How AI has become the foundation of digital advertising

The digital advertising landscape is undergoing a transformative shift, driven by the powerful capabilities of generative AI. In fact, said Ashima Singhal, principal product manager at Amazon Ads, “artificial intelligence is the bedrock of advertising.” 
  
It might sound like a bold statement—after all, generative AI only hit the mainstream two years ago. However, some companies have innovated with machine learning models and AI for years to help advertisers reach their target audiences and campaign goals. 

Singhai joined Ad Age Studio 30 to kick off its new "30 Seconds With Studio 30" video series to offer a primer on how new gen AI technology is turbocharging advertisers’ ability to create and optimize every component of their campaigns.

The videos in part one of the series (see below) collectively underscore three key themes: the transformative potential of AI in making advertising more efficient, creative and inclusive; the democratization of advanced tools that empower businesses of all sizes to produce high-quality, dynamic content; and Amazon Ads' commitment to innovation that is both impactful and responsible. Together, these insights reveal how generative AI is reshaping the industry by enabling advertisers to connect with audiences in more meaningful ways.

Gen AI: The not-so-new tech on the street

New gen AI tools significantly shift from previous approaches due to several groundbreaking characteristics:

First, there’s been a massive uptick in the amount of data and computing power available to create AI algorithms. As a result, these algorithms' accuracy and ability to take in real-time signals are unprecedented.

Second, “We have moved from optimization and classification problems to generation—now you can create new content,” Singhal said. 

Gen AI can produce a wide range of highly customized, creative content, including text, images and videos, that lends itself to more unique, engaging ads tailored to specific audiences or individual users. The technology’s capacity to rapidly generate diverse, high-quality content can help teams that lack in-house resources, enabling brands to launch more dynamic and varied campaigns.

Finally, gen AI is now mainstream, so everyone—not just data scientists—can use it and feel empowered.

“That unleashes a great amount of value,” Singhal said.

Why these tools are so crucial for the advertising industry

New gen AI tools help advertisers level up their campaigns in a few critical ways.

Using large language models and deep learning, gen AI can analyze vast amounts of information to understand what may be helpful for customers with unprecedented depth and accuracy. This helps in “creating content with low effort, so the ability to keep it dynamic and relevant will increase many-fold,” Singhal said, with ads more targeted to shoppers’ current needs and interests. 

LLMs are also good at instantly distilling insights from large customer reviews, creating enormous marketing opportunities. In addition, AI can optimize using real-time signals, which helps free up advertisers’ time to do what they do best, like creating new products and telling inspiring stories.

Connecting advertisers with their target customers

Since its inception, Amazon Ads has been using AI across its tech stack to deliver increased discovery and sales for advertisers.

With billions of first-party data signals, the company’s AI models can assist advertisers in everything from choosing the optimal bids to winning more placements, identifying complementary targeting options and automating decisions to help put time and money back in advertisers' pockets.

For example, in Sponsored Products, Amazon Ads vends related keywords and recommendations for hundreds of millions of advertised products and their bid suggestions using deep learning and LLMs that leverage historical shopping patterns and vast information in the Amazon catalog. These models also stay up-to-date so advertisers can identify varying trends. Similarly, while serving ads, Amazon Ads employs LLMs with billions of parameters to match relevant ads with the shoppers’ intent and rank them based on the response prediction scores, such as estimated click-through rate and conversion rate. The machine learning pipelines serving these LLMs are updated in near real-time and serve within a few milliseconds of latency.

For a company that has been using AI for years, Amazon Ads doesn’t think marketers have come close to seeing what the technology can do for advertisers.

“With the rapid advances in gen AI, we expect the innovation to keep up in the upcoming months,” Singhal said.

This is part one of Ad Age Studio 30 and Amazon Ads' "30 Seconds With Studio 30" video series. Check back in the coming weeks for part two. To learn more about generative AI solutions from Amazon Ads, click here.