The testing pyramid

Not all tests are the same shape

PawWalk now has real logic worth breaking: price math, a scheduling algorithm, distance calculations, networking, and a handful of screens. If you tested all of it the same way β€” by launching the whole app and tapping through it β€” you'd have a test suite that takes ten minutes to run and fails for reasons that have nothing to do with your bug (a flaky animation, a slow simulator, a network hiccup).

The testing pyramid is a shape, not a rule: many fast unit tests at the bottom, fewer integration tests in the middle, and a handful of slow UI tests at the top. Wide at the bottom, narrow at the top.