Comments on: Is estimation waste?

By: On Estimation « Paul Dyson’s Blog

On Estimation « Paul Dyson’s Blog — Mon, 12 Jul 2010 16:49:43 +0000

[...] has blogged about his thinking and who I hope will comment further on this subject. This seems to appeal to developers who find estimation difficult and teams that struggle with accuracy of estimation. Certainly [...]

By: Manuel Küblböck

Manuel Küblböck — Tue, 26 Jan 2010 07:18:30 +0000

I guess every process relies on it’s users to make reasonable decisions. Like everything else the urgent task slot can of course be misused/abused.

By: Dave

Dave — Sun, 24 Jan 2010 10:17:15 +0000

The formula for test estimation is:
drop dead date – date development is complete

If all goes well you will get a positive number

Thinking about the urgent slot I could see issues when there is more than one person who is the customer and want their own thing to be in there. I suppose there would have to be something in there to say once something is in that slot it has to be there until it is complete else the customer will cotton on to it and just change it every 5 minutes to suit their current needs. You have have 7 to some what control the context shifting problem

By: Manuel Küblböck

Manuel Küblböck — Thu, 07 Jan 2010 07:55:19 +0000

Thank you Aaron for the insights.

Seeing the wait time as a SLA is certainly interesting from a customer’s point of view.

I agree, that visibility is key in any process. I just didn’t touch it in this post, because it was already pretty long and achieving visibility in Naked Planning is not any different to Kanban. (Correct me if I am wrong.)

Trying to minimize the lead time as a motivator is something I haven’t considered, but certainly valid. Do you have experience with teams trying to do so? How would you compare it to the commitment to Scrum sprint goals?

By: Aaron

Aaron — Thu, 07 Jan 2010 05:24:03 +0000

Hi Manuel,

I really enjoyed reading this article, and would like to provide my own feedback to the issues you have.

The team is establishing an SLA with the customer which is measured by that wait time. There is an explicit promise made to deliver MMFs before their expiry date in the queue, whatever is the current wait time. A collaborative game begins to shorten the lead time, and up throughput of value with one piece flow. I also believe Arlo is using one as the limit now to encourage OPF. The Cumulative Flow Diagram (CFD) also shows visibly the amount of WIP, lead time and throughput. Mentioning visibility, this is a key element of Naked Planning. There is a board visible in the team room which shows information arrival and MMF progress, reinforcing the collaborative game. Other forms of motivation the team already uses can remain in place. Working agreements, 360 reviews, XP and pair programming, etc.

Cost is known with the lead time. If the customer is paying 1000 dollars a day and the lead time of an MMF on average is 3 days, the cost should be obvious. Flowing MMFs out incrementally should also help realize the value quickly and provide feedback from the market for the customer to check if they are on the right track.

sincerely,
Aaron