When Apple opened up its phone to outside developers—it’s easy to forget that “there’s an app for that” wasn’t part of the original iPhone pitch—Natalia Luckyanova and Keith Shepherd saw an opportunity to try something new. The husband-and-wife team designed their first video game—Imangi, a word-based puzzler—and had it ready in time for the App Store’s unveiling. This wasn’t a mere lark. Imangi Studios added an artist to the team, Kiril Tchangov, and kept producing iPhone games. Their first big success came in 2009 with the frenzied sea-traffic-control game Harbor Master.

Harbor Master qualified as a hit for Imangi, at least until Temple Run. The 2011 game takes the Indiana Jones golden-idol scene and stretches it to infinity. Forever sprinting, you turn tight corners and bound over flaming obstacles until you make a wrong move and die. After the game debuted to relatively slow sales, Imangi made Temple Run free (with in-game extras available for purchase), and suddenly the three-person studio was at the helm of a phenomenon: The game has been downloaded more than 70 million times. Luckyanova and Shepherd spoke to The Gameological Society about their modest expectations for Temple Run, internal debates over evil monkeys, and the decision to ignore pre-release feedback that would have changed the essence of the game.

The Gameological Society: Keith, you’ve said in interviews that you’d wanted to make games all your life, but until the iPhone’s App Store was launched, you’d never produced a game for the public. In 2008, though, on day one of the App Store, there you were. So obviously, there was something about Apple’s setup that made you guys finally take the plunge.

Keith Shepherd: I’ve always been an early adopter of Apple stuff. So I had an iPhone the first day it came out, and when they announced that you could make apps for it I was like, “Wow, this is really cool. Let’s try to make something.” Natalia and I came up with this idea for a little word puzzle [Imangi].

What was really so cool about Apple’s ecosystem was that it seems like a very approachable way to distribute games. Before, everything was online, and maybe you could make Flash games and sell them, or you could make PC games. But these things were really big in scope, a PC game or an Xbox game. It takes a huge team years and years and years to make something awesome. And with the web, it’s really hard to monetize things. Everybody’s expecting everything to be free, and you try to make it ad-supported. So this was a neat opportunity where I was like, “Oh awesome! You could make something and put it out there, and you could buy it as easily as you can buy a song on iTunes.”

Luckyanova and Shepherd's first game was Imangi, a word game that they developed for the launch of the iPhone App Store.

Natalia Luckyanova: I think we also got lucky because it was really the right time in our lives. We had just gotten married, and we were talking a lot about starting a business or doing something together outside of the traditional workforce. And then the App Store came along. We both had thought about starting a business before, but it’s much scarier to just quit your job and do something when you’re one person supporting yourself. With the two of us, you split up the responsibilities, and Keith was the one who quit his job and jumped in while I stayed working and supported us. So it was not quite as scary that way.

Gameological: Do you still have your day job, Natalia?

Luckyanova: Oh no. We’ve been full time for about two years now.

Shepherd: I think three years, actually. Yeah, three years. Natalia started full time about eight or nine months into Imangi’s existence.

Gameological: Which came first for Temple Run—the Indiana Jones-type theme or the infinite-sprint format?

Shepherd: We actually started with the infinite-sprint format. This is kind of how we develop all our games. We start with a nugget of gameplay, or a mechanic, or some idea related to the game before we ever really start thinking about, “What’s the theme of this going to be, and what’s the art going to look like?” The hardest thing about making a game is getting something that’s fun. There’s not really a formula or a set of rules you can follow to make something fun. You have to find the fun in every project by experimenting with different approaches and using your gut as a guide for, “Oh yeah, this is fun.”

We started Temple Run as kind of an offshoot of our previous game, Max Adventure, which was a completely different kind of game. We were trying to experiment with different ways to control a character in a three-dimensional environment. This was a big weak spot in our last game—we used these onscreen joystick kind of controls. And it wasn’t a great fit for the touchscreen. So we started experimenting with different ways to control it, and ultimately it led to this [in Temple Run]—the characters are always walking forward, like in every endless running game you have, but we’re in 3D, and we’re using these swipe gestures that would turn the character 90 degrees at a time.

Temple Run

That’s where everything from Temple Run started to fall in place. It’s like, “Okay. The character’s running. We can swipe left and right to turn him. In what kind of environment would this make sense? What would we be able to use this control scheme in? So we started thinking about this, like, endless maze. As we started building it out, using very rough shapes of blocks to make this maze that you’re walking on. It just really started looking more and more like a temple wall. The Great Wall of China was actually a big inspiration for that path. So we ran with that, and the theme had a link in our mind to a jungle environment. That was where the theme and the art started to come into play, once we had the basics of the game down.

