Functional vs Regression Testing: What's the Difference?

Functional and regression testing get lumped together as 'QA', but they answer different questions and you need both — especially with a closed test deadline staring at you.

What functional testing is for

Functional testing answers one question per feature: does it do what it's supposed to? You test with a defined expectation — this button opens checkout, this flow completes a purchase, this form rejects bad input — against a clean state. It's the first layer: the app's claims about itself, verified.

What regression testing is for

Regression testing asks: after you changed something, did anything else stop working? A fix to the login flow can quietly break password reset; a new permission can break an old feature. Regression passes re-run the scenarios that already passed, on the new build, hunting for exactly that class of 'unrelated' breakage.

Fitting both around a closed test

1. 1

   Functional pass before the window

   Before uploading the closed test release, one complete functional pass on the core flows — on the device mix your testers will use.

1. 1

   Regression pass after every fix

   Each time testers report a bug and you ship a fix, re-run the core scenarios in a short regression pass before the update goes to the track.

1. 1

   One final pass before applying

   Before production access application, run both passes on the exact builds testers have been using — the review is an application on the app you actually tested.