How do you increase survey response rate?
Make the survey relevant, easy to access and worth the respondent’s time, then test where participation is breaking down. Improve the invitation, questionnaire, collection method and follow-up together. Do not assume that more reminders—or a different software platform—will solve every problem.
The goal is useful coverage of the people or organizations whose views matter, not simply a larger number. Response rate alone does not establish whether a survey is biased. AAPOR’s Standard Definitions explains why calculating an appropriate rate is part of assessing survey quality, rather than a complete quality judgment.
First, measure the right part of the journey
Distinguish people invited, people starting, people completing and people answering a particular question. These measures reveal different problems. Define the denominator and handling of partial responses before comparing campaigns.
In a simple fictional example, a team invites 200 known eligible members. Eighty start the survey and 60 complete it. Completed responses represent 30% of the invited group, while completion among starters is 75%. The first figure points to overall participation; the second describes what happened after people began.
This is a simplified operational example. Formal response-rate calculations may require different treatment of eligibility, partial interviews and other dispositions. Use a method suited to the study and describe it. An open web link usually does not provide a known invited population from which to calculate the same rate.
Scroll horizontally to see all columns →
| Pattern | What to investigate |
|---|---|
| Few starts | Invitation delivery, trust, relevance, timing and access |
| Many starts, fewer completions | Length, confusing questions, technical problems and sensitive items |
| One group rarely responds | Language, channel, access and whether the invitation reaches that group |
| People stop answering later waves | Changed contact details, burden, relevance, timing and follow-up arrangements |
1. Explain the purpose in a useful invitation
Tell people why you are asking, what the survey concerns, how long it is expected to take and how the answers will be used. Use a recognizable sender and a clear deadline where appropriate.
For example: “Help us plan next term’s member workshops. The survey asks about topics and access needs and takes about five minutes. We will share a summary of the choices and what we can change.” Only use a time estimate supported by testing, and only promise a summary the team intends to provide.
Avoid vague requests to “complete our important survey.” The recipient needs to understand why this particular request is relevant to them.
2. Remove unnecessary work
For each question, identify the decision or analysis it supports. Remove repeated questions that add no useful information and simplify instructions that require explanation.
Keep necessary repeated measures. A follow-up may need to ask the same outcome question again to assess change. Reusing an old answer would defeat that purpose. Stable registration details can sometimes be reused or confirmed; current experience needs current evidence.
Use a pilot to estimate completion time and identify difficult sections. Shorter is not automatically better if it removes the evidence required for a useful result. Aim for the shortest survey that can answer the intended question well.
3. Show questions that apply
Use clear screening and branching where appropriate. Someone who did not use a service should not be required to rate it. They may have a separate reason for not using it that is worth understanding.
Test the paths and the resulting data. Do not interpret people excluded by design as unanswered or dissatisfied respondents. The survey logic guide explains how routing affects denominators.
Keep common questions visible to the groups you intend to compare. Asking only dissatisfied respondents for an explanation produces evidence about that selected group.
4. Make access practical
Test the form on a phone, with the expected connection quality and relevant accessibility needs. Check whether login, large uploads or complex layouts create avoidable barriers.
Offer suitable language or assisted-completion options when the audience needs them. If you combine web, telephone, paper or other modes, retain the collection method and consider its effect on comparability. See mixed-mode data collection.
Do not send the same channel repeatedly to a group it is failing to reach and assume that silence means lack of interest.
5. Explain privacy accurately
State whether the survey is anonymous, confidential or linked to a record, and explain the intended use. Do not call it anonymous if identifiable responses or links can be accessed by the team.
For employee or other sensitive feedback, people may need to understand who sees individual answers and how group reporting works. Avoid collecting unnecessary identifying details. Also consider whether free text could identify someone even when direct identifiers are absent.
Trust is not created by a reassuring label alone. The collection and access arrangements need to match the explanation.
6. Use reminders thoughtfully
Plan reminders with an appropriate cadence and stop when a person has declined or no further contact is appropriate. Where the system can identify completed responses under the stated arrangement, avoid repeatedly asking those people to respond again.
