Ticket Ticket Impleme nt Impleme nt Release Bug report UT UT UT E2E Automated test This test code is out of date When will I do exploratory test? What kind of test cases should we automate? I believe automated test verify our quality automatically
1) Out-of-date test codes Test script 1. Choose “Small” in “Size” 2. Type “1” into “Quantity” 3. Type tomorrow’s date into “Delivery date” 4. Click “Add to cart Old New Delivery date is removed in the new UI
to ambiguous specs Let’s add “Order Now” button to allow users purchase item immediately Gotcha (And I will miss the test pattern when the user has not entry their shipping address)
something weird • Site Performance is slightly decreased • It’s functional, but not so useful • What will happen if I click the button 100 times in a row? But I can find them!
we run them AFTER changing the app behavior • Test code does not follow the app behaviour automatically • But human is capable to update manual test script dynamically
Maintain test code and app’s source code at the same time Test script 1. Choose “Small” in “Size” 2. Type “1” into “Quantity” 3. Type tomorrow’s date into “Delivery date” 4. Click “Add to cart Test script 1. Choose “Small” in “Size” 2. Type “1” into “Quantity” 3. Click “Add to cart
separately • PO designs specs, Dev implements them, and QA implement the test cases ◦ PO may forget to write a part of spec ◦ Dev and QA implement without being aware of it
specs and edge cases during the ticket refinement • Create a concrete test cases and agree with them • Devs implement the feature and the automated tests according to the agreed test cases
tests Human eyes are capable to catch something weird • Site Performance is slightly decreased • It’s functional, but not so useful • What will happen if I click the button 100 times in a row? But trying to catch them at the test phase all at once is not sustainable
80% of network request and reduce 30% of response time • The closed beta test gets feedback from the • It turned out that there is a potential risk of data duplication from the exploratory test