As traditional targeting methods face increasing limitations, the industry is searching for more sophisticated approaches to understand user intent and deliver relevant advertising. RTB House has built its reputation on leveraging advanced technology to improve advertising performance.
IntentGPT emerged at a critical time for the digital advertising ecosystem. The industry has faced challenges in determining what users are actually interested in based on their online behavior. Traditional keyword matching casts too wide a net, leading to inefficient ad spend and poor user experiences.
What makes IntentGPT noteworthy is its use of large language models, the same underlying technology that powers advanced AI systems, to analyze and understand web content at a deeper level than previous methods allowed. Rather than simply matching keywords or analyzing basic page attributes, the system aims to comprehend the meaning and context of content across publisher sites, much like a human reader would, but at internet scale.
In our conversation, Gillespie explains how this technology addresses fundamental challenges that have persisted in digital advertising for years. He discusses the technical approach behind IntentGPT, including how it processes content across multiple languages and integrates with first-party data to create more precise audience targeting. The discussion also covers practical performance metrics, with early results suggesting significant improvements in cost efficiency compared to traditional methods.
The timing of our conversation is particularly relevant given recent developments in the browser landscape. While Google recently announced its decision to retain third-party cookies in Chrome, other major browsers continue to restrict or eliminate cookie-based tracking. This mixed environment requires advertising solutions that can perform effectively regardless of cookie availability—a key consideration in IntentGPT's design.
The following interview offers a look at how advanced AI technologies are being applied to solve longstanding challenges in digital advertising, and what this might mean for the future of audience targeting and ad relevance.
What key challenge in digital advertising does IntentGPT solve, and how does it outperform traditional targeting methods?
IntentGPT solves one of the most fundamental problems in advertising, which is understanding what users' real intent is across the wide variety of publisher pages throughout the internet. It goes way beyond prior methods of simply matching keywords, which tend to capture far too broad an audience. It creates hyper-specific audiences that are precisely defined for your advertisements and where those ads will drive your business outcomes. We're already seeing early results indicating that we can reduce the cost of engagement by up to 44% using IntentGPT, and we're just getting started.
How does IntentGPT use Deep Learning to predict user intent better and drive campaign results?
IntentGPT uses Deep Learning and what we call LLMs, or Large Language Models, which is a specific type of Deep Learning that understands not only the words on a page, but the actual meaning behind those words.
And if you deeply understand everything you're selling and you lean into true first-party advertising, such as leveraging an ecommerce company's product feed, you can join that with this LLM-based signal in order to find exactly which publisher pages on the internet are hyper-targeted to those same products. You can use it for targeting, and you can also use it to make sure that the ads that are shown have products that are hyper-relevant to that user at that moment in time. That's IntentGPT.
How does it differ from the existing solutions, such as ours, and competitors?
IntentGPT is really a step change compared to existing solutions, both the solutions we employ and the solutions that are employed across the internet. If you think about existing solutions, certain best-in-class approaches to understanding whether users will really engage on a publisher page are based on matching sets of words or keyword analysis. Or perhaps understanding basic sentiment analysis by looking at whether those keywords reflect certain emotional states, such as happiness or nervousness. That's a very prior-generation approach.
Even our existing approach at RTB House is based on more scientific lookalike models to better understand the kinds of people and web pages that would resonate with your product.
IntentGPT actually has the ability to read and understand a page in its entirety, deriving deep meaning from the content, the same way a human would, but without the core limitation that a human would have. We can implement this technology at scale. We can do it in 49 languages across the entire internet in near real time to ensure that we can drive as much scale and as much performance from optimally placed advertisements as possible.
Where does IntentGPT fit as a solution if cookies are actually here to stay?
IntentGPT is a solution that applies to every product we offer, and it's a solution that works in a world with or without cookies, and as you know, Google has recently decided that they actually won't be deprecating cookies in Chrome. But many other browsers out there, like Safari and potentially Opera and others, have taken a different path, where cookies are either less available or not available.
Some of the best solutions will be those that function equally well in a world where some users have cookies and some don't. If you think about what's constant in both those worlds, it's the idea that users traverse the internet based on what they're interested in, based on what they need. That's cookie-independent, and that's exactly the data that IntentGPT leverages in conjunction with your first-party data.
It's a first-party advertising solution that marries first-party data with any other relevant data that exists. There will still be users going to web pages, and those web pages will have words and images, white space, comments, layout—all sorts of signals that are strongly indicative of whether that user is a good match for your ads. IntentGPT considers all those signals, which is revolutionary compared to prior approaches, and drives advertising that'll perform with or without a cookie.
Interested in learning more about IntentGPT and how it could impact your advertising strategy? To discover how AI-powered intent detection might work for your specific advertising challenges, visit a dedicated IntentGPT page or contact RTB House directly to explore options for your brand.
