I’m just back from INMA’s Media Tech and AI Week in San Francisco with my notebook bursting with insights and my head bursting with many mind-expanding thoughts from every corner of the new information ecosystem.

We heard from a wide variety of voices — researchers who bring news organisations together; software developers who focus on content provenance and monetisation, trying to figure out the black box that is Generative AI; tech companies who are working on ways to increase engagement and the utility their GenAI products provide; and news publishers who are using AI effectively while bringing their newsrooms along. 

I had the good fortune to attend a panel discussion featuring four leaders from different U.S. news publishers examining their nascent AI projects. They were brought together by The Lenfest Institute, which focuses on building a collaborative community of practice for AI to help drive business sustainability and innovation in news. 

Here are two of those case studies.

The Seattle Times’ AI ad prospecting agent

The Seattle Times got started with AI in its advertising department by building a prospecting agent. Team members can query the tool using natural language and get leads for specific topics, areas of coverage, products and geographical areas.

“It’s proven to be really useful,” said Kati Erwert, senior vice president at The Seattle Times, with more than 90% of the team using it daily.

The tool has also provided benefits for existing advertising clients. Staff can plug in the advertiser and receive context and background, such as projected market spend, so they can enter meetings feeling well prepared to pitch advertising solutions. 

The sales team also has a relatively high turnover rate. The tool helps onboard and train sales reps, a process that used to take up to 90 days because The Times’ product portfolio is complex.

“What we built is essentially an info hub where people can go and ask questions. What’s the spec of a print ad? What’s the deadline for handing off banner ads? It’s very quick, intuitive, responsive. It captures our current go-to-market strategy,” Erwert said.

The Dallas Morning News and known users

The Dallas Morning News is working on an initiative to make unknown users known, “a huge, huge challenge for us and probably many of us in this room where we see 98% of the people that come to our site or the bots, and we don’t know who they are,” said Chris Patheiger, chief product and innovation officer.

“We don’t know anything about them, which makes our ability to turn those visits into somebody who’s brand loyal really, really difficult because we have to kind of treat them with a very generic set of offers and experiences.

“Our overall plan is to use a series of smart, AI-driven prompts in articles to be able to entice our users to declare first if they’re real people or not by engaging and asking questions and querying the articles themselves in exchange for some information along the way.”

At the moment, readers come to an article in search of specific information — and then get hit with a request from the news site, Patheiger said.

“The way that we do it right now is a really kind of ham-handed approach. You’ve come to this story to read something that’s of interest to you, but before you do that, here’s what we care about, like, sign up for a newsletter or subscribe to this article.”

The object of the AI project is to elicit an interaction with a reader about the article or the topic and get them to engage with it, Patheiger said.

“And then you can deliver any sort of smart registration opportunity there where they can maybe read more of the article or get a bigger meter” or similar options, he said, pointing out the conversion rate on a registered user is anywhere between 15x and 30x higher than that on a fly-by visit.

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

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