When researching users is required, many products and startups go to do surveys, and some even conduct focus groups. Conducting surveys may seem like a more effective way than conducting interviews, but at the start it is not a good way to do it.
1. Surveys assume that we know the right questions to ask.
How do we make a survey that covers all the right questions if we don't know them ourselves yet? For example, during user interviews, we may ask clarifying questions and explore areas that were beyond our initial understanding. An interview is about exploring what we don't know.
2. Polls suggest that we still know the right answers.
In a survey, we need to not only ask the right questions, but also provide the right answers. How many times have you answered [Other] in surveys? The best answers and insights come from asking open-ended questions.
3. We don't see people during surveys.
Body language, facial expressions, expression of emotions are the same full-fledged signals for Problem / Solution Fit as answers to open questions.
4. Focus groups are generally utter nonsense.
The problem with focus groups is that they reflect groupthink, which is of no use at all for most products.
5. Surveys are quite effective for testing what we have already learned in the interview.
An interview is a qualitative study that is effective for finding signals to confirm or disprove hypotheses with a small sample size.
A survey is a quantitative study. Once we have verified our hypotheses through interviews, we can use what we learned to design a survey and quantify those findings. The purpose of the survey will no longer be to learn something new, but to show the scalability or statistical significance of the results.