For an ROI discussion, I recommend Appendix A from Watt’s and Jim Over’s book Leadership, Teamwork, and Trust http://www.amazon.com/Leadership-Teamwork-Trust-Competitive-Engineering/dp/0321624505. I also recommend the book.
TSP is in investment in human and organizational capital.You absolutely should expect a positive return. Watts and Jim provide a framework, but make your own calculations.
Startup costs include
- Executive Seminar
- Team leader preparation and training
- Developer preparation and training
Operational costs include
- initial launch preparation
- cycle post mortem and relaunch
- ongoing coaching support
- engineering/developer work
Long story short and in round numbers; if you have 100 developers each with 200 work days, that’s 20,000 staff days effort. Your TSP overhead will be coaching support and training. The one time costs come to about 750 days effort. The operating costs are comparable.
Now the payback, we have lot’s of data showing that non-TSP teams spend about 1/3 of their total effort in test defect fixing. That’s 50 of every 150 days. TSP teams spend just 8% of their time in late rework, or 9 of 109 days. Round down that 27% saved effort to 25% In just one year, the gross savings becomes 5000 developer days. That’s a net of 3500 days including the one time training costs.
Use your own data and do your own math.
BTW the book also received a very positive review from CIO.