June 5, 2009
Pay me until you’re done

I always tell people that an ideal Agile project works like this: each iteration, the business picks things to get done. Those should be the most valuable things remaining to be done. Therefore, the ROI of the team decreases each iteration. Eventually, anything left to add to the product is worth less than the cost to add it. That’s the point at which the business should find something else for the team to do.

(Yeah, I know the business should be considering alternative uses for the money rather than just ROI>0. This way of describing it is punchier.)

It’s always seemed obvious the same thinking applies to consulting. My usual gig has me visiting a team one week per month. Each time, I pretty much sniff around, look for the most pressing problems, and apply myself to them. Assuming I help, each month the most pressing problems will be less pressing, so my value will decrease each month. At some point, I’m charging more than I’m worth. So that’s the moment the client should tell me not to come back. 

Is this an odd model?