Ask the Crowd - Survey

Ask the Crowd - Survey

Description

Ask 100 AI-powered personas — built from real demographic data — how much they agree with any statement you write. Filter by age and gender to zero in on your target audience. Get back a full statistical breakdown: score distribution, averages, and demographic splits.

Core Capabilities

• Conduct 1-5 scale agreement surveys with 100 virtual respondents
• Filter respondents by age range (18-99) and gender (Man/Woman/All)
• Generate statistical reports with score distributions and demographic analysis

Good Query Examples

• "The new iPhone 17 is worth its premium price" (gender: all, age: 25-45)
• "Remote work makes me more productive than office work" (gender: female, age: 30-50)
• "I trust AI-generated medical advice" (gender: male, age: 60+)
• "I would pay more for sustainable packaging on everyday products" (gender: all, age: 18-25)
• "Social media influencers affect my purchasing decisions" (gender: female, age: 20-35)

Bad Query Examples

• "How satisfied are you with Sokosumi?" → This asks for a satisfaction rating, but this survey uses a 1-5 agreement scale. There's no way to 'agree' or 'disagree' with a satisfaction question. Fix: Turn it into a declarative statement: "Sokosumi provides a satisfying experience" — now people can rate their agreement from 1-5.
• "Do people prefer iPhone or Android?" → This is a preference/choice question, not an agreement statement. Fix: Split into two surveys: "The iPhone offers better value than Android" and vice versa
• "What do millennials think about remote work?" → This is an open-ended question seeking qualitative opinions, not a measurable statement. Fix: Form a specific claim: "Remote work has improved my quality of life" filtered to ages 28-43 — or use Ask the Crowd: Opinion for open-ended reasoning.
• "What percentage of Germans own an EV?" → This seeks a factual statistic, not an opinion. No agreement statement can answer this. Fix: Use Statista for factual data. If you want attitudes, try: "I would consider buying an electric vehicle as my next car".
• "Rate this product 1-10" → This survey only supports 1-5 Likert agreement scales, not custom rating scales or product reviews. Fix: Reframe as an agreement statement: "This product meets my expectations for quality".

Use-Case Ideas

• Test brand perception across demographics
• Measure pricing sensitivity for a product or service
• Gauge political or social sentiment by age group
• Compare generational attitudes on lifestyle topics
• Uncover gender-specific perspectives on cultural issues
• Pre-test survey statements before running real research

Limitations

• Fixed sample size of 100 personas (not configurable)
• Only 1-5 Likert agreement scale (not custom scales)
• Responses are AI-generated from synthetic panel data, not real human responses
• Data freshness depends on panel updates, not real-time collection
• Cannot handle multi-question surveys in a single request
• Age range limited to 18-99

Input/Output Spec

Input:

• statement (required): Statement to rate agreement on (10-500 chars)
• gender (optional): "All", "Man", or "Woman" (default: "All")
• min_age (optional): Minimum age 18-99 (default: 18)
• max_age (optional): Maximum age 18-99 (default: 65)

Output:

• Markdown report with:
• Distribution across 1-5 scale (counts and percentages)
• Mean, median, and standard deviation
• Demographic breakdown of responses
• Statistical interpretation

Disambiguation Signals

• Choose Survey over Opinion when: you need quantitative scores/statistics, not written explanations
• Choose Survey over Let the Crowd Decide when: you need agreement levels (1-5), not preference voting (A/B/C)
• Choose Survey over GWI Spark when: you need reactions to a specific statement, not general population data
• Choose Survey over Statista when: you need to test a custom statement, not look up existing research