Blog · Board and management reporting
Ecommerce business intelligence connects orders, margin, repeat purchases, and returns by channel. Work through a channel comparison from exports, with formulas, customer definitions, reconciliation checks, and clear decision owners.
E commerce business intelligence connects order revenue, direct costs, repeat purchases, and returns so a team can explain which channels create value. Start with the exports behind the figures, agree on the definitions, and compare channels on the same basis. A sales increase alone cannot tell you whether orders became more profitable or customers became more loyal.
The useful question is specific: should the commercial team expand a channel, investigate a return problem, or change an offer? Each requires a different comparison. An order count explains activity, contribution explains the economics of that activity, and customer history shows who came back.
The broader business intelligence definition covers how business reporting supports decisions. For ecommerce, the practical unit is usually the order line, linked to its customer, refund, product cost, and channel. Keep those links intact even when the final output is a short summary.
Assign a decision owner before preparing the view. Ecommerce operations owns fulfillment and return exceptions. The commercial lead owns channel actions. Finance owns the revenue reconciliation and the agreed cost boundary. Customer marketing owns cohort definitions and the interpretation of repeat buying.
Bring an order export containing order ID, line ID, order date, customer ID, channel, quantity, gross merchandise sales, and discounts. Add refund records with their original order IDs and product costs for the items sold. Fulfillment and payment charges need the same order reference or a documented allocation basis.
Use a separate customer history to determine whether a buyer purchased before the reporting period. The first order visible in a short export is not necessarily the customer's first order with the company. Anonymous buyers and unresolved identities should stay in an explicit unknown group.
For this analysis, merchandise net sales exclude shipping revenue and sales taxes collected for tax authorities. The Census Bureau's retail sales definition deducts refunds and allowances and excludes sales taxes collected from customers and remitted directly to government. Document any difference between your management view and the ledger.
Match charges at order level before summarizing them by channel. An order with several lines must not acquire its full delivery charge on every line. If costs are allocated across lines, the allocated amounts must sum back to the original charge.
Juniper Home, an illustrative US online retailer, reviews September 2026 merchandise sales. Refunds below are positive deductions. COGS already reflects valid inventory cost reversals for returned merchandise. Variable order costs contain payment fees and outbound fulfillment, excluding acquisition marketing and corporate overhead.
| Channel | Gross sales (USD) | Discounts (USD) | Refunds (USD) | Net sales (USD) | COGS (USD) | Variable order costs (USD) |
|---|---|---|---|---|---|---|
| Direct website | 120,000 | 12,000 | 8,000 | 100,000 | 55,000 | 15,000 |
| Marketplace | 90,000 | 5,000 | 10,000 | 75,000 | 45,000 | 18,000 |
| Social checkout | 50,000 | 5,000 | 5,000 | 40,000 | 24,000 | 10,000 |
| Total | 260,000 | 22,000 | 23,000 | 215,000 | 124,000 | 43,000 |
Net sales = Gross sales − Discounts − Refunds
Order contribution = Net sales − COGS − Variable order costs
Contribution margin = Order contribution / Net sales
The IRS explanation of gross profit subtracts COGS from net receipts. Subtracting the separately identified variable order costs takes this management analysis below gross profit; it does not establish operating profit. Keep that distinction visible alongside the gross profit margin calculation.
| Channel | Net sales (USD) | Less COGS (USD) | Less variable costs (USD) | Order contribution (USD) | Contribution margin |
|---|---|---|---|---|---|
| Direct website | 100,000 | 55,000 | 15,000 | 30,000 | 30.0% |
| Marketplace | 75,000 | 45,000 | 18,000 | 12,000 | 16.0% |
| Social checkout | 40,000 | 24,000 | 10,000 | 6,000 | 15.0% |
| Total | 215,000 | 124,000 | 43,000 | 48,000 | 22.3% |
Percentages are rounded to one decimal place. Calculate the total margin as 48,000 / 215,000, using unrounded amounts. Averaging channel percentages would give each channel the same influence regardless of its sales.
The direct website leaves the most contribution in both dollars and percentage terms. That supports investigating what would happen to costs and acquisition spending before expanding it. It does not prove every extra website order would earn the existing margin.
For each channel, count distinct customers who ordered during September and flag those with an earlier completed purchase anywhere in the business. The working definition below is returning customer share, not a cohort retention rate.
| Channel | Customers purchasing in period | Customers with an earlier purchase | Returning customer share |
|---|---|---|---|
| Direct website | 800 | 240 | 30.0% |
| Marketplace | 600 | 90 | 15.0% |
| Social checkout | 320 | 48 | 15.0% |
Direct website share is 240 / 800 = 30.0%. Customers can purchase through multiple channels, so do not add these rows to claim a company customer count. Deduplicate across channels first. Missing marketplace identities also limit any comparison; report that coverage alongside the percentage.
The website combines higher contribution with a larger returning buyer share. That gives customer marketing a useful follow-up: compare acquisition cohorts over equal observation windows. It does not establish that the website caused loyalty. The broader ecommerce analytics workflow addresses those additional commercial questions.
The sales check is 260,000 − 22,000 − 23,000 = 215,000. The contribution check is 215,000 − 124,000 − 43,000 = 48,000, which also equals 30,000 + 12,000 + 6,000. Both identities must hold before anyone interprets a channel ranking.
Then reconcile net sales with the ledger using the same posting period and exclusions. Reconcile refunds to the refund register and variable costs to their charge files. Keep timing differences as named reconciliation items rather than forcing unmatched rows into a channel.
For the wider operating question set, see business intelligence for ecommerce. The analysis here establishes the order economics and buyer definitions those questions depend on.
Bring an order, customer, refund, and cost workbook, or a Google Sheet, with the identifiers used for the comparisons. Covirage maps the columns; its tools compute and check the figures, and the external AI model explains the results with every figure traced to its rows. Create a free Covirage account and bring a sample of your own export.
See customer profitability analysis for the supporting business context.
Start with order lines, refunds, product costs, fulfillment charges, and a customer purchase history. Keep order and customer identifiers in each relevant file. A channel summary can answer an initial question, but underlying rows are needed to investigate discrepancies.
Returning customer share describes the buyers active in a particular period who purchased before. Retention follows a defined acquisition cohort over a fixed observation window. Changing acquisition volume can move returning customer share even when the cohort's repeat behavior stays unchanged.
Use both, with an explicit cost boundary. Revenue shows commercial scale; contribution after product and variable order costs shows what that scale leaves to cover remaining expenses. Marketing costs and corporate overhead require separate treatment before calling the result operating profit.