The customer has just finished a support call when an email asks for a rating. Later, the product requests feedback on a new feature, and an account manager sends a quarterly questionnaire. Each request may seem modest to the team responsible for it. To the person receiving them, they are all requests from the same company.
Survey fatigue describes declining willingness or attention when answering surveys becomes burdensome. It can develop during one demanding questionnaire or across repeated invitations. Both are reasons to examine the experience of taking part, although neither should become a convenient explanation for every disappointing response rate.
Look for a pattern before diagnosing the cause
A drop in completion around a particular question is worth investigating, especially if that question introduces a long grid or asks people to retrieve information. It might indicate fatigue, but it could also reveal a broken form, an intrusive question or a routing mistake. Very fast answers and repeated selections need similar care: identical ratings may be sincere, while a concise comment can still answer the question perfectly well.
Compare completion time, abandonment and missing answers with the questionnaire itself. Where possible, watch a few people work through it. Their hesitation may show that the problem is not length but uncertainty about what the question means. Removing three easy questions will do little to repair one that requires a customer to reconstruct six months of activity from memory.
Reduce the work, not just the question count
A survey with five demanding questions can require more effort than one with ten straightforward ones. Keep the questions that serve the research decision, use routing to avoid irrelevant sections, and test the experience on the devices people will use. If the invitation promises five minutes, base that estimate on a pilot rather than the time it takes someone who already knows the form.
AAPOR’s account of poor respondent experiences includes participants’ frustration with long surveys and other avoidable difficulties. The practical lesson is to consider what completing the study asks of a person, including repeated screening, awkward interfaces and uncertainty about how much remains.
There is no universal safe duration. A relevant, well-compensated specialist study may justify more effort than an unsolicited request after a minor interaction. A change of format is not automatically a remedy either: recording a video can be more demanding, or less private, than writing a sentence.
Coordinate contact across the organisation
A shared record of research invitations helps teams see the cumulative demand on customers. Agree who can contact whom, how overlapping studies will be handled and when someone should be left alone. The appropriate interval depends on the relationship and the work being requested; an arbitrary monthly rule is less useful than a policy informed by actual contact patterns.
When analysing a study, describe who completed it and examine whether missing responses could affect the conclusion. Non-response bias arises when those who answer differ in relevant ways from those who do not; a low response rate alone does not establish its size or direction.
Sometimes the better next step is a smaller programme of interviews or a review of existing behavioural evidence. Good questionnaire design includes recognising when another survey would ask people to repeat something the organisation already has enough information to understand.