How Retailers Can Identify Where Margin Is Really Being Lost

Why gross margin does not show the full retail economics

Gross margin usually compares net sales with cost of goods sold. It is a necessary measure, but it does not automatically include every cost created by a product, order, store, customer, or channel. Two sales with the same gross margin can therefore produce very different contribution margins.

An in-store basket may use existing staff and leave with the customer. An online basket may require picking, packaging, payment fees, delivery, customer support, and reverse logistics. If management compares only the selling price and product cost, the channel economics look more similar than they really are.

“A product can have a healthy gross margin and still destroy contribution after the cost of discounting, returns, fulfilment, and store support is included.”

Where retail margin is commonly lost

• Discount leakage: stacked offers, markdowns, loyalty rewards, and manual overrides reduce realised price beyond the planned promotion economics.
• Unprofitable mix: volume moves toward products, stores, customers, or channels with lower contribution even while total sales rise.
• Returns and reverse logistics: handling, inspection, repackaging, write-offs, and delivery costs are separated from the original sale.
• Fulfilment cost: picking, packing, last-mile delivery, marketplace fees, and customer service make cost-to-serve different by channel and order type.
• Inventory loss: shrinkage, expiry, damage, and obsolescence reduce the margin that appeared available when stock was purchased.
• Supplier economics: rebates, payment terms, freight allowances, and promotional support are not matched consistently to the products or periods they affect.
• Store and shared costs: rent, labour, utilities, technology, warehousing, and head-office support are allocated with rules that may conceal the real cost drivers.

The dimensions a retailer should connect

Margin analysis software is only as useful as the model behind it. The starting point is a reconciled view of sales, product cost, commercial adjustments, and operating costs. The second step is to preserve the dimensions needed to explain why the result changed.

• Product, SKU, category, brand, and supplier for product profitability analysis.
• Store, region, format, floor area, and trading hours for store profitability analytics.
• Channel, marketplace, order type, delivery method, and return method for channel profitability and cost-to-serve.
• Customer segment, loyalty tier, basket type, promotion, and campaign for commercial decision analysis.
• Inventory movement, markdown, waste, damage, and shrinkage for stock-related margin loss.
• Finance account, cost centre, period, and allocation driver for reconciliation and governance.

A practical method for locating the loss

1. Start with a decision, not a dashboard. Define whether leadership needs to assess a store, product range, promotion, customer segment, channel, or fulfilment choice.
2. Reconcile net sales and costs to the general ledger. Keep discounts, returns, rebates, write-offs, and unassigned costs visible instead of forcing an artificial match.
3. Build a contribution bridge. Move from list sales to realised net sales, product cost, variable selling costs, fulfilment cost, and directly attributable operating costs.
4. Compare several dimensions together. A weak category average may contain profitable stores and unprofitable channels; a strong channel average may conceal costly customer or order segments.
5. Test the allocation logic. Use operational drivers where possible and show leadership how the result changes under reasonable alternatives.
6. Separate structural loss from temporary variance. A one-period logistics disruption requires a different action from a product or channel that is consistently uneconomic.
7. Assign an action and review date. Pricing, assortment, supplier terms, promotion rules, inventory policy, staffing, and fulfilment design each require different owners and evidence.

Questions retail leaders should be able to answer

• Which products create sales but lose contribution after discounts, returns, and fulfilment?
• Which stores are underperforming because of local demand, assortment, labour, occupancy, or inventory loss?
• Which promotions create incremental profitable demand rather than shifting existing demand to a lower price?
• Which customer and order segments have the highest cost-to-serve?
• Where are supplier rebates and commercial terms changing reported product profitability?
• Which channel should receive the next unit of inventory, marketing investment, or operational capacity?

How AI can support retail profitability analysis

AI can help a finance or commercial team query a governed profitability model, identify unusual movements, summarise margin bridges, and generate follow-up questions. It should not invent missing costs, silently choose allocation rules, or treat correlation as a commercial decision.

For AI search and internal decision support alike, the answer is more reliable when the underlying definitions are explicit: what net sales includes, which costs are variable, how returns are matched, how shared costs are allocated, and which period and business dimensions are in scope. Human review remains essential before action is taken.

Turn the analysis into a management routine

The strongest outcome is not a one-time list of low-margin stores or products. It is a repeatable routine that reconciles finance and operational data, explains movement through a margin bridge, records approved assumptions, and connects each finding to a commercial or operational action.

MIZAN by NEXEL is the profitability-analysis path for retailers that want to apply this multi-dimensional method to their approved data. A working session starts with one priority decision and establishes the measures, dimensions, allocation rules, access, and review responsibilities needed for a controlled analysis.