So John. Can you explain a little bit about your career thus far, and how your journey has lead you here?

My career has been many faceted. I’ve been in newspapers, radio, online publications over the years. Recently, before coming to Quartz I was at the public radio station in New York City, WNYC. I was the news director there for many years. Then moved to head up the data news team there. That was a lot about bringing visualizations, maps, data to people directly on the web, which today seems like, “Yeah. Everybody does that.” But back when we started it, only the really big players were doing it, and it was something new to bring to a public radio station. That was a lot of fun and kind of groundbreaking.

I came to Quartz a little over a year ago mainly because what Quartz’s been up to from its inception is user-centered service and a lot of innovation.

Really exploring different ways to communicate, to bring journalism to users, to the audience, and that’s really exciting. When there was an opening in the Bot Studio, which is brand new here at Quartz, I reached out to Zach Seward who is the head of product here at Quartz and asked him if he needed any help with that, and he said, “We should get some coffee.” I started working here at Quartz.

Don’t feel like reading? Check this interview out in video format below!

Yeah. That’s great! Now you’re here. What does Quartz do exactly, and what’s your role within all of that?

Quartz is a guide to the global economy for people who are excited by change. My role in that is really looking at how conversational interfaces might play a role in being a guide to the global economy.  Because our audience is really about reaching people who are excited by change, exploring different ways that we might reach them is really important. Quartz is known, in part, for its groundbreaking app. It’s an iPhone/Android app that is not like any other news app, especially when it first came out. It’s very chat based. The news comes at you in bubbles that are written by amazing people, and it’s a very chatty interface. Exploring what can go beyond that is something that the Bot Studio has been up to.

What happens when the audience starts talking back to you? How do you handle that?

How could a conversational interface and or in a voice interface like an Amazon Alexa or a Google Home; How could that play a role in delivering news and information to our audience? That’s what we have been experimenting with.

 

Can you tell me a bit about the value of prototyping and why developers should be prototyping before they code?

I am in love with prototyping. It is one of my favorite things to do. I find it super engaging and important. I learn whenever I prototype. Prototyping can be anything from drawing things, building stuff with tape and paper, to even using very simple applications, bot platforms and what have you to just try things out. Prototyping is to just play with it. To see how the flow feels

An example of an idea, here we have a workshop where we have a 3D printer, right here at Quartz. Why would you ever think that that would be something a media company would need? Then, along comes Apple making it easy to put 3D objects inside the phone using AR. Well, we actually have people here who have been printing 3D objects, so the idea of taking virtual 3D objects and putting them inside a phone wasn’t that much of a stretch, because we had already been playing with that. To me, any of that play and prototype that you do, that you’re interested in, it will always have some application. Either directly, or just in your brain.

 

How can conversational interfaces rework the way the general public consumes content? What does this mean for journalism?

That’s a good question. We’re still exploring how conversational interfaces might apply to journalism as a whole. I think what we’re finding right away, is that it’s more about the intensity of that relationship and less about the breadth of their relationship. If we make a bot for journalism, and we have. We’ve done this a couple of times, in a couple of different ways. It might not be that we suddenly get hundreds of thousands of people coming at that bot, but the people who come there are really deeply engaged in the content and remain so.

They will actually go through our entire conversations usually to the end. 90 percent or more of the people who start these conversations with us, go through to the end of whatever we’re offering them that day. I don’t know that 90 percent of the people who are going to articles are going through to the end of articles, even. That would be something to look at. That close relationship is super valuable, we think. That’s what we’re exploring.

Yeah, I know we kind of talked briefly about classroom, kind of from an educational standpoint, but how can prototyping conversationally or conversational experience provide value in classrooms?

I think the first thing it does is it demystifies what’s possible. Right? I mean you can walk up to an Amazon Alexa or even a messenger chatbot and you can think that that’s too complicated to do yourself. How would you code that and in some ways you would be right. Coding some of these things can be hard, except that now there are all sorts of platforms out there that allow you to play with this kind of thing.

