The Report Said 30% Margin. The Bank Said Otherwise. Here's Where E-commerce Margin Actually Goes.

"Every product page said 30% gross margin. Every month, my bank account said 8%. It took three months to find the 22 points that disappeared between the two — and they weren't hiding, we just weren't looking."

An owner of a mid-sized clothing brand — ₴14M annual revenue, three sales channels (own site, two marketplaces), no physical retail — described the audit that changed how she reads her P&L.

Her product margin math was correct. Cost of goods per SKU averaged 42% of retail. That left 58% gross margin. Subtract 28% for the "usual" operating costs — marketing, salaries, rent for a small fulfillment space — and the model said 30% net. Every product decision had been made on that math for two years.

Actual net margin, once the full picture was reconstructed: 8%.

The 22 missing points were not in one place. They were distributed across five leaks that individually looked minor and cumulatively ate two-thirds of expected profit.

The Five Leak Points

The audit — done over a weekend and a Monday with her bookkeeper — traced revenue and cost month by month for the previous six months. Each of the five leaks was quantified. Leak one — returns and refunds (−6 points). Reported "returns rate" was 12% of orders. But when the audit counted the full cost of a returned item (the refund, the return shipping her business absorbed, the restocking labor, the discounted resale of returned inventory that couldn't be sold at full price), the effective cost per return was 1.6× the original margin on that unit. Applied across the volume, this alone was −6 points of net.

Leak two — marketplace fees and commissions (−5 points). Marketplace commissions of 15–18% were tracked. What wasn't: payment processing on top, promotional fees for placement, and the fee on returned orders that the platform kept. Total marketplace channel cost was closer to 22% of gross revenue on those channels.

Leak three — last-mile delivery subsidy (−4 points). Her site offered "free delivery over ₴800" to compete. Average delivery cost per order: ₴95. The subsidy averaged ₴58 per free-delivery order across her order mix. Applied across half her volume, another −4 points.

Leak four — packaging and fulfillment labor (−3 points). Padded envelopes, tape, labels, filler, thermal labels, thank-you cards, plus one-third of a full-time fulfillment person's time attributed to picking and packing. Sitting inside "operating costs" as a single line, invisible per-unit.

Leak five — ad spend booked as generic "marketing" (−4 points). She spent ₴180K/month on paid social and search. Attributed on the P&L to marketing. Not attributed per product line or per marketplace, so the low-margin SKUs and low-margin channels were being subsidized by high-margin ones without her knowing.

How She Fixed Each One

The response wasn't to cut everything. It was to see each leak clearly and decide, per leak, whether it was strategic, fixable, or wasteful.

• Returns: installed a size-guide overhaul and a fitting-video for each SKU. Six months later, return rate fell from 12% to 7.5%. Net recovery: +3 points.
• Marketplace fees: dropped the lowest-performing marketplace (the one with the highest total cost and lowest incremental sales). Consolidated ad spend onto own-site direct. Net recovery: +2 points.
• Delivery subsidy: raised the free-delivery threshold from ₴800 to ₴1,200. Average order value went up (customers added to reach threshold). Fewer subsidized orders. Net recovery: +2 points.
• Packaging & labor: renegotiated packaging supplier for larger volume, redesigned pack station for efficiency. Net recovery: +1.5 points.
• Ad spend attribution: re-cut spend by SKU and channel. Killed two SKUs that were structurally unprofitable at scale. Reallocated to the best margin SKUs. Net recovery: +3 points. The rebuild lifted net margin from 8% to 19%. Not the projected 30% — some of the drops were structural (marketplace fees, delivery expectations) and couldn't be eliminated without breaking the model. But 19% is a business with room to invest in growth. 8% is a business one bad month away from crisis.

The Framework You Can Use

The audit reproduces in any e-commerce business. It takes 8–12 hours of focused work. Step one. Pull three months of orders with full detail: gross revenue, discounts, marketplace fees, refunds, shipping charged to customer, shipping paid by business, packaging cost, return status.

Step two. Compute true net per SKU. Retail − COGS − marketplace fee − delivery subsidy per order − packaging per order − return allocation per order − ad spend allocated by SKU.

Step three. Rank SKUs by true net margin. Identify the bottom quartile. Decide per SKU: kill, reprice, redesign, or accept.

Step four. Rank channels by true net margin. Identify the channel where ad spend is being wasted. Reallocate.

Step five. Set up a monthly review of the same numbers so the leaks stay visible.

📌 Want to see where the margin is actually leaking in your e-commerce business? Send us three months of orders and channel data — we'll build the per-SKU true-margin waterfall and walk you through it in 15 minutes. Request your free Finmap diagnostic →

Read also

• How Finmap Helps Retail Businesses Establish Financial Order
• Sales ≠ Profit: How to Stop Margin Bleed in 10 Minutes a Day

Topic foundation: Unit Economics for a Small Business: Do You Actually Make Money on Each Sale?