Survey Fatigue

Your response rate on the second or third send to the same cohort is materially lower than the first. A drop of 10+ percentage points between sends is a fatigue signal.

Average completion time is dropping but you have not shortened the survey. Users are rushing - selecting answers without reading, which produces random data that looks real.

Users select the same answer option (e.g., always "3" or always "Neutral") for every question. The response pattern is too consistent to be honest. Common in surveys longer than 5 questions.

Users open the survey but do not reach the final question. Drop-off before the last question in a 3-question survey suggests the survey is too long, poorly timed, or the questions are unclear.

Open-text responses are getting shorter and less specific over time. Users who previously wrote two sentences now write "ok" or leave it blank. They are still technically responding but no longer engaging.

NPS or CSAT scores are moving but nothing changed in your product or support quality. Fatigued users give arbitrary scores - which creates noise in your trend data and masks real signals.

Set a frequency cap per user

Enforce a maximum of one survey per user per 60-90 days across all survey types. If NPS fires and CSAT is also scheduled, the second survey waits until the cap resets. This one change produces the largest improvement in response rate for teams running multiple survey types.

Shorten surveys to 1-2 questions

One rating question (NPS, CSAT, or CES) plus one optional open-text follow-up. That is your baseline survey. Only add a third question if you have a specific, actionable need for that answer that cannot be answered by the first two. Remove questions you have never acted on.

Trigger on behavior, not calendar

Replace monthly batch sends with event-triggered surveys. CSAT fires after support resolution. CES fires after onboarding completion. NPS fires 30 days after signup. Behavioral triggers produce 2-3x higher response rates than calendar sends because the survey is relevant to what the user just did.

Pick the right moment in the session

Never show a survey mid-workflow or immediately on login. Show it after the user has completed an action - they just resolved a support ticket, finished onboarding, or used a feature. The task is done. The experience is fresh. They have 30 seconds. That is your window.

Rotate survey types across cohorts

Not every user needs to receive every survey type. Run NPS on your full active user base quarterly. Run CSAT only on users who interacted with support or a specific feature. Run CES only on users who completed a high-effort workflow. Segmenting who receives which survey reduces per-user frequency without reducing total coverage.

Close the loop visibly

When you fix something users flagged in a survey, tell the users who flagged it. A single email - "You told us onboarding was confusing. Here is what we changed." - does more to sustain future response rates than any survey copy optimization. Users respond to surveys they believe are read.