Scrumban

[Michael Coll]

I would really like to use Scrumban with Jira.

I know that there is now a kanban backlog but it's not the same for various reasons:

1) No timeboxed sprints

2) Which means no sprint goals, burndowns etc

3) the 'Done" column will just grow indefinitely

Ha sanyone any resoution to this?

Cheers

[Maarten Cautreels]

Hi Michael,

At De Persgroep, some teams work Scrum and some use Scrumban. Basically they work Kanban but since we'd like the reporting for all teams to be the same we are working Scrumban. 

So these Scrumban teams do have two weekly iterations with a demo and retrospective at the end and a backlog grooming session in the middle of the sprint. They don't have a sprint planning session anymore as they don't commit on a set of features but they commit on an amount of story points (usually the average of the 5 last sprints taking into account the availability of the team).

We do have a burnup which we built with lego and we keep track of all the numbers in a Google Sheet, where we have a lead time chart, a velocity chart and an overview of all the unexpected features (read bugs) we've been working on each sprint. Simply to track evolution of the team and find out if there are any particular things we can do differently.

We're sharing this data with the rest of our company using a two weekly Sprint Report on Confluence.

You can view an example of such a sheet here: https://docs.google.com/spreadsheets/d/1iIatT_TRmhgH3U2mMLd8qbZNyTkpfPbI8Rm4SWEAKWI/edit?usp=sharing (feel free to make a copy)

I've also created a dummy version of the last Sprint Report for the same team:

It uses the HTML macro (not enabled by default in Confluence) to include an iframe to the Google Sheet Charts.

And we have a quick filter to hide the issues in the Done column as well as adjusted our filter to not show issues that haven't been updated in the last 2 weeks:

(status = DONE AND updatedDate > -2w)

We've tried filtering issues that haven't been put in Done in the last 2 weeks but while those History queries are very powerful, they become slower in time (if there's more issues). And you won't want your board to become slow obviously.

Best,

Maarten

[Michael Coll]

@Maarten thanks for that in depth reply. It really helped and I will definitely be taking some of your practices and implementing them.

Since asking this question I came across the Jira "version" and "release" functionality. This does in effect allow for timeboxing as when you create a release you are asked to enter start and end dates - and after the end date, everything in the "done" column is removed.

I'll persevere for a couple of weeks and see how we go.

@jira - it would be relatively simple to implement proper Scrumban. Is it on the roadmap?