Try It Yourself: The Pilot
If you are still interested, it is now appropriate to see if empirical software development solves your problems and meets your needs. This is the time to conduct a pilot—a small-scale preliminary study—to evaluate feasibility, time, and costs and to uncover adverse effects. In this chapter we tell you:
1. How to run a pilot study of this new approach to software development.
2. What information you can gain from this pilot.
3. How this has worked for others who have done it (the issues they discovered and what they needed to address).
You are critical. In this chapter we describe to you what you need to do to help this succeed, step by step. We also provide a more thorough description and some examples; these can be read later when the pilot is under way.
The process followed during the pilot is quite simple:
1. Form the team.
2. Figure out what you want to pilot.
3. Do a small piece of it, completely.
4. Evaluate what you want to do next.
5. Assess what can be improved and do it.
6. Continue iterating Steps 3 through 5 until satisfied.
Before starting, figure out what will happen if the pilot fails or succeeds. How much time and money are you willing to spend on something that isn't working for you before giving up? What will you do if valuable functionality is delivered, increment by increment, and you want to proceed? By the end of the pilot, you will know if you want to engage in empirical software development on a larger scale. Of ...