How the Sequel game character bot platform will drive believable conversations with players

When Facebook unveiled its chat bot platform for Messenger in April, Omar Siddiqui felt like a new industry was born. His company, Kiwi, adapted a platform called Sequel to create conversation game character bots to interact with game characters.

The San Francisco company made its platform available to other developers, and thousands are now playing around with how to make characters come alive and appear more immersive to players. This is just the latest character-and-story focus for Kiwi, which has had five titles in the top-25 grossing mobile games in the last five years. Before cofounding Kiwi in 2011, Siddiqui was the vice president of product at Playdom.

He hopes to awaken a new platform for gaming with Sequel. Siddiqui is one of our newest speakers at GamesBeat 2016. He’ll speak in our “lightning round” talks at our event at the Terranea Resort in Los Angeles.

You can register for the event here. Check out the new agenda here.

I caught up with Siddiqui at our recent MobileBeat 2016 conference. Here’s an edited transcript of our interview.

GamesBeat: What year did the company start?

Omar Siddiqui: The company is Kiwi, but the product is Sequel. The company was founded in 2011 and we started working on Sequel about a year ago. As I mentioned, we came to Sequel with a background in story-based gaming. The question we asked initially was, “In a world where you can build entertainment experiences in a conversational format, what matters?” For us it was immersive character design and the ability to connect people emotionally to personas. That was really interesting.

GamesBeat: Did you go directly from Playdom?

Siddiqui: We went from Playdom to Kiwi and we’ve been in this venture since, yes. The team is pretty much intact. Our studio team more or less started Kiwi.

The thing we had become enamored with and interested in was interactive storytelling. That was the category of games we’d enjoyed making. When we anticipated and saw that messaging platforms were going to start opening up, the opportunity to create compelling characters for this new format was what was interesting to us.

GamesBeat: That was the craze around Kakao and Line and so on?

Siddiqui: We were one of the first developers on Kakao as well. We’d been familiar with the Asian messaging apps as developers. We knew that dynamics in the U.S. would be unique and different, but it was inevitable that you’d be able to create conversational experiences here.

Now, conversational — let’s not fool ourselves. It’s still a combination of UI and text. When I say
conversational experience,” it’s happening in the context of a conversation. But you still need to use both UI and text to design well-crafted experiences. And so we started developing Sequel as a platform to allow any creator to be able to come in and craft a compelling conversational experience.

That was the endeavor we launched about a year ago. It suited our creative interests, but it also took advantage of a lot of infrastructure and thinking we’ve done through the years. Guards of Time, for example, had tens of millions of MAUs. We were pushing content three times a week. The content was a really complex interactive soap opera. We had to have a platform that would allow game designers and writers and artists to work on it, to craft those experiences, and push them live without having to talk to a bunch of engineers. Engineers don’t want to talk to them about all that stuff either.

That was the infrastructure and technology we already had. Cut to the present. Sequel today — we have thousands of creators and developers building bots on it. You’re able to create an interactive bot through a visual GUI that makes it easy to use without giving up capabilities or technology. You can author and then publish to Facebook, Kik, and Telegram. We’re adding support for other platforms. We were a launch partner with Kik as a part of their bot platform launch.

On Kik we launched initially with a celebrity trivia quiz bot, in partnership with J14, which is a leading teen magazine in the U.S. It was like a game-ey trivia experience. We also launched Sequel Stories, which is a collection of interactive stories people can play through, like Choose Your Own Adventure.

Above: Omar Siddiqui is working on a game bots platform, Sequel.

Image Credit: Kiwi

GamesBeat: With each book, you were authoring it, or they were authoring it?

Siddiqui: We have third parties who have authored it. Again, our interest — we’ll build showcase bots to help with strategic partners and keep pushing the medium. But our goal is to facilitate the emergence of compelling bots by breaking down the barriers that anyone would have in building and launching a significant experience. That’s our goal.

GamesBeat: I wonder how similar it might be to something like Episode, from Pocket Gems, their storytelling platform.

Siddiqui: I’m not intimately familiar with it. Storytelling is definitely one aspect of what we do. We have narrative story bots, Choose Your Own Adventure story bots. We have trivia bots, quiz games. With Now Your See Me we have a puzzle adventure game. That was in partnership with Lion’s Gate, to support their movie and tie in with the movie launch. We’ve built a bot with Awesomeness TV for Kik. Awesomeness wants to connect their personalities and their video content in an effective way in this medium.

