In our previous post we suggested a new collaboration curve was emerging in which the more participants and interactions you bring together in one place, the more performance and learning improve.

We pointed to World of Warcraft, and touched on SAP’s Software Developer Network (SDN), and the world of big wave surfing. And we suggested that collaboration curves, characterized by their dynamics of increasing returns, might be gaining importance relative to the decreasing returns found in the experience curves that characterize the world of push.

But how do collaboration curves come into being in the first place? Can they be intentionally created? In our experience the answer is yes–but you can’t simply amass a whole bunch of participants just anywhere and expect one to arise. One of the most persistent myths about the open source movement, in fact, is that somehow open source activity comes into being on its own and then thrives through self-organization alone. Emergent behavior from participants is, of course, essential. But it rarely takes hold unless it’s organized and supported within carefully designed (and tended) “creation spaces.” By the same token, one has to guard against overdesigning these spaces, too. As with gardens, it’s best to design creation spaces minimally at the outset and then let them create a life of their own, layering in additional design elements over time.

This balancing act combines three essential elements to set in motion the increasing returns dynamics that makes creation spaces successful: participants, interactions, and environments.

Participants: The first challenge is getting a critical mass of participants. How can creation space organizers attract the people they need? First by keeping barriers to entry low; next by giving them the real-time feedback and performance measures they need to advance quickly within the community. In World of Warcraft, for example, the early game play is fairly simple and requires relatively limited investment of time. For most creation spaces, initial participation is even simpler – it consists of observing others. Many big wave surfers began by visiting surf breaks and watching from a bluff above the beach as surfers tackled the waves offshore. In the case of SAP, the vast majority of participants register and log into discussion forums to observe and learn from the discussions taking place–without initially contributing anything themselves.

Participants advance past the “lurker” stage by posting questions and comments–and getting feedback from other participants. Many then begin seeking out people with complementary skills and shared interests. Eventually they coalesce into teams with sustained membership. The “guilds” that form in World of Warcraft are one example of the kinds of teams that emerge spontaneously within a broader creation space.

Interactions: But participants alone are not enough. Achieving increasing returns in a creation space requires the co-evolution of two forms of interactions between participants – team interactions, and looser interactions across a broader range of participants.

Team interactions are where participants share and develop new tacit knowledge relating to difficult performance needs. Whether in a World of Warcraft raid or while developing a Twitter-like service for corporate users within SAP’s SDN, these teams build shared understanding and trust that helps them make the most each other’s knowledge and experiences as they innovate new approaches and generate tacit knowledge.

Teams, however, can only get so big. Yes, you can design a space where many teams work in parallel, but creation spaces go one step further by facilitating peer-to-peer networks that cut across teams. These learning networks organize around shared resources like discussion boards, video repositories, archives of previous contributions, or in physical gatherings like conferences or competitions. Big wave surfing competitions, for example, provide more than just a chance to watch and learn from other world-class surfers. The beaches often become spawning grounds for new practice ideas as surfers watch each other and talk with other surfers.

Environments: To make it all work, the creation space must encourage and support various layers of activity–even if they’re not all present at the beginning. One layer might foster team interactions, whether its raiding parties within a World of Warcraft guild, or interactions within a guild as a whole. Another layer supports cross-cutting peer-to-peer networks. Yet another layer reaches beyond the creation space to engage outside participants.

Getting the balance right between these layers requires the right blend of governance protocols (although in World of Warcraft these function within the guild layer itself), incentive structures, and activity-supporting platforms. More broadly, creation space designers must be careful to create the right blend of external incentives and intrinsic motivations.

Do any of you participate in creation spaces? In what ways have they drawn you in–or put you off? Do you think they have the potential to support and increase performance and innovation for big institutions like corporations?

Read other posts by John Hagel III, John Seely Brown, and Lang Davison.

The original post is here, by John Hagel III, John Seely Brown (JSB), and Lang Davison @ Harvard Business Publishing