It’s not enough to simply liberate big, messy datasets and hope other journalists will pick them up and run with a series of investigations. But help is here, and more is on the way, as an emerging model of support spreads and matures in 2025.
Is it “concierge data journalism”? Is it “white-glove collaborations”? No matter what you call it, it’s a smart idea and one that solves a problem we’ve all heard of by now: How might we help small, local newsrooms produce more and better accountability journalism in their communities, even if they lack the people, the bandwidth, or the expertise in documents and data?
Of course, there has been foundational work from newsrooms like ProPublica (and from my colleagues here at The Marshall Project) to put more data journalism into the hands of local reporters and newsrooms, some of this work dating back nearly a decade. But now, teams of data, FOIA, and beat topic experts are hoping to not only put datasets into the hands of local journalists, but also guide them through localizing the data and story angles, offering editing and infrastructure support as partners.
My colleagues at The Marshall Project are expanding our Investigate This! model, and similar work is coming through Big Local News at Stanford. These teams work directly with local journalists to clean and analyze datasets, evaluate story ideas, and share assets. Further emerging in this space is work from The Trace’s gun violence data hub, which launched as a help hotline in late 2024 with big aspirations to come to market in 2025.
Three smart teams with a smart and similar mission: help support and create more data journalism from local newsrooms.
For The Marshall Project and The Trace, we’re both single-subject newsrooms. This means our teams of journalists tend to be subject-matter experts on our particular beats. But the truth is criminal justice systems and incidents of gun violence happen at the local level, in towns and communities. It’s important for the journalists in those places to be able to tell these in-depth and crucial stories, even if they aren’t deep-dive experts — yet.
Big Local News works to put datasets in the hands of journalists, but also creates webinars to onboard folks into a story angle or subject area. P. Kim Bui, a project manager with BLN, said the coaching aspect of their work is one of the most important pieces of the model. The team at BLN is largely made up of expert data journalists who invest their time in helping the smaller newsrooms navigate and execute on data stories.
All three groups aim to find or obtain datasets, help clean them, then work with local journalists to specialize the topics to their own communities — building on the existing local expertise to strengthen each investigation.
It’s not enough to just drop something into the laps of local teams and hope they’ll use it exactly as-is, my colleague Michelle Billman told me. And she would know: In a past life, she ran a local public media newsroom, just like the ones she’s trying to partner with now. But teams like hers and BLN’s (and soon, The Trace) are working to help journalists in small shops not just by sharing the data, but also by offering things like specialized beat knowledge, illustrations, and data demonstrations.
And it’s that real-time, individual level of support that The Trace hopes to provide, in addition to access to datasets in the future. For now, their help desk is open and journalists in need of support can file tickets to the team. The hotline exists so that “any journalist can reach out and ask questions about gun violence coverage,” George LeVines, editor of the data hub, told me.
Although The Trace will start making datasets available in the spring, LeVines told me his team is currently working through a number of collaborations with other newsrooms, offering support like data analysis.
“I am hoping that this product can serve as a model that more broadly can be applied to data and investigative journalism overall,” he said.
Emma Carew Grovum is director of careers and culture at The Marshall Project.
