Saturday, November 04, 2006

When would agile fail ?

Following are some of the situations where agile methods would fail:

1. When Development team is interested in agile methods, and the product owner and his team is not aware of anything about agile. For example, consider an offshore project where the development team is practicing waterfall model. Suddenly one day, one of the leads feels they should move towards agile. The offshore development team start practicing some of the practices bit by bit. They would try explaining the benefits of agility to the customer, but in vain. But the team continues practicing in bits and pieces without informing PO. This would lead to failure. Basically, one needs complete support from the management, PO and other stakeholders for a successful implementation of agile methods.

2. When development team is managed by "traditional" project managers. For example, consider a situation where the development team wants to practice agile methods. If the management is not aligned with it, the team would fall apart. I have heard of a story, where the development team was practicing some of the agile practices. Due to change in project management, the new project manager with a traditional/CMM/Waterfall background started demanding the team to do more work, and started giving them deadlines to finish work. The team almost fell apart.

Ensure that the management is convinced about the benefits of agile methods.

3. Don't practice without a coach. Passion is a critical aspect to implement the practices of agile methods, but one needs to be aware of the value also. One should know the dependency of some of the agile practices. For example, practicing "just" refactoring without TDD cycle would add more problems than any benefits. In such situations, coach could add lot of value to the team

4. Don't try to implement all agile practices at once. Take one practice at a time, play with it for some time before taking on the new ones.

Practice which works in one organization may not work in the other.


Narayana.D said...

Thanks for sharing this, indeed these are a few of the critical points that need to be addressed

Artem Marchenko said...

There is one more not that obvious reason - collaboration with the neighboring waterfall teams.

Whatever good results, product owner or coach you've got, the agile pilot can be shut down by politics. Agile teams often outperform waterfall guys by any measurable metric, but the ability to *say* the completion date on the first day. No surprise, that surrounding teams might not like their faults to be highlighted.

Agile project not to fail, should get the top management sponsor eager to support the approach until the results are measurable.