Use AI to Design Surveys That People Want to Answer
Survey Science: The AI Prompt Framework For Questions That Get Answered
Hey there, 👋
Survey design isn't about asking questions—it's about getting answers. Learn how the pros do it 🎯
Have you ever sent out a survey and watched tumbleweeds roll by? Or worse—got plenty of responses but data that tells you absolutely nothing useful? Yeah, me too!
And we’re not alone. Most surveys fail before the first question is even written.
Today, I'm sharing the ASKS Framework—a systematic approach to survey design that not only gets responses but delivers insights you can actually use. I've used this approach for many years, and it works every time. Now, with AI, I can bang out credible surveys in minutes, not hours.
5 Problems With Most Surveys
Before we dive in, let's address why most surveys fail:
They're too long
They ask the wrong questions
They're structured poorly
They're difficult to analyze
They don't lead to actionable insights
Introducing the ASKS Framework
ASKS stands for:
Audience: Who are you surveying?
Scope: What exactly do you need to learn?
Knowledge: What format will get honest answers?
Structure: How should you organize the questions?
Let's break down each component and see how to use them.
The Framework in Action
Step 1: Audience Definition
Use this prompt to define your audience:
For [survey purpose], identify:
1. Primary respondent characteristics
2. Their key motivations
3. Potential response barriers
4. Time availability
5. Preferred communication style
Format as clear bullet points with specific details.
Step 2: Scope Mapping
Use this prompt to clarify your objectives:
For [survey goal], outline:
1. Must-know information
2. Nice-to-know information
3. Success metrics
4. Required action points
5. Decision dependencies
Prioritize and rank each item by importance.
Step 3: Knowledge Format Selection
Use this prompt to choose question types:
For each must-know item, determine:
1. Best question format (multiple choice, scale, open-ended)
2. Response validation requirements
3. Data analysis needs
4. Potential biases to avoid
5. Follow-up opportunities
Provide specific examples for each format.
Step 4: Structure Organization
Use this prompt to organize your survey:
Create a survey flow that:
1. Builds trust progressively
2. Groups related questions
3. Maintains engagement
4. Prevents survey fatigue
5. Ends with clear next steps
Include transition text between sections.
Example: SaaS Customer Satisfaction Survey
Let's see how this works in practice.
Step 1 Output (Audience):
Primary: Active users who've used the product >3 months
Motivation: Desire for product improvement
Barriers: Limited time, survey fatigue
Available time: 5-7 minutes maximum
Style: Direct, professional, value-focused
Step 2 Output (Scope):
Must-Know Information:
Feature usage frequency
Pain points in current workflow
Likelihood to recommend
Priority improvement areas
Integration needs
Step 3 Output (Knowledge Format):
Question Design Examples:
1. Feature Usage (Multiple Choice)
"How often do you use [feature name]?"
- Daily
- Weekly
- Monthly
- Never
- Not sure what this is
2. Satisfaction Scale (Likert)
"How satisfied are you with [specific aspect]?"
1 (Very Dissatisfied) to 5 (Very Satisfied)
3. Pain Points (Open-ended)
"What's your biggest frustration when using [product]?"
[Text box with clear size indicator]
Step 4 Output (Final Structure):
1. Welcome & Purpose (2 seconds)
"Help us serve you better. This survey takes 5 minutes."
2. Quick Wins (1 minute)
- Usage frequency
- Primary use case
- Key features used
3. Deep Dive (3 minutes)
- Satisfaction metrics
- Pain points
- Feature requests
4. Future Vision (1 minute)
- Improvement priorities
- Integration needs
- Additional feedback
5. Close & Next Steps
"Thanks! Here's what we'll do with your feedback..."
Survey Design Principles
1. The 3-3-3 Rule:
3 seconds to understand
3 minutes to complete
3 clear insights gained
2. Question Flow:
Easy → Complex
General → Specific
Closed → Open-ended
3. Response Optimization:
Mobile-first design
Progress indicators
Save-and-resume capability
Common Pitfalls to Avoid
1. Leading Questions
❌ "How amazing was your experience?"
✅ "How would you describe your experience?"
2. Double-Barreled Questions
❌ "How satisfied are you with our features and support?"
✅ "How satisfied are you with our features?"
✅ "How satisfied are you with our support?"
3. Unclear Scales
❌ Rate from 1-5
✅ 1: Very Dissatisfied to 5: Very Satisfied
Next Steps
1. Download the ASKS Framework template
2. Run your next survey through it
3. Test with a small group first
4. Analyze and iterate
Remember: The best survey is the one that gets completed and provides actionable insights. Everything else is just noise.
Until next time.
Andy
P.S. If this was helpful, share it with someone who's struggling with survey design. Good data leads to better decisions.
P.P.S. If you want to up your game with AI image generation, check out my Best Selling Substack - AI Art Daily