Mike Cohn compares upfront planning with insurance in a recent blog: https://www.mountaingoatsoftware.com/blog/upfront-thinking-is-like-insurance.
Some extension of my reflections and comments.
It’s a great areat analogy. After alle, who would enter the UNKNOWN-UNKNOWN un-insurred?
I always saw a project as a long sequence of questions you need to find and answer. Your inductive challenge is to figure out which questions you benefit most from answering early – and which you can safely leave for later – the final answer to which you will only know in the end.
No matter how long you search we know from hard-learned experience that you never find all the questions you eventually will need to answer up front.
You need to start answering some before you know them all.
UNKNOWN-UNKNOWNs that surface or questions that change in some way are only bad if they cause rework, excessive new unexpected work or new unrealized risks for the project.
Your systemic objective is to minimize the refactoring and risk, while delivering as much value as you can as soon as you can withoutcompromizing final objectives.
Especially when applying agile at scale how you chunk up and sequence your work – or your approach to backlog replenishment – is one of the most important determinants for final project lead time.
As Søren Kierkegaard puts it the eternal dilemma is: life is lived forward, but understood backwards
You can play with thinking as all the questions you need to answer as a string of pearls. What are all the possible combinations and what kind of final lead time, risk exposure and value creation would each combination drive?
It’s not so much about how you burn the backlog as it’s about what order you enter things into it … how you chunk up and sequence your work means everything.