Gameological: It reminds me of some of those tracks from the Super Nintendo Mario Kart game.

Luckyanova: We’re definitely big fans of Mario Kart. I don’t think anyone’s ever compared Temple Run to Mario Kart because it has those sharp turns which would be hard to do with a kart.

Shepherd: Maybe the older Mario Kart games on the Super NES where’s there’s less high-fidelity tracks and curves.

Gameological: How did your first working prototypes differ from the product that you released?

Shepherd: It differed in very subtle ways but very impactful ways. Originally, the character couldn’t tilt left and right on the track. But when everything was on that center line, it wasn’t challenging enough to pick up the coins. Turning [around corners] was challenging, but collecting coins just wasn’t challenging at all. And when we first started adding in the tilt mechanic, that made the coins more interesting because you can have coins on the left side or on the right side.

It was a challenge because, well, the game before was all about the swiping. And when you threw in that tilt, too, it complicated the controls a bit. We really worried that it would be too difficult. You’d see it often happen that somebody would be running, and they’d come up to a turn and if you had both tilt and the swiping enabled, often they’d try to tilt to go around the turn instead of swiping. And then they’d be like, “Oh! I need to swipe.” They’d remember it, and they’d know they messed up, but we were really worried that it’d be too complicated from a control standpoint.

We got a lot of that feedback from our first beta testing that it was a little too difficult or confusing to control. But we decided that that was just the way the game is, and that was what made it unique. We went with it. It turned out to be a gamble that paid off in a big way. It adds a lot of uniqueness to the game. And people got it. Like, people got over that initial hump, and once you get the hang of the controls, people love that.

We got a lot of that feedback from our first beta testing that it was a little too difficult.

Luckyanova: The other thing that we wrestled with quite a bit was the evil monkeys. That was a big point of contention. It’s a running game, and there were lots of running games in the genre. But we were always thinking, “Why are you running? Why can’t you just stop? Slow down and walk? What is this that’s keeping you running into walls and trees and everything.” So we added that chase element. But we had a lot of disagreement on, you know, is it good to have that extra thrill of adrenaline? Or are the monkeys just really terrifying, and you want to put the phone away because they’re so scary? I kept saying that the monkeys were way too scary. But fortunately, Keith and [artist] Kiril [Tchangov] stuck with it.

Gameological: You were against the evil monkeys!

Luckyanova: I was. I was a monkey denier.

Shepherd: This is a theme, too, that really surprised us. Because if you look at a lot of the casual games that are out there that have been super successful, the Angry Birds, and Fruit Ninja, and Doodle Jump—all of them have this very casual air to them. Bright, colorful graphics. Very, very kid-friendly stuff. Temple Run is not really like that. It’s dark and dreary looking. It’s got these scary, evil monkeys. Even our icon is kind of terrifying. Before we launched it, we were super-concerned about that. We thought it was cool, but we were like, “There’s no way this is going to have mainstream appeal.”

Gameological: So when you were making this, you weren’t saying, “Okay, this is going to be our smash hit.”

Shepherd: [Laughs.] You always hope that happens, and you always think that you have something cool when you’re working on it. You hope that it’ll do great, be a commercial success. But Temple Run has surpassed our wildest imagination as far as success and how popular it’s gotten.

Luckyanova: Also, you know, having been making games for four years, you realize that making a huge hit is close to impossible. So you try to look at it more realistically.

Gameological: It’s interesting to hear about this notion of tilting to move and the refinements that came out of that. Your beta testers pushed back, but you made a gut call, and to me, that is an important part of the story. Sometimes you do have to make those gutsy calls, I would think.

Shepherd: Yeah, and I think you do have to listen to your gut a lot when doing creative things. I think our previous game, Max Adventure—we were trying to make a game for someone else. We were trying to make a game for a target audience. We were like, “Okay, we want to make a kid’s game for the gamer kids who like more hardcore games.” We were trying to make a game that we wouldn’t necessarily have made for ourselves.

With Temple Run, I think we decided, “You know what? We wanna make a game for ourselves.” So we really focused on making the kind of game that we want to play. We gave Kiril a lot of flexibility on the look for Temple Run. He’s always been—from an artistic standpoint, he always does kind of darker, scarier stuff than I think Natalia would play, but we gave him more creative freedom in that direction. We’re making a game that is more personal to us instead of trying to make it for someone else.

Little Red Sled

Gameological: One of your earlier games, Little Red Sled, has some similarities to Temple Run. You’re sliding down this mountain on a sled that can’t be stopped; you’re tilting left and right. Did you take some lessons from Little Red Sled in making Temple Run?

