Metric

What is survey completion rate?

Definition

Survey completion rate is the share of people who started a survey and reached the end. Where response rate measures whether the invitation worked, completion rate measures whether the questionnaire itself holds attention — and drop-off by question shows exactly where it doesn't.

Formula: Completion rate = completed responses / started responses

What data do you need?

  • Response records with started and submitted timestamps
  • A completed flag or last-question-reached field
  • Question count and position metadata per survey
  • Device field — completion differs sharply on mobile
  • Answer-grain data for drop-off-by-question analysis

SQL pattern

Completion rate and median duration by surveyPostgreSQL
SELECT
  s.name AS survey,
  s.question_count,
  COUNT(r.id) AS started,
  ROUND(
    100.0 * COUNT(r.id) FILTER (WHERE r.completed)
    / NULLIF(COUNT(r.id), 0), 1
  ) AS completion_rate,
  percentile_cont(0.5) WITHIN GROUP (
    ORDER BY EXTRACT(EPOCH FROM (r.submitted_at - r.started_at))
  ) AS median_seconds_to_complete
FROM surveys s
JOIN survey_responses r ON r.survey_id = s.id
GROUP BY s.name, s.question_count
ORDER BY completion_rate ASC;

Common pitfalls

Comparing completion across surveys of different lengths.→ Completion falls with every added question. Segment by question count, or compare drop-off per question instead.
Counting partial responses as complete because answers exist.→ Define completion from the tool's submitted flag or last-page-reached, not from the presence of any answer.
Ignoring suspiciously fast completions.→ Sub-minimum durations usually mean straight-lining. Flag them, and exclude them from quality-sensitive analyses.

Where does this metric apply?

This metric commonly uses data from Typeform, SurveyMonkey, Qualtrics, Tally, plus any warehouse models that provide the same grain.

Dashboards

Integrations

Metrics

Analytics

FAQ

What is a good completion rate?
Short in-app surveys can complete above 80%; long research questionnaires often finish under half. Length, incentive, and audience dominate the number — trend each survey against itself and fix the questions where drop-off concentrates.
How do I find which question loses people?
Land answer-grain data with question positions, then compute the share of started responses that answered each position — the biggest step-down is your leak. The survey performance dashboard includes the SQL pattern.