I've got a client that I would like to try pair programming with, but they've been a little reluctant, and with good reason. His fears:
- I'm paying two people to do the same work - is the task they are working on going to be done twice as fast or twice the quality, or a mixture?
- Which person is responsible for an item and how do I keep track of it?
I guess all valid and normal points. My rebuttal is with pair programming - yes it might cause a task to be done a little slower, however the quality will be much higher and less likely need to be revisited due to bugs, but that doesn't seem to justify the cost to him. He's agreed to try it for a week - but I don't think that is long enough to see the benefits.
So, my questions are..
Does anyone have a resource, online or otherwise, that has done case studies with pair programming? Metrics such as a percent change in amount of time spent fixing bugs would be cool.
Does anyone else have any tips to convincing a client that pair programming is a practical approach to software development?
What is, in your opinion, the minimum length pair programming should be done before you can measure it's effectiveness?
