Join 34,000+ subscribers and receive articles from our blog about software quality, testing, QA and security.
 

Understanding the progress/forecast calculation for Test Runs

We have started our TestRail project without any estimates for Test Cases (and now have several hundred testcases in our project without any estimates).
We thought that it should be possible to start recording the duration in future test runs, and by doing so we will also get more and more accurate forecasts (without having to estimate all the testscases).
We now have already collected some duration data, but the forecast is months aways from being at least a little bit accurate.

Example of a very small TestRun, in which all tests have been executed:
Completed: 100% (11/11)
Elapsed: 2h 10m
Tests / day: 11
Hours / day: 334

Can anyone explain to me, how the big number “Hours / day” was calculated? None of the 11 test cases has recorded more than 20 minutes duration (also in past test runs). The total test execution time for all 11 test cases was 130 minutes.
But the forecast is 334 hours …?

Any insight and explanation is appreciated!
Thank you!
Best regards,
Juergen

Update:
For a Test Suite I have added estimates to all test cases.
Now the metric for a Test Run of this Suite (Test Run not started yet) looks like this:

testrail_forecast01

Still don’t get it why 7 testcases each with estimate and duration of previous test runs of 20 min or less (7*20 = max. 140 min) get a forecast of 200 hours…