Supercell (Ilkka Paananen), 2010

the problem

Small teams kept pushing mediocre games toward launch because admitting failure felt like a career-ending embarrassment.

background

Supercell, the Finnish mobile game studio founded in 2010 by Ilkka Paananen and Mikko Kodisoja, organized itself into small, autonomous 'cells' — teams of just a handful of people empowered to conceive and build their own game ideas with minimal oversight. That structure produced genuine creative freedom, but it ran into the same problem every product organization eventually hits: once a team has spent months building something, the people closest to it are the least willing to declare it isn't working, because doing so looks and feels like admitting personal failure in front of the company.

The conventional response to underperforming projects is a top-down kill decision — leadership reviews metrics and cancels what isn't working, which solves the sunk-cost problem but does it by force, from outside the team, and does nothing to change how the team itself will behave on its next project. It keeps failure a shameful, externally-imposed event rather than something a team learns to recognize and act on early, on its own.

what everyone would do

The standard fix was a top-down review process where leadership cancels underperforming projects based on metrics — which stops bad launches but does it by external force, leaving the team's own instinct to hide or excuse a failing project completely unchanged for next time.

what they saw

Teams avoid killing bad projects not for lack of data but because cancellation reads as personal failure. Supercell moved the celebration onto the cancellation itself, flipping the emotion from shame to status.

the move

Supercell instituted a practice, discussed publicly by Paananen and documented in a Harvard Business School teaching case, where a cell that cancels its own game presents what happened at a company-wide Friday meeting — and the room toasts the decision with champagne, treating the cancellation itself as the thing worth celebrating, not a failure to apologize for. The ritual made killing a bad project, rather than quietly limping it toward a weak launch, the socially rewarded move.

why it works

The champagne ritual doesn't change the facts of any single project — it changes what behavior earns social approval company-wide. By making the team that cancels its own game the one that gets applauded and gets to present, Supercell converted the fastest route to positive attention into early honesty rather than late-stage denial, which is precisely the behavior a purely top-down kill process can never install, since it only ever punishes, never rewards.

the payoff

Supercell has killed 30-plus games against five live hits, redirecting teams off failing projects faster instead of shipping them.

where it breaks

It depends on genuine organizational trust that a cancelled project won't quietly hurt someone's compensation, promotion or standing — if the ritual is public praise layered on top of private punishment, teams will see through it fast and revert to hiding failure. It also only functions where teams have real autonomy to conceive and kill their own projects; imposed on a top-down organization, there's no decision left for the team to be honest about.

what came after

The champagne ritual became one of the most widely cited examples in tech and games-industry management writing of building psychological safety around failure directly into company ritual rather than policy documents, studied in the Harvard Business School case on Supercell and referenced across the games industry as other studios adopted similar postmortem-celebration practices.

references

[1]Celebrate Learning From Failures (With Champagne!)Corporate Rebels, 2022corporate-rebels.com
[2]"This is bullsh*t": Supercell opens up on cancelled projects and the role of failure in creativityPocketGamer.biz, 2023pocketgamer.biz

Supercell popped champagne whenever a team killed its own game, so bad games died faster