A/B Testing Plan Prompt

Create an A/B testing plan for [WEBSITE/LANDING PAGE/EMAIL/AD].

Current situation:
What you're testing: [PAGE/EMAIL/AD]
Current conversion rate: [PERCENTAGE]
Traffic volume: [VISITORS PER WEEK/MONTH]
Goal: [WHAT YOU WANT TO IMPROVE]
Testing tool: [GOOGLE OPTIMIZE/OPTIMIZELY/BUILT-IN/MANUAL]

Generate a complete testing plan:

**Testing Fundamentals**

Sample size needed:
Current conversion rate: [PERCENTAGE]
Minimum detectable effect: [PERCENTAGE IMPROVEMENT YOU WANT]
Statistical significance: 95%
Estimated visitors needed: [CALCULATE OR ESTIMATE]
Test duration: [WEEKS BASED ON TRAFFIC]

Traffic requirements:
If under 100 conversions/week: Test only major changes
If 100-1000 conversions/week: Test medium changes
If 1000+ conversions/week: Can test small changes

**Test Priority Framework** (What to test first)

High Priority (Test these first):
Elements above the fold
Headline
Primary CTA
Value proposition
Price presentation
Form fields

Medium Priority (Test after high priority):
Images
Testimonial placement
Feature order
Page layout
Color schemes

Low Priority (Test last):
Button colors alone
Minor copy tweaks
Footer elements
Font choices

**Test 1: Headline Variation**

Current headline: [EXISTING]
Why test this: Headline is first thing people see
Hypothesis: [WHAT YOU THINK WILL HAPPEN]

Variation A (Control): [CURRENT HEADLINE]
Variation B: [BENEFIT-FOCUSED ALTERNATIVE]
Variation C: [CURIOSITY-DRIVEN ALTERNATIVE]

What to measure:
Time on page
Scroll depth
Click-through to CTA
Conversion rate

Success criteria: [PERCENTAGE] improvement
Confidence level: 95%
Run for: [DAYS/WEEKS]

**Test 2: Call-to-Action**

Current CTA: [EXISTING BUTTON TEXT/PLACEMENT]
Why test this: CTA drives conversions
Hypothesis: [WHAT YOU THINK WILL HAPPEN]

Elements to test:
Button text: [CURRENT] vs [ACTION-ORIENTED] vs [BENEFIT-FOCUSED]
Button color: [CURRENT] vs [HIGH CONTRAST]
Button size: [CURRENT] vs [LARGER]
Button placement: [CURRENT] vs [ABOVE FOLD]

What to measure:
Click-through rate
Conversion rate
Bounce rate

Success criteria: [PERCENTAGE] improvement
Confidence level: 95%
Run for: [DAYS/WEEKS]

**Test 3: Form Optimization**

Current form: [NUMBER OF FIELDS]
Why test this: Forms create friction
Hypothesis: Fewer fields = more conversions

Variation A (Control): [CURRENT NUMBER] fields
Variation B: [FEWER] required fields
Variation C: Multi-step form

What to measure:
Form start rate
Form completion rate
Time to complete
Conversion rate

Success criteria: [PERCENTAGE] improvement
Confidence level: 95%
Run for: [DAYS/WEEKS]

**Test 4: Social Proof Placement**

Current placement: [WHERE TESTIMONIALS ARE]
Why test this: Social proof builds trust
Hypothesis: [WHAT YOU THINK WILL HAPPEN]

Variation A (Control): [CURRENT PLACEMENT]
Variation B: Testimonials above fold
Variation C: Video testimonials
Variation D: Specific metrics in testimonials

What to measure:
Trust indicators (time on page)
Scroll depth
Conversion rate

Success criteria: [PERCENTAGE] improvement
Confidence level: 95%
Run for: [DAYS/WEEKS]

**Test 5: Pricing Presentation**

Current presentation: [HOW PRICE IS SHOWN]
Why test this: Pricing affects purchase decision
Hypothesis: [WHAT YOU THINK WILL HAPPEN]

Variation A (Control): [CURRENT FORMAT]
Variation B: Annual pricing vs monthly
Variation C: Value stack shown first
Variation D: Comparison table

What to measure:
Time on pricing page
Click to checkout
Conversion rate
Average order value

Success criteria: [PERCENTAGE] improvement
Confidence level: 95%
Run for: [DAYS/WEEKS]

**Testing Calendar** (When to run each test)

Month 1:
Week 1-2: Test 1 (Headline)
Week 3-4: Analyze, implement winner
Month 2:
Week 1-2: Test 2 (CTA)
Week 3-4: Analyze, implement winner
Month 3:
Week 1-2: Test 3 (Form)
Week 3-4: Analyze, implement winner

Rule: Only test ONE element at a time
Why: Multiple changes = can't identify what worked

**What NOT to Test**

Don't test if:
Traffic under 1000 visitors/week
Current conversion under 1%
You haven't fixed obvious broken things
Sample size too small for significance
You can't commit to full test duration

Fix these first, then test:
Page load speed issues
Mobile responsiveness
Broken forms or links
Major usability problems

**Analysis Framework**

For each test, measure:
Conversion rate (primary metric)
Statistical significance (95% minimum)
Time to significance
Secondary metrics (engagement, revenue)

Declare winner when:
95% confidence reached
Minimum sample size met
Lift is meaningful (>10%)
Results are consistent over time

Declare inconclusive when:
No clear winner after 4 weeks
Results flip-flop
Insufficient traffic

**Common Testing Mistakes to Avoid**

Mistake 1: Stopping test too early
Fix: Wait for statistical significance

Mistake 2: Testing too many things at once
Fix: Isolate one variable

Mistake 3: Changing test mid-flight
Fix: Commit to full duration

Mistake 4: Ignoring external factors
Fix: Note holidays, promotions, seasonality

Mistake 5: Testing insignificant elements
Fix: Prioritize high-impact changes

**Documentation Template**

For each test, record:
Test name: [DESCRIPTIVE TITLE]
Start date: [DATE]
End date: [DATE]
Hypothesis: [WHAT YOU PREDICTED]
Variations: [WHAT YOU TESTED]
Winner: [A/B/C OR INC