Once you can start playing with it, then the coding of it and the technology of it sort of falls away. You can start really thinking about how do you make a really good conversation. Then you can try it really quickly and you can see if it works or it doesn’t work. You can turn to the person next to you and say, “Hey, try this. Ask it a question.” You can see right away whether it works, whether it clicks, whether the language is off, how best to do that. Being able to iterate on something like that really quickly is really important.

 

What do you think makes a good bot conversation?

I think there are several things that make a really good bot conversation. One is for it to be very conversational. That seems like it would be an easy thing to achieve. It turns out though, and I know this from being in radio, when you sit down to write out what’s supposed to be conversational, not always easy. I think humor when used well, and surprise, is super important. Emily Withrow, who is also in our bot studio and leads up our writing and editing folks, she likes to talk about how when you get a button or a choice in the chat platform, so you might get a question and you get a couple of possible answers you can just click, to put funny things in those options is actually a great way to engage the audience, engage the individual user, so she likes to put humor and clever things inside the selections.

I think the other thing is to really acknowledge how people are responding. Right? If they respond that, “Do you agree with this? Yes or no?” They say “Yes”, then really acknowledge and address their direct answer. Even if you’re going to keep moving on to something else. I think the last thing I would say is that this is something again, Emily and the bot team here at Quartz has really worked on, there’s a perception out there that the way you make a great bot is to have a very big decision tree where you could choose your own adventure or go any way you want, but we find that that’s not actually the best way to go about it.

We give people choices, but they’re really all the choices are heading toward the same point.

Instead of saying, “Do you want to go into the cave or into the forest,” if you’re at that point, do I want to go in the cave or the forest, maybe I want to do both. Maybe I want to do neither. It creates this problem. It also creates a problem that you can’t easily go back in a conversation, but so what we might do instead is say, “We’re going into the cave. Are you excited or afraid?” Then respond to that. Then, “Okay, now we’re in the cave. It’s dark. What do you want to do?” We might have an option there, but the whole idea is, no, we’re going in this direction. We’re taking you on this journey. You might have some options along the way, but our goal is to get you through that journey. That journey may be a bread recipe. It might be a news experience. It might be an explainer. That’s our trick I think, much more than the choose your own adventure infinite branching. The other thing about that is that if you’re building it, building an infinite branching or a very complicated branching storyline takes a lot more work than building something that’s going to get you to one location. That allows us to generate much more content much more quickly.

 

Yeah, you’ve talked a lot about your team, so can you tell me how essential is the collaboration between production manager, the development team, your designers, in the success of these branded chatbots or mobile apps?

The collaboration of the team is essential. This is not something where you can just say, “Hey, developer, here’s the ticket. Solve this.” Our developers, our designers, our editors, our writers, all work very, very closely to make this a great experience. That’s important because remember in a chat experience, the conversation is the user experience. Right? The user interface is maybe just the chat window, but really the UX design is how do you run this conversation? What are the words? What are the phrases? What’s the humor? What’s the storyline? In that case the designer, whether it’s graphic design or conversation design, the developers, coders, the writers, the editors, we all have to work really, really closely to make that happen. The way we tend to work in the bot studio, it’s a very collaborative system. The developers have just as much say over how things might work as the writers and editors because we’re all building this together.

We also have played with, and had some success around, actual physical events. So we had an event at South by Southwest where a bot actually talked to the people who were at the event, depending on what was happening in the event, over the course of the evening. So with these kinds of experiences, this is the niche that we’ve carved out. And we’re really excited about the possibilities. And we’ve had some really good successes, so far.


What type of conversational interfaces are you working with, specifically? So, like Chatbot, voice or a hybrid of them both? Where are you connecting?

 

Yeah. So the interfaces that we’ve been using mostly, are, in areas where you can either type to or talk to verbally your bot. Or our bots. So in the typing to area, that’s been in Facebook Messenger, in SMS texts. We actually have been doing some development for Slack. So in those kinds of platforms, we’ve been experimenting. There are many, many more platforms on which people communicate by text. Especially in other countries. And, of course, it’s a global guide to the global economy. So we wanna be able to reach out to people. And have them communicate with our bots in many different platforms. So we’re exploring those.

 

Of course, on the voice side, for things you can talk to, we’ve been looking, primarily, at Amazon Alexa and Google Home. But also are very interested in what’s coming out from Apple and Siri.

