A/B Testing
Test different descriptions to find what converts best.
How A/B Testing Works
Copyly's A/B testing lets you compare two or more description variants to see which performs better:
- Generate multiple description variants
- Create an A/B test and assign variants
- Deploy variants to your store (traffic is split automatically)
- Track views, clicks, and conversions
- Copyly declares a winner when statistically significant
Creating a Test
- Go to Dashboard → Analytics → A/B Tests
- Click "New Test"
- Select the product to test
- Add 2-4 description variants
- Set traffic split (default: equal split)
- Set minimum sample size (recommended: 100+ views per variant)
- Start the test
Metrics Tracked
Views
How many times each variant was displayed
Add to Cart Rate
Percentage of viewers who added the product to cart
Conversion Rate
Percentage of viewers who completed a purchase
Revenue per View
Average revenue generated per description view
Statistical Significance
Copyly uses statistical analysis to determine a winner:
- • 95% confidence level: Default threshold for declaring a winner
- • Minimum sample: At least 100 views per variant recommended
- • Test duration: Run for at least 1-2 weeks for reliable results
A/B Testing Best Practices
- • Test one thing: Change tone OR length OR structure—not all at once
- • Run long enough: Don't end tests early based on initial results
- • Document learnings: Apply winning patterns to other products
- • Test high-traffic products: Results come faster
Shopify Integration
When connected to Shopify, A/B tests automatically deploy variants to your store. Copyly handles the traffic splitting and tracks conversions through Shopify's analytics. Learn more →