Surveys and questionnaires let design teams gather structured data from large numbers of users simultaneously, something no qualitative method can match at scale. They are excellent for quantifying attitudes, validating hypotheses formed during qualitative research, and segmenting user populations by behaviour or preference. The challenge is that surveys can only answer the questions you already know to ask, which is why they work best after exploratory research has shaped your understanding.
What It Is
A survey is a set of standardised questions delivered to a sample of users, typically online. Questions range from multiple-choice and rating scales to short open-text responses. The goal is to collect data that can be aggregated, compared across segments, and tracked over time. Well-designed surveys use validated scales where possible, such as the System Usability Scale (SUS) or the Net Promoter Score (NPS), to enable benchmarking against established norms.
How to Run It
Begin with a clear research objective: know exactly what decisions the survey results will inform. Draft questions starting with the most engaging and ending with demographics. Keep the survey under ten minutes to protect completion rates. Pilot the survey with three to five colleagues to catch ambiguity before it reaches your real audience. Distribute through channels where your target users are already active. Aim for a sample large enough to detect meaningful differences, typically at least one hundred responses for basic segmentation.
When to Use It
Use surveys to validate patterns identified during qualitative research, to measure satisfaction or usability at scale, and to track changes in user attitudes over time. They are especially useful for prioritisation exercises: ask users to rank features or rate pain points to guide product roadmaps. Surveys are not suited to understanding why users behave in certain ways or to exploring problems you have not yet identified. Pair them with interviews to get both depth and breadth.
Tips for Success
Avoid double-barrelled questions that ask two things at once. Each question should address one idea only. Use consistent scales across a survey and label every point on a rating scale to avoid ambiguous interpretation. Be cautious about social desirability bias: people often answer how they think they should behave rather than how they actually do. Anonymous surveys tend to produce more honest responses, especially when asking about sensitive topics like frustration with a product.
