Skip to content

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