BI for E-Commerce Companies: Metrics, Dashboards and Tools

Shopify analytics and Google Analytics show you the funnel. They do not show you whether you’re making money. An e-commerce business that tracks conversion rate and cart abandonment but not contribution margin per channel, gross margin per SKU, and inventory turnover is optimising the wrong things.

Mid-market e-commerce companies face a specific data problem: their operational information is fragmented across an ERP, a WMS or 3PL system, multiple ad platforms, marketplace seller accounts, and a web analytics tool. Each system gives a partial view. None of them answer the question that actually matters: which channel, which product, and which customer segment is profitable?

A BI layer over your e-commerce stack answers that question — and changes how you make inventory, marketing, and fulfilment decisions.

Key Takeaways

• 67% of e-commerce companies track marketing metrics but not contribution margin by channel — optimising spend without knowing profitability
• Multi-channel e-commerce companies that unify data sources grow revenue two times faster than those operating data silos
• GA4 consistently understates actual revenue — never use it as the authoritative source for financial metrics
• Mid-market e-commerce companies using BI report 12–18% improvement in inventory turnover

What Business Intelligence Means for E-Commerce

Native platform analytics — Shopify analytics, Amazon Seller Central, Google Analytics — are designed to show you what happened in their specific domain. Shopify shows orders and revenue processed through Shopify. Amazon shows performance within Amazon. GA4 shows traffic behaviour. None of them cross boundaries.

Beyond Native Platform Analytics

BI adds the integration layer that connects all these sources. The result is metrics that cross boundaries: total customer CLV across channels, gross margin by SKU factoring in returns and fulfilment costs, ROAS by ad channel attributed to actual revenue (not GA4-reported revenue).

For small DTC brands on Shopify, native analytics may be sufficient. For mid-market companies with multi-channel distribution, wholesale accounts, multiple warehouses, and complex order management, native analytics leaves critical questions unanswerable.

Descriptive vs Predictive Analytics

Descriptive analytics — what happened — is the baseline. Revenue by channel last month. Top-selling SKUs last quarter. These are reporting-level questions that most e-commerce analytics tools can answer.

Predictive analytics — what will happen — is where BI creates additional value. Using historical sell-through rates to forecast inventory requirements for a seasonal peak. Identifying customers whose purchase patterns suggest they’re about to churn. Modelling the margin impact of a proposed promotional discount. These capabilities require a data foundation that goes beyond native platform reporting.

Why Mid-Market E-Commerce Needs BI More Than DTC Startups

A single-channel DTC brand with 20 SKUs and one warehouse has simple data needs. A mid-market company with 1,000+ SKUs, multiple fulfilment centres, wholesale and direct channels, and marketplace presence on Amazon and eBay has data spread across a dozen systems. The complexity of the business justifies — and requires — a dedicated BI layer.

The cost of not having that layer is specific: over-investing in low-margin marketing channels, carrying excess inventory in the wrong locations, missing the margin impact of high-return product lines, and making pricing decisions without visibility into product-level profitability.

The E-Commerce Metrics That Actually Matter

Most e-commerce companies track too many metrics and the wrong ones. Here’s the set that informs actual decisions.

Revenue and Financial Metrics

• Net revenue by channel — gross revenue minus returns and discounts, segmented by direct, marketplace, wholesale
• Gross margin by SKU — the most important product-level metric; identifies which products actually make money after COGS and fulfilment
• Contribution margin by channel — gross margin minus channel-specific costs (marketplace fees, ad spend, fulfilment cost per channel)
• Return rate by product — high return rates compress margin significantly; track by SKU and product category
• Average order value (AOV) by channel and customer segment

Customer Metrics

• Customer lifetime value (CLV) — the total profit generated by a customer over their relationship with the business; the denominator for evaluating customer acquisition cost
• Customer acquisition cost (CAC) by channel — not just cost per click or cost per conversion, but total channel cost divided by new customers acquired
• Repeat purchase rate — what percentage of customers buy a second time within 90 days; the primary indicator of whether you have a retention problem
• Cohort retention — how revenue and purchase frequency evolve for customers acquired in the same period over time

Operational Metrics

• Inventory turnover — how many times you sell through average inventory in a period; the primary measure of inventory health
• Stockout rate — the percentage of time SKUs are unavailable due to stock depletion; directly reduces revenue
• Days on hand — average inventory days remaining at current sell rate; informs reorder timing
• Fulfilment cycle time — from order placement to shipment; affects customer satisfaction and repeat purchase
• Return processing cost — the fully loaded cost of handling a returned item

Marketing Metrics

• Return on ad spend (ROAS) — attributed revenue divided by ad spend, calculated against actual margin not reported revenue
• New vs returning customer split by marketing channel — many channels over-report new customer acquisition by counting repeat buyers
• Attribution accuracy — is your attribution model producing plausible numbers, or is multi-touch attribution inflating every channel’s contribution?

