# The Survey Question Rewrite Sheet Before-and-after for the questions SaaS teams keep asking, and why the rewrite gets a real answer. From saasfeedback.ai — https://saasfeedback.ai/blog/saas-survey-questions-that-work --- ## The four failure modes Every dead survey question fails in one of four ways. Name the failure before you rewrite, or you will just write a different broken question. 1. **Answerable without thinking.** The respondent can satisfy the question with a plausible answer that costs them zero recall. Multiple choice does this by construction. 2. **Asks for a prediction.** People are poor forecasters of their own behaviour. "Would you use X?" measures politeness, not demand. 3. **Asks for a summary.** "How satisfied are you overall?" compresses a hundred experiences into one number and discards all the information. 4. **Leaks the answer you want.** "How much did our new dashboard improve your workflow?" contains its own conclusion. --- ## The rewrites ### Cancellation | Instead of | Ask | Why | |---|---|---| | "Why are you cancelling?" (dropdown: Too expensive / Missing features / Switched / Other) | "What was the last thing that happened before you decided to cancel?" | The dropdown makes "too expensive" the cheapest click for every reason. The rewrite asks for an event, which cannot be faked cheaply. | | "How could we have kept you?" | "What would have had to be true for you to still be a customer today?" | The first asks them to design your product. The second asks for a condition, which is what you can act on. | | "Would you consider coming back?" | "If we fix [specific thing], would you want to hear about it?" | The first is a prediction. The second is a low-cost commitment with a real follow-up. | ### Onboarding and activation | Instead of | Ask | Why | |---|---|---| | "How easy was setup? (1–5)" | "What was the first thing that confused you?" | A 4/5 tells you nothing you can fix. The confusion has a location. | | "What features would you like to see?" | "What did you try to do this week that you couldn't?" | Feature requests are solutions from people who have not seen your codebase. The rewrite gets the problem. | | "Are you finding the product useful?" | "What did you use it for most recently, and what happened?" | Behaviour, not opinion. See NN/g on why self-reported claims are unreliable. | ### Value and price | Instead of | Ask | Why | |---|---|---| | "Is our pricing fair?" | "What else did you consider spending this budget on?" | The first invites agreement. The second reveals your real competitive set. | | "How much would you pay for X?" | "What are you paying for [adjacent thing] today?" | Willingness-to-pay questions get fantasy numbers. Current spend is a fact. | | "How valuable is [feature]? (1–5)" | "When did you last use [feature], and what were you doing?" | Rated value and used value diverge wildly. | ### Satisfaction | Instead of | Ask | Why | |---|---|---| | "How likely are you to recommend us? (0–10)" | "Have you recommended us to anyone? Who, and what did you say?" | NPS asks for a prediction of a social act. The rewrite asks whether it happened. | | "How satisfied are you overall?" | "What's the most annoying thing about using [product] right now?" | Summary versus specific. The second gets a fixable answer from a happy customer too. | --- ## Design rules **Length.** Every question after the fourth costs you completions. If you have twelve questions, you have not prioritised — you have deferred the decision to your respondents, and they will decide by abandoning. **Order.** Open question first, while attention is highest. Demographics and multiple-choice last. Most surveys do exactly the reverse. **Timing.** Ask within 48 hours of the event you are asking about. Memory of a specific frustration decays fast, and what replaces it is a tidy narrative that is not what happened. **Channel.** In-app reaches people still using the product — which excludes, by definition, everyone who left. If your survey is in-app only, you are sampling survivors. **The "other" box.** If more than 20% of respondents pick "other" and write something, your options are wrong. That is a finding, not noise. --- ## The one-question survey If you only get one, make it this one: > "What almost stopped you from [signing up / upgrading / staying]?" It is answerable from memory, it is about a specific moment, it does not leak the desired answer, and it surfaces the objection your funnel is quietly losing people to. It outperforms every rating scale we have tested it against. --- ## When to stop surveying A survey is the right instrument when you already know the shape of the answer and need to size it. It is the wrong instrument when you do not know what you are looking for — that is what conversations are for. If your open-text answers are still surprising you, stop writing surveys and go have ten conversations. Come back to the survey once you know which two things you are trying to count.