Survey template

Product feedback survey

Generic 'how are we doing?' surveys collect noise. This one anchors every answer to a real task the user just attempted, so you learn where the product helped, where it got in the way, and the single change that would matter most.

Best for: PMs and founders collecting in-product feedback to prioritize the next build.

The questions

1. What were you trying to get done in the product today?

   Long text

   Anchors feedback to a concrete task instead of a vague overall mood.

2. What worked well while you were doing that?

   Long text

3. Where did it get in your way or slow you down?

   Long text

   Friction points are easier to recall than abstract likes and dislikes.

4. If you could change one thing about the product, what would it be?

   Long text

   Forcing a single change surfaces the top priority instead of a wish list.

5. How easy was it to do what you came to do?

   Opinion scale

   1-7 ease scale — gives you a trend line alongside the verbatims.

6. How often do you use the product?

   Single select

   Daily / Weekly / Monthly / Rarely — lets you weight feedback by engagement.

7. What's your role?

   Short text

   Lets you slice feedback by segment rather than averaging everyone.

How to run it well

• Trigger it after a real action, not on a timer — feedback tied to a fresh task is specific; feedback fired at random is mood.
• Read the 'one change' answers as a frequency table, not a wish list; the change named most often is your next ticket.
• Segment by usage frequency before you act — daily users and triers are reporting on two different products.
• Quote the friction verbatims directly into the ticket; paraphrasing strips the detail that tells you what to fix.

FAQ

How is this different from an NPS or CSAT survey?

NPS and CSAT give you a score; this gives you the task-level reasons behind it. Run them together — the score tells you the temperature, this tells you what to ship.

How many responses before I act on it?

Open-ended themes usually stabilize around 15-20 responses per segment. Below that, treat a pattern as a hypothesis worth a follow-up, not a roadmap commitment.

Won't a single 'one change' question miss things?

It's deliberate — asking for everything gets you a flat wish list. Forcing one change ranks priorities for you, and the open friction question still captures the rest.

Related templates