How to Create a Customer Feedback Form (with Examples)

A customer feedback form should ask one rating, one open-ended "why," and little else, then get out of the way. The most common mistake is treating the form as a data-entry exercise: ten fields, three rating grids, and a wall of text that nobody finishes. The fix is to keep the form short and let the follow-up happen conversationally. Koji turns a static form into an AI-moderated conversation that asks the rating, then automatically probes the reasoning, so you capture the "why" without adding a single extra field.

This guide walks through what to put on the form, examples by use case, where to place it, and how to upgrade it from a one-way collection box into a two-way conversation.

Form vs. conversation: pick the right tool

A feedback form is one-directional. The customer fills boxes; you read them later. It is perfect for a fast pulse, "rate this and add a comment", but it cannot ask "what do you mean by clunky?"

A feedback conversation adapts. When an answer is vague, it digs; when an answer is complete, it moves on. Historically that required a human moderator, which does not scale. Koji closes that gap: its AI interviewer runs form-like questions and follows up in real time, giving you the structure of a form with the depth of an interview.

Rule of thumb: if you only need a number, use a form. If you need to know why the number is what it is, use a Koji conversation.

The anatomy of a form that gets completed

A high-converting customer feedback form has four parts and no more:

1. A single primary rating. Satisfaction (1-5), effort ("how easy was this?"), or recommendation (NPS 0-10). One number, clearly the main question.
2. One open-ended follow-up. "What is the main reason for your score?" This is where the value lives.
3. An optional identifier. Email or order ID, only if you intend to follow up. Never required.
4. A clear purpose statement. One line: "We read every response and use it to improve [product]." It measurably lifts completion.

That is it. Every additional field costs you completions. The reason Koji works is that it does not force you to choose between brevity and depth: the AI adds the follow-up questions dynamically, per respondent, instead of you stacking them on the form up front.

Question examples by use case

Support / service feedback

• How easy was it to get your issue resolved? 1 (very hard) to 5 (very easy) (scale)
• What could we have done to make that faster? (open_ended)
• Was your issue fully resolved? Yes / No (yes_no)

Product / app feedback

• How satisfied are you with [feature] today? (scale 1-10)
• What is the main reason for that rating? (open_ended)
• Which part of this workflow felt slowest? (single_choice)

Website / landing-page feedback

• Did you find what you were looking for? Yes / No (yes_no)
• What were you trying to do on this page? (open_ended)
• What almost made you leave? (open_ended)

Post-purchase feedback

• How was your buying experience? (scale 1-5)
• What nearly stopped you from completing the purchase? (open_ended)
• How likely are you to buy again? (scale 0-10)

Each example pairs a quick quantitative question with an open one. In a static form the open question often gets a one-word answer. In Koji, the AI responds to that one word with "tell me more," and you get a usable sentence instead.

Choosing question types

Koji exposes six first-class question types so each answer is analyzable the moment it arrives: open_ended, scale, single_choice, multiple_choice, ranking, and yes_no. Pick scale for ratings, yes_no for quick gates, single or multiple_choice when options are known, ranking to force trade-offs, and open_ended whenever you need reasoning. Because every question carries a stable type, results roll up into the right chart automatically, no manual tagging. The structured questions guide breaks down when to use each.

Placement and timing

The best feedback form is the one that appears right after the relevant experience:

• In-product, immediately after a key action (embed it; see in-app AI surveys)
• Post-purchase, on the confirmation page or in the receipt email
• End of a support interaction, while the resolution is fresh
• Targeted email, to a segment you want to hear from

Timing beats placement: a perfectly designed form shown at the wrong moment underperforms a plain one shown in context. And never block core functionality behind a form, friction here poisons the relationship you are trying to measure.

Design best practices

• Keep it to 3-5 fields. Length is the number-one killer of completion.
• Lead with the easy question. Start with a one-tap rating to build momentum.
• Write neutrally. "How was your experience?" not "How great was your experience?"
• Make it mobile-first. Most feedback is given on a phone.
• Show progress and an end. People finish what they can see the bottom of.
• Close the loop. Tell respondents what changed because of their feedback; it earns you the next response.

From static form to AI conversation with Koji

Here is the upgrade path. Take your 3-question form and run it through Koji instead of a form builder. The AI asks the same questions, but when someone writes "the checkout was confusing," it immediately asks "what specifically confused you?" By the end you have a short transcript, not a single ambiguous line.

Then Koji does the analysis for you: open-ended responses are clustered into themes with verbatim quotes and counts, so "checkout confusion" arrives as a quantified pattern instead of scattered comments you have to read one by one. Reports update as responses come in. Koji's quality gate means only substantive conversations (scoring 3 or higher) consume credits, so junk submissions never inflate your data or your bill. Text conversations cost 1 credit each, and new accounts start with 10 free credits, enough to pilot a real feedback loop before you pick the Insights (€29/mo) or Interviews (€79/mo) plan.

Compared to a traditional form tool, the difference is not cosmetic: a static form gives you what you asked; Koji gives you what you needed to know.

Acting on what the form tells you

Collecting feedback is the easy half. The teams that compound value follow a simple loop:

1. Categorize. Group responses by theme (bug, confusion, missing feature, pricing, praise). Koji does this automatically; with a static form you do it by hand.
2. Quantify. Count how many people raised each theme. One angry comment is noise; twelve is a roadmap item.
3. Prioritize. Weigh frequency against impact and effort. Not every request deserves a build.
4. Close the loop. Reply to respondents, or post a public "you asked, we shipped" note. This single habit measurably increases your next response rate.

A quick before-and-after

A static form asks "How was checkout?" and stores "confusing." You are left guessing. The same question in Koji yields a 20-second exchange: "What specifically confused you?" -> "I could not tell which shipping option was selected." That second answer is a design ticket. Multiply it across 200 respondents and you have a ranked, quoted, quantified backlog instead of a column of one-word replies, which is the entire difference between collecting feedback and acting on it.

Related Resources

• Structured Questions Guide — the 6 question types that make feedback analyzable
• Customer Feedback Questions — a deeper bank of questions to draw from
• How to Get Customer Feedback — channels and tactics beyond the form
• Customer Feedback Analysis — turn responses into themes and action
• In-App AI Surveys — embed feedback collection where the experience happens
• Customer Feedback Management Guide — build a closed-loop program