As we follow The Anaplan Way methodology, after completing our implementation phase, the next phase is testing (UAT), you may have two UAT round totally depends on the timeline, during UAT, we suppose to test each user stories and Acceptance criteria with the user.
**(during implementation phase itself we are doing our internal testing based on user stories & ACs)
Few below points, I would like to recommend you before you start your unit testing 1. First, prepare your Dev model * Make sure all the necessary list must be marked as production lists. * Set-up all the roles * Set -up Access which include (Module, lists, Actions and dashboard) based on each role 2. Create a test model from the revision tag. 3. Prepare and upload the data (if you are using any Anaplan integration then you can test here as well) Make sure after uploading the data file, check the totals. 4. Based on the test script, test each acceptance criteria of each user stories one by one based on the roles. ( If possible try to test all the cases without workspace admin, this will really help in the case of process and workflow)
There are two different types of testing - one is unit testing & other one is SIT
As @Akhtar.shahbaz pefectly described these two and the timings of these two. However I would like to add couple of points here
1.Be very careful when marking Lists as Production Lists. If you mark List as Production List and later you realized the list needn't be Production one then all your data related to that Production list will be wiped out. But when your list is not marked as production list you can always come back and make that as a production list with no consequences whatsoever.
2. This Point is open for everyone to comment upon - Sorry to hijack your thread here 🙂 -
When you are in implementation phase you can create Test model from Dev model and then start testing your scripts and thorough testing of your user stories. But once you go live you may have to create another test model from Production model and you may want to keep your test model as close to Prod as possible.