E-commerce Analytics – What to Focus On

E-commerce platform analysis is a complex topic — bounce rates, average order value, transaction flow, conversions, SEO impact. These are useful performance metrics, but from a strategic perspective, they are just part of a larger puzzle.

True e-commerce analytics begins where standard reports end. It is about understanding why customers behave in certain ways, what barriers they encounter, and what drives them to purchase or abandon. Only combining quantitative data with qualitative insights provides the complete picture.

Setting Up GA4 for E-commerce

Google Analytics 4 replaced Universal Analytics and introduced a fundamentally different data collection model — event-based instead of session-based. For online stores, proper Enhanced E-commerce implementation is critical, tracking the full purchase path from product view to order completion.

Key E-commerce Events in GA4

1. 1view_item_list — product list view (category page, search results)
2. 2view_item — product detail page view
3. 3add_to_cart — adding item to cart
4. 4begin_checkout — starting the checkout process
5. 5add_payment_info — entering payment details
6. 6purchase — order completion with transaction value
7. 7refund — order refund (partial or full)

Proper implementation requires collaboration with your technical team. The most common mistakes include missing transaction value parameters, duplicate purchase events on confirmation page refresh, and inconsistent product category naming.

Questions Worth Answering

• Why do visitors come to the site? What brought them here?
• Why aren't they clicking the CTA on the product page?
• What are they looking for? How can we help them find it?
• Why do so many people abandon their carts?
• Which products generate the most returns and why?
• How does the purchase path differ for returning vs. new customers?

Key E-commerce Metrics — Deep Dive

Not all metrics are equally important. The best e-commerce teams focus on a few key indicators that directly impact business profitability. Here are the most critical ones:

CAC — Customer Acquisition Cost

Customer acquisition cost is the sum of all marketing and sales expenses divided by the number of customers acquired in a given period. A healthy CAC should be 3-5x lower than customer LTV. Monitor CAC separately for each channel — this enables budget allocation optimization.

LTV — Customer Lifetime Value

Customer lifetime value is the projected revenue generated by a customer over their entire relationship with your business. For e-commerce, calculate it as: average order value × average purchase frequency × average retention period. LTV is the most powerful metric because it informs acceptable acquisition cost decisions.

AOV — Average Order Value

Average order value equals total revenue divided by number of orders. Increasing AOV by 10% with constant traffic directly translates to 10% revenue growth. Strategies for raising AOV include cross-selling, upselling, free shipping thresholds, and product bundles.

ROAS — Return on Ad Spend

Return on ad spend measures how much revenue each advertising dollar generates. A ROAS of 400% means every dollar spent generates $4 in revenue. The minimum acceptable ROAS depends on product margins — at 50% gross margin, 200% ROAS is break-even.

Attribution Modeling

Attribution modeling answers the question: which marketing channel deserves credit for the conversion? In a world where customers interact with a brand 7-12 times before purchasing, attributing all credit to the last click is inadequate.

• Last click — simple but favors bottom-of-funnel channels
• First click — credits discovery channels but ignores nurturing
• Linear — equal credit split across all touchpoints
• Data-driven (GA4) — algorithmic model based on actual conversion data
• Marketing Mix Modeling — advanced econometric analysis for large budgets

GA4 uses data-driven attribution by default, analyzing actual conversion paths and assigning credit based on statistical analysis of each touchpoint's impact. This is a significant improvement over Universal Analytics' simplified models.

Cohort Analysis

Cohort analysis groups customers based on a shared characteristic (e.g., month of first purchase) and tracks their behavior over time. It is a critical tool for understanding retention, seasonality, and the long-term value of different customer segments.

Example: the cohort of customers acquired in December (Black Friday) may have a completely different profile than the March cohort. Cohort analysis reveals whether "deal-seeking" customers return at regular prices or if their LTV is significantly below average.

User-Centered Approach

Traditional analytics only partially helps answer "why" questions. A user-centered approach reveals how customers landed, navigated, and ultimately left the platform — the complete customer journey.

Funnel Visualization

The e-commerce conversion funnel shows where you lose the most customers. GA4 offers Funnel Exploration reports that let you build custom funnels and segment them by various dimensions (device, traffic source, product category).

1. 1Category page → product page (typical drop-off: 60-70%)
2. 2Product page → add to cart (typical drop-off: 85-90%)
3. 3Cart → begin checkout (typical drop-off: 30-40%)
4. 4Checkout → payment (typical drop-off: 15-25%)
5. 5Payment → order confirmation (typical drop-off: 5-10%)

Complementary Tools

• Heat maps — visualizing page interactions (Hotjar, Microsoft Clarity)
• Session recordings — observing user journeys in real time
• Surveys — understanding customer expectations and tracking CX metrics
• Feedback widgets — instant user feedback on specific pages
• A/B testing — Optimizely, VWO, Google Optimize for hypothesis testing

Real-Time Dashboards

The best e-commerce teams build dashboards that present key metrics in real time. Tools like Looker Studio (formerly Google Data Studio), Tableau, and Metabase allow you to combine data from GA4, CRM, order systems, and ad platforms in a single view.

• Operational dashboard — orders, revenue, conversion (hourly refresh)
• Marketing dashboard — ROAS, CAC, channel performance (daily)
• Strategic dashboard — LTV, cohorts, category profitability (weekly/monthly)

Predictive Analytics

GA4 introduced built-in predictive models that forecast purchase probability and churn risk. This data can be used to create remarketing segments — for example, targeting users with high purchase probability with increased ad budgets.

Advanced teams go further, building custom predictive models: demand forecasting, optimal pricing, return probability, and RFM segmentation (Recency, Frequency, Monetary). These models require an integrated data warehouse and data science capabilities.

Privacy-Compliant Tracking

GDPR, ePrivacy, and browser changes (third-party cookie blocking) are fundamentally changing e-commerce analytics. Companies must transition to first-party data and server-side tracking models to maintain measurement accuracy.

• Consent Mode v2 — required by Google since March 2024
• Server-side tagging — GTM Server Container for more accurate tracking
• First-party data — building a strategy based on owned data
• Enhanced Conversions — sending hashed user data to Google Ads
• Conversion API — direct integration with Meta Ads (Facebook)

Data can measure site performance quantitatively, but it is not the knowledge that will help you understand problems at the UX and CX level. Combine quantitative and qualitative data to make sound business decisions.