Early-stage research needs different methods than validation research. When exploring problems, use interviews and contextual inquiry to understand user needs. When evaluating solutions, use usability testing and A/B testing to measure effectiveness. Match your method to your question.

The mistake many teams make is over-researching or under-researching. You don't need 50 user interviews to identify common pain points—5-8 well-conducted sessions will surface most issues. But you do need quantitative data when making high-stakes decisions that affect millions of users.

We've seen companies waste months on elaborate research that could have been answered with a simple usability test. And we've seen companies launch features based on a hunch that user research would have instantly disproven. Find the balance. Research smart, not just hard.