I'm sure my friend Jack Ricchiuto's new book, The Power of Circles, will do a much better job of answering the "how" of engagement than I could ever do. But I have a theory on why it rarely produces the desired results: The wrong kind of organizations are conducting the engagement.
Most organizations that conduct engagement are designed to address complicated issues. As Will Allen (and so many others) explains, complicated issues are dramatically different than complex issues. Simply, complicated issues can be resolved with a technical solution and complex issues can only be solved with solutions that emerge from the system involved.
Organizations designed to deal with the complicated (let's call them "complicated organizations") tend to design processes that rely on a command-and-control framework to implement a technical solutions. Such processes do not work on complex issues, and complicated organizations aren't comfortable designing engagements that enable emergent solutions.
Complicated organizations are accustomed to engaging stakeholders in environments where they control the conversation. In a complex system, there is no control. Most complicated organizations are used to choosing who they engage with. In a complex system, it is the system that determines who needs to be engaged.
To design effective engagements in complex systems we need to design frameworks, processes and tools that promote emergent solutions. This is the heart of FSG's framework for Collective Impact. Engaging stakeholders is a primary role of what FSG calls a "backbone" entity. To be successful, backbones need to be designed for complexity.
Organizations designed for complexity are designed for engagement. Complicated organizations are designed for control. Asking complicated organizations to conduct engagement in complex systems is like sending a hammer to do a screwdriver's job: It can be done, but the damage done usually exceeds the benefits.