Customer onboarding survey questions: 12 for SaaS teams
Find where new customers get stuck, understand what they need, and choose which onboarding problem your product or UX team should investigate first.
Start with the three-question check-in below. Add a question when its answer will help you decide what to do next. Replace [Product], [setup milestone] and [task] with terms your customers recognize.
The 12 questions at a glance
Scroll sideways to see all columns.When to ask
After a setup attempt, ask about clarity and obstacles. Invite customers who stopped partway as well as those who finished. Their replies may reveal different problems.
After someone has tried the task they signed up to do, ask about completion, confidence and usefulness. Connecting a data source is a setup step; producing a useful report is a separate outcome.
Send the relevant questions while that experience is fresh, through a channel your team already uses. Try the wording with a few customers first and ask them to explain what each question means to them. Pew Research Center recommends pretesting new questions.
Start with a three-question check-in
Ask questions 1 and 2 first. Then show question 3 to everyone, including people who haven’t started.
1. What were you hoping to do when you signed up for [Product]?
Use a short text box. Ask this before offering a list of use cases so customers can describe their goal in their own words. Compare later answers among customers trying to do similar things.
2. Which best describes your progress with [setup milestone]?
Use a single-choice question. Name a concrete milestone, such as connecting a first data source.
- I haven’t started.
- I’ve started and plan to continue.
- I’ve started and am unsure whether to continue.
- I’ve started and decided not to continue.
- I’ve finished.
- I’m not sure which option applies.
Use this answer to choose relevant follow-ups. If their status and intent are unknown, bypass questions that assume them.
Skip this question if you already know both the relevant progress and the customer’s intention to continue. Records of completed steps alone do not tell you whether someone plans to return.
3. What, if anything, has made getting started difficult?
Keep this optional and open. Ask it before questions naming specific obstacles, so your list does not steer the first answer. “If anything” leaves room for a customer to say setup went smoothly.
Find the step that needs attention
Choose from these follow-ups when you need to understand a setup problem. Use branching to show relevant questions, or send separate survey versions to the appropriate groups. Test each intended survey path before sending.
4. How clear were the instructions for [first step]?
Show this to everyone, including people who haven’t started. The last two options cover people who saw no instructions or haven’t reached the step. Offer: Very unclear; Somewhat unclear; Neither clear nor unclear; Somewhat clear; Very clear; I didn’t see any instructions; or I haven’t seen this step.
If replies point to confusion, watch customers attempt the step. Check whether the problem is the instruction, the screen layout, or a prerequisite they lack.
5. Which step did you most need help with?
Ask people who started setup. List the steps in their usual order, using labels customers recognize. Include “Another step (please describe),” “I didn’t need help,” and “I’m not sure.” Ask for one choice.
If several steps were difficult, this identifies the one where help was most needed. Look at the instructions, examples and support available there.
6. What, if anything, would most help you finish [setup milestone]?
Show this to people who started and either plan to continue or are unsure whether to continue. Ask for one choice.
- Access or permissions.
- A file or data.
- Input from someone on my team.
- More time.
- Clearer instructions.
- Something else (please describe).
- Nothing else right now.
- I’m not sure.
Use the answer to decide what to send: steps for requesting access, a sample file or clearer instructions. An optional text box can clarify an overlapping or unclear answer.
7. What is the main reason you decided not to continue?
Ask people who started setup and decided not to continue. Use their answer to question 2 or information you already have about both their progress and intention.
- I couldn’t complete a step.
- The product doesn’t meet my needs.
- Getting set up would take more effort than I want to spend.
- I no longer need the product.
- Another reason (please describe).
- I’m not sure.
Review the reason alongside the goal in question 1.
Check whether customers completed a useful task
Use these questions after customers have had a chance to try a meaningful task. Replace [task] with that specific activity, such as “create your first report.”
8. Have you completed [task]?
- Yes, I completed it.
- I tried it but haven’t completed it.
- I haven’t tried it.
- I’m not sure.
Skip question 8 if question 2 covers the same task and establishes whether the customer tried and completed it, or reliable records provide that status. A missing completion record doesn’t show whether someone tried.
- Completed: show questions 9 and 10.
- Tried but not completed: show question 10 only.
- Not tried: skip questions 9 and 10.
- Unknown, including a skipped or “I’m not sure” answer with no reliable records: show question 10 only. It includes “I haven’t tried [task].”
9. How confident are you that you could [task] again without help?
Ask people who completed the task. Offer: Not at all confident; Slightly confident; Moderately confident; Very confident; Extremely confident; or I’m not sure.
10. How useful has [Product] been for [task]?
Ask people who tried the task, including those who did not finish, and people whose participation is unknown. Offer: Not at all useful; Slightly useful; Moderately useful; Very useful; Extremely useful; I haven’t used it enough to say; or I haven’t tried [task].
Compare answers from people with similar goals and stages of progress.
Close with two optional questions
11. What would you like help with next, if anything?
Use a short text box. Look for an example, explanation or next step your team could provide. If you intend to reply personally, tell customers who will contact them.
12. May our team contact you about your experience?
Offer Yes or No. Request contact details only when needed and only after Yes. If answers can be linked to the person, for example through contact details or a personal survey link, say so before they start.
Use the answers to choose an investigation
Treat the patterns below as starting points to check against the original replies and observed behavior.
Scroll sideways to see all columns.| If the replies show… | Investigate… |
|---|---|
| Unclear first steps, including among people who have not begun | Whether customers understand the first screen and have what they need to start. |
| Plans to continue or undecided, with access or teammate dependencies | Who needs to act next and what would unblock the customer. |
| A decision to stop because the product does not meet the stated goal | Product fit and expectations set before signup. |
| Finished setup but an unfinished first task | The transition from account setup to real work, including examples and guidance. |
| Completed tasks with low confidence | What help was needed and whether customers can repeat the task independently. |
| Low usefulness among people with a similar goal | What those customers expected and what the product actually helped them accomplish. |
Turn feedback into a change you can test
Read each reply alongside the customer’s goal and progress, when available. Group related comments while keeping the original wording accessible.
Several comments about “a difficult import” may describe different problems: finding the file, understanding the required format, or encountering an upload error. Separate those problems before deciding what to change.
For the issue you choose, record the customer’s words, the step involved and what you will try. You might clarify file requirements, provide a sample file, or observe another import attempt.
Keep the number of replies and who responded visible. A few replies can identify a problem to investigate; they cannot establish how common it is across all customers.
When comparing feedback over time, keep wording, question order, timing and audience as comparable as you can. A change in answers alone does not prove your product change caused it. Pew Research Center explains why consistent wording and question context matter for comparisons.
Build your onboarding survey in Validata
Choose the questions that fit your current onboarding decision and adapt them to your product. Browse the Onboarding Check-in example in Validata’s template gallery for more ideas.
Use Validata’s conditional logic to show follow-up questions based on each customer’s answer to question 2. The Free plan includes logic and branching, unlimited surveys, and up to 50 responses in total per account. AI analysis is not included.
Validata+ includes your first analysis. Analysis requires at least eight responses. Additional analyses use purchased AI credits. See current plans for details.