Most e-commerce teams obsess over conversion rate. But here's the uncomfortable truth: conversion rate can go up while revenue goes down. This guide shows you how to optimize for what actually matters — revenue per visitor.
The Conversion Rate Trap
Imagine you run an A/B test on your product page. Variant B increases conversion rate by 15%. You celebrate and roll it out. But then you notice something strange in your revenue reports...
Control (A)
Conversion Rate: 2.0%
Average Order Value: $85
Revenue/Visitor: $1.70
Variant B (Winner?)
Conversion Rate: 2.3% (+15%)
Average Order Value: $65 (-24%)
Revenue/Visitor: $1.50 (-12%)
Variant B attracted more bargain hunters with smaller carts. You optimized for the wrong metric and lost 12% of your revenue.
The Right Metric: Revenue Per Visitor
Revenue Per Visitor (RPV) combines conversion rate AND order value into a single metric. It's the only metric that tells you the complete story.
The Formula
RPV = Conversion Rate × Average Order Value
or simply: Total Revenue ÷ Total Visitors
| Metric | Formula | Limitation |
|---|---|---|
| Conversion Rate | Purchases / Visitors | Ignores order value |
| Average Order Value | Revenue / Purchases | Ignores conversion rate |
| Revenue Per Visitor | Total Revenue / Visitors | The complete picture |
5 High-Impact Revenue Tests
Price point testing
Test $49 vs $59 vs $69 — higher prices may reduce CR but increase revenue
Bundle offers
"Buy 2 get 10% off" can increase AOV 20-40%
Free shipping threshold
Set threshold 20-30% above current AOV
Upsell placement
Test cart page vs checkout vs post-purchase
Premium tier visibility
Make higher-priced options more prominent
Deep Dive: Price Testing
Price testing is the highest-leverage test you can run, but most stores are afraid to try it. Here's how to do it right:
The Math of Price Testing
A 10% price increase that causes a 5% drop in conversions still increases revenue by 4.5%:
Before: $50 × 2% CR
RPV: $1.00
After: $55 × 1.9% CR
RPV: $1.045 (+4.5%)
How to Run Price Tests
- Test on a subset of products first, not your entire catalog
- Run for at least 2 weeks to account for weekly patterns
- Track revenue per visitor, not just conversion rate
- Consider long-term effects on returns and customer satisfaction
⚠️ Important Note
Price testing requires revenue tracking per variant. Most A/B testing tools only track conversions. ExperimentHQ automatically tracks Shopify revenue so you can run true price tests.
AOV Optimization Strategies
Free Shipping Threshold
Set your threshold 20-30% above current AOV. "You're $15 away from FREE shipping!" with a progress bar.
Typical result: +15-25% AOV
Bundle Discounts
"Buy 2, get 10% off" or curated bundles. Test different bundle sizes and discount levels.
Typical result: +20-40% AOV
Cross-Sell Recommendations
"Frequently bought together" or "Complete the look." Test AI vs curated recommendations.
Typical result: +10-30% AOV
Post-Purchase Upsell
One-click upsell on the thank you page. Customer has already committed — low friction to add more.
Typical result: +5-15% revenue
Revenue Testing Framework
Define Your Primary Metric
Always use Revenue Per Visitor as your primary metric. Track conversion rate and AOV as secondary metrics to understand the "why."
Calculate Required Sample Size
Revenue tests need larger sample sizes due to higher variance. Plan for at least 2-4 weeks of testing.
Segment Your Analysis
Look at revenue impact by device, traffic source, and customer type. A test might win overall but lose on mobile.
Wait for Statistical Significance
Don't call tests early. Aim for 95% confidence on your primary metric (RPV).
Monitor Post-Launch
Track revenue for 2-4 weeks after rollout. Some effects (like increased returns) take time to appear.
Revenue Tracking for A/B Tests
Most A/B testing tools only track conversions, not revenue. This makes true revenue optimization impossible. You need a tool that:
ExperimentHQ: Built for Revenue Optimization
ExperimentHQ automatically tracks Shopify revenue per variant. No manual setup required — just create a "Shopify Purchase" goal and revenue tracking is enabled. Start free →
Stop Optimizing for the Wrong Metric
Conversion rate is a vanity metric. It feels good when it goes up, but it doesn't pay the bills. Revenue Per Visitor is the only metric that tells you if you're actually making more money.
Start tracking revenue per variant in your A/B tests. You might be surprised to find that some of your "winning" tests are actually costing you money. And some tests you dismissed might be your biggest revenue opportunities.