With the deprecation of ad identifiers, advertisers are turning toward alternative solutions, including well-established ones such as contextual targeting. This renewed interest means that these well-established products are having a bit of a renaissance to incorporate the newest advances, including predictive modeling and generative AI.
Contextual targeting has always allowed advertisers to place ads aligned with relevant content that consumers are viewing in real time. Amazon Ads has redefined what contextual means and moved beyond a simple “if keyword is present on page, then serve ad” heuristic by leveraging Amazon’s unique shopping insights and AI. Amazon DSP analyzes the context in a nuanced and scalable manner, accounting for complex relationships between words, images, and video content with the intent of being aligned with a consumer’s shopping journey. The AWS-powered AI’s ability to parse through vast amounts of unstructured data, recognize semantic themes, and understand audiences’ current intent, elevates contextual targeting beyond traditional keyword matching. As a result, advertisers are now able to target categories and contexts they previously couldn’t, or could do so using behavioral signals only.
This innovation aligns with broader industry trends, as 60% of marketers now use AI in advertising, with nearly half focusing on contextual applications. With the AI advertising market growing 35% annually, the demand for identifier-independent strategies continues to rise. Investments in AI-driven contextual targeting are yielding measurable results, including a 25% increase in consumer engagement, highlighting the value of Amazon Ads advanced approach.
This paper details how AI is transforming contextual targeting and showcases Amazon DSP innovative solutions to deliver value while minimizing the necessity of third-party cookies.
In this technical paper, you’ll learn:
- How we leverage Large Language Models (LLMs) for sophisticated content understanding.
- Advanced techniques for analyzing open web and mobile supply content.
- Our unique product category-based taxonomy methodology to understand demand.
- Our framework for evaluating model relevance and performance.
- Real-world impact through customer case studies and results.
Download our technical paper to learn more about AI-powered contextual strategies.
_
Daniele is an Applied Science Manager at Amazon Ads, where he leads a team building machine learning models for ads sourcing within Amazon DSP. With a PhD and experience across Deloitte, Teradata, and Capgemini, he brings both academic rigor and practical expertise to advertising technology.
Anurag is a Machine Learning Scientist at Amazon Ads, where he develops advanced advertising models for Amazon DSP. With a PhD and experience ranging from Microsoft’s recommendation systems to physics research for the European Space Agency, he brings a deeply technical perspective to solving advertising challenges.