For us the platform allows media owners, game designers, anyone who wants to connect emotionally through bots, to author and publish that experience easily. Now we’re working on social media and individual personalities as a new segment that we’re hoping to figure out and drive.

GamesBeat: The intersection of bots and what you guys were doing, games and bots, how early did that come for you?

Siddiqui: We actually started developing this last June. We developed our own client app, because there was no way to plug these interactive experiences into anything else. It went into private beta in January with our platform plugged into our client app. In March Kik launched their bot platform, and then in April Facebook launched theirs. It was timely for us because we’d been working on the problem in terms of providing a platform that would allow creating these types of experiences.

We’d previously been anticipating plugging away on our own, figuring out the medium, in anticipation of when there would be a broader opportunity to distribute through these platforms. Frankly, we got lucky that the platforms started opening up at the time they did. It’s helped catalyze a lot of interest.

Now, the delta between interest and reality, like any other platform launch — if you went through Facebook’s original app platform launch, through iOS and Android, there’s an accelerated hype cycle, then disenchantment and the trough of sorrow, followed by actual experiences. People are forgetting that this is all three to four months old. In 2008 nothing was figured out on mobile. Three or four months in we were still doing food fights and hugs and sending viral spam at each other on Facebook. So the medium is still emerging.

In all the games we’ve ever build, real magic only happens when you can connect the designers directly with the act of creation. A creative process is tweaking. It’s building something, adjusting it, seeing how it feels, adjusting it again. As soon as you intermediate that, it’s hard to craft excellent stuff. Similarly, for conversational experiences, we want to provide that level of access for creatives to build stuff. That’s our version 1.0. Now the pieces we’re adding to it are, where can AI and the latest techniques in auto-generation and machine learning aid you in bringing down the required investment to still build a compelling conversational experience?

That’s how the promise will be delivered on. Not by thing that, out of the box, you’ll have some fully automated experience that suddenly solves everyone’s problems.

Above: Kiwi has had five top-grossing games.

Image Credit: Kiwi

GamesBeat: Were there some things that helped along the way, helped you to crystallize what you wanted to do? Like the experience of Talking Tom and the like.

Siddiqui: It’s true. The other thing that people are not fully appreciating yet is that we view ourselves — we’ve reluctantly added bots as our moniker. Otherwise, our tagline is, “Build personas, not bots.” When you build personas that can help you and that connect with you and help you connect with things you care about — when you have that mentality you realize that this is way bigger, as a macro trend, than just text messaging-based experience.

Talking Tom is an interesting example of that. One can argue that Talking Tom was an interactive video bot. It is, because you can interact with it. It responds. It gives you feedback. The broad-based attraction people have had to that speaks to something really core that I think is going to happen now. If you could take Talking Tom and make it rich in how you interact with, suddenly you’re talking about video bots. You’re talking about audio bots, even. These are experiences that have yet to be defined and built.

For the next three to five years, bringing automated conversations to wherever people want to have them is going to be a really interesting problem. That’s what we’re excited about helping to facilitate, but from the point of view of media and entertainment and content.

Games are obviously an example. But even news is an example of that. I wouldn’t think that a CNN bot is as interesting an endeavor as an Anderson Cooper bot. I don’t connect to CNN, but I connect to Anderson Cooper. If he was conveying the news to me with his personality and his spin and his background and I felt like I was connecting to him, that’s differential value in a conversational format. I can get CNN news on the web. I don’t need a bot for CNN’s voice.

GamesBeat: It sounds like that part of the challenge is the immersive part. You have to feel like you’re in a game or a game world, rather than just doing something on top of Facebook chat. It seems like the challenge for the original Facebook desktop. At some point you realize you’re playing FarmVille on top of an extra piece of software.

Siddiqui: It’s true. To be honest, though, it took time. But people were open to that. Facebook launched their platform in July of 2007. I think FarmVille launched in October or November of 2009. More than two years before FarmVille emerged. You’re going to see accelerated timelines in the evolution of all these platforms, but still, thinking that we understand what’s going to be compelling here just three months in is premature. We have a lot of experimentation left.

Above: Omar Siddiqui envisions a day when lots of game character bots hold conversations with players.

