UX survey checklist: before, during, and after data collection
This checklist covers the full survey cycle — from question design through analysis to reporting. The “Before” phase takes the most effort, especially pilot testing: a five-person pilot catches problems that would otherwise corrupt hundreds of real responses.
Before the survey
- Define the research questions and the decisions they support
- Create a question-to-research-question mapping table
- Choose question types for each question (single-select, Likert, multi-select, open-ended)
- Write questions following the five principles: neutrality, specificity, singularity, answerability, clarity
- Include standardized instruments (NPS, SUS, CSAT) where applicable
- Write the introduction text (purpose, time estimate, data usage)
- Set up branching logic and skip patterns
- Randomize answer options for multiple-choice questions
- Add a progress indicator
- Run a pilot test with 5-10 people using cognitive interviews
- Fix all issues found in the pilot
- Calculate target sample size and plan distribution
During the survey
- Launch to the planned distribution channels
- Monitor completion rate daily — investigate if below 50%
- Check for technical issues (broken logic, display problems) in the first batch of responses
- Send a reminder if responses plateau below target
After the survey
- Clean data: remove straightliners, speeders, and spam
- Calculate response distributions, means, and standard deviations for closed-ended questions
- Cross-tabulate by segments and run significance tests
- Code open-ended responses into themes (manually or with AI)
- Write the report using “What / So What / Now What” for each finding
- Include a limitations section (response bias, sampling gaps)
- Prioritize recommended actions by impact
- Share the report with stakeholders and agree on next steps