In continuous process innovations, many of which are outlined in Managing Agile Open-Source Software Projects with Microsoft Visual Studio Online, we evolved the rolling triage and the fire-and-forget project type.
In the past we processed quarterly triages of project ideas, many originating from UserVoice, creating four main waves of projects per annum. As shown, some side-effects were last-minute spikes of activity as teams realised they needed to wrap-up, and troughs during which resources are on the bench, waiting for new projects.
We introduced the concept of a rolling project idea triage, triggered whenever a project enters its last scheduled sprint. The rolling triage allows us to investigate new project opportunities, and resources to engage, while we wind up a project. While we have not been able to nudge the last-minute spike to extinction, the occurrence and impact hereof are on the decrease.
In Managing Agile Open-Source Software Projects with Microsoft Visual Studio Online we share our default and Program Manager (PM) managed process, with geographically distributed, and part-time resources. It starts with an idea, planning, a fairly detailed roadmap, planning and proceeds with the development and release of a solution, aligned with a common iteration / sprint plan cadence. We often refer to this type of project as a formal project, or project that comes with a lot of ceremony, and a weekly scrum.
- Predictability, based on roadmap.
- Near real-time insight into status and impediments.
- Can work with one or more feature teams, each with 6+-3 team members.
- Volunteers with only 15min to spare, find it difficult to engage on projects with pre-defined milestones and priorities.
- We are limited to ~five concurrent projects, based on PM and resource bandwidth.
The new fire-and-forget project begins with the same triage and investigation, but less planning, process, ceremony and typically no detailed roadmap and milestones. Teams self-organise, work through their backlog at their own pace and meet with the PM less frequently, typically every second week or monthly.
- We can support an infinite number of concurrent projects, constrained only by the available part-time bandwidth.
- Teams can self-organize and work when and how they prefer.
- Insight into status and impediments is less frequent.
- Does not scale well beyond one team with 6+-3 team members.
As the Program Managers never “forget” a project or associated team, we should really call these projects fire-and-self-organise, but fire-and-forget is simpler and more common.
If you have been keeping an eye on our status board (aka.ms/vsarflightplan) you would have noticed that our first two fire-and-forget projects have taken off, under the leadership of Gordon Beeming and Utkarsh Shigihalli.
What do you think?