RJI Fellow’s tool measures ‘impact with context’

What will happen to this article about an impact tracking tool once it’s published?

Will readers do more than just read it?

Will it inspire them to start a GoFundMe campaign to develop the tool further or start a discussion board about the need to track the real-life impact of stories?

RJI 2019–20 Fellow Leezel Tanglao hopes to answer questions like these with her new online story impact tracker, for which she began developing at the Donald W. Reynolds Journalism Institute and is now seeking additional funding.

With newsrooms becoming leaner, time and resources are limited to manually track the full impact of stories beyond page views and clicks, says Tanglao, who is also the owner of the consultancy firm StatFury, LLC. She heard this repeatedly from journalists who beta tested her tracking tool, which tracks impact automatically using artificial intelligence and natural language processing.

“Being able to track the impact of stories is a way to articulate what we do in a very intentional and purposeful way to other industries,” says Tanglao. “Unfortunately, many organizations rely too much on vanity metrics such as clicks and time spent on site.”

According to Tanglao, the current tracking used now “does not give a greater holistic view of impact, and when you do start to track and measure impact, that opens up a lot of different opportunities on the editorial side, as well as the revenue side.”

How the tool works

When the tool is finished, journalists will be able to run a story URL through it. From there, the tool looks at key phrases within the story and automatically categorizes types of impact a story could have, whether that is an economic, social or policy impact. With this information, it will then ping open-source APIs and databases to find related reference materials, research and other articles.

Tanglao says the tool will dig deeper than an internet keyword search where the content you retrieve may or may not be related to what you are looking for. Instead, her tool’s natural language processing allows the user to look for relationships between words and phrases in the article.

How funding will help improve the tool

Additional funding is needed for development work to improve the tool, says Tanglao. For example, the tool is only able to access content from free sites. Funding would allow Tanglao to get licenses to more databases and APIs, which would mean the tool would have access to more content that highlights impact.

Funding would also help her continue to improve the tool’s functionality and have more time and resources to focus on the development. Currently, there is some lag time between when a user inserts a URL and gets responses because the tool is on a small server.

Tanglao also wants to improve the front end of the system to make sure it’s more user-friendly, which would include an easy way to provide feedback on the results that are surfacing. She also wants to make the system accessible to consumers, as well as news outlets.

“News consumers only see the finished product — the story,” Tanglao says. “But they don't see the time and work it took to put that story together and the subsequent impact of it after the story publishes. By showing this in a way that’s easy to digest, it will go a long way toward restoring trust in the media.”

What she learned

Tanglao says developing the tool taught her that when you are developing something as complex as a computer model or algorithm that requires testing, journalists should factor in extra time into the timeline. She says there are many variables to think about when someone is trying to figure out how to scrape data and scrape it in a way where it can continue to evolve and teach the computer.

“It’s always going to probably take more time than you think,” Tanglao says. “It’s not impossible, but it’s just going to take more time and resources to figure it out.”


Comments are closed.