Question 1 of 3

What if your
survey expert
was always on call?

Survey Craft is an AI you chat with. Ask it anything about survey design, question wording, sample sizes, or result analysis — and get methodology-backed answers in seconds.

Ask Your First Question
Step 1

Bad surveys waste everyone's time

You know the pain. Here are the sticky notes from every survey project gone wrong:

"Is this question leading? I genuinely can't tell anymore."

"We got 2,000 responses but the data is basically useless."

"How many people do we actually need to survey for this to mean something?"

"The boss wants a 50-question survey. I know that's too long."

"Likert scale? Semantic differential? Why are there so many options?"

"Our response rate was 4%. Something went very wrong."

Sound familiar? ↓

Step 2

Ask Survey Craft. Get real answers.

It's a conversation, not a course. Here's what talking to Survey Craft actually looks like:

Question Design

I'm writing a customer satisfaction survey for a SaaS product. Here's my draft question: "Don't you think our product is easy to use?" — is this okay?

That's a textbook leading question — the "Don't you think" phrasing nudges respondents toward agreement. Try instead: "How would you rate the ease of use of [Product]?" on a 5-point scale from Very Difficult to Very Easy. This removes the bias and gives you measurable data.

Sample Size

We have 5,000 employees. How many responses do I need for statistically significant results?

For a population of 5,000 with a 95% confidence level and ±5% margin of error, you'd need roughly 357 responses. Want tighter results at ±3%? That jumps to about 880. What margin of error works for how you'll use this data?

Bias Detection

Our employee engagement survey always shows 90%+ satisfaction. Is something off?

Classic signs of social desirability bias. If surveys aren't truly anonymous (or employees don't believe they are), people tell you what's safe. Some fixes: use a third-party collection method, include indirect questions, add a "social desirability scale" to detect it, and consider randomized response techniques for sensitive topics.

Scale Selection

Should I use NPS or a regular satisfaction scale for our post-purchase survey?

Depends on your goal. NPS (0–10 "would you recommend") is great for benchmarking loyalty over time, but it's a blunt instrument — it won't tell you why. If you need actionable detail, pair it with a 5-point CSAT question plus one open-ended follow-up. What are you planning to do with the results?

These are real conversation types, not canned responses
Step 3

Everything you can ask about

Survey Craft knows methodology. Here's the whiteboard of topics it covers:

📝

Question Writing

Reword leading, double-barreled, or loaded questions. Get alternatives that reduce bias and improve clarity.

📊

Scale & Format Selection

Likert, semantic differential, NPS, ranking, matrix — know which format fits your research question.

🔢

Sample Size & Significance

Calculate required sample sizes based on your population, confidence level, and margin of error.

🎯

Bias Identification

Spot response bias, selection bias, acquiescence, social desirability — and learn mitigation strategies.

📬

Distribution Strategy

Plan how to reach your target population, boost response rates, and time your survey right.

📈

Results Analysis

Interpret your data: cross-tabulation, significance testing, open-ended coding, and what the numbers actually mean.

Works for all survey types ↓
Academic Research Customer Feedback Employee Engagement Market Research UX Research Healthcare Surveys Political Polling Course Evaluations

Notes from the board

What people say after chatting with Survey Craft

"I pasted my 20-question draft and it immediately flagged three double-barreled questions I missed. Rewrote them on the spot."

— Rachel, graduate researcher

"Asked it to help with sample size for an internal survey. Got the number with a clear explanation of the math. Took 30 seconds."

— Daniel, HR manager

"It asked ME what my research objective was before giving advice. That's exactly what a good methodology consultant would do."

— Maria, product researcher

Frequently asked questions

(about Survey Craft, not about survey design — though it handles those too)

Final Question

Ready to make your next survey
actually good?

Describe your survey project. Survey Craft will ask the right questions, catch your blind spots, and help you design something people will actually want to complete.

Try asking something like:

"Review my survey questions for bias"
"Help me design an NPS survey"
"What sample size do I need?"
Meet Survey Craft

Related assistants