Image Credit: Kiwi

GamesBeat: You guys have been quick in this space. Is that important, like it was important to be Zynga back in 2008 or so?

Siddiqui: I’ve participated in ecosystems as a content creator in the past. I’ve tried to build franchises and products that could stand on their own as examples. Here, we’re trying to facilitate other people’s success through our platform. It’s a slightly different vantage point. On the one hand, it can be frustrating as a content creator when you’re like, “Come on, do it this way, that’ll be more compelling.”

Being quick here, I think there’s a danger. One of the dangers I’m seeing from people who are wading into it is that they think because there’s no UI, this is the promised land. It’s so easy to build a compelling experience here because it’s just conversation. The reality is that you’re stripped away from a lot of the user aids that you have in a traditional experience, to where actually crafting a compelling conversational experience is even harder. You have even less to work with. The scarcity of immersion makes it even more important that you nail what you do and you do it well.

You can be quick and launch and launch a lot of stuff here, but the meaningful stuff that connects with people will require real iteration and real work. I don’t think there are any shortcuts. There was a lot more of a land grab mentality when there was massive virality, too. Here, discovery is the bottleneck.

GamesBeat: Microsoft already proved that it’s a treacherous area as well. A lot of people find their entertainment in twisting a bot to their own purposes.

Siddiqui: Right. You do something quick, you end up with something like that. For Now You See Me, we experimented with a puzzle game brought to this format. It’s still character-driven. It’s not NLP-based. We used the native button format. We were comfortable making it more game-ey. You can succeed or fail. If you fail there’s a fail state and you start over. There are visual puzzles layered in. We’re seeing a ton of engagement from people playing it. It’s definitely an interesting experiment. Those kinds of experiments will lead to figuring out what people really want here.

The thing I’m seeing as a common theme across all the experiences that pick up a critical mass and have real user engagement is that they’re connecting people to things that they care about. They’re doing it in a way that has personality. Again, I’m not privy to the utility-based side. I’m talking about things people want to engage with from a media and entertainment perspective.

GamesBeat: You guys are kind of like a platform on top of a platform.

Siddiqui: We’re a conversational platform on top of — there’s a many to many opportunity here. Yes, Facebook matters. Yes, Kik matters. Yes, Telegram matters. But over time, we believe these conversations are going to happen on multiple platforms in multiple ways. We just want to keep our eye on the ball of, how do we make it easy to create and distribute compelling conversation media?

GamesBeat: What’s the readiness level now?

Siddiqui: We’re public. We officially launched in June. We’re obviously layering in updates and functionality all the time. In June we launched a native natural language capability as well. You can build experiences that are either button-based or natural language-based.

We’re now layering in additional functionality like bringing in web services to inform your bot through the same visual interface. You can have something like a recipe bot that’s helping teach you to cook, but you can integrate a web service to bring down recipes from the web and deliver them to people as part of the conversational experience. Just because it’s entertainment-based doesn’t mean you can’t execute commerce or how-to with this approach as well.

GamesBeat: How many employees are you at now?

Above: One of Kiwi’s games

Image Credit: Kiwi

Siddiqui: About 30 people.

GamesBeat: If somebody comes to you saying, “Hey, I have an idea for a game,” how do you direct them?

Siddiqui: They can log in and make it. They can use our platform to build it. Obviously right now we’re being as supportive as we can, because our platform and what we do is defined by what people build on top of it.

GamesBeat: How many people are coming in?

Siddiqui: We’re getting registrations at a pretty decent clip, people playing with the bots. We have thousands of developers at this point. But like any other medium that people are playing around with, there’s a drop-off between playing with it and actually polishing and launching quality executions.

GamesBeat: How can they start making money, too?

Siddiqui: That’s to be determined. None of the platforms have yet exposed monetization hooks. It’s inevitable, though. Everyone’s going to open up monetization hooks. We have a going concern ourselves. It continues to be the same venture.

We’re focused on media, entertainment, content. That’s the order in which we’ve taken it down. We’re now attacking some interesting ideas related to individual personalities — celebrities, bloggers, and so on. Anything with an authentic voice plays well in this format. Especially for games and game developers, there’s a thing here as well. At this point, if you’re trying to build a game as a mobile app, it’s over. Unless you’re Pokemon.