What is the biggest problem with surveys?

Responses don't always reflect what people actually think or do - that's the core issue. The people who bother to respond are often different from the ones who don't, which introduces bias before you've even asked the first question. On top of that, people misread questions, forget details, give the answer that sounds good rather than the true one, or share opinions about experiences they haven't really thought through.

You can end up with numbers that look precise and are still measuring the wrong thing - usually because the questions lead somewhere, the sample is off, or the timing is wrong. Surveys work best when you're honest about their limits and pair them with behavioral data, customer interviews, support conversations, and whatever other evidence you have.