Fresh from INMA’s Media Tech and AI Week in San Francisco, I am looking back at several case studies that serve as solid examples of AI use by news brands and what they have learned from their experiments.

I wrote about two of them earlier this week — The Seattle Times using AI for ad sales and The Dallas Morning News using the tech to identify unknown users.  

Here are two more I wanted to share, examples of AI helping in the newsroom and targeting ads:

The Philadelphia Inquirer uses AI to simplify archive search

The Philadelphia Inquirer is running RAG applications over their deep archive to make it much simpler for reporters and editors to go back and find content.

The previous search process was “very clunky and cumbersome,” said Matt Boggie, chief product and technology officer. “It required knowing roughly which day a particular story was published on. You needed to know exactly what kind of terms were being used to describe it.

“Now we’re finding that not only can we much more easily retrieve those stories and identify that coverage, but we can also ask for basic summaries or different presentations of that information analysis. For example, you can take two different competing candidates for an office and have the system compare their stances on various issues to each other. You could identify a politician and see how their stances have changed over the course of their career. So it’s been really powerful for our reporters and editors.”

The Inquirer’s journalists are saving, on average, a day or two per week compared to previous ways of searching. “Who wouldn’t appreciate a 20% or 40% time savings, particularly on work that isn’t your highest level sort of work, and you can focus on more of the analysis than the collection of the information,” Boggie said.

An unexpected benefit has been for the sales department, he said.

“One of the use cases that popped up really quickly that we hadn’t anticipated was: We’re about to go on a pitch and we’re going to go talk to this brand. They’re a storied brand here in the area. We must have done some coverage in the past. Let’s go back and take a look and see what kinds of things that we talked about, both in terms of being able to promote the fact that we are promoting them but also potentially to prepare for an objection that may come up in that meeting.”

The system is updated every night, so the information contained is up to date.

The Boston Globe uses AI to better target ads, audiences

The Boston Globe is working on improving their data so they can better target ads and subscriptions. 

“We have a ton of data around our stories,” but it needs clear categorisation, said Shira Center, vice president of innovation and strategic initiatives at Boston Globe Media. “When I was an editor, I saw this all the time: The reporter would go to file their story, and they would attach their own taxonomy, which is a very subjective process.”

How does AI help? If an advertiser wants to place ads against the Boston Marathon, a large annual event, but not next to any story that mentions a bombing at the marathon 12 years ago, “we can facilitate that if we use AI, which is better than the human at picking the label and the sentiment for the story,” she said.

“Ideally, it goes from our advertising shop, where they can monetise it, to our subscription shop as well, and we can develop much better insights and ultimately answer the question: Why does someone subscribe to The Boston Globe?”

The Globe has more than 250,000 paying digital subscribers and is one of the most expensive regional newspaper subscriptions in the country.

If you’d like to subscribe to my bi-weekly newsletter, INMA members can do so here.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .