Today I posted, in Spanish, about the four characters that show up in almost every Salesforce project. One of them is the client who promised three weeks ago to test everything and still has not logged into the Sandbox.
I do not read that as lack of interest. Testing competes with everyone's day job, and the day job usually wins. The problem is that the project finds out on Go Live day. The full piece, with the other three characters (the Admin fixing Production on a Friday evening, the consultant who shows up on Go Live day, and the sponsor who asks what a Flow is), is here: https://discover.egafutura.com/los-cuatro-personajes-de-un-proyecto-salesforce/?utm_source=devto&utm_medium=referral&utm_campaign=amplify
What I want here is evidence. The sandbox already records who logged in and who created data. Three checks you can run halfway through UAT, while there is still time to do something about it.
Before anything: did the testers get in at all?
When a sandbox is created or refreshed, Salesforce appends .invalid to user email addresses so the sandbox does not email real people by accident. That is the right default, but it also means a tester whose email was never fixed may have never received a login link. Check that first, for the users who are supposed to test.
1. Last login of every UAT user
Group your testers with a permission set (here called UAT_Testers) and query their last login in the sandbox:
SELECT Name, Username, Email, LastLoginDate FROM User WHERE IsActive = true AND Id IN (SELECT AssigneeId FROM PermissionSetAssignment WHERE PermissionSet.Name = 'UAT_Testers') ORDER BY LastLoginDate ASC NULLS FIRST A null LastLoginDate means that user never logged in. An Email still ending in .invalid usually explains why. If half of the list is null at the midpoint of the testing window, that is the day to raise it, not the Go Live meeting.
2. Did they do anything once they were in?
A login is not a test. If the test cases say "create a quote" or "approve a discount", the records should exist, created by the testers:
SELECT CreatedById, COUNT(Id) total FROM Opportunity WHERE CreatedDate = LAST_N_DAYS:21 AND CreatedById IN (SELECT AssigneeId FROM PermissionSetAssignment WHERE PermissionSet.Name = 'UAT_Testers') GROUP BY CreatedById Swap Opportunity for the objects in your test plan (Quote, Order, or the custom objects your ERP uses). A tester with zero records on the object they were supposed to test has not tested it, whatever the status column in the spreadsheet says.
3. Write the test cases where the testers live
The spreadsheet of test cases is usually the first thing to go stale. A small custom object in the sandbox works better: one record per case, with the tester, the expected result, a status and the date it was run. Then the same kind of query gives you a real picture:
SELECT Tester__r.Name, Status__c, COUNT(Id) total FROM UAT_Case__c GROUP BY Tester__r.Name, Status__c It also gives the sponsor something short to read every week: how many cases passed, how many failed, how many nobody ran.
What the queries cannot do
They tell you who did not test. They do not make anyone test. That part is a calendar slot, booked like any other meeting, with concrete cases written together with the people who will use the system.
At EGA Futura we build a native ERP on Salesforce for mid-sized companies. When an implementation adds invoicing, inventory or accounting, the UAT list gets longer, with people from finance, purchasing and the warehouse, and a null in LastLoginDate becomes the earliest warning you will get.
