“Done” is often “Waste In Progress”

My experience is that most Scrum teams define "Done" as a "potentially shippable" story/feature. However, Lean has taught us that this definition actually introduces waste in the software development process.

I like to think of such a "Done" column as "Waste In Progress" and should thereby also have a limit. So, once the waste limit has been reached, it is time to merge shippable stories to the current release branch and deploy a new version to production.

Hopefully I will be able to test this idea in my current team. 🙂

10 responses on ““Done” is often “Waste In Progress”

  1. A bit harsh, in my opinion. Why can’t we have “Waiting for production deploy” (W4PD) and then “Done”? Thus, when shipped then it is done.

  2. Absolutely, but waste is waste, no-matter what you call it 🙂 My point is that it is not very Lean to add all stories in a sprint to a Done column just waiting for a planned production deploy.

  3. I agree that having “Done” in a column that really means “Ready for production” is wrong but “Waste in progress” sounds like you’re developing the wrong features to me. Maybe “Value in progress” would be better 🙂

  4. Hello, and thanks for your comments.

    “Definition Of Done” (DoD) varies between teams and R4P could very well be a valid one – at least in Scrum teams. “Waste In Progress”
    sounds a bit odd, I agree. I was just playing around with the acronym WIP (Work In Progress) :-). If you think of “Work In Progress”
    as valuable work in progress, then “Waste In Progress” means that no valuable work is in progress. Or, relating to Value Stream Maps,
    think “Work In Progress” as “Value Added Work” and “Waste In Progress” as “Wait Time”.

  5. Good idea. I remember spending two months on a feature that ended up molding on a branch because QA never had time to test it. Now it’s long forgotten and I have moved on. That’s true waste.

  6. Done is a state. It indicates story is ready for production. DoD is on other hand definition, how the state Done looks like. However, I dont agree with idea of considering Done as a waste. Where in LEAN have you found it?

  7. Thanks for your comments.

    It is waste since no actions are taken to deliver the shippable feature to the end customer. We’ve entered wait
    time with respect to that feature. Wait time equals waste. It like ordering a computer that are assembled
    in the factory and then put on the shelf, waiting for the delivery truck that arrives only once a month 🙂

  8. I think its missinterpretation. Even in LEAN waiting times and small batches are allowed, if they make technological sense. If minimal batch to dispatch your computers is the size of the truck, then waiting for filling up the truck is perfectly ok. Same with software. If you use e.g. Scrum, there is size of the batch defined by the process (one iteration). Make no sense for any customer having features delivered constantly and having constant changes of the product under their hands. So I would practically say there is no such thing as “Waste in Progress”.

  9. It depends, as always, on the context. If technical problems hinder you from delivering faster, you may try to overcome that problem (smaller trucks maybe :-)). If all stories in the sprint depend on each other, well, then you either wait for the sprint to end and deliver the whole batch, or you could try to make some stories independent in order to deliver them earlier.

    My point is, that in some contexts, it just doesn’t make sense to put done stories on the shelf, just waiting for the sprint to end. If it is possible to eliminate this waste by delivering some stories earlier, we should all be happy 🙂

    Regards,
    Magnus

  10. Hi again,

    I agree that using “DoD” instead of “Done” in the title would be more correct. However, I felt that the current title would catch more interest from readers 🙂

Leave a Reply to Viktor Nordling Cancel reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.