News bot

RJI Fellow developing bot-driven engagement to build community, extend news coverage

Coverage. It’s what we promise as news organizations. We cover stories, communities, events and people. Every news organization should make one promise to its community: We’ve got you covered.

But we don’t, not by a stretch. Our coverage is actually becoming more gap-filled and uneven by the day. I won’t spend much time here writing about why this is happening. The decline in local news has been well documented, and there’s not much to add to John Oliver’s must-watch jeremiad.

I prefer to focus on what’s working and what might work in the future. My fellowship work at the Reynolds Journalism Institute is all about using technology to improve coverage and build community, given even the morbid economics of local journalism.

So what’s this magic technology that might help us enhance our coverage, extend our reach, build community and even develop a foundation for a healthy and robust journalism business?

Robots, naturally. More precisely, natural-feeling conversations with messaging applications that automate both messages and questions.

In the last several months, “bots” — or some semblance of them — have proliferated. CNN launched its first bot on Facebook Messenger. Quartz’s new news app mimics a conversation bot. You can message Poncho on Facebook for weather updates tuned to your location. And there’s no shortage of efforts to move customer service to bots. Beyond messaging apps, whenever you interact with Amazon Echo you’re engaging a bot. 

In the case of news, bots are very early stage and experimental — and largely underwhelming. Every one I’ve seen so far (as with several texting and WhatsApp experiments) treats messaging bots as a way to distribute and discover content.

But messaging was not designed purely as a broadcast tool. By nature, it’s a two-way conversational channel. Users expect to be able to respond naturally and to ask questions. So, yes, bots should help personalize content based on questions and other prompts, but it should also ask people to respond with substance — not just with instant polls, buttons or other simplistic feedback mechanisms.

When people respond with substance, at scale, vast potential opens up for gaining eyes and ears in all parts of our community, enabling newsrooms to achieve total community coverage.

That, in a nutshell, is what I’ll be doing as an RJI Fellow: exploring how GroundSource, the community engagement platform I’m building, can refine the application of bot-driven messaging that broadcasts as well as listens — distributing and collecting information in equal measure.

Early-stage projects with several news organizations including Univision, WWNO in New Orleans and GPB Media, suggest that when properly deployed, people really enjoy corresponding with bot-driven messaging; and as they get comfortable, will message back as if talking with another person.

I’ll be working with University of Missouri news publications, a few select newsrooms, and my colleague RJI Fellow Connor Sheets to streamline the process by which journalists use GroundSource to spin-up bot-driven engagement and spread it through a range of outreach and community development tactics.

Given the moribund state of local news, the promise of technology that can not only beef up our coverage but also support personal conversational engagement might just be the formula we need to strengthen local news coverage.


Comments are closed.