After a wave of high-profile police shootings in other parts of the country, two Texas Tribune journalists wondered: Was there any way to gauge how often police shot at people in Texas?
"As soon as we started looking into it, we realized there was just this huge information void out there," said Alexa Ura, a reporter who covers demographics.
As they reported, they also found departments don't have one way of keeping track of such shootings, despite a law passed to fix that.
So a project that was supposed to come out in the spring was pushed to late August. The results, though, were worth waiting for.
"Unholstered" digs into the data and the stories behind it, helping to fill a void of information in Texas while being transparent about what's still missing. The project, which debuted last week, includes several stories, a visual primer, a video tailored for Facebook and a Reddit question-and-answer session as a follow-up.
During the year-long project, Tribune journalists learned several lessons about reporting on police shootings.
Prepare for the unexpected
Ura and Jolie McCullough, a reporter and news developer, knew they couldn't take on each of the 2,000 cities in Texas to build a database of police shootings. But they could capture some of them. They decided to gather data from 2010 to 2015 from cities populations of 100,000 or more. That left 36 cities which, as the project's about page explains, make up about half of the state's population.
The two were joined by Tribune criminal justice reporter Johnathan Silver. Together, they started filing carefully detailed open-records requests.
They asked for spreadsheets with the names of individuals shot by police, the shooting victim's gender, race, the location of the shooting and the officer's race.
"We wanted to be really specific in what we were requesting to get uniform data," Ura said. "We were almost too specific in that way."
A lot of departments were helpful and transparent, McCullough said. But some invoked exceptions. Others had the information stored in a way that wasn't searchable. So Ura, McCullough and Silver had to modify their requests.
The Tribune initially took for granted that every shooting was accounted for, said Ayan Mittra, managing editor at the Tribune. He doesn't think those department had bad intentions, he added, but rather a lack of resources.
"There were a lot of shootings that were supposed to be sent to the state database that didn't get sent or didn't show up in the database," he said.
Be willing to pivot
In January, after months of collecting data and working their daily beats, one editor looked at what they'd found and thought it was worth taking the time to go deeper.
Ura, McCullough and Silver sent another round of records requests, this time asking for specific incident reports.
"It really ended up making the project a lot better," Ura said, "even though it made it longer."
Giving the team time to write narratives around the data allowed them to look into trends they saw, including a number of shootings involving people with mental health issues. Those trends make up several sections of the finished project.
"I don't think we intended to write six stories when we set out to do this," Mittra said.
But as they gathered what they found and teased out the issues, they saw the shootings as standalone incidents with their own sets of circumstances worth exploring.
Use presentation to help reach the most people
The Tribune didn't just produce a lot of words with this project. It also published a data visualization that boiled down the main findings. And guess which page got the most traffic?
"I think presentation was something that we thought of in terms of reaching people and making sure that the information provided wouldn't get lost," Ura said
"We wanted to lay them out as simply as possible," McCullough agreed.
The fast look at the main findings makes them digestible and allow people to sample the story before deciding whether to dig deeper.
Think beyond the platform
For big projects, the Tribune wants to prepare the audience and follow up after the big project comes out. It's part of an effort to "expand the arc of story beyond when we hit publish," said Amanda Zamora, the Tribune's chief audience officer.
To do that with "Unholstered," they created a trailer for Facebook that invited people to sign up with their emails to get the story as soon as it came out.
"It does tease the investigation, but it also stands alone," Zamora said. "You're going to learn something by watching that video in and of itself."
That video was one of the Tribune's most watched on Facebook this year, with more than 30,000 views. It brought in a few hundred email signups, Zamora said, and Facebook was one of the biggest drivers of traffic to the story.
"So it was a win for social, and it was a win for email."
And think beyond the moment you publish
After "Unholstered" came out, the Tribune did its first "Reddit IamA," a question-and-answer session with a former police chief now teaching criminal justice at Texas State University.
"I think it's great whenever we as a news organization can connect people with journalists," Zamora said, "but also newsmakers and sources."
Reddit helped the Tribune keep the dialogue going, Mittra said.
"This isn't just about us putting a project out and forgetting it in a week. This is an issue we need to keep addressing from multiple perspectives. This can't be seen as an us-versus-them issue. It's not," he said. "We need to keep working together."
Then, mobilize your sources
With big projects, Zamora's trying to get Tribune journalists to activate their source network by sending an email linking to their projects.
It's a new effort, and it's a basic thing, she said, but it can be effective. Reporters can play a big role in making sure their work makes it to the communities it involves.
"It takes five or 10 minutes to write out a note for those folks," Zamora said. "It's a simple step in making sure that your story is not just widely read but read by the folks who are able to make sure that the right community sees your work."
Correction: An earlier version of this story contained a duplicated paragraph. It has been corrected.