Having just read through all of the other threads on this matter, it seems its a common point of confusion with GTD!

At the moment I’m swaying towards having project pages, with a defined outcome, and a brainstorm/mindmap, once I’ve entered a project, I’ll extract the next action from the mindmap, and enter it in next actions.

When I complete that, I can go back to the mindmap again and determine the next action.

I don’t quite like the idea of putting all next actions on the project page, as I want to keep them together in next actions, also I don’t really want to list actions (note, not next actions) on the project page, and then moving them to next actions as they become applicable because thats a lot less flexible to change then a mindmap is?

Think I’m starting to get my head around it, could still do with some clarification/feedback/advice though.

Thanks, Matthew

The way I do it is very much like Option 3 above. However, I don’t actually move actions onto NA lists because my software task tracker (the web-based SimpleGTD) just lets me “star” actions to make them show up as NAs. Keeping full action lists for each project makes it very easy to review my progress weekly.

However, you’ll need to determine a GTD implementation that works for you. I recommend picking the method that seems best to you and just using it initially, tweaking things as you go. I took me quite a while to end up with a stable system; now that I have something that works for me, all of GTD just melts away into doing actual work.

Don’t forget to separate your NAs by context, not by project. Creating your personal set of contexts is another very tricky thing, but again that’s part of the process of developing your own GTD-based system. The key is just to pick something and start doing it without worrying whether it will be initially perfect (because it won’t, no matter what). Refinements will come later, but you can’t refine a process you haven’t already started using!