I read the following article Global CIO: A Global Project, With No Big Project Plan which reinforced my opinion that the set of valid reasons for NOT adapting agile practices for a particular type of technology project is steadily shrinking.
The project was global (so not all team members and stakeholders were co-located), spanning multiple offices with different stakeholders (i.e. not just one customer), involving a change that would have different impacts on different offices.
Here are some of the practices used by the team.
- Where attention goes, energy flows: Project sometimes suffer from too many goals. We tend to work best when we are able to focus on achieving one objective. Instead of defining multiple objectives for their project, the main one was to improve their Google search ranking. In a globally distributed environment, this was a good overarching, culturally-neutral goal for everyone to rally behind.
- Keep executive reporting simple: Instead of overwhelming senior management with complex burndown charts, dizzying arrays of stoplights or spaghetti-like Gantt charts, a simple heat map that provided executives with a quick understanding of the completion status at each office was used.
- Get some early wins to boost morale & productivity: A common practice with agile projects is to tackle architecture & risky deliverables early. While not a bad practice, there is also merit in the approach used by the project team to knock off some smaller, easier offices – this validated their project on a micro scale, and acted as a motivator for the entire team. It can also provide the project team with a steady exposure to issues which is preferable to being overwhelmed by them as might happen if a highly complex or risky scope area was targeted.
- Build stakeholder relationships early and sustain them: Too many times we meet our stakeholders early in the project life cycle, but neglect to circle back with them on a regular basis. With agile global projects, out of sight means its a whole lot easier to get out of mind.
I’d be interested in learning if there are other such “sacred cows” where agile practices were historically avoided that can now be successfully managed using these methods.