The "what happens when an approver is away" test is the one that gets skipped most. For purchase approvals in a small plant, I'd add two more cases: a PO that gets amended after approval (does it re-trigger approval, or does the old sign-off silently carry over?) and a PO just under the threshold split into two lines. Both tend to slip through in real use. Delegation is worth testing too, with an expiry date so a holiday substitute doesn't keep approval rights for months. Short written test scripts per role also keep the checks repeatable after upgrades. Do you test with copies of real user roles, or with dedicated dummy users?
I’d use dedicated test users with the same roles and record permissions as real buyers, approvers, and delegates. That way we test the access each person actually has without using anyone’s login.
Amended PO: Check whether a change that requires fresh approval sends it back for sign-off.
Split lines: Check that the approval threshold applies to the full PO total.
Delegation: Confirm the substitute can approve during cover and loses that ability when it expires.
I’d save those as short test scripts so the team can rerun them after an upgrade.
Dov at HireDov
The "what happens when an approver is away" test is the one that gets skipped most. For purchase approvals in a small plant, I'd add two more cases: a PO that gets amended after approval (does it re-trigger approval, or does the old sign-off silently carry over?) and a PO just under the threshold split into two lines. Both tend to slip through in real use. Delegation is worth testing too, with an expiry date so a holiday substitute doesn't keep approval rights for months. Short written test scripts per role also keep the checks repeatable after upgrades. Do you test with copies of real user roles, or with dedicated dummy users?