Your Best-Selling Product Is Actually Losing Money. Here's the Math Most Sellers Never Do.

Let me tell you about a product I will call "The Widget." The Widget sells on Amazon for $24.99. It moves 2,000 units per month. That is $49,980 in monthly revenue from a single SKU.

The Widget is the hero product. The one the seller points to in every conversation about the business. "We do $50K a month on our best seller alone." Revenue growing 15% year over year. 4.4-star rating with 2,800 reviews. Top 3 in its subcategory.

The Widget is also losing $0.40 on every unit sold. The seller did not know this. They had never calculated the true per-unit profit. They looked at revenue, subtracted COGS, and assumed the rest was margin.

It was not. Let me walk you through the math.

The Math Everyone Does

Here is how most sellers calculate profitability:

Line Item	Amount
Selling price	$24.99
COGS (product cost)	-$6.50
Amazon referral fee (15%)	-$3.75
FBA fulfillment fee	-$3.42
"Profit"	$11.32

$11.32 per unit. A 45% margin. Looks incredible. With 2,000 units sold per month, that is $22,640 in monthly "profit." This seller thinks they have an extremely healthy product.

They do not. They have a product that looks profitable because they stopped counting costs too early.

The Math Nobody Does

Here is the full unit economics, every cost that touches this product:

Cost #1: True Landed COGS, $7.80 (not $6.50)

The seller quoted $6.50 as their COGS. That is the ex-factory price: what the manufacturer charges per unit. But the product does not teleport from the factory to the customer. There is freight, duties, and inspection to account for:

Component	Per Unit
Manufacturing cost (ex-factory)	$6.50
Ocean freight (40ft container, $4,200 / 3,200 units)	$1.31
Import duties (6.5% on declared value)	$0.42
Customs brokerage (amortized)	$0.08
Inland freight to warehouse	$0.14
Quality inspection (amortized)	$0.04
True Landed COGS	$8.49

The "real" cost of the product is $8.49, not $6.50. That $1.99 difference, applied to 2,000 units/month, is $3,980/month in costs the seller was not tracking.

Cost #2: Marketplace Fees, $3.75

The 15% referral fee is correctly calculated. $24.99 x 0.15 = $3.75. No surprises here.

Cost #3: FBA Fulfillment, $3.42

Also correctly calculated. Standard-size, 10-12 oz weight tier. $3.42 per unit. Correct.

Cost #4: Monthly Storage Fees: $0.28

This product sits in FBA for an average of 45 days before selling. At standard storage rates of $0.56-$0.87 per cubic foot per month (January-September average), a product occupying 0.35 cubic feet costs:

$0.56 x 0.35 cubic feet x 1.5 months average = $0.29/unit

Not huge. But it is not zero. On 2,000 units, that is $580/month.

Cost #5: Inbound Placement Fee, $0.27

Amazon charges $0.27 per unit for inbound placement (distributing inventory across fulfillment centers). Straightforward. Another $540/month.

Cost #6: Returns Cost, $1.87

This is where the math gets ugly. The Widget has a 12% return rate, about average for its category. Here is what each return costs:

Return Cost Component	Per Return
Lost revenue on returned unit	$24.99
Return shipping label	$4.20
Returns processing fee	$1.49
Original FBA fee (already paid, non-refundable)	$3.42
Restocking/inspection labor	$1.50
Condition downgrade (35% of returns unsellable as new, sold at 50% discount)	$4.37
Total cost per return	$15.58

Wait: $24.99 in lost revenue? Not exactly. The refund means revenue is reversed, but you still incurred the original COGS, fulfillment, and advertising costs. The net cost per return after accounting for the refund is approximately $15.58 in costs that are not recovered.

Blended across all units sold (returned and not returned): 12% return rate x $15.58 = $1.87 per unit sold.

On 2,000 units: $3,740/month in return-related costs. And this is an average return rate. Products in apparel, electronics, or fitness equipment see 15-25% return rates, which pushes the blended per-unit return cost to $2.50-$4.00.

