Everyday Kanban

Discussing Management, Teams, Agile, Lean, Kanban & more

Cost of Delay with CD3

I am at LKNA 14 this week — Lean Kanban North America, aka Modern Management Methods. There has been, as there should be, a lot of discussion about cost of delay with more yet to come.

The importance of knowing as much as you can about the business value is obvious, but its easy to stop at a certain level of info and not realize you should go further. For people who haven’t ever done cost of delay calculations before they might start with some bucket categories like Expedite, Fixed Date,  Standard & Intangible. I know I certainly did and just that classification made a world of difference for my team’s well-being. However, I could have gone farther as @OzzieYuce and @JoshuaJames pointed out in their session entitled “Black Swan Farming using Cost of Delay.”

One expedite may not be as urgent as others. The general classification is rarely enough to make a great decision. It is best if we can get to the specific cost of delay per unit of time, such as cost of delay per week. This formula is noted as

CD3 = Cost of Delay / Duration

If you’re asking why we’re not dividing by cost, that was asked here at the conference as well and the answer was that time is a much scarcer resource than money and thus a better factor to calculate upon.

Economics is often misunderstood to be about money. Its actually about scarcity and how that affects our choices.


To circle back around, you get 2 expedite tickets (if such a thing were ever to happen) or 2 fixed date tickets without enough time to process both… How do you choose. Or you are using the Moscow Method for requirements classification and you have to choose which MUST HAVE to work on first. You have to look at the CD3 together with how long you will incur that cost if you do the other item (and vice versa). Only then you can truly know what the best monetary choice is.

But, I don’t have cost of delay in $ terms 🙁

Let’s start with an exercise:

If you were going to a desert island for five years and could only take one thing, what would you take?

OK, now the stupid airline lost your luggage. How much would you pay to get it back?

So, the point of that exercise is to show that if you have a scenario in which cost of delay seems intangible, ask how much would you pay to have this done?

Seriously though, don’t quit just because you don’t know how to figure out. First, pretend you know have some figures for the information you don’t know. Calculate what the cost would be if your numbers were right. Note your assumptions. Now you at least know what you need to determine to get the right cost of delay. I am not 100% sure that being extremely accurate matters, getting closer to knowing the value / cost of feature, project or task will allow you to really understand what the value of all of those “MUST HAVE’s” are and let everyone really appreciate the value they enable for your company.

This is just a small part of one topic I observed during the day 1 of the conference. There’s much more from this topic and many more topics going on. So far, so great at LKNA14!

1 Comment

  1. Great post. I just want to emphasize that getting a 100% accurate number is not necessary, and getting a sense of relative value between this and other items to choose from is good enough for the purpose at hand.

Leave a Reply

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

© 2024 Everyday Kanban

Theme by Anders NorenUp ↑