When I here discussion of "managing" development projects using agile, I always come back to the response that "managing" software development is not the same "developing" software.
While still in a meeting yesterday, the conversion came to the point...
...for us, the big issue is foresting when a project will finish up, and that requires looking into the future a year to 18 months from now. It's a challenge to recognize the correct revenue and forecast cost for completion.
The challenge we're working is how to "manage the business side" of the project. The famous "business value" that the thought leaders speak about, but fail to provide any method and units of measure. In the end writing software for money, internal money or external money, means "managing" the business of writing software for money.
When we mix writing software with managing the writing of software, the conversation about process gets confused as well. Agile provides useful process for writing software in response to the needs of the business. But agile assumes there is a process for defining those business needs, measuring the "business value" of the software from the projects.
Without the actual physical connections between the "writing" of software and the management can be connected with an approach like this. See of this adds any value to the conversation. It has for our domain (enterprise mission critical systems).
