Is it advisable to estimate using story points in Kanban?

[Ernest Nji]

Is it advisable to estimate using story points in Kanban in order to measure the teams velocity monthly? What other methods can be used to measure Kanban teams velocity?

[Nic Brough -Adaptavist- · Accepted solution]

No.

That's too short an answer.  A lot depends on context though.

The "strict" bit is that Kanban does not have estimates.  Kanban is interested in throughput and forecasting, not velocity.  Estimates don't help you with Kanban and you can't do velocity without estimates, so that goes out the window.

But. Context.

Jira implements "plain" Scrum and Kanban, Atlassian does not try to be clever with them.  But they do leave us free to play around with them a bit and don't block Agile at scale thinking too much.

When you try to scale Agile, you often run into teams doing a mix of Scrum and Kanban, and other ways of working, and you need to think about how to estimate consistently across methods.  (I have a bit of a bugbear that people don't understand story points in the slightest and then do things like "roll-up story points to the Epic", but that's another conversation)

The most simple scheme for "Kanban estimates" is to let people work, one team Kanban, one Scrum, and then look at what has happened over the last 3ish sprints.

Let's say your Scrum team averaged 60 SP per sprint.  While Kanban doesn't time-box it, you can still look at the start date of the first sprint and the end date of the third and then say "the Kanban team fixed 90 cards in that time".  So you can make an equivalence that 1 Kanban card = 2 Scrum SP.

TLDR: your Kanban "velovity" is "how many cards you get done"

[Ernest Nji]

This is helpful. Thank you!