Blended SKU and Bundle Margin Analysis for Multi-Product Sellers

A store-wide or category-wide average margin is one of the most reassuring and least reliable numbers in e-commerce reporting. It is easy to calculate, it moves slowly, and it can look perfectly healthy while several individual SKUs are quietly losing money on every unit sold, subsidized by a handful of high-margin products carrying the average upward. The average does not identify which products those are, and a seller managing to the blended number alone has no way of knowing where to intervene.

This matters more, not less, as a catalog grows. A five-SKU store can often keep the whole picture in someone’s head. A fifty- or two-hundred-SKU store cannot, and blended reporting becomes actively misleading rather than just incomplete once the catalog is large enough that no one is mentally tracking every product’s individual economics.

Why the Blended Average Hides Losing SKUs

A blended margin figure is a weighted average of every SKU’s individual margin, weighted by sales volume. That means a small number of high-volume, high-margin SKUs can pull the average up enough to fully offset a larger number of lower-volume SKUs that are break-even or losing money, and the resulting blended number gives no visibility into which group each product falls into.

The distortion gets worse in three common situations:

• Bundles and multi-packs priced at a discount without recalculating COGS. A bundle is often priced below the sum of its individual component prices to create a perceived deal, but if the cost of goods for the bundle is not separately recalculated as the sum of the actual component costs, the bundle’s true margin can be materially lower than assumed, or negative, even while the individual components sell profitably on their own.
• Promotional and discount SKUs carried at the same fee-and-freight assumptions as full-price SKUs. A product frequently sold at a discount, whether through a platform promotion, a coupon, or a clearance listing, absorbs the same per-unit fulfillment fee and advertising cost as a full-price sale, but on a lower revenue base, which compresses its margin more than the discount percentage alone would suggest.
• New or loss-leader SKUs added for catalog breadth or ranking purposes. Products added to round out a category, improve search ranking, or support a bundle, without a clear intention to run each one profitably on a standalone basis, quietly accumulate as margin drags if no one is checking them individually.

Per-Unit Landed Cost Plus Per-Unit Fee Allocation

Getting to a true SKU-level margin requires two components that are each often approximated storewide rather than calculated per SKU: landed cost and fee allocation.

Landed cost per unit is the purchase price plus freight, duties, and any per-unit prep, packaging, or inbound handling cost, allocated down to a single unit basis. The Landed Cost for Canadian E-Commerce Sellers guide covers how to build this figure; the point relevant here is that landed cost varies meaningfully by SKU (different weights, different supplier terms, different freight allocation) and using a single storewide average landed cost percentage instead of the actual per-SKU figure is one of the most common sources of blended margin distortion.

Per-unit fee allocation covers platform referral or transaction fees (which are usually already percentage-based and therefore self-allocating), plus fixed or semi-fixed costs that need to be spread across units: fulfillment fees that vary by size tier, storage fees allocated by average space occupied, and payment processing fees. Where a cost genuinely varies by SKU (a large, heavy item costs more to fulfill than a small, light one), applying a single average fee rate across all SKUs understates the true cost of the heavier item and overstates the cost of the lighter one.

The Bundle and Multi-Pack Pricing Pitfall

Bundling two or more SKUs together at a combined price below the sum of the individual listing prices is a common and often effective way to increase average order value. The pricing decision, however, is frequently made by comparing the bundle price to the sum of individual selling prices, without re-deriving the bundle’s cost of goods sold from the sum of the individual landed costs of its components.

A worked example illustrates the risk:

In this example the bundle still contributes positive margin, but at a materially lower rate than either standalone product, and if the combined fulfillment fee for the packed bundle runs higher than CAD $9 (a realistic outcome if the bundle is heavier or requires different packaging), the bundle’s contribution margin can compress further, or turn negative, while both standalone products remain healthy. Without a bundle-specific cost calculation, this compression is invisible in blended reporting; the bundle simply looks like additional revenue at an assumed margin that was never actually verified for the combined SKU.

How Discount SKUs Distort the Average

A SKU sold at a 20% discount does not simply lose 20% of its margin; it loses 20% of revenue while most of its cost structure (landed cost per unit, fixed per-unit fulfillment fee, payment processing on the lower amount) stays largely fixed or falls only slightly. On a product with a 40% gross margin at full price, a 20% price discount can reduce contribution margin percentage by considerably more than 20%, and on a lower-margin product the same discount can turn a marginally profitable sale into a loss. A blended average across full-price and discounted units in the same period masks this compression unless discounted units are tracked and analyzed as their own segment.

SKU-Level Margin Review Workflow

1. Pull a per-SKU sales report for the period, including units sold and gross revenue by SKU
2. Attach the true per-unit landed cost to each SKU, not a storewide average landed cost percentage
3. Attach per-unit fulfillment, storage, and payment processing costs using actual size/weight tiers or rates where they vary by SKU
4. Recalculate bundle and multi-pack COGS as the sum of actual component landed costs plus any bundle-specific packaging or fulfillment cost, rather than assuming the bundle’s cost mirrors its discounted price
5. Segment discounted or promotional unit sales separately from full-price sales of the same SKU
6. Calculate contribution margin (revenue minus landed cost, fees, and attributable advertising) at the SKU level, not just gross margin
7. Rank SKUs by contribution margin percentage and by contribution margin dollars, since a low-percentage but high-volume SKU can matter more to the bottom line than a high-percentage, low-volume one
8. Flag any SKU or bundle with negative or near-zero contribution margin for a pricing, cost, or catalog decision
9. Repeat the review on a recurring basis (monthly or quarterly) rather than as a one-time exercise, since fee schedules, landed costs, and discount frequency all shift over time

Scope of This Guide

This guide covers why blended margin reporting hides SKU-level and bundle-level losses, and how to build a true per-SKU margin view. It does not cover:

• Landed cost calculation methodology in detail, covered in Landed Cost for Canadian E-Commerce Sellers
• Inventory costing method selection (FIFO vs. weighted average) and its effect on COGS timing, covered in Inventory Costing Methods
• Channel-specific true profitability mechanics, covered in Shopify True Profitability and Amazon FBA True Profitability
• Pricing strategy or recommendations for specific discount or promotion tactics

A blended average is a useful headline number, but it should never be the only number a multi-product seller is looking at. The SKUs and bundles dragging the average down are identifiable, and once identified, they become a pricing or catalog decision rather than an invisible ongoing cost.

Get in touch if you have not run a SKU-level margin review since your catalog grew past a handful of core products. It is common to find at least one bundle or discounted SKU quietly running at a loss once the numbers are broken out.