I like Lada's outline.
I think another great way to work is just x-functionally.
This is something that has been working really wellf or me lately.
Design with your eye on the prize and let your developers tell you "No" for
all the possible reasons that they can say no.
What this leaves you with is the following:
1. You have communicated to your x-functional team (business, opps,
development, etc.) a vision of what is to come.
2. As Lada suggested, often designers under shoot the possibilities when we
consider technology too soon in the process.
3.