Experiment & Research Method
Surveys
Surveys collect structured, quantitative feedback from many people to measure attitudes and behaviors at scale.
- Fidelity
- Medium
- Effort
- Low
- Time to run
- 1–3 weeks
What is a Surveys?
Surveys gather standardized responses from a large number of people through a fixed set of questions. They are the workhorse of quantitative attitudinal research: fast, scalable, and ideal for measuring how widespread an attitude, preference, or self-reported behavior is across a population.
Their power is scale; their risk is bias. Poorly worded questions, leading options, and skewed samples can quietly produce confident but wrong conclusions, so careful design matters more than survey length.
When to use it
Good fit
- You need data from many people to size or quantify an attitude or behavior.
- You want to track a measure (like CSAT or NPS) consistently over time.
- You already know the right questions to ask (often from prior qualitative research).
Reach for something else when
- You are exploring an unknown problem — interviews will teach you what to ask first.
- You need to understand deep "why" and context (self-reports are shallow and sometimes inaccurate).
- You cannot reach a representative sample, which would bias the results.
How to run it
Define the decision and questions
Start from the decision you need to make and write only the questions that inform it — every extra question costs response quality.
Write unbiased, answerable questions
Avoid leading and double-barreled questions, provide balanced options, and prefer concrete behavior over hypotheticals.
Sample the right people
Target a representative group and be mindful of who is likely to respond, since non-response can skew results.
Pilot before you launch
Test the survey with a few people to catch confusing wording and broken logic before it reaches everyone.
Analyze and segment
Look beyond averages: segment responses, watch for bias, and pair surprising results with qualitative follow-up.
What you'll learn
Quantified, comparable data on attitudes, preferences, and self-reported behavior across a large sample — ideal for sizing problems and tracking sentiment over time.
Frequently asked questions
When should you use a survey instead of interviews?
Use a survey when you already know the right questions and need to measure how common an attitude or behavior is across many people. Use interviews when you are exploring an unfamiliar problem and need depth and context. A common pattern is interviews first to learn what to ask, then a survey to quantify it.
How do you avoid bias in surveys?
Write neutral, single-idea questions with balanced answer options, avoid leading language, ask about concrete behavior rather than hypotheticals, and recruit a representative sample. Pilot the survey first, and remember that only-the-motivated responding can skew results even when the questions are well written.