There’s often confusion in companies transitioning to Agile about how to “cost” an “Agile project.” Projects are irrelvant; focus on the product as a whole. You fund teams, and they work on the most important aspects of the most important product. Here’s how that works:
Every story is effectively a two-day-long single-team project. That's trivially easy to cost. A team of 6 is burning roughly $10,000 per day. Two days is $20K. The question is where is the best place to put that $20K? What will give you the best return on that $20K investment? 2/
You answer that question by asking which product is more important to the customer and the company, and then move that product forward. You assess that dynamically by looking at the actual (moving average) revenue generated by the product. 3/
You don't assign work to the teams. Instead, use your assessment to rearrange the backlog to move work associated with the critical product up. The teams just pull work and do it. They are not assigned to a specific project. They just continuously pull the most valuable work. 4/
You can't do any of this up front---before the product exists---because you have no way to assess actual value at that point. It's all just guesswork. So, costing a product up front is a waste of time. 5/
Instead, build the smallest piece of the product possible, so small that you're a bit embarrassed by it (the so-called MVP). Then decide whether or not to allocate more work to that particular product. Constantly reassess that evaluation & adjust the backlog accordingly. /end