Metrics to Stop Tracking

Website sessions, page views, and email open rates are activity metrics. They don’t drive decisions. An e-commerce company that replaces three of these with contribution margin by channel and inventory turnover by SKU is making better use of analytical capacity.

The Five Dashboards Every Mid-Market E-Commerce Company Needs

1. Executive Overview Dashboard

Four to six headline metrics: net revenue vs target (current period and YTD), gross margin, top-performing and worst-performing channels by contribution margin, and a cash flow position summary. This view answers: “Is the business on track?”

Update cadence: daily for in-season periods, weekly in off-season.

2. Marketing Performance Dashboard

ROAS by channel, CAC by channel, new customer count by acquisition source, and campaign-level contribution margin (not revenue). Include a comparison of marketing spend mix vs revenue contribution to surface channels where spend is disproportionate to return.

The critical integration here: connect ad platform cost data to actual revenue data from the ERP, not GA4. This usually shows significantly different channel efficiency than GA4-attributed analysis would suggest.

3. Inventory and Operations Dashboard

Inventory position by SKU and warehouse, days on hand by category, stockout alerts, return volume by product line, and fulfilment cycle time by channel. This dashboard is used daily by the operations and merchandising teams.

The insight this enables: proactive reorder decisions before a stockout, identification of high-return SKUs before they consume more marketing budget, and fulfilment bottleneck detection before it affects customer experience.

4. Customer Analytics Dashboard

CLV by acquisition cohort, repeat purchase rate by channel, days to second purchase distribution, and revenue concentration (what percentage of revenue comes from top 20% of customers). This dashboard informs CRM strategy, loyalty investment, and win-back campaign targeting.

5. Financial Dashboard

P&L by channel with gross margin and contribution margin, inventory value on balance sheet by category, seasonal cash flow projection based on inventory commitments, and accounts payable aging for merchandise suppliers. This is the CFO’s view.

Case study — Natalie Huang, COO at a multi-channel apparel brand with $28M in revenue:

Natalie’s company was growing revenue at 30% per year but margin was declining. The marketing team was increasing ad spend — ROAS looked strong in the ad platforms. But nobody was connecting ad spend to actual margin. After building a contribution margin by channel dashboard, Natalie discovered that their highest-revenue channel (Amazon) had a contribution margin of 8% after marketplace fees, fulfilment, and return processing. Their lowest-volume channel (direct) had a contribution margin of 34%. The business had been optimising for revenue while inadvertently degrading profitability. Rebalancing channel investment over the following two quarters improved total contribution margin by 11 percentage points.

Connecting Your E-Commerce Data Sources

The integration architecture is the hardest and most important part of e-commerce BI.

ERP: The Source of Truth for Revenue and COGS

Your ERP is the authoritative source for financial data. Order management, inventory cost, fulfilment costs, and returns processing all run through the ERP. Any financial metric — revenue, margin, COGS — should trace back to ERP data, not to platform-reported figures.

For companies running Netodin’s ERP, order and inventory data is structured for downstream analytics, making the connection to a BI layer straightforward.

Marketplace Data Normalisation

Amazon and eBay report revenue differently from each other and from your own website. Fees are structured differently. Return handling works differently. Normalising across platforms — establishing a common revenue and cost definition — is one of the most time-consuming integration tasks in e-commerce BI.

Define the normalisation rules before building: what counts as “net revenue” for each channel, how marketplace fees are classified, and how returns are handled. Document these definitions. They become the foundation of every cross-channel comparison.

Ad Platform Cost Data

Google Ads, Meta Ads, and TikTok Ads each have native connectors to most BI platforms. Pulling cost data is usually straightforward. The challenge is attribution: connecting an ad spend record in Google to an actual revenue record in your ERP.

Do not use GA4 as the attribution bridge. GA4 uses last-touch attribution by default, systematically over-credits certain channels, and understates revenue (due to ad blockers, browser privacy settings, and direct navigation). Use order data from your ERP as the revenue source. Map it to first-touch or data-driven attribution based on your business model.

WMS and 3PL Data

Fulfilment data — pick/pack cycle time, shipping costs, return processing — comes from your WMS or 3PL provider. This data is essential for calculating true contribution margin by channel: a channel with high ROAS but high fulfilment cost per order may be less profitable than a lower-ROAS channel with efficient fulfilment.

Most WMS and 3PL providers have data export capabilities. Automated nightly exports to a central data store are sufficient for the metrics above.

BI Tool Options for E-Commerce

Power BI

Best fit for companies already using Microsoft infrastructure — Microsoft 365, Dynamics ERP, Azure. Strong native connectors for most data sources. Relatively low per-user cost. Limited in advanced data modelling without significant technical investment. The right choice for most mid-market companies with an IT team familiar with Microsoft tools.

Looker

Best fit for companies on Google Cloud infrastructure with a dedicated data team. Looker’s modeling layer (LookML) is powerful but requires technical expertise to configure. Strongest for companies with complex data transformations and a data engineer on staff.

