Teams Of Rivals: A New Approach To Agility

A model for competition-based delivery

If you need (and who doesn’t?) complex delivery against immovable deadlines, you should consider competition-based delivery. The old red team/blue team approach is back in vogue. We have seen it save teams, firms, and careers. Here, we look at a mini case of a gaming company with an aggressive, scalable, and agile multi-team approach to product development. The model has lessons for all kinds of companies and next-generation problem-solving, development, and delivery.

Supercell, a Finnish e-gaming company, was started in 2010. It posted profits of $1 billion on sales of $2 billion in 2016. That year, China’s Tencent acquired a majority stake (84.3 percent) with an $8.6 billion investment at a valuation north of $10 billion. A nice story, but what has really caught the eye of key thought leaders is the delivery model that produced these results.

With around 200 or so employees, compared to roughly 3,000 at Zynga and 10,000 at Electronic Arts (competitive e-gaming companies), Supercell offers a radical new model for agile content development that has made it the highest grossing iOS (iPhone and iPad) game developer with only two game titles (versus Electronic Arts’ 970 games). More than 100 million people play Supercell games every day; on average, players of Clash of Clans (released in 2011, a war strategy game in a Medieval-style fantasy world) play the game 10 times per day! This is in striking contrast to the large-console game industry, which appears to have largely forgotten the value of great game play, devoting massive resources and Hollywood-class budgets to games with immersive, cinematic production values.

In most companies, leadership comes from the top — here’s the plan, now execute. At Supercell, leadership comes from the cells — top management facilitates and enables teams

As its name implies, Supercell is organized as a collection of small, independent teams called cells, each tasked with developing new games or building deep features for existing games. The company as a whole is merely an aggregation of these cells — a Supercell. Cells are given complete autonomy in terms of how they organize themselves, prioritize ideas, distribute work — and what projects (e-games) they ultimately launch (or kill). “Each cell operates almost as an independent company,” says founder and CEO, Ilkka Paananen. Tencent’s WeChat mobile chat-and-payment app, which now has 1 billion active monthly users, emerged from a similar multi-team competition in 2011.

In most companies, leadership comes from the top — here’s the plan, now execute. At Supercell, leadership comes from the cells — top management facilitates and enables teams. Describing himself as the “world’s least powerful CEO”, Paananen prides himself on having no creative control over cells once they are constituted.

At the SLUSH meetup in Finland in 2017, which attracted 20,000 tech startups, investors, and journalists, Paananen said that Supercell was only interested in publishing games that would rank in the top 10 — and that nine out of every 10 games were killed by the cells. Each of its four games has spent a significant amount of time in the U.S. top 10 highest-grossing game chart. Hay Day has been in the top 10 for 884 days, Clash of Clans for 1,960 days, Boom Beach for 616 days, and Clash Royale for 701 days.

This track record of success cannot be a coincidence. Nor can the speed of delivery. Top-grossing games Clash of Clans and Hay Day were built and launched in six months, by teams of five and six people. A cell building a game drives to a deadline, which the cell defines, when the game is released internally. Both Clash of Clans and Hay Day reached this first milestone in three months. Based on feedback from this initial internal release, the game is either killed or sent back into development to be refined before another internal release — when the cell determines whether the game is killed or launched.

Unlike many studios that brush their failures under the rug, Supercell deliberately celebrates “kill” decisions because it recognizes that by building games quickly with small teams, it can have multiple shots at developing a blockbuster game in the time it takes others to build a single so-so game. It is also psychologically and economically easier to write off a project that was built by six people in six months than one that took 100 people 18 months to build.


The Supercell “rocket” brings a new perspective to product delivery that consumer-facing companies are likely to find thought provoking and potentially game changing. Companies may think that organizing multiple teams is a duplicative and costly endeavor, but the result is speed to market and significantly higher quality of consumer resonance. In the long run, getting that right is likely a more efficient way to do business.

Supercell has multiple teams working on different games, one or none of which may make the cut. But you can also task multiple teams with the same problem. And the red team/blue team approach works for issues that may or may not be product related. It works to deliver an outcome that a client or manager wants — whether it’s a product, a remediation, a process fix, an AI algorithm, a new market entry, a non-organic play. Two examples:

  • I am head of wealth and need to remediate a customer base before October 31 or get fined $1bn. My division has a track record of late delivery — or on time but poor-quality delivery. I set up a challenger team. I now have two teams aiming for the same outcome. Who will win? I don’t care. I just want the fix before October 31.

  • I’m VP-business development for a tech company. We’re launching a major upgrade to our popular mobile phone, with an advanced operating system. I need a suite of cutting-edge healthcare apps ready for release with the phone. I don’t actually have a clear idea of what we need to deliver — it’s open to the imagination, but I will know it when I see it. Instead of relying on my one existing team, I set up a challenger team (or two). Ideally, we ready several apps for launch. But as long as we get one great one to set the bar, I don’t care as long as it’s ready to demo at launch.

The Supercell model challenges the common approach to agility, which tasks one (cross-functional) team with delivery — on the theory that multiple teams mean higher costs, resource contention, and complexity. In fact, in many cases, the competitive model may mean higher quality delivery that meets critical deadlines. And, the model may even save money — assuming you kill less viable products before expending years on development and excess dollars. On top of this, there is less need for management “push” if you let small, high-energy, and competitive teams battle it out to prove their mettle. In fact, we know of one large company that is now running three rival teams on an AI project.