Do I need a data warehouse for e-commerce analytics?

Not initially. Most stores under $5M revenue can rely on their analytics platform's built-in storage and reporting. A data warehouse becomes valuable when you need to combine data from many sources for custom analysis, run complex queries across millions of rows, or build machine learning models on historical data.

How much should I spend on analytics tools?

A reasonable budget is 1-3% of revenue for analytics and data infrastructure. For a store doing $1M, that is $10-30K per year. The ROI should be measurable - better attribution leading to better ad spend allocation typically pays for the analytics stack many times over.

Should I build custom analytics or use off-the-shelf tools?

Use off-the-shelf tools for 90% of your needs. Custom analytics makes sense only for unique business logic that no platform supports. Building and maintaining custom analytics infrastructure is expensive and distracts from your core business. Start with platforms, and only build custom where they genuinely fall short.

The Modern E-commerce Analytics Stack

The tools, integrations, and data flows you need to build a complete analytics infrastructure for your online store.

The analytics landscape for e-commerce has fragmented. Google Analytics handles web traffic. Your ad platforms report campaign performance. Shopify has its own dashboards. Email platforms track opens and clicks. Each tool sees one piece of the picture, and stitching them together manually is a full-time job.

A modern analytics stack connects these pieces into a unified view where customer journeys flow across platforms, attribution spans channels, and decisions are based on complete data rather than platform-specific fragments.

Anatomy of an Analytics Stack

Layer 1: Data Collection

Data collection is the foundation. Everything built on top depends on capturing complete, accurate data at every customer touchpoint.

**On-site tracking** captures visitor behavior: pageviews, product views, add-to-cart events, checkout steps, and purchases. Your tracking script should capture UTM parameters, click IDs (fbclid, gclid), referrer data, and device information alongside behavioral events.

**Server-side events** ensure critical conversions are captured even when client-side tracking fails. Server-side tracking via platform APIs and webhooks provides a reliable backup for purchase events.

**Ad platform data** comes from API integrations with Facebook, Google, TikTok, and other advertising platforms. This data includes spend, impressions, clicks, and the platform's own attribution - which you will compare against your unified attribution.

Layer 2: Data Storage

Collected data needs a home. The choice of storage affects query performance, cost, and what analysis is possible.

For most e-commerce businesses, a PostgreSQL database handles transactional data (events, sessions, conversions) well. For larger volumes, a data warehouse like BigQuery or Snowflake enables complex analytical queries across large datasets. Cloud storage handles bulky data like session recordings.

Layer 3: Processing and Attribution

Raw data becomes useful through processing. This layer handles identity resolution (connecting anonymous sessions to known customers), attribution modeling (assigning conversion credit across touchpoints), metric calculation (conversion rates, LTV, CAC, ROAS), and anomaly detection (alerting when metrics deviate from expected patterns).

Layer 4: Visualization and Action

Processed data feeds dashboards, reports, and automated actions. The output layer includes real-time dashboards for monitoring, periodic reports for trend analysis, alerts for anomalies and opportunities, and integrations that push insights back to ad platforms and email tools.

Essential Tools by Function

Web Analytics

The baseline is Google Analytics 4 for traffic analysis and basic conversion tracking. For deeper behavioral insights, tools like Arktis provide session-level detail, real-time monitoring, and first-party data collection that works regardless of ad blockers.

Attribution

GA4 offers basic data-driven attribution within its ecosystem. For cross-platform attribution that includes Facebook, Google, TikTok, and direct traffic in a single model, you need a dedicated attribution platform. Revenue attribution connects ad spend to actual revenue, not just conversions.

Customer Data

A customer data platform (CDP) or unified customer profile system connects data from all touchpoints into a single view per customer. This enables segmentation, personalization, and lifecycle marketing based on complete behavioral data.

Session Recording

Session recordings show exactly what visitors experience on your site. Watching real user sessions reveals UX issues that quantitative data alone cannot explain - confusing navigation, broken layouts, or unclear calls to action.

Building Your Stack

For Early-Stage Stores (under $500K revenue)

Keep it simple. Use your e-commerce platform's built-in analytics (Shopify Analytics, WooCommerce reports) plus Google Analytics 4 for traffic insights. Add a tracking script that captures UTM parameters and click IDs. This covers 80% of your analytics needs with minimal cost and complexity.

For Growing Stores ($500K-5M revenue)

Add dedicated attribution tracking, session recording, and ad platform integrations. Connect Facebook and Google ad data to on-site conversion data for true ROI measurement. Implement customer segmentation based on behavioral data. This is where a platform like Arktis delivers the most value - connecting the dots between ad spend, on-site behavior, and revenue.

For Established Stores ($5M+ revenue)

Layer in a data warehouse for complex analysis, predictive models for revenue intelligence, and automated workflows that act on analytics insights. At this scale, the analytics stack becomes a competitive advantage that enables faster and better decisions.

Common Stack Mistakes

Too Many Disconnected Tools

Adding a new tool for every need creates data silos. Prefer platforms that integrate natively and share data through APIs. Five well-connected tools beat fifteen disconnected ones.

Tracking Everything, Analyzing Nothing

Collecting data is easy. Deriving insights and taking action is hard. Focus on the metrics that actually drive decisions: key e-commerce KPIs like conversion rate, CAC, LTV, and ROAS. Track these well before adding complexity.

Ignoring Data Quality

Garbage in, garbage out. If your tracking has gaps, your attribution is wrong, and your decisions are based on bad data. Invest time in proper tracking setup and regular data quality audits.

Sources

1. Segment, "The State of Customer Data"

2. Gartner, "Marketing Analytics Technology"

3. Google Cloud, "Analytics Architecture"