Data Analysis
A/B Test Design
A/B Test Design analyzes supplied material for patterns, evidence, and decision-ready findings with clear constraints and an explicit output.
The prompt
A/B Test Design guides the model through a defined task while preserving the source prompt's useful structure and constraints. Use it when exploring data and explaining what the results support and you want a response that is easier to evaluate and act on.
text prompt
You are a senior product analyst.
Task:
Design an A/B test for {{change}} with hypothesis; primary metric; guardrails; sample sizing approach; stop conditions.
Inputs:
- Change: {{change}}
Output requirements:
- Return JSON only. Do not wrap in Markdown or code fences.
- Do not add keys beyond the schema.
- Populate clarifying_questions if any critical input is missing or ambiguous, but still produce a best-effort deliverable using assumptions.
- Add a short, realistic next_steps list that would be performed after using this deliverable.
- Include deeper reasoning artifacts as structured fields (assumptions, risks, tradeoffs) while keeping the final deliverable usable.
JSON schema (example values only, keep the same keys):
{
"title": "A/B Test Design",
"deliverable_type": "Structured experiment design document",
"clarifying_questions": [],
"assumptions": [],
"deliverable_markdown": "",
"quality_checks": [
"Every metric includes definition, formula, data source, cadence, and caveats",
"Metrics are grouped logically and avoid duplicates",
"Assumptions are explicitly listed"
],
"next_steps": []
}
Tune the prompt, not the plumbing.
Every control comes from this prompt’s content schema. Changes stay in your browser and update instantly.
Customized prompt1128 characters
You are a senior product analyst.
Task:
Design an A/B test for a specific change with hypothesis; primary metric; guardrails; sample sizing approach; stop conditions.
Inputs:
- Change: a specific change
Output requirements:
- Return JSON only. Do not wrap in Markdown or code fences.
- Do not add keys beyond the schema.
- Populate clarifying_questions if any critical input is missing or ambiguous, but still produce a best-effort deliverable using assumptions.
- Add a short, realistic next_steps list that would be performed after using this deliverable.
- Include deeper reasoning artifacts as structured fields (assumptions, risks, tradeoffs) while keeping the final deliverable usable.
JSON schema (example values only, keep the same keys):
{
"title": "A/B Test Design",
"deliverable_type": "Structured experiment design document",
"clarifying_questions": [],
"assumptions": [],
"deliverable_markdown": "",
"quality_checks": [
"Every metric includes definition, formula, data source, cadence, and caveats",
"Metrics are grouped logically and avoid duplicates",
"Assumptions are explicitly listed"
],
"next_steps": []
}
Useful structure, room to move.
A/B Test Design guides the model through a defined task while preserving the source prompt's useful structure and constraints. Use it when exploring data and explaining what the results support and you want a response that is easier to evaluate and act on.
The prompt establishes the job first, then supplies concrete decisions a model can act on. The variables preserve that structure while letting you change the subject, context, or output.
Related prompts
Cohort Analysis Plan
Cohort Analysis Plan analyzes supplied material for patterns, evidence, and decision-ready findings with clear constraints and an explicit output.
ChatGPTClaudeGemini
Dashboard Audit
Dashboard Audit analyzes supplied material for patterns, evidence, and decision-ready findings with clear constraints and an explicit output.
ChatGPTClaudeGemini
Forecasting Model Outline
Forecasting Model Outline analyzes supplied material for patterns, evidence, and decision-ready findings with clear constraints and an explicit output.
ChatGPTClaudeGemini
High-Value Data Questions
High-Value Data Questions analyzes supplied material for patterns, evidence, and decision-ready findings with clear constraints and an explicit output.
ChatGPTClaudeGemini