Tableau

Best for organisations with dedicated data analysts who need rich visualisation capabilities. Higher per-user cost than Power BI. Stronger visualisation depth, weaker on data modeling. More common in retail and e-commerce analytics teams with analytical talent.

Purpose-Built E-Commerce Analytics Tools

Tools like Glew, Polar Analytics, and Triple Whale are built specifically for DTC Shopify brands. They connect to Shopify, ad platforms, and GA4 out of the box, with pre-built e-commerce dashboards. They’re excellent for single-channel DTC businesses. They’re not built for multi-channel complexity, ERP integration, or the operational analytics needs of mid-market companies.

Seasonal Demand Forecasting with BI

One of the highest-value applications of e-commerce BI is inventory forecasting for seasonal peaks.

Using two to three years of historical sell-through data by SKU and channel, a BI platform can project the inventory required for Q4, summer peak, or a promotional event. The model combines historical velocity, projected traffic (from marketing plans), and seasonal adjustment factors.

The output informs purchase order quantities, warehouse space allocation, and supplier lead time planning — weeks before the season starts, instead of after stockouts have already occurred.

Companies that integrate their BI inventory projections with their purchasing workflow report measurably lower stockout rates and lower end-of-season inventory overhang. Both directly improve cash flow and margin.

Case study — Derek Osei, Inventory Director at a home goods e-commerce company with 180 employees:

Derek’s company had recurring stockout problems on their top 50 SKUs every Q4 despite building in what the buying team believed were conservative buffers. The team was using year-over-year sell-through rates and manually adjusting for new product launches and promotional plans in a spreadsheet. After connecting their ERP and WMS data to a BI platform and building a demand forecast model that incorporated channel growth rates and promotional calendar data, stockouts in Q4 2025 fell by 62% compared to the prior year. Overstock at season end also fell 28%, recovering $400,000 in working capital.

Common BI Mistakes E-Commerce Companies Make

Using GA4 as Revenue Source of Truth

GA4 consistently understates actual revenue — typically by 15–30% depending on the business — due to ad blockers, browser privacy settings, and data sampling. Every financial analysis built on GA4 revenue data is systematically inaccurate.

Use your ERP or order management system as the authoritative revenue source. Use GA4 for behavioural data — traffic, page engagement, session analysis — where it is appropriate.

Tracking Marketing Metrics Without Connecting to Margin

High ROAS is meaningless without margin context. A channel with a 4x ROAS on products with a 20% gross margin generates less profit than a 2x ROAS channel selling products with 50% gross margin. Track contribution margin by channel, not ROAS alone.

Not Accounting for Returns in Revenue Calculations

Returns can be 20–40% of gross revenue in apparel and footwear categories. A revenue analysis that doesn’t net out returns overstates performance significantly and leads to inventory and marketing investment decisions made on inflated numbers.

Calculate all revenue metrics on a net-of-returns basis. Track return rate separately and monitor it by SKU and channel to identify product-level quality issues and category-level fulfilment problems.

FAQ

What e-commerce data sources should we connect first? Start with your ERP (revenue, orders, COGS) and your primary ad platform (cost data). These two sources enable contribution margin by channel calculations, which is the highest-value insight for most mid-market e-commerce companies. Add GA4 for traffic data and your WMS for fulfilment cost in phase two.

How do we handle multi-currency in e-commerce BI? Define a reporting currency and a conversion rate policy — typically average monthly rate for P&L and spot rate for balance sheet items. Apply this consistently across all channels. Multi-currency e-commerce businesses that don’t standardise conversion rate policy produce comparable metrics that aren’t actually comparable.

How long does it take to build an e-commerce BI environment from scratch? For a mid-market company with multiple data sources, allow three to six months from project kickoff to a stable, trusted dashboard environment. The majority of that time is spent on data integration, normalisation, and quality validation — not on the visualisation layer. Companies that rush the integration phase have adoption problems because users don’t trust the numbers.

Should we build the BI environment in-house or use a specialist? For the initial build, a specialist with experience in e-commerce data integration will move significantly faster than an in-house team learning on the job. Bring the work in-house for ongoing maintenance and iteration once the foundation is stable. Hybrid approaches — specialist-led build, in-house team trained during the project — produce the best long-term outcomes.

Conclusion

Effective e-commerce BI connects the front-end funnel to the back-end operation and financials. The companies getting the most value from it are not those with the best dashboards — they’re the ones that connected their ERP to their analytics layer and started making decisions based on contribution margin by channel rather than platform-reported revenue.

Start with the integration work. Connect your ERP. Define net revenue and contribution margin consistently across channels. Build the executive overview and marketing performance dashboards first. Then add inventory analytics, customer analytics, and financial detail.

The insight that comes from seeing the full picture — marketing, inventory, fulfilment, and financial data in one place — is qualitatively different from what any single-platform analytics tool can provide.