Project Dependency Mapping

This is a stub, re-capping a process I recently used to map a complex process at the VA. I borrowed this tool from Michael Keeling, when we worked together at LendingHome. It's similar to the Event Storming method, from DDD. This write-up is incomplete, but hopefully there is enough here to prompt questions. Enjoy, and as always, YMMV!


  • On the far right of the visible area of the board, we made a tall (full-screen height at the start) box labeled DONE.
  • Beginning with DONE, we asked "What needs to happen for the work to be considered DONE?", then we created blue stickies for each independent deliverable, arranging them vertically to the left of DONE (more on this later; don't worry about perfection).
  • For each blue sticky, we then worked our way right to left (backwards), asking "What needs to happen for us to be able to work on that?". We identified dependencies for each, and dependencies for the dependencies, until the answer was "nothing, we can start now".


  • Use rows for work threads or streams or whatever you call sets of interdependent work that are themselves independent of other sets of work.
  • Use columns for iterations: sprints if that's the term you use.
  • Use blue notes for independent deliverables for a particular work stream.
  • Use purple to denote the expected, roughly predictable dependencies.
  • Use pink notes for risks.
  • Use yellow notes for time-boxed activities.