Trust Games

For most companies, succeeding in the mobile game market is like catching lightning in a bottle. It takes the right game at the right time with the right user acquisition strategy. It's hard enough to do properly once, much less repeatedly. But there are some companies that make it look at least a little bit easier. Supercell, for example, has released four games--Clash of Clans, Hay Day, Boom Beach, and Clash Royale--all of them hits.

Supercell's Timur Haussila, lead developer on Hay Day and Boom Beach when those games launched, offered some insight into how the studio has managed to churn out nothing but successful games in his keynote address this morning at the Montreal International Games Summit. While one might expect a company with such a consistent track record to insist on a tightly controlled and predictable development methodology, Haussila said Supercell's approach is anything but.

Before getting into Supercell's model, Haussila recapped a bit of the more traditional industry model, which he saw first hand when he began his career with Digital Chocolate's Helsinki studio as a product manager on Facebook games just seven years ago. Like many other companies, Digital Chocolate had teams of developers that answered to management that answered to more management and so on all the way up to the top of the company. There was a decision-making and greenlight process that took the big calls out of the hands of the game team and pushed them further up the management chain. Periodically, there would be product reviews where those same higher-ups would determine the fate of the game.

"It all starts from trust. If the team's working independently, it can't work without trust"

Haussila said the system worked to some extent, but it had issues. Before long, the goal became not to make the best game for the audience, but to get the product through the review process. It also led to a less innovative approach to game design that was ill-suited for a free-to-play world. When consumers have nearly unlimited choices before them, simply cloning the same style of game without any new angles didn't work as well as it might have in the more traditional console or PC industry.

There were moments of respite, Haussila said, where he was on a small team without much oversight. But that was more a matter of nobody paying attention to what the team was doing rather than any great corporate revelation about developer autonomy. As soon as his project began returning promising results, the old controls came back in. Haussila was discouraged enough with this turn of events that he sought new opportunities, and wound up with Supercell.

At Supercell, Haussila said the decision-making hierarchy is flipped with what he called a cell structure. Each team of developers is an autonomous cell, typically with no more than 15 people. The developers conceive game ideas and act on them, and the rest of the company exists to serve those teams, hiring the best people around them and taking care of logistical issues. Within each cell, Haussila said there is a lead, but responsibility is shared among the team. It can take some work to get the model working, but once things click, it begins to feed back into itself and create better results.

"It all starts from trust," Haussila said. "If the team's working independently, it can't work without trust."

That trust needs to be mutual, he added. The management needs to trust their developers to make great games, but the developers also need to trust the management to have their backs. If developers are afraid of failing because there will be consequences, then they won't take risks.

It hasn't always gone smoothly, even for Supercell. After development of Hay Day and Clash of Clans, there were a number of teams and projects in a sort of limbo, where the projects had stagnated. They didn't want to kill the projects, but they were also worried about their projects living up to the success of the studio's first two titles. The company had created an atmosphere where the developers didn't trust the company enough to think failure was acceptable. To change that, Supercell set up a prototype team to give employees a safety net where they knew it was acceptable to try things out and throw them away if it wasn't coming together. That helped, but they still needed to make changes to ensure even developers on those prototype teams felt like their contributions were helpful.

Part of that has to do with Supercell's greenlight process, which is a cycle starting from design to implementation to testing to analyzing. In the traditional model of game development, the analysis stage is conducted by management, which then makes the decision on whether to kill the project or keep it going. At Supercell, after analyzing the project, the developers take another crack at the design phase to see whether they have ideas to fix whatever might be wrong before killing it or greenlighting it through another round of the cycle.

Again, this process runs on trust. With Clash Royale, Haussila said there was a point where almost nobody in the company believed in it. There were definite issues with it, but the game team still believed in it. It's hard sometimes to communicate the merits of what you're working on to the rest of the company, Haussila said, so it was crucial that Supercell placed enough trust in the team to continue on with development.

"If we deliver something to a global audience, I think we're also responsible to update it and maintain it for years to come"

At Supercell, Haussila said less is more. That extends not just to its catalog of released games, but to its headcount of under 100 people. More games and more people means more management and bureaucracy, he said. In turn, that means people have less ownership over their projects, less independence, and ultimately, less quality in the games that get made. Supercell doesn't even have an HR department in the usual sense, with those duties split between a recruitment staff and the various team leads.

While the decision-making hierarchy is set up to support the individuals on each team, it only works at Supercell because the priorities guiding the decisions are made with an inverted focus. The top priority is always Supercell as a company. Given the company's focus on staying small, that can mean that resources and developers can be moved around to get behind a key title. That might rub people the wrong way if their project loses out in the process, but the well-being of the company has to come first. The next priority is to the teams, and then to individuals.

Supercell also has a "culture of killing," Haussila said. While the company launched four games, it has started on plenty more. There are tons of projects that start as prototypes, far less that become playable to the company, less that go into beta (only about 10% make it into company playables, Haussila said), and plenty of those never get a full release. Failure is tough to deal with, but the company wants to celebrate the things that were learned when projects are killed. That means when a game actually makes it through to launch, Supercell is utterly committed to it.

"If we deliver something to a global audience, I think we're also responsible to update it and maintain it for years to come," Haussila said.

If Supercell's track record to date is anything to go by, the audience will be there to support those games for just as long.

Disclosure: MIGS has a media partnership with GamesIndustry.biz, and is paying for our accommodation during the event.