EIDR APM: Can AI Traffic Predict Box Office Results?

PLAYA VISTA, Calif. — Pre-sales, historical data, audience surveys, social media and sentiment analysis. Studios have a wide range of tools for predicting how a film will perform at the box office.

But what if AI agents — and what they’re digging through online — are the best indicator yet of whether a film booms or busts on opening weekend?

Bruce Nash, founder and president of Nash Information Services, which includes The-Numbers.com, a database of movie financial information, and its associated database, OpusData, argues that by following what AI agents are up to, content stakeholders can find out six months in advance which limited edition popcorn buckets to stock, and which films to put on more screens.

He shared his insights at the recent EIDR Annual Participant Meeting.

The-Numbers.com receives upwards of two million page requests every day, with 80 percent from web crawlers, 10 percent from humans, and about 10 percent from live search AI agents.

Most AI systems feature both crawlers and live search agents, and those AI agents generate a whole lot of ton of online traffic. According to Nash, monitoring what those AI agents are requesting provides a real-time, unfiltered view of what people are asking about.

Just looking at some recent releases, the AI agent traffic for The Super Mario Galaxy Movie and Project Hail Mary were both massive leading up to opening weekend, and both performed extremely well in theaters. And just based on ChatGPT traffic, Nash predicted a $98 million opening weekend for The Devil Wears Prada 2 (domestic ended up being $145 million) and a $59 million opening for Mortal Kombat II ($41 million actual).

“The interesting piece we’ve seen in the last year or so is the traffic of AI agents,” Nash said. “It gives a real-time, unfiltered view. We’re getting very close to predicting, with a great deal of confidence, [box office performance].

“We’re beginning to see signs that we can do this six months in advance. There are strong messages you can get by examining this traffic.”

And, since AI agents seek out frequently updated and information-rich sources, by embedding EIDR IDs in page URLs, content companies can link newly found market intelligence to other sources of data, completing the prediction picture for new releases.

“Growing EIDR Together” was the theme of this year’s APM event, with the organization and its members discussing how to register more content, have more companies participate, and include more global voices, languages, and alternate IDs.