What were you trying to complete?
The intended task establishes what the failure interrupted and which context a recovery message must preserve. Keep private task data out of the response.
Product and software research / Survey guide
Product and software research · Practical survey guide
Application error recovery survey
Improve guidance after a task fails.
Opens product registration. Example questions are not imported automatically.
BEFORE YOU INVITE PEOPLE
Local planning aid. No answers are submitted or saved.
Read the answers carefully
Keep the conclusion within the evidence.
Do not collect stack traces containing private data.
Errors ask users to retry without explaining consequences. This is a synthetic example, not a customer result. Do not collect stack traces containing private data.
A useful starting point
Questions to understand the actual experience.
Improve guidance after a task fails. Use the prompts relevant to your audience and allow people to skip situations they did not encounter. Keep response handling consistent with what you explain in the invitation.
If nothing was unclear or difficult, say so. Skip questions about steps you did not experience; use not applicable where appropriate.
What next action did the message suggest?
The suggested action reveals whether the failure message offers an understandable recovery route instead of merely announcing a problem. Clarify actual supported steps, without inventing fixes or implying the underlying failure has been resolved.
What recovery information was missing?
Missing recovery information identifies the uncertainty preventing an appropriate next action. Explain verified retry or assistance routes without promising that the underlying failure is resolved.
If nothing was unclear or difficult, say so. Skip questions about steps you did not experience; use not applicable where appropriate. What were you trying to complete? What next action did the message suggest? What recovery information was missing?
Download the questions as text ↓
Examples to adapt, rather than a validated research instrument.
A worked interpretation example
From a comment to a practical next step.
Illustrative scenario. These are not customer results.
Keep the finding specific
Errors ask users to retry without explaining consequences.
Choose an action to test
Explain whether retrying duplicates the action.
Check the experience again
After trying the change, repeat “What were you trying to complete?” with people who experienced it. Check whether they still describe the original issue: errors ask users to retry without explaining consequences. Compare explanations and sample context, rather than claiming the change caused an improvement.
Before you send it
Check what your wording assumes.
Would you agree that everything about application error recovery was clear and easy?
What were you trying to complete?
The first wording combines an assumed positive outcome with two different judgments. The revised prompt asks about a specific experience and permits an inconvenient or uncertain answer.
Choose the response format
Make room for the explanation.
| Question | Suggested format | What to preserve |
|---|---|---|
| What were you trying to complete? | Optional short text | Retain the specific point or condition described; avoid replacing it with an unexplained rating. |
| What next action did the message suggest? | Optional written explanation | Keep context that distinguishes different experiences. Do not collect stack traces containing private data. |
| What recovery information was missing? | Optional improvement suggestion | Keep the suggested change separate from whether it has been tested. Explain whether retrying duplicates the action. |
Make questions optional where appropriate. If you add a rating scale, label its endpoints, keep one idea per question, and retain a follow-up for reasons. These examples are not a validated measurement instrument.
Build with SurveyTeams
Draft your own application error recovery questionnaire.
Use these examples to draft your questionnaire. Create a survey draft, add questions, and check the respondent preview in SurveyTeams. Additional features depend on your account and plan.
Create your survey in SurveyTeams
Opens product registration. Example questions are not imported automatically.
Learn about the product ↗From question to next step
Make the feedback useful.
Set the decision
Start with the decision: Improve guidance after a task fails.
Invite relevant voices
Invite people with relevant experience. Ask after someone has encountered the situation described in “What were you trying to complete?”, while they can still recall the details. For material testing, show the actual draft first; for planning, ask before the next relevant activity.
Choose an action
Explain whether retrying duplicates the action.
Before you send it
Questions, answered.
Who should receive this questionnaire?
Product users who attempted the described task. Ask only about steps each person encountered; do not treat a voluntary response sample as representative.
When should I ask these questions?
Ask after someone has encountered the situation described in “What were you trying to complete?”, while they can still recall the details. For material testing, show the actual draft first; for planning, ask before the next relevant activity.
What should I investigate in the answers?
Errors ask users to retry without explaining consequences. This is an illustrative pattern to look for, not a claim about your respondents. Do not collect stack traces containing private data.
What would a practical follow-up look like?
Explain whether retrying duplicates the action. After trying the change, repeat “What were you trying to complete?” with people who experienced it. Check whether they still describe the original issue: errors ask users to retry without explaining consequences. Compare explanations and sample context, rather than claiming the change caused an improvement.
How should I ask “What next action did the message suggest?”?
The suggested action reveals whether the failure message offers an understandable recovery route instead of merely announcing a problem. Clarify actual supported steps, without inventing fixes or implying the underlying failure has been resolved. Use an optional written explanation when predefined choices would hide relevant context.
Can I use the questions in SurveyTeams?
Copy the prompts or download TXT, CSV, JSON, or a Markdown worksheet, then adapt them in your own survey. The CTA opens registration; it does not import this example automatically. Verify available features in your account.
Continue exploring