The Problem: You're Measuring the Wrong Things
"I have 50,000 monthly sessions" – sounds impressive. But if conversion is 0.5%, you earn less than a store with 10,000 sessions and 3% conversion.
More important than traffic is what you do with it.
This guide focuses on metrics that directly impact revenue – not vanity metrics.
Part 1: Key E-commerce KPIs
Conversion Rate (CR)
Formula: Orders / Sessions × 100% Benchmark: E-commerce: 1–3%, top 25% of stores: >3.5%
The most important metric. If your CR is below 1%, you have a problem with UX, trust or offer – not with traffic.
Average Order Value (AOV)
Formula: Revenue / Number of orders How to increase: Upselling, cross-selling, free shipping thresholds, product bundles
Increasing AOV by 20% at constant CR = 20% revenue growth with no additional acquisition costs.
Cart Abandonment Rate
Formula: (Carts - Purchases) / Carts × 100% Benchmark: 65–75% (above 80% = problem)
Customer Lifetime Value (LTV / CLV)
Formula: AOV × Purchase frequency × Retention period Significance: Determines how much you can spend on customer acquisition (CAC)
Rule: LTV should be minimum 3× higher than CAC.
Customer Acquisition Cost (CAC)
Formula: Marketing spend / New customers Goal: Reduce CAC through campaign optimization and CR improvement
Return Rate
High returns (>15%) signal problems with product descriptions, photos or quality.
Part 2: How to Measure Conversion Correctly
Avoid Configuration Errors
- Make sure the order confirmation page is the only conversion goal
- Exclude test transactions and your own sessions (IP filter)
- Use Enhanced Ecommerce in GA4 to track the entire funnel
Micro-Conversions
Alongside main conversion (purchase) track:
- Add to carts
- Newsletter signups
- AI chat clicks
- Product page views > 30 seconds
- Search usage
Micro-conversions help identify funnel bottlenecks.
Conversion Funnel in GA4
Homepage → Category → Product → Cart → Checkout → Purchase
Check where you lose the highest percentage of users. If 40% drop between product and cart – you have a problem with the "Add to Cart" button or price.
Part 3: A/B Testing in E-commerce
What is an A/B Test?
You show 50% of users version A, 50% version B and measure which converts better.
What to Test (from highest impact):
- CTA button (color, text, size) – easy, quick results
- Product page headline – high conversion impact
- Price and discount presentation – direct impact
- Product images (white background vs lifestyle) – often underestimated
- Checkout layout (one-page vs multi-step)
- Shipping policy (free shipping thresholds)
Rules for a Proper A/B Test:
- Test one change at a time – otherwise you don't know what worked
- Minimum test duration: 2 weeks or until 100 conversions in each variant
- Statistical significance: 95%+ (use a calculator, e.g. abtestguide.com)
- Test at the same time (not: version A at Christmas, version B in January)
A/B Testing Tools:
- Hotjar – heat maps + tests
- VWO (Visual Website Optimizer) – popular in e-commerce
- Optimizely – for larger stores
- Convert.com – flexible and powerful
Part 4: Heat Maps and Behavior Analysis
Heat Maps
Visualize where users click and how far they scroll.
What to look for:
- Do they click on elements that aren't links? (Frustration – add link)
- How far do they scroll on the product page? (Do they see the "Add to Cart" button?)
- Which elements are ignored? (Remove or improve)
Session Recordings
Ability to replay a specific user's session. You'll literally see what they did on your site.
Look for: Where do they hesitate? Where do they go back? Where do they click multiple times without effect?
Tools:
- Hotjar – most popular, free plan available
- Microsoft Clarity – free, no limits
- FullStory – advanced, paid
Part 5: LTV, CAC and Margin – How to Calculate Them
Sample Calculations for a Store:
-
AOV = $90
-
Purchases/year/customer = 2.5
-
Retention = 3 years
-
LTV = $90 × 2.5 × 3 = $675
-
Monthly marketing spend = $2,500
-
New customers/month = 50
-
CAC = $50
-
LTV/CAC = 13.5 – very good ratio (target: min. 3)
Net Margin
Don't think in revenue. Calculate margin:
Net margin = Revenue - Product costs - Shipping - Marketing - Platform fees - Returns - Operational costs
Part 6: Sales Dashboard – What Should It Contain?
Daily KPIs (monitor every day):
- Revenue vs daily goal
- Number of orders
- Conversion rate (CR)
- Cart abandonment rate
- Number of AI assistant conversations
Weekly KPIs:
- AOV trend
- New vs returning customers
- Top 10 products
- Traffic sources and their conversion
Monthly KPIs:
- CAC per channel
- LTV cohort (how customers behave after 1, 3, 6 months)
- Return Rate
- Net Promoter Score (NPS)
Dashboard Tools:
- GA4 – free, basic
- Looker Studio – free, advanced visualizations
- Metabase – open source, for more advanced users
- Klaviyo Analytics – if using Klaviyo for emails
Part 7: AI Assistant Analytics
If you use SellsAI, you have access to unique data:
- Most common customer questions – know what they're looking for (and can't find)
- Most recommended products – what AI is selling
- Chat conversion – how many conversations end in a purchase
- Problem topics – what customers ask support about
This is invaluable product intelligence – customers are literally telling you what they need.
Checklist: Analytics Setup for E-commerce
Tracking:
- [ ] GA4 with Enhanced Ecommerce
- [ ] Meta Pixel (Facebook)
- [ ] Google Tag Manager
- [ ] Own session filter (IP)
Conversion:
- [ ] Goal: order confirmation page
- [ ] Micro-conversions: add to cart, signup
- [ ] Checkout funnel in GA4
Behavior:
- [ ] Hotjar or Microsoft Clarity
- [ ] Heat maps on top 5 product pages
- [ ] Session recordings at checkout
Testing:
- [ ] A/B test plan for Q1/Q2
- [ ] Statistical significance calculator
- [ ] Test results journal
Reporting:
- [ ] Daily dashboard (revenue, CR, orders)
- [ ] Weekly report (top products, sources)
- [ ] Monthly report (CAC, LTV, margin)
Summary
E-commerce analytics is not about collecting data – it's about making better decisions based on facts. Stores that systematically run A/B tests and optimize conversion grow 2–3× faster than those focused only on traffic.
Start with three steps: configure GA4 correctly, install Hotjar, and run your first A/B test on the CTA button. You'll see effects within 30 days.