I have not attempted to calculate ROI. We are still entering tests from our manual testing system and have yet to determine how much personnel time it will take to run all the tests during a production release.

I do not have insight into ROI. I can tell you that since we distribute our tests using the Host/Target model we saved a great deal of execution time after we switched from SilkTest. Since our execution time decreased substantially,

We reduced the regression cycle from 10 days to three days, though it could be reduced further but we had dependency on configuration systems. It was a significant achievement from both the offshore and the onshore teams. Management really appreciated us and was happy.

The ROI was quick, and is based on the following: 

1. Developed fifteen regression tests in less than four months, executing the same tests twice per week.
2. Manual testers took three days of effort to perform testing on a huge volume of data in Excel, whereas with automation was achieved in an hour.
3. ROI was realized from the sixth month, onwards.

ROI was generated within the first year as we developed the test cases within a short period of time. Months effort was brought to few days effort.

We had a pretty decent ROI for this venture of investing in AscentialTest. We were able to find out many crucial bugs in our regression tests developed using AscentialTest, which improved the quality of the software decently.