For anonymous collection, explain any limits on targeting reminders. Do not imply that the team cannot identify answers while using response content to personalize a follow-up.
Test timing with your own audience. A universal “best day” or a fixed number of reminders is not a reliable strategy across employees, members, customers and field programs. Monitor complaints and opt-outs as well as starts.
7. Evaluate incentives as part of the design
An incentive may make participation more feasible or attractive, but it is not automatically appropriate. Consider the audience, burden, study requirements, fairness and administration.
Explain the terms accurately. Keep contact details collected for an incentive separate from answers when the privacy arrangement requires it. Review duplicate or ineligible participation without assuming every fast response is invalid.
Test whether the incentive improves useful coverage, not just the raw count. Do not assume it is always less important than showing feedback was used; the effect depends on the setting.
8. Show what happened after the last survey
Share a short account of what the team heard, what it decided and what remains unresolved. People should not have to guess whether anyone examined their answers.
Be specific without overpromising. “Several respondents found the registration instructions unclear; we revised them and will review the next cycle” is more credible than saying every suggestion will be implemented.
This practice can support a useful ongoing relationship, but do not promise a guaranteed response-rate increase. If the team cannot act on a request, explain the constraint and the next review point.
9. Check who is still missing
Look beyond the overall rate. Where appropriate information is available, compare participation across relevant groups or locations. A strong overall response can still leave an important group poorly represented.
Suppose a fictional network has 100 members in each of two regions. One contributes 70 responses and the other 20. The total is 90 out of 200, or 45%, but the regional coverage differs substantially. Investigate the second region’s invitation reach, language and access before treating the combined result as equally representative of both.
Do not remove hard-to-reach members from the denominator merely to improve the metric. Document legitimate eligibility changes separately from nonresponse.
10. Test one change and inspect the result
Choose the likely bottleneck and define success before changing several things at once. A practical test could compare two clear invitation messages with otherwise similar collection arrangements.
Use a suitable comparison design and enough evidence for the intended conclusion. A larger completion count after changing the invitation does not by itself prove the invitation caused it; timing or audience differences may also matter.
Track starts, completions, item missingness, relevant group coverage and complaints. Record the change so the next team member understands what was tried.
Plan recurring surveys as a continuing relationship
Coordinate requests across departments to avoid unnecessary overlap. Define which information is stable, which must be updated and which outcome questions need repeating. A member may legitimately need different questions for registration, an annual return and an event.
In a federated organization, local teams can tailor collection while sharing the core definitions needed for comparison. Record eligible population, period and method so the central report can interpret the results.
Sopact’s focus is self-managed collection, analysis and governance of continuing evidence. Evaluate whether the proposed workflow reduces unnecessary work and makes findings usable. No platform can guarantee participation or remove the need to review the data.
Start with the weakest point
Review one recent survey: who was invited, who started, where people stopped and which groups were underrepresented. Choose one practical improvement and record what you will measure.
For the wider collection plan, use survey software requirements. For repeated measurement, see baseline survey planning. For making the responses useful after collection, read how to analyze survey data.
Frequently asked questions
What is a good survey response rate?
There is no universal threshold. Consider the audience, recruitment method, purpose and potential nonresponse bias. Define the rate and assess who is missing.
Is completion rate the same as response rate?
No. Completion among people who start answers a different question from participation among an invited eligible group. State the denominator for each.
Should every follow-up ask fewer questions?
Remove unnecessary repetition, but keep measures needed to assess current conditions or change. Reusing an old outcome answer is not a follow-up measurement.
Do reminders always improve results?
They may help, but their value depends on the audience and collection arrangement. Monitor unwanted contact and fix relevance or access problems rather than relying only on repetition.
Can a high response rate still produce biased findings?
Yes. The rate alone does not show how respondents differ from nonrespondents on the questions that matter. Examine coverage and the study design.
Can software guarantee a higher rate?
No. Software can support collection and follow-up, but participation depends on several factors. Test the workflow and report observed results without guaranteed uplift claims.
