Most of what is published about agile/scrum methods doesn’t talk about this area very much. It tends to be focused mainly on a team dedicated to getting a single product to market. As an internal team, we often don’t have that luxury. We run multiple concurrent projects, some with different members, as well as numerous operational tasks & enhancements.

We manage our work, we looked at having multiple Product Backlogs (basically a prioritised list of work the sponsor would like us to complete). This naturally led to multiple Sprint Backlogs (which contain the work the team has committed to doing in the next 2 weeks). This was going to be very difficult to manage. Particularily when it comes to the team keeping focus on their activities. Multiple projects and backlogs all over the place.

What we really needed was multiple Product Backlogs that each project sponsor could manage, but one pipeline for the team to work off during their 2 week sprint.

Mingle to the rescue!

We’ve been using this tool by Thoughtworks for a few months now, and the more I use it, the more I love it. We’ve pretty much ditched Microsoft Project completely as this tool is powerful and simple enough to do all that and more (albeit very differently).

It was quite easy for us to create a number of “card walls” to filter out the work for each project or operational area. We then use these separate lists when we sit with the sponsor to have them prioritise the work.

From that, the team picks off the top ‘x’ stories and commits to delivering them during the next sprint. Mingle makes this very easy for us to manage.

The tricky part is knowing how much from each project we can commit to. Currently, I communicate the number of hours we should be dedicating to each project each week. Hours are one thing, but I’d like to get to the stage where we track our overall velocity. This way, we can more easily determine how many “points” we can take off the backlog from each project and commit to doing.

We’re not there yet, but should begin getting some metrics in the next few weeks. I’m keen to see how this pans out and whether the idea above will work in practice.