Now Playing

Pokémon Pandemonium

The inside story of the biggest out-of-nowhere smash hit of 2016—and what life’s like riding aboard a runaway train.

SLIDESHOW

Employees at Niantic Labs figured Pokémon Go would turn into a popular game—not a cultural phenomenon.

(1 of 5)

Tatsuo Nomura, senior project manager.

(2 of 5)

 

Mike Quigley, chief marketing officer.

(3 of 5)

Archit Bhargava, product marketing manager.

(4 of 5)

Phil Keslin, chief technology officer.

(5 of 5)

 

Editor's Note: This is one of many stories about local influencers, insiders, and rabble rousers that San Francisco is publishing over the next month, all part of the December 2016 Power Issue. To peruse the rest of the issue’s contents, and to read stories as they become available online, click here.


For several weeks
this summer, the international cultural zeitgeist seemed to emanate from an anonymous suite inside an equally anonymous office building off the Embarcadero. It was here, on July 6, that Niantic Labs unleashed Pokémon Go, which almost immediately became one of the top-grossing apps of the year and the agent of frenzied mob scenes worldwide. By combining mobile phones’ geolocation features with “augmented reality,” the game not only took the world by storm; it also heralded the future. And it turned Niantic into the city’s hottest unicorn almost overnight.

The numbers are stupefying: more than 500 million downloads in its first two months; $10 million a day in revenue; $200 million during the game’s first month. Quartz reported that a Niantic memo dated July 19, two weeks after launch, valued the company at $3.65 billion. One back-of-the-envelope calculation by bloggers at Money Nation suggested the game could make as much as $180 billion in the next 10 years.

Niantic couldn’t possibly have foreseen the game’s blockbuster success. Engineers were blindsided when traffic the day of the launch was more than 50 times what they’d forecast. The company, originally launched as a startup inside Google in 2010, had released two previous apps, Field Trip and Ingress, the latter of which pioneered some of the location-based and augmented-reality software later used in Pokémon Go. Both of Niantic’s previous releases had relatively small but devoted followings.

Pokémon Go, of course, was another story altogether. But even now that the initial hysteria has mostly passed, the game remains a hit. Rollouts in South Asia, China, Russia, and Africa are imminent. The company says it has hundreds of new characters to release in subsequent versions of the game. Much is left to monetize.

On a heat-wave afternoon in September, I sat down with a number of Niantic employees to hear firsthand what it’s like to ride the comet tail of a phenomenon that became the runaway pop culture event of 2016. It was the first time the company had invited a reporter to speak with the engineers, designers, and marketers who made it all happen.


The Birth of a Blockbuster
 

Tatsuo Nomura, senior project manager: I was a software engineer on the Google Maps team. I did a couple of April Fools’ projects in the past. The first one was this Google Maps 8-bit version: Like, turn all the Google Maps to the classic Nintendo 8-bit theme. Then, after that, I kind of became the default guy to ask for April Fools’ [projects]. The next year, I did a Google Maps treasure hunt. And in 2014, I came up with this idea to show Pokémon on the Google Maps

Phil Keslin, chief technology officer: It was pretty much right after that April Fools’ joke that we thought the idea sounded pretty interesting—that the whole Pokémon in the real world was kind of a cool thing. And luckily we had the guy who created the April Fools’ joke working at Google.

Nomura: [Niantic CEO] John Hanke reached out to me to ask if I knew about Ingress, and if we could collaborate—Ingress and Pokémon—on a new project. So I introduced Ingress to the Pokémon Company, which is the intellectual property holder of Pokémon [based in Japan]. That’s how the whole conversation started. 

Carlie Fischer-Colbrie, project manager: I was in a lot of the meetings in the beginning. We would fly back and forth to Tokyo. When you go in [to the Pokémon Company], there is a full wall of all the Pokémon plush you can imagine. They have all these special-edition ones that you probably wouldn’t find in the store. There’s this businessman Pikachu that I think is based on the CEO. He’s wearing glasses and has a suit and a little briefcase. 

Mike Quigley, chief marketing officer: We heard—and [this was] one of the ways we started to chat with the Pokémon Company—that their CEO in Tokyo, Mr. Ishihara, was a high-level Ingress player. His wife, Mrs. Ishihara, was an even higher level player. They really appreciated the innovation and the things we were doing with Ingress, although, again, it was a relatively small-use case. 

Archit Bhargava, product marketing manager: My first reaction [after learning of the potential Pokémon partnership] was, “Holy shit.” Pokémon is one of the biggest game franchises on the planet. My generation grew up with it—and I grew up in Mumbai. So my initial reaction was, “Wow, this could really be something.” When we finally heard that everything went well, the deal was signed and we could start development, it was this huge sign of validation for the real-world gaming genre and the platform. 

