1 of 37

Planning Chat for Scale by Predicting a Two-Way Conversation

September 15, 2016 #ONA16chat

The Audience Engagement + Analytics track is generously sponsored by:

2 of 37

Planning Chat for Scale by Predicting a Two-Way Conversation

Ariana Tobin, Guardian News & Media @Ariana_Tobin

Emily Withrow, Knight Lab @emilywithrow

Zach Seward, Quartz @zseward

Special Guest Appearance: John Keefe, WNYC @jkeefe

September 15, 2016 #ONA16chat

3 of 37

Infomagical (WNYC)

4 of 37

We made a little short-run text message chatbot for a specific project. THe project – called ‘infomagical’ , which was meant to be tongue in cheek tbh, asked participants to fight information overload. It followed a similar project we’d done last year called Bored and Brilliant, about being creativity and smartphone fatigue. That project was great… but frankly totally exhausting. It was zillions of emails and big push on every platform imaginable.

This one we framed as the next step – how to actually use technology to help yourself do what you really mean to do with your devices, rather than just getting sucked into the vortex. We did it by text which wasn’t simple (hi John) but, we hoped, more personal, more easily accessible, and more streamlined for participants…. Important when your project is about feeling overwhelmed by information.

5 of 37

6 of 37

7 of 37

8 of 37

9 of 37

10 of 37

11 of 37

12 of 37

13 of 37

14 of 37

15 of 37

16 of 37

Knight Lab + Vox Product

on Amazon Echo

17 of 37

18 of 37

19 of 37

~ there is no screen ~

20 of 37

For us: “Flash Briefing”

21 of 37

Set out to build an interactive news “skill”—aka an app— for the Echo

22 of 37

Graphic: Sanette Tanaka

Story Flow

23 of 37

How do people talk to this device?

What “legacy” behavior for speech already exists?

How do we write for this device?

What does rapid prototyping look like in this environment?

Do people *want* to ask questions? What kind of questions?

Can we predict what questions people will ask?

...with 90 percent accuracy?

Questions we loved answering

24 of 37

User Testing: Dashboard solution

25 of 37

Quartz for iPhone

1 of 37

2 of 37

3 of 37

4 of 37

5 of 37

6 of 37

7 of 37

8 of 37

9 of 37

10 of 37

11 of 37

12 of 37

13 of 37

14 of 37

15 of 37

16 of 37

17 of 37

18 of 37

19 of 37

20 of 37

21 of 37

22 of 37

23 of 37

24 of 37

25 of 37

26 of 37

27 of 37

28 of 37

29 of 37

30 of 37

31 of 37

32 of 37

33 of 37

34 of 37

35 of 37

36 of 37

37 of 37