Qualitative test methods are great for discovery or early validation (when assumptions are still quite risky). You'll gather relatively few results, but dig deep into your customers motivations, uncovering the WHY behind their actions

Quantitative test methods show you WHAT users do in an (almost) real-world situation. In some cases, you'll be able to get data for a large enough sample size to give you some degree of confidence that your test results reflect your target audience's behavior (statistical significance, or something close to it).

The key is in combining both types.

The matrix in the image shows my favorite test methods divided by phase (pre- or post-launch) and qual vs. quant.