Dennis Hwang, art director: When I was doing Google Doodles, I had some experience working with other IP [intellectual property] rights holders. Whenever I celebrated the 80th anniversary of Astro Boy or something, I’d have to work with art directors on that side. They were really picky—“It should have five eyelashes,” that level of thing. I was kind of going into this project expecting a similar level of scrutiny. [The Pokémon Company] gave us a lot of freedom, almost a surprising degree, given that this is a 20-plus-year franchise.

Nomura: Before we released it to the public, we did field testing in New Zealand, Australia, and Japan. That was in February. Tens of thousands of people participated. We had this Google Plus community, so the game evolved while getting real feedback from real people.

Hwang: We had a lot of play-testing runs where we just walked around. We probably looked really silly. A bunch of geeky-looking folks looking at the phone in weird ways. I was constantly worried that the iconic red-and-white Pokéball would catch someone’s eye, but, luckily, we were pretty discreet.

Kei Kawai, product management director: We were actually pretty depressed by the brutal feedback from the field testers. One of the big changes we made was the level curves. Initially, we set it to much bigger steps. But people wanted more continuous, kind of positive feedback—if you do something, you get to the next level. So we made the levels far more granular. We designed it in a way that you can actually get to level 20 reasonably fast, for example—so you can see your contributions get rewarded.

Quigley: We were starting to get feedback from the community in terms of what we can do—not just bug smashing, but also the feel of the Pokéball when you make a capture; things like the UI [user interface] of the Pokédex in terms of where things are and how they get categorized. A lot of that feedback was useful. 

Nomura: I was excited but also a little bit scared—like how the public would react to it. I saw the very first version. I’ve been evolving this project from the very beginning. It came a long way from crappy graphics and everything to a polished version.


The Day It All Blew Up

Bhargava:
Pretty much all of us were working through the July 4 holiday weekend, just making sure that all our t’s were crossed and i’s were dotted. When launch day arrived, there was a sense of urgency to get these steps done, but there was a sense of calm. There was this excitement that our baby that we’d been working on for the last several months—couple of years, pretty much—was finally going to see the light of day.

Keslin: The first day, we did a staggered rollout. We didn’t really know how big this was going to be. We launched initially in Australia and New Zealand, and the uptake there was good, probably better than we expected. It really wasn’t enough to make us feel that our brains were fried. Then two days later we launched in the U.S., and it was probably within the first four to six hours that we knew that we had the distinct possibility that this thing could run off the rails.

Kawai: We were in this room. Everybody got together. We pushed the button. Nothing happened for a while because it got propagated to Google’s and Apple’s servers. It was a slow ramp-up when we launched in New Zealand on the first day. It took until the night or the morning of the next day when we opened it up, and we were number one in the App Store. Number one in downloads. We were like, “Wow.” It was an amazing moment.

Keslin: We ran out of server resources. We blew out our quota. We didn’t have enough machines. We didn’t have enough underlying infrastructure support to support the load we were carrying. The estimate that we put together if the trajectory continued—which it actually did—was that we would need orders of magnitude more machines than what we started with.

Luke Stone, director of customer reliability engineering at Google: When Pokémon Go launched in the United States, there was an incredible surge in demand. Actual game play exceeded our most aggressive estimates by more than 10 times. The engineering challenge was to scale the application in real time while new users poured into the game. It’s like changing engines on an airplane while it’s already in the air. To do this, you need engineers working extremely closely together. We were working alongside Niantic engineers and brought in over 20 teams of engineers and product managers to help. For nearly two months, we worked as an embedded unit within Niantic’s engineering team until we were satisfied with the game’s performance levels.

Quigley: I don’t think we’re saying publicly what our internal forecasts were, but we thought it would be a strong-performing app. We certainly thought it would be bigger than Ingress. We thought it would be a strong performer, that’s all I can say. 

John Hanke, CEO and founder: It wasn’t until probably a week later or whenever that it sort of blew up on social media. I remember my wife and her friends would send tweets from Jimmy Fallon talking about it, being referenced on Colbert and this pop culture main stage—that’s when it was clear to me that it had taken on a life of its own.

Bhargava: We did not expect the kind of phenomenon Pokémon Go would become.

Kawai: Sometimes the product that you build, you feel and know that it’s going to be super big, and it doesn’t get big. This one, definitely we knew it was going to be big, but not this big. 

Keslin: I think every day of those first two weeks, up until the Japan launch—I would say that most people on the team felt this, too—we could see this thing just continually going up and up and up, and although we’re working our butts off trying to ensure the system stayed up and running, every time we took a step back, we were all kind of amazed that this thing took off as well as it did.

Fischer-Colbrie: John [Hanke] and I were in Japan [later in the week of Pokémon’s launch]. We were taking really early meetings. And we were sad because it hadn’t launched in Japan yet, so we couldn’t play. The craziest thing was people were sending me photos. I don’t know if you’ve walked around Cupid’s Span [on the Embarcadero], but there are hundreds of people just walking around. If you saw a cord coming out of their pocket, you knew that they were playing and there was an external battery attached. 

