What is a Questionnaire?

"Questionnaire" is a term researchers throw around casually, but it can feel a bit unfamiliar if you're the person in your team asked to run a survey for the first time. This page explains what a questionnaire is, in plain English.

In one sentence

A questionnaire is the design spec for a survey.

Just as you'd draw up blueprints before building a house, you write down "what to ask, whom to ask, and in what order" before sending out a survey. That document is called a questionnaire.

In other words, a questionnaire is not the web form your respondents see — it's the specification document that the web form is built from.

Questionnaire vs. web form

The two words are often used interchangeably in everyday speech, but they play different roles.

	Questionnaire	Web form (survey)
Format	Word / Excel / Markdown document	Screen respondents see in the browser
Role	Defines the questions, options, and branching	Collects the actual responses
Authored by	Survey owner or researcher	The system (kicue generates it automatically)
Ease of editing	Easy — it's just text	Changes after launch can affect response data

Put simply, the questionnaire is the design, and the web form is the implementation. With kicue, you upload the questionnaire file and the AI generates the web form for you.

What a questionnaire should contain

No matter the industry or purpose, a solid questionnaire usually includes the following elements.

1. Purpose of the study (one or two lines at the top)

State what the survey is for — "to evaluate the concept of new product A," "for quarterly customer satisfaction tracking," etc. Respondents don't see this, but it keeps you focused while drafting questions.

2. Target audience

Specify who you're asking — "women aged 20–50," "anyone who purchased product A in the last three months," etc. This becomes the basis for screening questions later.

3. Question text

The actual wording shown to respondents. By convention, questions are numbered (Q1, Q2, …).

4. Response format

For each question, decide how people should answer. The main formats are:

Single Answer (SA): pick one from a list
Multiple Answer (MA): pick all that apply
Open Answer (OA / FA): free text
Number (NUM): numeric input
Matrix (MTX): several items rated on the same scale
Scale (LIKERT / NPS): 1–5, 0–10, etc.

See Question Types for a detailed guide to each.

5. Options (answer choices)

For SA, MA, and MTX questions, list all options. Decide here whether to include "Other (free text)" and "None of the above" at the end.

6. Branching conditions (when needed)

For conditional logic like "only respondents who answered 'Yes' in Q3 should continue to Q4," write the rule in plain English next to the question. kicue's AI will parse it correctly.

What format to write a questionnaire in

Supported File Formats covers this in detail. The short version:

You already have one → use it as-is (Excel / Word / PDF all work)
Writing from scratch → Excel is best
Drafting with an AI → Excel or Markdown works best

Excel's rows and columns make it easy to organize questions, options, and response types — and it's equally convenient for reviewing or sharing with colleagues. If you want multiple people to edit at the same time, the same layout works in Google Sheets too.

Sample questionnaire (Excel layout)

Here's a small customer satisfaction survey mocked up as an Excel sheet. If you build a .xlsx with the same layout, you can upload it directly to kicue.

Column A (Question #)	Column B (Question text / options)	Column C (Response format)
Q1	Overall, how satisfied are you with our service?	SA (single answer)
	Very satisfied
	Somewhat satisfied
	Neutral
	Somewhat dissatisfied
	Very dissatisfied
Q2	Which aspects are you most satisfied with?	MA (multiple answer)
	Price
	Quality
	Support
	Ease of use
	Other
Q3	Please rate each of the following on a 5-point scale	MTX (matrix)
	Rows: Price / Quality / Support
	Columns: Very satisfied / Somewhat satisfied / Neutral / Somewhat dissatisfied / Very dissatisfied
Q4	Any comments or suggestions about our service?	Open text

Layout tips:

Put the question number (Q1, Q2, …) in Column A
Put the question text in Column B, then list one option per row in the cells below
Put the response format (SA / MA / MTX / Open text, etc.) in Column C
For matrix questions, list the row items and column items in Column B below the question text

Build this Excel file, upload it to kicue, and you'll have a working web form in about 30 seconds.

Drafting a questionnaire with generative AI

If coming up with questions feels hard, generative AI (ChatGPT, Claude, Gemini, etc.) is a great way to produce a first draft you can refine. The prompts below ask the AI to generate a downloadable Excel (.xlsx) file. Run them in an environment that supports Code Interpreter (ChatGPT) or Artifacts (Claude).

Prompt 1: writing from scratch

Copy the template below and swap in your own purpose, audience, and topics.

