Earn early, spend late, turn on a dime.

"We need to make our software development economically more valuable by spending money more slowly, earning revenue more quickly, and increasing the probable productive lifespan of our project. But most of all, we need to increase the options for business decisions." K. Beck. Extreme Programming Explained: Embrace Change. Addison-Wesley, 1999.

It seems to me that the economics of investment in IT continues to drift more and more towards the general principle of "Earn early, spend late" -- many short releases, delivering highest value features earliest, traversing the value gradient until further feature delivery offers no further return.

There are many economic benefits to this approach: early value delivered to the market can generate revenue sooner; spending less up front to achieve a business outcome is desirable; and prioritizing features against this business outcome is a very effective means of eliminating unnecessary work.

However, the real benefit to this approach is the opportunity to lead the market. Short queues of features for delivery means less waiting and greater responsiveness to change. The true value of a feature may not be clear until the feature successfully helps someone to do something valuable. The faster we receive this feedback the better we can serve our customers, the more we can lead the market.

It's not cheap to achieve low cost monthly production software releases. Substantial testing and integration automation is essential for speed and to keep costs down. But. Can you put a price on market leadership and responsiveness to market changes?