 

Very cool. Very. So how can companies ease worries with their employee’s around all this innovation?

I think there’s a lot of anxiety around bots and computers taking over people’s jobs. And I think a lot of those are founded concerns. Right? We’ve seen that happen over the years. And we’ve had people here at Quartz visionaries come and talk about it. It’s basically saying that certain jobs that are out there, even the job of a driver, if autonomous cars are out there, the role of a driver becomes obsolete. Those kinds of things, I think, are something that our economy is gonna have to deal with. And our society will have to deal with when those things happen.

I can say that, in terms of the bot building space, so here we are at Quartz. At a media company full of reporters and editors. In those roles, there’s been a lot of talk about, “Well, are bots gonna take over the news business?” And I think what we’re finding is that people are really important for those roles. And, in fact, one of my mantras is that the secret to a good bot is really, really amazing people. And that goes for the people who are building and coding them, for sure. But also, at Quartz, we have an entire team of people who are writing the content for … And the news … For the bots. For the news bots that run on Facebook Messenger. And in the Quartz app.

They are amazing people. They write in a format that is incredibly unique. And they are really, really good at it. And frankly, our bots would not be successful without great writing. And so, we actually have hired some more writers, recently. So, in our particular case, the bot space is creating jobs. And not just coder jobs and product jobs. But also writer jobs and journalism jobs. So that’s pretty exciting. I don’t know if that would be the case for everybody. And I don’t think I can speak more broadly to that. There’s certainly folks out there who are gonna end up having their roles automated. And I guess I don’t have a particular answer to that, except that in our media space, for sure, we’re … Right now, we’re making a couple of jobs here and there.

What’s a new design trend that’s really caught your attention here?

I think the biggest design trend that I’ve seen that has surprised me this year has been in our ability to write stories, in a conversational way, that really loop people in. That is a little bit unexpected. I expected that when we embarked on looking at how bots could play a role in news and information, that it would be more of a call and response thing. You know, “Hey bot. What did the President do today?” Or, “Hey bot.” You know, we already do, “Hey Alexa. What’s the weather today?” More of this kind of call and response. Or maybe even a quiz kind of situation.

What I think has surprised me is that, done right, the threaded conversations that can actually last up to five, ten minutes, are incredibly successful. 90% completion rate we’re seeing on these things. That, to me, is incredibly surprising. And also very exciting. Because it means that we can actually, as long as we do it well, and we do it in a spirit that respects that intimate conversation between us and the audience member. As long as we do that well, we get people really, really attentive to what we’re providing them. And we think that there’s a lot of opportunity there. But that, frankly, was kind of a surprise. This threaded storyline is not something that I expected would see a completion rate that high.

 

That’s still ridiculous and, with that completion rate, I bet there’s a ton of data that you guys can use to improve your bots. So, can you tell me a bit about how you’re doing that?

I think the way we’re using data right now for our bots is really about making the experience the best kind of experience possible for each individual user. And so, we have a global audience. So if we just send a broadcast out to everybody, like a message saying, “Hey. We have something new.” And we send that as a push notification or an SMS text message to everybody at the same time, that’s not gonna be really nice for people who are on the other side of the planet who might be asleep. And then, you have the situation where actually that might have been a perfect time because they’re on the other side of the planet, but they work a night shift. And that’s perfect. And so we can’t really know that. We can’t guess that in the same way you might do for dropping a new story, or a new article.

If we’re gonna have this relationship with somebody, and we’re gonna be pushing stuff to them, we don’t want to make them despise us. We want to engage them and delight them. And not annoy them. In order to do that, we think the best thing to do is to watch how each individual user is using our bot. Especially when. What time of day, and what days make the most sense. And then, really use some machine learning, use some math to figure out when is the best time to let you know about something that we have that’s new? In a way that will delight you, and not annoy you. And if we can get good at that, right now I think that is the best way to use data.


Want more juicy insights from your favorite UX thought leaders? Check out the interview section of our blog for more of the content you love!

Comments

Categories: Interviews

Bianca Nieves

Digital Marketing Manager at Botsociety