What my terminal looks like while testing a prototype.

Customer Care in Journalism? Yes, please.

Epsilon is a messaging product for engagement, feedback and onboarding.

After my previous post explaining how I ended up with a product idea by deconstructing the fundamentals of my initial project, it’s time to introduce you to the MVP I’ve been developing. The working title of the project is EPSILON. I want to build an intelligent agent that connects journalists to their community and engages with their audience. In other words, a messaging product for engagement, feedback and onboarding.

As Nic Newman, an experienced journalist and research associate at the Reuters Institute for the Study of Journalism in Oxford University, mentions in his paper “Media, and Technology Trends and Predictions 2018:”

“Personal Assistants for journalists disappeared in a round of cost cutting in the mists of time, but they’re making a comeback in the form of AI bots that can manage diaries, organise meetings, and respond to emails.”

With their conversational interfaces and increasingly larger databases of news stories, chatbots may become a more popular means of connecting newsrooms with new audiences in 2017. In its latest report, the Reuters Institute for the Study of Journalism predicts that this year the media will keep investing in innovation to develop chatbots, or what it’s calling “conversational journalism.”

With a bit of training chatbots can do all sorts of things for you, even tweeting while you are sleeping. I want EPSILON to go in a totally different direction.

I’m able to be part of this years cohort due to the generous support of the Onassis Foundation.

Can publishers meaningfully connect with their audiences with the help of chatbots?

Customer Care Is everywhere in the U.S., even in the subway. So why not in journalism?!

Historically newsrooms have lacked incentive structures and useful systems for journalists to be in efficient touch with their audience members for the benefit of their journalism. According to the Membership Puzzle Project, this is changing.

So far news organizations have used chat bots in order to do what they did for years: publish the news. What if newsrooms used chatbots in an absolutely reverse way? To ask for feedback? What if newsrooms embraced more customer care? What if digital agents asked for feedback and emails and brought them back to you, the journalist, in a spreadsheet?

The problems are still there

  • Social platforms and email remain the only channels for reaching out to journalists. While the communication channels are multiplying, they are platform-based, so you need to be a member of a network. Or if you want to call the reporter or the editor, it’s up to you to wait for hours on the line.
  • The comment section is a toxic environment. Toxic conversation online seems like a wicked problem: Online harassment is pervasive, silencing and siloing.
  • While everyone in journalism is using tools like using Google docs, CRMs, tools for payments and invoicing, very few are using customer care tools! And as community-based journalism grows, “customer care” will be more necessary.

So how will EPSILON look and function?

If you are a reader following a beat or a personal blog, or just reading a publication, you probably get something like this when reading.

This is a hypothetical article. Pay attention to the chatbot ☝️

And here’s how the screen of your device looks like after the interaction with the chatbot.

This is how a dialogue with the bot looks like.

Pretty much like any other messaging product out there right? Well, this is why it’s called “Engagement” and it’s up to the journalist to create it with the help of our software and “Engagement Management System.”

So how can journalists create their own personal assistant for interacting with their community?

The interface: These are the very basic wireframes for our “Engagement Management System.”

The first attempts to wireframe.

Our FUGLY way to allow journalists to create their own chatbots. Here is the first attempt.

In other words, this the very first version of our Engagement Management System.

After the interaction with the bot, you receive the feedback in your email!

The tools that took us here

Balsamiq and Keynotopia: Very easy tools for wireframing and designing screens.

Dexter: The easiest way to create your own chatbot.

Dialogflow: You can build text-based conversational interfaces powered by AI and can connect to any API.

Mailgun: The user should receive the results of the interactions directly in their inbox.

Atom and TextWrangler: The most common code editors out there.

Google Docs: Notes, thoughts.

Skype: A vital tool to communicate with my developer in Athens.


(I’d like to thank Jennifer Brandel for introducing me to the term “EMS” that hopefully doesn’t mean “Emergency Medical Services” but “Engagement Management System” and David Cohn for introducing me to the term “FUGLY”)