Cost #7: Advertising: $5.40

Here is the number that kills the most products. The Widget spends $4,800/month on Amazon Sponsored Products and Sponsored Brands advertising. At first glance, the ACoS looks reasonable: $4,800 / $49,980 = 9.6% ACoS. Healthy, right?

But the per-unit calculation tells a different story. $4,800 / 2,000 units = $2.40 per unit in direct ad cost.

Except that is not the full advertising cost. The seller also runs:

• Off-Amazon advertising (Google Shopping, Meta) driving traffic to the Amazon listing: $1,200/month
• Amazon brand campaign to protect the brand name from competitors: $600/month
• Video ads for product awareness: $800/month
• Deal/coupon subsidies (coupons, Prime Exclusive Discounts): $1,200/month

Total advertising and promotional spend on this single product: $8,600/month. Divided by 2,000 units: $4.30 per unit.

But wait: there is one more advertising cost most sellers forget. The 12% of units that get returned? The advertising cost that acquired those sales is entirely wasted. So the effective ad cost per successfully sold unit is even higher: $8,600 / 1,760 units actually kept = $4.89 per unit.

We will use $4.89 as the true advertising cost.

Cost #8: Prep and Handling: $0.35

Each unit needs to be inspected, poly-bagged, labeled with FNSKU barcode, and packed for FBA shipment. Whether you do this in-house or use a prep center, the cost is approximately $0.30-$0.40 per unit. We will use $0.35.

Cost #9: Customer Service, $0.18

Not every unit generates a customer service interaction. But enough do, questions about sizing, shipping status, product usage, complaint handling, that the blended cost is real. At 15% of orders generating a customer service touch (average 8 minutes each at $20/hour): 0.15 x (8/60) x $20 = $0.40 per customer interaction, or $0.18 blended per unit including non-contact orders.

Cost #10: Software and Tools: $0.12

Inventory management tools, repricing software, analytics platforms, email marketing, review solicitation tools. Total software spend allocated to this product based on its share of total revenue: approximately $0.12 per unit.

The Full Picture

Line Item	Per Unit	% of Revenue
Selling Price	$24.99	100%
True Landed COGS	-$8.49	34.0%
Amazon Referral Fee	-$3.75	15.0%
FBA Fulfillment Fee	-$3.42	13.7%
Monthly Storage	-$0.29	1.2%
Inbound Placement Fee	-$0.27	1.1%
Returns (blended at 12%)	-$1.87	7.5%
Advertising and Promotions	-$4.89	19.6%
Prep and Handling	-$0.35	1.4%
Customer Service	-$0.18	0.7%
Software/Tools	-$0.12	0.5%
Net Profit Per Unit	+$1.36	5.4%

Hold on, $1.36 profit? That is actually positive. Where does the -$0.40 loss come in?

Two more costs that most sellers never allocate to individual products:

Cost #11: Refund Processing and Chargebacks: $0.52

Amazon does not refund your referral fee in full on returned items. They keep a portion. Plus, credit card chargeback costs, partial refunds for damaged-in-transit claims, and Amazon's restocking adjustments. Blended per-unit cost: $0.52.

Cost #12: Inventory Shrinkage and Loss: $1.24

Amazon loses and damages inventory in their fulfillment centers. The reimbursement you receive (now at sourcing cost, not selling price) does not cover the full value. Plus, there is natural shrinkage: units that disappear during transit, receiving errors, miscounts. For this product, the combined shrinkage rate is approximately 2.5%. At $8.49 COGS + lost profit per unit, the blended per-unit shrinkage cost is $1.24.

The REAL Final Number

Previous subtotal	$1.36
Refund processing/chargebacks	-$0.52
Shrinkage and loss	-$1.24
True Net Profit Per Unit	-$0.40

Negative forty cents. On every single unit. On the bestselling product.

At 2,000 units per month, this product is losing $800/month or $9,600/year. The more it sells, the more money the business loses.

And the seller thought this product was making them $22,640/month.

