Ecommerce Analytics Setup: What to Track and Why

Most ecommerce brands are drowning in data and starving for insight. They have Google Analytics installed, a few marketing pixels firing, and a Shopify dashboard they glance at occasionally. But their analytics setup is rarely configured properly, the data they collect is often inaccurate, and the metrics they focus on frequently lead them in the wrong direction.

Over 20 years of building and scaling ecommerce brands, we have learned that the difference between brands that grow consistently and brands that plateau is rarely about traffic or product. It is about measurement. Brands that measure the right things make better decisions. Better decisions compound into sustained growth.

This guide covers how to set up ecommerce analytics properly from the ground up. Not the theoretical frameworks you will find in textbooks, but the practical, battle-tested approach we use across every store we build and manage.

Why analytics setup matters from day one

Analytics is not something you add after launch. It is infrastructure that needs to be planned, implemented, and validated during the build phase. There are three reasons this matters:

First, you cannot improve what you do not measure. Every pound you spend on marketing, every design change you make, every product you add to the catalogue — the impact of each decision should be measurable. Without proper analytics, you are making expensive decisions based on intuition rather than evidence.

Second, historical data is irreplaceable. If your analytics are misconfigured for the first six months of trading, that data is lost permanently. You cannot retroactively fix tracking issues or recover data that was never collected. Every day with broken analytics is a day of insights you will never recover.

Third, accurate data improves over time. Machine learning models in GA4, advertising platforms, and email tools all improve with more data. Starting with clean, accurate tracking from day one means these systems learn faster and provide better recommendations sooner.

The essential analytics stack

Before discussing metrics, let us establish the tools. A properly configured ecommerce analytics stack includes:

Google Analytics 4 (GA4)

GA4 is the foundation. It tracks user behaviour across your site, measures conversion events, and provides the data needed for audience analysis and channel attribution. For ecommerce, GA4's enhanced ecommerce tracking is essential — it captures the full purchase funnel from product view through to completed order.

Key configuration requirements:

• Enhanced ecommerce events properly configured (view_item, add_to_cart, begin_checkout, purchase)
• Cross-domain tracking if you use subdomains or external checkout flows
• Internal traffic filtering to exclude your own team's activity
• Custom dimensions for business-specific data (customer type, product category, subscription status)
• Conversion events defined for all meaningful actions (not just purchases)

Google Search Console

Search Console provides data that GA4 cannot — actual search queries driving traffic, indexation status, Core Web Vitals data, and crawl statistics. For SEO performance tracking, it is indispensable.

Google Tag Manager

Tag Manager provides the control layer between your site and your analytics and advertising tools. Rather than hardcoding tracking scripts into your theme, Tag Manager lets you manage, test, and deploy tracking configurations without code changes. This is particularly important for:

• Deploying and updating marketing pixels without developer involvement
• Testing tracking configurations in preview mode before going live
• Managing consent-based tracking for GDPR compliance
• Implementing custom event tracking without theme modifications

Shopify Analytics

Shopify's native analytics provide accurate revenue data that serves as the source of truth for financial metrics. GA4 will always undercount revenue due to ad blockers, consent management, and tracking limitations. Shopify's server-side data captures every transaction.

Use Shopify Analytics for: revenue figures, order counts, average order value, and product performance. Use GA4 for: traffic analysis, channel attribution, user behaviour, and conversion funnel analysis.

Heatmap and session recording tools

Quantitative data tells you what is happening. Qualitative data tells you why. Tools like Microsoft Clarity (free) or Hotjar provide heatmaps showing where users click, how far they scroll, and where they abandon. Session recordings let you watch actual user journeys, revealing friction points that aggregate data cannot surface.

The metrics that actually matter

Not all metrics are equal. Some drive decisions. Others are noise. Here is the hierarchy we use for every ecommerce brand we work with.

Tier 1: Revenue metrics (review daily)

Metric	What it tells you	Benchmark
Revenue	Total sales performance	vs. same period last year
Conversion rate	Percentage of sessions that convert	1.5-3% for most UK ecommerce
Average order value (AOV)	Revenue per transaction	Category-dependent
Sessions	Total traffic volume	Trend vs. previous period
Revenue per session	Combined measure of traffic quality and conversion	£1.50-£5.00 typical

Tier 2: Acquisition metrics (review weekly)

These metrics tell you where your customers are coming from and how efficiently you are acquiring them:

• Customer acquisition cost (CAC): Total marketing spend divided by new customers acquired. The most important metric for assessing marketing efficiency.
• Return on ad spend (ROAS): Revenue generated per pound spent on advertising. Target varies by margin, but 4:1 is a common benchmark for healthy brands.
• Organic traffic share: The percentage of traffic from non-paid sources. Higher organic share means lower blended CAC and less dependence on paid channels.
• Email revenue percentage: Revenue attributed to email marketing flows and campaigns. Healthy brands typically generate 25-40% of revenue from email.
• New vs. returning customer ratio: The balance between acquisition and retention. A healthy ratio depends on your business model, but if over 80% of revenue comes from new customers, your retention strategy needs attention.

