How to deal with re-planing a lot and missing milestones over and over,

After finishing the project an issue is found and the project is restarted from the beginning to improve the output.

how to plan the project so that it gets out of the planning phase to the production phase?

  • Welcome to PMSE!
    – nvogel
    Sep 11, 2022 at 15:48

3 Answers 3


Much depends on what kind of product or output you are trying to accomplish. The textbook answers are rolling-wave planning, prioritisation, iterative/incremental delivery, and, if the product is software, a DevOps style approach to delivery.

Don't feel constrained by project-thinking and predictive planning. Manage the product or value-stream, never the project. To see the difference, just consider that the kind of continuous improvement you described is actually normal and is often the desirable state in some fields. Millions of technology professionals build, develop, fix and improve software products continuously, and most of the time they can do that work without planning anything called a "project". What matters is to deliver frequent improvements that are of real business value.

If too much project planning is getting in the way then focus on working out what can be done this week or this month, start delivering and just accept that longer term plans will adapt and change as you go. That may not be easily achieved for every kind of activity, but it is very often the case that complex challenges are best tackled by breaking them down into smaller problems that can be delivered piece by piece.


Sometimes the client or sponsor isn't taking the time to understand how the project plan will reach their goal.

A friend of mine in sales told me this "A confused mind says 'no'."

  • If you have the relationship, take the time to talk things through with the sponsor/team, letting them ask questions.
  • And if you sense they're not understanding, stop and help them ask their question.

This is a wrong answer from someone so horrible at PM that writing a graduation thesis takes years.

How about going bottom-up?

  • identify functionalities that will surely be needed (e.g. logging)
  • implement
  • test the code while implementing a tiny demo(the demo code is not discarded; it's a part of the codebase from now on)
  • report to superiors
  • repeat

This differs the decision of "how do we make the product work" to the latest time possible. When we have the most information. While still producing deliverables(see: we have a printf() that prints in green :) ).

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.