Hwang: I think we were all pretty bleary-eyed from working to make the launch. I don’t think any of us could have foreseen the kind of response we got, this instant explosion of the online and offline reaction, and suddenly everybody’s out on the street all at once. I was just kind of joking internally, having internal bets on when the first tattoo was going to show up. I think it showed up within a couple days.

Keslin: I had trouble sleeping for pretty much the first three weeks of launch.

 
The Aftermath—and the Future

Nomura: The thing I remember most was that the afternoon we launched, we saw things going up and the server was stable, so I went home. And on the way, I saw a woman holding her phone like this [sideways], and I knew she was playing it. That’s the first time I saw someone play the game I worked on for almost two years. That was the happiest moment. One block away from here.

Keslin: My first encounter was in my own neighborhood in San Jose. That was the first time in a long time that I saw packs of teenagers, from 15 to 20 years old, running around the neighborhood playing Pokémon Go. The second was the San Francisco crawl [on July 20], when like 9,000 people showed up. That was pretty phenomenal. I didn’t go. I was trying to keep the servers from burning down. We were terrified that day because the server went down about 2:30 in the afternoon, and the crawl was supposed to start at 6 p.m. We’d heard through the grapevine—total rumor mill—that if we didn’t get the servers back up in time for the crawl, they were going to take all 9,000 people and march on our office. So we were frantically trying to get the servers back up.

Fischer-Colbrie: We weren’t expecting that kind of attention, so we had to go incognito. That’s why we have blackout screens on the office windows. Niantic’s name is not on the [directory] board. We’ve been here for almost a year. We still had people coming to our door. They were just really excited and wanted to meet Niantic.

Hwang: After we launched, I started getting text messages from my friends in Korea, like, “Hey, congrats on launching Pokémon Go. My coworkers and I are going to drive to Sokcho City to play.” First, I was really confused because I didn’t know what Pokémon had to do with Sokcho City, which is this tiny city in the northeast of South Korea. So I did a Google News search, and of course there are countless articles in the Korean media about how there was apparently this mass migration of people taking long-distance bus rides to Sokcho City, really in this corner of the country, close to the DMZ. The whole city was lit up with Pokémon signs, and restaurants were like, “Hey, come on in, we’re having Pokémon discounts.” So I texted one of our engineering managers, and I said, “Did you do anything with the Korean map distribution?” And he said, “Oh yeah, I had to block [access to the game for] a good chunk of the Asia region for various technical reasons.” But of course the portion that he [didn’t] block out was this tiny corner of South Korea. So people figured out that you can’t play Pokémon Go except for in that city. The bus tickets apparently sold out. It was just hilarious. There were Pokémon-themed tour packages. To this day, everyone in Korea is calling that city the Pokémon Mecca.

Bhargava: There was one email that the GM sent that just said, “Hey guys, we’re all barely sleeping, we’re extremely tired. I don’t know about all of you, but personally I’m having the best time of my professional career right now.” And everyone replied back and said, “Amen.”

Kawai: We hadn’t even celebrated properly yet. We finally started talking about it—I think it was the first Friday. We have this ritual TGIF, where everybody comes to the same room and talks about things. I think it was the first weekend that we finally found some time to pull ourselves together and, one by one, talk about how amazing it was. Somebody talked about their family. I talked to my mom, who installed it and tried to play. It was not a big celebration. It was more like personal moments shared, and that was—I liked it. It feels more Niantic to me.

Hanke: People are excited about ways to enhance the game. The Pokémon universe is obviously very large in terms of the number of Pokémon that exist. We launched our first 150 or so. The second generation of Pokémon is something people are interested in seeing. I think that’s going to be a big event when they show up in the game. The second thing is events: We’ve been running these events with Ingress for the past several years in cities around the world. We get crowds of 10,000 for an all-day gaming event. So we’re looking at staging similar events for Pokémon Go. And there are features like trading and head-to-head battling that we’re looking at. There are certainly plenty of things within the game itself, in terms of game mechanics, that we’re working on, and players can expect to see some of that coming next year.

Quigley: Not to sound arrogant, [but] we truly believe in what we’re doing. We see the impact it has on people’s lives. We see the genuine fun that people have. And yeah, all right, winter’s coming to the northern hemisphere, school’s back in session—there’s going to be those things that will always compete for people’s discretionary time and income, but we feel pretty confident that as long as we’re continuing to add new features and respond to issues, we have a real shot to become a daily habit in people’s lives. If it’s in a fun way, maybe while you’re getting a little exercise along the way, that’s a good thing.

Hanke: Whenever someone used to ask what I did, there was this long explanation. But now it’s pretty short: Pokémon Go.

 

Originally published in the December issue of San Francisco

Have feedback? Email us at letterssf@sanfranmag.com
Follow us on Twitter @sanfranmag
Follow Jeremy Lybarger on Twitter @jeremylybarger