#Agile vs. #ProjectManagement, #Agile vs. #DesignThinking, #Scrum vs. #Kanban, #Forecasting vs. #Estimating, etc. etc. I’ve seen it all, with friends on all sides of these debates.
I’m an Agile guy and have been since my career began as a RIP software developer and TQM / Lean facilitator. And, along the way, I ran a PMO. In my shop, “guestimating” could cost you your job. Because our estimates were data-driven, empirical, and rarely wrong. Because we used statistical approaches such as Monte Carlo, the technique of choice by today’s #NoEstimates advocates, which has been included in the Project Management Body of Knowledge (PMBOK) since 1996.
Much estimating failure is due not so much to poor techniques but to poor practitioners. Some is due to poor management–and good luck selling them on any solution. And the rest is due to the complex nature of the work.
Finally, yes, of course we should produce forecasts vs. estimates, when we’re doing repeatable, predictable work with a strong data baseline. But then it wouldn’t be complex work and simplistic, linear, conventional approaches would work just fine.