Where does revenue fail to become profit—and which problems should a retail manager investigate first? A deeper look at product mix, regional economics and discount exposure across four US regions.
The decision: which product–region combinations deserve attention before making blanket discount cuts or dropping whole categories?
The main finding: Central’s weak margin survives adjustment for its broad product mix. National totals also hide losses in individual regions. Start with targeted checks of pricing, cost and product mix within those groups.
Data basis: rounded aggregates embedded in the original dashboard, representing 9,994 source records. The transaction file, date coverage and original SQL are unavailable in the public repository. Unique orders, dataset provenance and cost definitions are not verified.
Region buttons update the overview, loss review and segment table. Cross-region comparisons remain visible for context.
Recorded profit by product sub-category. A negative total does not mean every transaction lost money.
Recorded profit divided by sales within each source discount tier. These are associations, not estimated effects of changing prices.
Discounts may coincide with different products, customers or inventory conditions. These aggregates do not isolate those explanations. A margin below zero is a loss relative to recorded sales, not proof that every sale was unprofitable. The original tier labels overlap at 40%; their exact boundaries need the missing SQL or raw records.
Sales and recorded profit margin for the selected region, or all four regions.
Total recorded profit by customer segment. Compare margins below before making allocation decisions.
Regions sell different combinations of products. To compare them on a common basis, assign every region the same sales shares across the 17 sub-categories, while retaining each region’s observed margin in each sub-category.
| Region | Observed | Common mix | Change |
|---|---|---|---|
| West | 14.9% | 14.8% | -0.17 pp |
| East | 13.5% | 12.4% | -1.10 pp |
| South | 11.9% | 14.3% | +2.33 pp |
| Central | 7.9% | 6.9% | -1.05 pp |
This is an accounting comparison, not a forecast of what would happen if the assortment changed. It controls for broad sub-category sales shares only. Product models, discount levels, customer mix, shipping, timing and quantities can still differ within a sub-category.
For each sub-category, divide its sales summed across the four regions by total regional sales. Multiply that common weight by each region’s sub-category margin (profit ÷ sales), then sum across all 17 groups. All margins in this extension use the underlying displayed dollar totals, rather than rounded percentage fields. A percentage-point change is the difference between two percentages.
Central’s margin is 5.81 percentage points below the other three regions combined. A decomposition separates differences in product shares from differences in margin within those products.
| Sub-category | Central margin | Other regions | Gap contribution |
|---|---|---|---|
| Binders | -1.8% | 21.3% | -2.63 pp |
| Appliances | -11.2% | 24.7% | -1.69 pp |
| Furnishings | -25.6% | 22.2% | -1.45 pp |
| Storage | 4.3% | 10.9% | -0.60 pp |
| Machines | -5.5% | 3.0% | -0.46 pp |
Binders, Appliances and Furnishings contribute −5.78 percentage points together. That is about 84% of the net within-sub-category component; better results in Phones and Copiers offset part of the weakness elsewhere. Binders are especially worth examining: the national total is profitable, but Central’s total is negative.
This identifies where to investigate, not who or what is responsible. Possible explanations include deeper discounting, different items within the same group, unfavorable costs or customer mix. The aggregate data cannot distinguish them.
Let w be a sub-category’s share of regional sales and m its profit-to-sales margin. “C” is Central and “O” combines West, East and South. Mix = Σ(wC − wO)mO; within = ΣwC(mC − mO). Their sum equals Central’s margin minus the others’ margin. This uses the other regions as the reference; a different reference can change the split. Dollar-level rounding causes small differences from headline KPIs.
Across the four regions, 14 sub-category–region groups have negative aggregate profit totaling $36,971. Looking only at nationally negative sub-categories identifies $22,387. Regional gains offset $14,584 of losses in the national view, making local problems less visible.
These figures are net losses within groups, not the sum of all losing transactions. Profitable transactions inside each group may offset additional losses. Neither figure is automatically recoverable profit.
| Region | Sub-category | Sales | Profit | Margin |
|---|
Illustrative arithmetic, not a forecast. Hold sales constant and assume a selected share of this view’s net negative group profit is eliminated. Subtract an incremental intervention cost.
Net change = selected loss pool × assumed recovery share − intervention cost. Demand, cross-selling and cost avoidability are unknown; real results can be worse, including beyond this range. Do not add this loss pool to discount-tier losses: they overlap.
In the two negative-margin source tiers, 1,393 records—13.9% of the dataset—sum to $135,376 in net losses. They offset 32.1% of the $421,773 earned in the other two tiers. The resulting total is $286,397.
| Source tier | Records | Recorded profit | Margin |
|---|---|---|---|
| 0% | 4,798 | $320,988 | 29.5% |
| 1-20% | 3,803 | $100,785 | 11.9% |
| 21-40% | 460 | −$35,817 | -15.3% |
| 40%+ | 933 | −$99,559 | -77.4% |
The deepest tier accounts for $99,559 of these losses; Central contributes $40,793, or 41.0%, of that tier’s net loss. That makes it a sensible place to audit exceptions. It does not establish a universal safe discount threshold, and the source labels do not resolve how exactly 40% was classified.
The next question is incremental profit: would customers still buy with a smaller discount, and at what quantity? Simply removing a discount assumes away the demand response. The current aggregates cannot estimate that response, inventory holding costs or the value of clearing stock.
Consumer contributes the most total profit nationally, but Home Office has the highest profit-to-sales margin. Neither proves which segment would respond best to more spending: acquisition costs, retention and future demand are not available.
| Segment | Sales share | Profit | Margin |
|---|
| Priority | Evidence | Decision test |
|---|---|---|
| Audit Central’s Binders, Appliances and Furnishings | Largest negative contributions to its within-sub-category margin gap | Compare the same products, quantities and periods across regions; reconcile costs and discount exceptions. |
| Review East Tables and local loss groups | East Tables alone records a $11,025 net loss; national totals hide additional regional weakness | Identify repeated losses by product and customer before repricing or removing an assortment. |
| Pilot a narrower discount policy | Negative tiers offset 32.1% of positive-tier profit | Test eligible offers against current policy; judge incremental profit with demand and inventory effects included. |
Obtain transaction records with order and line identifiers, date, product ID, region, customer segment, quantity, discount, sales and profit, plus the definitions of cost and returns. Separate unique orders from line items and verify dates, duplicates, missing values and returns before fitting a model.
Within products that appear at multiple discount levels, compare profit and quantities within similar time periods and regions. Report sample sizes and overlap: products observed at only one discount cannot support a within-product discount comparison. A model with product and time controls can describe conditional associations, but selective discounting still prevents a causal interpretation.
For eligible products and customers, predefine a smaller-discount policy and compare it with current practice through a randomized pilot where feasible. Randomize at a level that limits spillovers, such as store or customer group. Use incremental contribution profit per eligible customer or store over the test window as the primary outcome, including non-purchases, returns and implementation cost. Set guardrails for conversion, volume, inventory clearance and repeat purchase. Choose test length and sample size from expected variability and a business-relevant minimum effect before launch.