Shepherd: Yeah, a lot of our technology over the years stems from Little Red Sled. Little Red Sled was our first 3D game and we built a lot of technology—a 3D engine and all of that. Over time, that’s been the basis of all of our games, including Temple Run, from a technology standpoint.

The tilting is always something that’s been really compelling about mobile devices. Just that form factor in your hand and being able to tilt it. It feels like when you were a kid playing video games and would kind of yank the controller to get it to go in the direction, but back then it didn’t do anything. Now it does.

Gameological: How do you two divide the labor? Do you both work on the same stuff, or do you have different skill sets that complement each other?

Luckyanova: We have similar skill sets, but they’re slightly complementary. We’re both programmers. Keith works a little more on the back end and on the engine stuff. He wrote the whole 3D engine. And I’m working more on the front end. Kiril is the one who does all of the artwork and makes everything look pretty. I also do the music and a lot of the sound effects. And then we both do the business stuff.

Shepherd: As a small team, you have to wear a lot of different hats. We identified pretty early on that art just wasn’t something that either one of us was good at, and we brought Kiril on board to help with that. He’s done a tremendous amount of great work for us. I think Natalia and I are fortunate in that we do have the technical skills—Natalia having the music skills, and I’ve got some sound engineering and music background as well. We have some business savvy. Being able to have some experience with all those things helps with that a lot, in that we’re not just, “I’m a programmer, and that’s all I can do.” It really is a perfect job for Natalia and I because we both have very wide interests and skill sets. I don’t think either one of us was ever really one to specialize at our previous jobs. We were always more of the generalist kind of person. I think that really helps us in a small games studio.

The Imangi team, from left: artist Kiril Tchangov, Luckyanova, and Shepherd.

Gameological: Yeah, because if you were developers at a large studio, you might be working on just a single aspect of the game, and you guys would find that boring.

Shepherd: Exactly. That was why I didn’t get into the game industry. I knew I always liked games and that’s what got me into computer science—wanting to understand how to make games is why I learned how to program when I was young. But by the time I graduated from college, the game industry was really different than it is right now. It was a place where you had to go work for a big studio, and you had to go work on a console game that took three years with 200 people working on it. And the roles in that kind of environment are very specialized.

Gameological: Natalia, had you been interested in getting into the game industry for a while as well?

Luckyanova: It was definitely a surprise to me to end up in games, but it’s actually a perfect job for me, because it is so varied and so multi-disciplinary. Obviously, I work on writing the code, but also there’s the creative side—creating a world and a story. I’m really happy that I’m doing this. It’s games, so it’s satisfying that the things that we produce—they make people happy. They put smiles on people’s faces. There’s not a lot of jobs where you get to say that.

Gameological: Temple Run is by far the most successful game you’ve made in terms of sales. But putting the financial success aside, is it your favorite game that you’ve made?

Luckyanova: That’s a good question. I think it is my favorite game.

Shepherd: I think it’s my favorite, too. This goes back to the “following your gut” thing. When you’re making a game, you end up playing it a lot to test out new features. I think Temple Run is the game where I spent the most time just—I find myself getting lost in playing it, and having a good time, even when it wasn’t finished. That says a lot, if you’re playing with something every day for six months while you’re making it, and you still think it’s fun.

Luckyanova: I would find myself being like, “All right, I’m gonna just test a couple things.” And then 20 minutes have gone by, and I’m still playing, and I don’t need to be playing it anymore, but I am.

Harbor Master

Gameological: I get coin-obsessed when I play. But that’s my approach—do you see different play styles?

Shepherd: That’s a good question. I’ve heard of people that go back, and they’ll delete their game [save file]. They’ll start again from scratch, and they’ll power up certain aspects of their players. So when they’re playing, they’ll only power up the boost thing. And then when they get the power up they’ll only get the boost. People strategize in that way once they’ve played through. Like, people love the coin magnet. If you have all the power ups leveled up, you get proportionally less of the coin magnet [during a run]. So people will go back and be like, “Okay, I’m not going to [level up] anything except for the coin magnet.” You’re like super coin-magnet man all the time.

It’s kind of funny to see, when you have a product that’s out there and being played by 70 million people, they find every little way you can possibly imagine to take advantage of the system in your game. All these little things that you just never really thought about. They’re little flaws in the design that people figure out how to take advantage of, which is cool.

Gameological: It’s like “min-maxing” a role-playing game, except with Temple Run.

Shepherd: Oh yeah, totally. You know, we don’t publish anywhere exactly how the scoring system works or how many points every little discrete thing is. There’s a Wikipedia page, and people have postulated, “Oh, I think this is how many points that thing gets. This is how the scoring algorithm works.” They’re really, from the outside, tearing apart and trying to piece everything together to figure out how it works.

