In 2003, Vin Crosbie wrote a Poynter column suggesting that news organizations “should get ready to use mobile telephone Short Messaging Service (SMS) for news delivery and as a micro-transactions processor.”
This prescient column was published four years before the iPhone launched; Crosbie based his prediction on foreign news organizations, which were already using text messages to connect with readers, and the success of FOX’s "American Idol" — which inspired people to send 2.5 million SMS messages to vote and respond to polls.
In 2012, SMS usage peaked when messaging platforms like Kik and WeChat began to outpace SMS services in volume. Over the past decade, news organizations have experimented on both SMS as well as messaging platforms like Facebook Messenger, Slack, and What’s App. (Not to mention Twitter, Amazon Echo, Telegram, Kik, Line, and plain vanilla push notifications.) Below are some of the creative ways news and other organizations have used these platforms, with some ideas for how we could extend these mediums even further.
- Share behind-the-scenes information from a person or group of people. In 2016, The New York Times deputy sports editor Sam Manchester shared real (and human) personal dispatches and photos from the Rio Olympic Games via text message, and the political desk experimented with something similar using Facebook Messenger. Over on Slack, the paper experimented with a chatbot that gave people the ability to ask political questions using the command ‘/asknytelection.’ BuzzFeed experimented with feedback loops and bots during the Republican National Convention (as did Mic.) And the New York Public Library allows you to text with a librarian if you have a question. (And the Brooklyn Museum has an entire list of questions that they recommend as starting points when corresponding with museum staffers.) More ideas and questions: How do you then create audience engagement in real time? This SRCCON session from 2015 has some ideas. And how do you cultivate audiences around even more niche topics? Is there a way to hand control over to the audience?
- Do one thing really, really well. John Keefe, who leads the Bot Lab at Quartz, made a Twitter bot that helped journalists navigate NICAR 2017. His colleague Zack Seward calls these “worker-bee bots” — they’re good at doing one thing well. (Other examples include the single-stream coverage Alexa bots that The Washington Post has created, and their Olympics Facebook messenger stream, which was powered by a bot.) More ideas and questions: How can these help people navigate ongoing new stories?
- Become the source on a single topic. ProPublica’s election data bot shares campaign data in near real-time. The bot, which pulls data from several sources, “makes it easier for journalists, researchers and citizens to quickly find newsworthy information about the presidential race and congressional campaigns in their states.” Another example? Univision sent text message fact checks during the election. More ideas and questions: Eventually, there will need to be a directory of these single-source streams. Is there a way to find them via text?
- Help reach a specific audience. The BBC experimented with Facebook Messenger to send daily digests of material on Facebook Messenger and Telegram to reach a wider audience and to reach audiences where the BBC is censored or blocked, respectively. The Wall Street Journal, BuzzFeed, and the BBC are all experimenting on Line to reach a younger audience based in Asia. Infowire, a project from Michigan Radio, “provides high quality information about education, health, services, food, jobs and community” entirely through SMS. (Its founder, Sarah Alvarez, now has a company that provides more news through texts.) And Listening Post at WWNO in New Orleans uses text messages to survey people in the community about local issues, which are then shared on the radio. More ideas: Outside of news, personalized text messages have nudged the number of low-income high school students to attend college. What other ways can behavior be nudged?
- Surface interesting stories based on topic or domain. Digg’s Slack bot allows people to search for articles by domain, keyword or randomness. TechCrunch has a similar bot for Telegram. BreakingNews.com has a Slack integration that shows when breaking news has occurred. CNN sends out news alerts on Facebook messenger; the BBC does so via text. More ideas: How can this be combined with other methods of discovery?
- Help your newsroom out internally. Al Jazeera has a Slack bot for web analytics and one for breaking news. The New York Times has a bot for what stories to push to social.
- Surface interesting stories from your archives. The Guardian’s Sous-chef finds recipes from their archives, based on what you have in your fridge. (They also have a news bot.)
- Experiment with different modes of responses and inputs. Fusion’s chatbot “Emoji News” allowed people to send an emoji to Fusion, and then receive a story that reflected that emotion. The Washington Post asked people how they felt via Facebook Messenger during election season, and then reported back interesting answers. And SFMOMA asked people to text them an emoji or feeling and they immediately text back with a piece from their collection.
- Experiment with different modes of engagement. The Washington Post uses Kik to experiment with “games, quizzes, and chat adventures.” The BBC is using What’s App to ask for user-generated content from its audience. The WNYC show “The Takeaway” connected with listeners who had tips through text messages. And Mic is keeping readers aware of advocacy issues on Facebook Messenger.
- Cultivate sources and find information. Foreign correspondents use chat apps “to cultivate sources and gather news” by gaining access to group conversations on chat apps, but find relevant conversations can be a challenge, says a recent Tow Center report. ProPublica used the GroundSource platform to survey New Yorkers about increasing rents. Also: The Virginian-Pilot surveyed their audience about light rail plans, and Note to Self from WNYC sent 300,000 texts as part of an experiment measuring information overload.