Tier 3: Behavioural metrics (review weekly)

• Add-to-cart rate: Percentage of product page views that result in an add-to-cart action. Typical range: 5-12%.
• Cart abandonment rate: Percentage of carts that do not complete checkout. Industry average is 65-75%, but high-performing stores achieve 55-65%.
• Product page bounce rate: How often users leave after viewing a single product page. High bounce rates on key products signal content or pricing issues.
• Site search usage and results: What customers search for (and whether they find it) reveals gaps in navigation and product catalogue.
• Mobile conversion rate: Tracked separately from desktop because mobile conversion is typically 40-60% lower and requires specific optimisation.

Tier 4: Strategic metrics (review monthly)

• Customer lifetime value (CLV): The total revenue a customer generates over their relationship with your brand. This is the metric that determines how much you can afford to spend on acquisition.
• Repeat purchase rate: Percentage of customers who buy more than once. Directly impacts CLV and long-term profitability.
• CLV to CAC ratio: The relationship between what a customer is worth and what it costs to acquire them. A ratio below 3:1 typically indicates an unsustainable business model.
• Net promoter score (NPS): Customer satisfaction and likelihood to recommend. Leading indicator of retention and word-of-mouth growth.

Setting up GA4 for ecommerce

GA4 requires specific configuration for ecommerce tracking. The standard installation captures pageviews and basic events, but meaningful ecommerce analysis requires enhanced ecommerce events.

Essential ecommerce events

These events map to the purchase funnel and must be configured with the correct parameters:

// Key GA4 ecommerce events
view_item          // Product page viewed
view_item_list     // Collection/category page viewed
add_to_cart        // Product added to cart
remove_from_cart   // Product removed from cart
begin_checkout     // Checkout initiated
add_shipping_info  // Shipping method selected
add_payment_info   // Payment details entered
purchase           // Order completed
refund             // Order refunded

Each event should include structured product data: item_id, item_name, item_category, price, quantity, and any custom parameters relevant to your business (subscription type, product line, colour, etc.).

Custom events for deeper insight

Beyond the standard ecommerce events, consider tracking:

• Size guide interactions (indicates sizing uncertainty — a leading indicator of returns)
• Image gallery engagement (how many images customers view before adding to cart)
• Filter and sort usage on collection pages
• Review interactions (reading reviews, filtering by rating)
• Email signup source and location
• Discount code application success and failure

Data validation

After implementation, validate thoroughly. Compare GA4 transaction data against Shopify's order data for a representative period. A discrepancy of 10-15% is normal (due to ad blockers and consent management). Over 20% indicates a tracking issue that needs investigation.

Use GA4's DebugView in real-time to verify events fire correctly with the right parameters. Test every step of the purchase funnel across devices and browsers. This validation is a critical part of our pre-launch QA process.

Attribution modelling explained

Attribution is how you assign credit for a conversion to the marketing touchpoints that contributed to it. It is also one of the most misunderstood aspects of ecommerce analytics.

Consider a typical customer journey: a customer sees a Facebook ad on Monday, clicks a Google search result on Wednesday, receives an email on Friday, and purchases on Saturday. Which channel gets credit?

Common attribution models

• Last click: The final touchpoint before conversion gets all credit. Simple but misleading — it undervalues awareness channels.
• First click: The first touchpoint gets all credit. Overvalues discovery and undervalues nurture.
• Data-driven (GA4 default): Uses machine learning to distribute credit based on the actual contribution of each touchpoint. The most sophisticated and generally the most accurate model.

The practical reality is that no attribution model is perfectly accurate. What matters is consistency — pick a model, understand its biases, and use it consistently so you can compare performance over time.

Cross-platform attribution challenges

Every advertising platform (Meta, Google, TikTok) reports its own attribution, and they all overclaim. A single conversion might be claimed by Meta Ads, Google Ads, and your email platform simultaneously. Understanding this overlap is essential for accurate budget allocation.

We recommend maintaining a blended view using your own GA4 data as the primary source, validated against platform-reported data. When there are significant discrepancies, investigate rather than simply trusting the platform that reports the best numbers.

