Generate A/B Test Ideas
Generate 20 A/B test ideas for any landing page, email, or product feature ranked by expected impact.
The Prompt
Generate A/B test ideas for the following page/feature. For each test: 1. What you're testing (element) 2. Control variant (current state) 3. Test variant (what you're changing) 4. Hypothesis: why you expect the change to improve the metric 5. Primary metric to measure 6. Expected impact (High / Medium / Low) 7. Estimated traffic needed for significance Prioritize by impact/effort ratio. Flag tests that can run simultaneously vs. must be sequential. Page/feature to optimize: [DESCRIBE YOUR PAGE OR FEATURE] Current conversion rate: [IF KNOWN] Monthly visitors: [TRAFFIC VOLUME] Key friction points: [WHAT DO YOU THINK IS CAUSING DROP-OFF?] Goal: [WHAT METRIC TO IMPROVE]
Example Output
20 A/B test ideas prioritized by PIE score (Potential, Importance, Ease). Top test: change CTA from 'Get Started' to 'See My Cost Estimate' (+estimated 15-25% CTR). Second: add social proof counter ('Join 4,200 teams comparing AI costs') above the fold. Each with metric, sample size estimate, and conflict flag.
FAQ
Which AI model is best for Generate A/B Test Ideas?
GPT-4o or Claude Sonnet 4 — both generate creative, evidence-based test hypotheses.
How do I use the Generate A/B Test Ideas prompt?
Copy the prompt, replace the [BRACKETED] placeholders with your specific information, and paste into your preferred AI assistant (ChatGPT, Claude, Gemini, etc.). 20 A/B test ideas prioritized by PIE score (Potential, Importance, Ease). Top test: change CTA from 'Get Started' to 'See My Cost Estimate' (+estimated 15-25% CTR). Second: add social proof counter ('Join 4,200 teams comparing AI costs') above the fold. Each with metric, sample size estimate, and conflict flag.
Model Recommendation
GPT-4o or Claude Sonnet 4 — both generate creative, evidence-based test hypotheses.