusivity feature scores content based on established definitions of inclusive language.
Similarly, PrimeAudience evaluates articles and assigns confidence classifications (“Yes” or “Probably”) to ensure marketers target the right audience. It provides insights into the content’s relevance, allowing for strategic ad placements.
Once the desired URLs are identified, marketers can target them through demand-side platforms (DSPs). Both companies facilitate private marketplace deals (PMPs) with various supply-side partners, including major players like Magnite and PubMatic.
Privacy Considerations in Contextual Targeting
To maintain user and publisher privacy, PrimeAudience implements technologies such as SharedID, which assigns pseudonymous IDs to users while respecting publisher data preferences. Cognitiv also prioritizes privacy by using cookieless and ID-less technologies, ensuring no sensitive information is compromised.
Building a Robust Content Library
For these generative AI tools to function effectively, companies first develop extensive content libraries. Cognitiv scrapes millions of web pages daily to create a continuously updated database, while PrimeAudience relies on bid requests as a form of user consent.
The integration of AI models with these content libraries enables the precise matching of marketer prompts with relevant segments. Both companies utilize a combination of public and proprietary models, allowing for advanced analysis of content across multiple dimensions.
Future Implications
As both Cognitiv and PrimeAudience gear up for the end of third-party cookies, interest in their generative AI solutions is growing among brands, especially in sectors like consumer packaged goods (CPG). Opportunities also arise around significant events, such as the Olympics, where timely content related to athletes can drive engagement.
By leveraging GPT-powered tools, marketers can navigate the complexities of contextual targeting more effectively, ensuring their ads resonate with the right audiences and enhance overall campaign transparency.