Server-side tracking

Browser-based tracking is becoming increasingly unreliable. Ad blockers, intelligent tracking prevention in Safari and Firefox, cookie consent requirements, and third-party cookie deprecation all reduce the accuracy of client-side analytics.

Server-side tracking addresses this by sending conversion data directly from your server to analytics and advertising platforms. The data never touches the browser, so it is not affected by ad blockers or cookie restrictions.

Implementation options

• Google Tag Manager Server-Side: A server-side container that processes and forwards tracking data. Requires hosting (Google Cloud is the typical choice) and technical configuration.
• Meta Conversions API: Sends purchase and event data directly to Meta from your server. Essential for accurate Meta Ads optimisation and reporting.
• Shopify Web Pixels: Shopify's native approach to tracking that runs in a sandboxed environment, providing better data accuracy than traditional JavaScript-based tracking.

The impact on data accuracy

We typically see a 15-30% increase in tracked conversions after implementing server-side tracking. This is not "new" revenue — these are conversions that were always happening but were not being captured by browser-based tracking. More accurate data means better advertising optimisation, more reliable analytics, and more confident decision-making.

For any brand spending significantly on paid advertising, server-side tracking is no longer optional. The improvement in data quality directly translates to better campaign performance through more accurate optimisation signals.

Building a reporting framework

Data without structure is noise. A reporting framework defines what you look at, how often, and what actions the data should trigger.

The daily pulse

A five-minute daily check of key indicators: revenue vs. target, sessions, conversion rate, and any anomalies. The purpose is not analysis — it is early warning. If something drops significantly, you investigate. If everything is normal, you move on.

The weekly review

A 30-minute weekly review covering channel performance, product performance, and funnel metrics. This is where you identify trends, spot opportunities, and make tactical adjustments to campaigns, merchandising, and on-site experience.

Structure the weekly review around questions:

• Which channels are performing above or below target?
• Are there products with high traffic but low conversion? (Content or pricing issue)
• Are there products with high add-to-cart but low checkout completion? (Checkout friction)
• What is the email performance trend? (Open rates, click rates, revenue attribution)
• Are there any SEO performance changes worth investigating?

The monthly deep dive

A comprehensive monthly analysis covering strategic metrics, customer behaviour trends, cohort analysis, and competitive positioning. This is where you evaluate whether your strategy is working and make significant adjustments.

Monthly reporting should include:

• Cohort analysis: How are customers acquired in different months behaving over time?
• Channel efficiency: What is the true blended CAC, and how is it trending?
• Product analysis: Which products drive the most profit (not just revenue)?
• Customer segmentation: How do different customer segments behave differently?
• Funnel analysis: Where are the biggest drop-offs, and what has changed?

Common analytics mistakes

After working with hundreds of ecommerce brands, we see the same analytics mistakes repeatedly. Here are the most damaging ones:

1. Tracking everything, analysing nothing

More data is not better data. Tracking hundreds of custom events without a clear purpose creates noise that makes it harder to find signal. Track what you will act on. Everything else is a distraction.

2. Ignoring data quality

A dashboard showing inaccurate data is worse than no dashboard at all because it creates false confidence. Validate your tracking regularly — compare GA4 data against Shopify data, check for duplicate transactions, verify that refunds are being tracked, and ensure internal traffic is filtered.

3. Optimising for vanity metrics

Pageviews, time on site, and social media followers feel good but do not pay bills. Focus on metrics that directly correlate with revenue and profitability: conversion rate, AOV, CLV, and CAC.

4. Single-channel attribution

Evaluating each marketing channel in isolation leads to bad budget allocation. A channel might appear unprofitable on a last-click basis but be essential for driving awareness that other channels convert. Always evaluate channels in the context of the full customer journey.

5. Not segmenting data

Aggregate data hides insight. Your overall conversion rate might be stable, but when you segment by device, you discover that mobile conversion has dropped 30% — masked by a corresponding increase in desktop conversion. Always segment by device, traffic source, customer type, and geography.

6. Forgetting about consent and compliance

GDPR and the UK's PECR regulations require informed consent before setting analytics cookies. A poorly implemented consent mechanism can suppress 30-40% of your tracking data. Implement consent management properly, use server-side tracking to improve data capture within consent boundaries, and account for the data gap in your analysis.

---

Analytics is not a one-time setup exercise. It is an ongoing discipline that requires regular validation, refinement, and adaptation as your business, your tools, and the regulatory landscape evolve. The brands that treat analytics as infrastructure — with the same rigour they apply to their product and their marketing — are the brands that consistently make better decisions and grow faster.

If your ecommerce analytics need a proper setup or audit, we would be happy to help. We will assess your current tracking, identify gaps, and implement a measurement framework that actually drives decisions.