You are a professional survey researcher. Please produce a questionnaire as an Excel file (.xlsx) based on the conditions below, and provide it as a downloadable file. Use Python (openpyxl, etc.) to generate the file.

# Purpose
To assess concept acceptance for a new product "XYZ"

# Target audience
Women aged 20–50 who buy skincare products at least once a month

# Topics to cover
- Current skincare brand(s) used
- Appeal of the new product concept (5-point scale)
- Purchase intent (5-point scale)
- Reasons they would or would not buy (free text)
- Demographics (age band, region)

# Excel layout
- Column A: Question number (Q1, Q2, ...)
- Column B: Question text, with each option listed one per row beneath it
- Column C: Response format (SA / MA / MTX / LIKERT / NPS / Open text, etc.)
- Place all questions on a single sheet
- For matrix questions, clearly list row items and column items
- Keep the total between 10 and 15 questions
- Include screening questions or branching conditions in a notes column if relevant

Finish by providing a download link for the generated .xlsx file.

Prompt 2: when you haven't decided what to ask yet (a two-step flow)

When you know you want to run a survey but haven't nailed down what to ask, don't start by writing questions. Instead, use Step 1 to brainstorm the study design with the AI, then pass the outcome to Step 2 to generate the Excel file. This produces fewer false starts.

Step 1: brainstorming the study design

First, use the prompt below to work with the AI on what to measure and how to analyze it.

You are a professional survey researcher. I want to run a survey, but I haven't decided exactly what to ask yet. Based on the background below, please help me shape the study design together.

# Background
- Business / product: (e.g. a new SKU from a D2C skincare brand)
- Problem we're facing: (e.g. we want to gauge target-audience reaction before launch)
- What we already know: (e.g. target is women in their 20s–30s, competitors are Company X and Company Y)
- Decisions we need this survey to inform: (e.g. go/no-go, price point, messaging direction)

# What I want you to propose
1. Three to five hypotheses this survey should validate
2. For each hypothesis, the question themes that would test it (rough topics + likely response formats)
3. Screening conditions for respondents (age, usage, purchase frequency, etc.)
4. Design pitfalls to watch for (sample size, bias, respondent burden, etc.)

# How we'll iterate
I'll push back on your proposals with additional context or requests, and you'll refine the plan accordingly. At the end, output a concise "study design memo" with the following as bullet lists, so I can paste it into Step 2 to generate the Excel questionnaire:

- Purpose of the study (1–2 lines)
- Target audience conditions
- List of question themes (with the response format expected for each)
- Any branching logic that will be needed

Iterate with the AI — reject hypotheses, add new angles ("let's also explore post-purchase experience"), etc. — until the plan feels solid.

Step 2: generate the Excel questionnaire from the finalized plan

Paste the "study design memo" the AI produced at the end of Step 1 into the prompt below.

Based on the "study design memo" below (produced in Step 1), please produce a questionnaire as an Excel file (.xlsx) and provide it as a downloadable file. Use Python (openpyxl, etc.) to generate the file.

# Study design memo
(paste the final memo from Step 1 here)

# Excel layout
- Column A: Question number (Q1, Q2, ...)
- Column B: Question text, with each option listed one per row beneath it
- Column C: Response format (SA / MA / MTX / LIKERT / NPS / Open text, etc.)
- Place all questions on a single sheet
- For matrix questions, clearly list row items and column items
- Include screening questions or branching conditions in a notes column if relevant
- Keep the total between 10 and 15 questions

Finish by providing a download link for the generated .xlsx file.

Upload the downloaded .xlsx to kicue and you'll have a working web form in about 30 seconds.

Treat the AI's output as a first draft. Always review it against the pitfalls below before sending it out.

Common pitfalls

Whether you write it yourself or have an AI draft it, these mistakes come up often.

Leading questions: "Why did you like it?" presumes they liked it — rewrite neutrally.
Missing options: always include "Other" or "None of the above" when relevant.
Double-barreled questions: "How satisfied are you with the design and price?" asks two things in one — split them.
Too many questions: if completion takes more than 10 minutes, drop-off rises sharply.
Jargon: avoid internal or industry vocabulary respondents may not understand.

For practical design tips, see our blog post AI-powered survey design guide.

Next steps

Once your questionnaire is ready, it's time to upload it.

Supported File Formats — if you're unsure which format to use
How to Upload — step-by-step upload instructions
How AI Parsing Works — what happens after you upload