Hyperlocal AI with a Million Subscribers

Patch, according to Warren St. John, its CEO, is "a little shop that's trying to do this kind of antiquated thing that everybody is giving up on." That "antiquated thing"—hyperlocal news—became less old-timey when, last spring, the company announced the rollout of AI newsletters that could reach as many as thirty thousand communities. Give the site your zip code and it would produce a daily or twice-weekly newsletter customized for your town, for a minimum readership of one subscriber. "Once we get a sign-up, then we start generating it," St. John told CJR. Since then, PatchAM, as it's known, which carries local and national advertising, has been activated in fourteen thousand communities and has accrued nearly a million subscribers.

"This is a utility. This is not the high church of journalism. This is about creating a foothold in a relationship and meeting a need."
— Warren St. John, CEO, Patch

PatchAM provides the most value to places with few or zero outlets providing local news. The newsletters rely heavily on aggregation, automated event calendars, and posts from Nextdoor. They ask users to submit photos of their pets and throw in a riddle. The newsletters try to avoid niche or divisive topics such as local politics. "People don't want everything about their community, they want the most important things," St. John said. "They don't want to know that tomorrow night in town is third-grade parent-teacher night. That's not broadly relevant. But they would want to know that the Greek food festival is on Main Street this weekend and everybody's invited."

Patch is far from the only local news publisher experimenting with AI for tasks ranging from scraping meeting minutes to producing articles. But there's a particular reason to watch Patch: the company knows how to turn a profit. It has mostly been in the black since its 2014 purchase by Charles Hale, an investor specializing in tech turnarounds. Under Hale Global's ownership, Patch has been willing to experiment and fail. As a for-profit, Patch is always trying to solve for "some component of the riddle of how to make this sustainable," St. John said.

Patch's AI does occasionally pull news from the wrong city—what St. John calls the "wrong Springfield" problem. But the company sees its aggregation approach, including links to paywalled local outlets, as a feature that supports the broader local news ecosystem.

Patch's next major AI endeavor is developing a tool to recap town meetings, which would be used directly in its newsrooms—scanning lengthy public meeting recordings for transcripts and news nuggets. "If people are sending rockets into space and cars are driving themselves, I think we can apply this technology to do our jobs better and more efficiently," St. John said.