It turns out that what was common across these near failures was that no one had ever really defined what the innovation team was supposed to do, or what innovation means to the organization. This is one of those sad but funny items that make you scratch your head.
On one hand, everyone agrees that innovation is important. On the other hand, everyone has a slightly different definition of what innovation is. Therefore, the team gets many important requests that pull it from one idea to another with little overlap or respect for other ideas or previous work. So one manager's important incremental product idea is worked, then another manager's disruptive service idea. At some point the management team shows up and says - what have you done for me lately, and the innovation team has a real grab-bag of efforts that are disjointed and not complete.
What these folks need is a charter. A charter for the team details what the team's goals and responsibilities and roles are - what it will and won't do. If the organization can't get behind a definition of innovation, then the team can do the next best thing and define its scope and boundaries. This lets the team say "yes" to some projects and "no" to others, and create a more rationalized work effort and portfolio.
This seems evident, right? Yet you'd be amazed at how many teams just start out working on the first project that comes along, and then get sucked into the second, and third, and fourth projects, none of which have any relation or bearing to the others. Once you're in the whirlwind, it's tough to stop and realign.
If you don't have one now, create a charter which details what your innovation team plans to do - it's focus, priorities, resources and responsibilities. It's as important to say what you won't do as it is to define what you will do.