An automation testing checklist helps you assess what a workflow does when real work is less tidy than a demonstration. Missing fields, repeated requests and unavailable services all deserve attention. Define the expected business result before testing individual steps.
Start your automation testing checklist with outcomes
Write down what should happen to the input, the destination record and any messages. Specify what must not happen, such as sending a customer two confirmations or creating a record before an approval. Give each test an expected result that a reviewer can assess.
For an illustrative website enquiry workflow, success might mean one assigned record and one accurate acknowledgement. A run showing as completed is not sufficient if the record went to the wrong team.
Include the main exception cases
- Required information is missing or invalid.
- The same request arrives twice.
- A connected service is unavailable.
- An earlier step succeeds before a later step fails.
- Access credentials expire or change.
- A human pauses or corrects the process.
Use safe test data and controlled recipients. Keep the test environment separate from live work where possible, and confirm how connected applications are prevented from taking unintended actions.
Check error handling and investigation
The n8n error-handling documentation explains error workflows and reviewing failed executions. These features can support investigation, but they need configuration and an owner who acts on the result. Other platforms have their own mechanisms, which should be tested in the actual setup.
Ask what the alert contains, who receives it and how they identify the affected business record. Avoid putting unnecessary personal information or secrets into notifications and logs. Decide how long diagnostic information is retained.
Prove that retrying is safe
If a customer message was already sent before a later step failed, rerunning everything could send it again. Ask the builder how completed actions are recognised and how retries avoid repeating them. Test this deliberately in a controlled environment.
Also consider a workflow that never starts because its trigger is misconfigured. Compare expected inputs with recorded outputs rather than relying exclusively on execution error alerts.
Write the operating handover
Name the owner, support route and authorised people who can pause or retry work. Document the manual fallback and the checks needed before restarting. Have someone other than the builder follow the instructions during the test.
Record test results, unresolved issues and the approval to launch. Repeat relevant tests after changes to forms, fields or connected services.
For help planning dependable website workflows, contact Kilwhiss Automate and explore its automation integration services.