Gameological: I saw a talk at the Game Developers Conference this year by another iPhone developer who mentioned the social-media aspect of Temple Run’s success. He said that because you didn’t include all these Facebook buttons or a bunch of social-media gadgets in Temple Run, people ended up manually taking screenshots of their high scores and posting them to their Facebook page. He said that it felt more original and authentic that way—that you had taken this ingenious “less is more” approach to social media. What do you think of that theory?

I don’t think we have the goal of, “Can we make something that’s even bigger than Temple Run?”

Shepherd: Yeah, that was totally our theory! [Laughs.] No, maybe there’s some truth to it, in retrospect. The screenshot thing is pretty interesting. You’re right, we don’t have screenshot functionality in the game. So people actually go out of their way to take a screenshot and post it. That’s a testament to how proud the people are of their scores and how much they enjoy the game. I think when people are doing that in an organic way, it does lend some authenticity to it.

Social media has been really a big part of what’s made Temple Run successful. People talk about it a lot, and they tweet about it. They post on Facebook about it. Just in person, people talk it about it too. That’s how Temple Run has spread. It being free, there’s such a low barrier for anyone to download it. We actually spent almost no money on marketing Temple Run. Really, that’s just how strong the word of mouth is. If you look a lot of games and a lot of traditional freemium games, a lot of companies spend a ton of money on marketing their games. They have ads, they have user acquisition, they have all these fancy buzzwords and acronyms for all the stats that they can measure about their users and how they can optimize that. We just don’t have that. We don’t have to get programs that analyze users to death like that.

Gameological: It seems like you guys have a pretty healthy sense of how hard it is to have a big App Store best seller. Do you feel any pressure now to follow up Temple Run with another big hit?

Shepherd: [Laughs.] Yes, you know, there is always that pressure. Harbor Master was kind of our first hit. We definitely felt some pressure then. We had this hit, and that was what sustained our company for a few years, and we’re like, “Well, is our next game going to be as good?” You do think about that. You just can’t let that be the thing that holds you back from doing more. After Harbor Master we made, let’s see—I think we made three games after Harbor Master, and then we made Temple Run. Not too many people know about those three games in between. None of them were very successful. Every one of those was important to us, and we learned a lot from those individually that helped us make Temple Run.

It feels a little bit better now that it’s like, “Oh, we made Temple Run. We’ve now had two successful products.” At least we’re not a one-trick pony. But yeah, Temple Run, this is a once-in-a-lifetime kind of hit. I think it would be tremendously hard for us, no matter what we do next, for it to achieve that level of success.

Luckyanova: Yeah, so it’s two-sided. There is some pressure to produce something else that’s awesome, but at the same time it’s almost like there isn’t, because Temple Run is just such a ridiculous pop-culture phenomenon at this point that to expect we’d be able to achieve that [again] is completely unrealistic. So we would be happy to continue making games and continue making things that people enjoy, but I don’t think we have the goal of, “Can we make something that’s even bigger than Temple Run?”

Shepherd: In a lot of ways, it’s opened up a lot more freedom for us to pursue whatever we want. I think we can continue to make games if that’s what we wanna do, or, we’ve got a lot more in store for Temple Run, still. It’s become such a phenomenon that it’s what we’re spending all our time on. More Temple Run. We’ve expanded out beyond mobile games, and we’re getting Temple Run on other platforms now. It’s on Android, and we’re going to be launching on the Amazon Android market soon. We’re hoping to get it on the Mac App Store and Facebook, and pretty much every platform we can. We want it to be available everywhere.

We’re also working on a bunch of merchandising and licensing stuff. I think everyone’s seen what Angry Birds has done with that. You can hardly leave your house without seeing Angry Birds merchandise somewhere. So when Temple Run got so successful early on, we started getting contacted by a ton of groups that wanted to license Temple Run to make T-shirts, or plush toys, or board games, or whatever. That’s been one of the crazy things. When you have a pop culture phenomenon like this, people come out of the woodwork wanting to talk to you. This summer we should see some T-shirts and apparel starting to pop up in malls around the country. We signed a deal to do a board game and a card game for Christmas. We’re working on all sorts of these other things that we never imagined we’d be working on.

Gameological: Are you designing the board game and the card game?

Shepherd: What’s interesting about all this licensing stuff—we learn a lot as we go—you can be as involved as you want. We’ve been pretty involved with the folks that we’ve been working with so far. I think they do the bulk of the work, but we provide a lot of our creative input for everything, and everything goes through our approval before it ever gets sent out for production. We were doing some prototyping with them early on and playing it over video chat with them, just walking through the steps of how the mechanics of the game work. That was pretty cool because we’ve never made a board game before.