How This Happens to Smart Sellers

This seller is not stupid. They are running a $50K/month product. They know their COGS. They know their referral fee. They know their FBA fee. The problem is that 7 of the 12 cost categories are either invisible, averaged away, or allocated to "overhead" instead of tracked at the product level.

Nobody thinks to allocate per-unit advertising cost against individual products. Nobody calculates the blended return cost across all units including the non-returned ones. Nobody tracks the shrinkage rate by SKU. Nobody amortizes prep labor to the unit level.

So the product looks profitable in every report except the one that matters, the one that tracks every dollar in and every dollar out at the unit level.

The Five-Minute Profitability Check

You can do a rough version of this analysis in 5 minutes for any product. Here is the formula:

True Profit = Selling Price - (COGS x 1.3) - (Selling Price x Total Fee %) - (Total Monthly Ad Spend / Monthly Units Sold) - (Selling Price x Return Rate x 0.62)

Where:

• COGS x 1.3: Adds 30% to your ex-factory cost to approximate landed cost (freight, duties, prep)
• Total Fee %: Marketplace referral fee + fulfillment fee as a percentage of selling price (typically 28-35% on Amazon)
• Ad Spend / Units: Your total advertising cost divided by total units sold (not just ad-attributed units)
• Return Rate x 0.62: Return rate multiplied by 0.62 (approximate blended cost coefficient) x selling price

For The Widget: $24.99 - ($6.50 x 1.3) - ($24.99 x 0.30) - ($8,600 / 2,000) - ($24.99 x 0.12 x 0.62)

= $24.99 - $8.45 - $7.50 - $4.30 - $1.86 = $2.88

The quick formula gives $2.88: slightly positive but dangerously thin. The detailed analysis shows -$0.40. The difference is the costs the quick formula cannot capture (shrinkage, customer service, software). But even the quick formula tells you this product is in trouble: a 11.5% margin before overhead is not healthy.

Run this formula on your top 10 products right now. If any show margins below 10%, do the full analysis immediately.

What to Do When Your Best Seller Is Underwater

Option 1: Raise the Price

This is the highest-impact, lowest-effort fix. The Widget at $24.99 loses $0.40/unit. At $26.99 (an 8% increase), the math changes dramatically:

• Revenue per unit: +$2.00
• Referral fee increase: -$0.30
• Net improvement: +$1.70/unit
• New per-unit profit: $1.30

Most sellers assume a price increase will tank sales volume. Data says otherwise. A 5-10% price increase on Amazon typically reduces unit volume by 2-5%, not proportionally. The profit improvement on remaining sales far exceeds the revenue loss from fewer units. Test it for 14 days. Use a split test if your category supports it.

Option 2: Cut Advertising Spend

At $4.89/unit, advertising is the largest cost after COGS and marketplace fees. If the seller reduces total ad spend from $8,600/month to $5,000/month and units sold drop from 2,000 to 1,700 (a 15% volume reduction for a 42% cost reduction), per-unit ad cost drops to $2.94 and the product becomes profitable at $1.55/unit.

The fear: "If I cut ads, I will lose ranking." Maybe. But you are currently paying to lose money on every sale. Less volume at a profit beats more volume at a loss.

Option 3: Shift Volume to Lower-Fee Channels

The same product on Shopify with Shopify Payments:

Amazon	Shopify
Referral fee: $3.75 (15%)	Payment processing: $0.73 (2.9% + $0.30)
FBA fee: $3.42	Shipping cost: $4.50 (you ship)
Storage: $0.29	Storage: $0.15 (your warehouse, cheaper)
Inbound placement: $0.27	N/A: $0.00
Channel fees total: $7.73	Channel fees total: $5.38

Shopify saves $2.35 per unit in channel fees. That alone flips the product from -$0.40 to +$1.95 per unit. The tradeoff: you need to drive your own traffic to Shopify, which means marketing costs. But even with $3-$5 in customer acquisition cost per sale, the Shopify unit economics are better than Amazon's.

The operational challenge of selling on both channels simultaneously is inventory management. You need accurate, real-time inventory sync to prevent overselling across channels. This is a solvable problem: tools like Nventory keep counts synchronized across Amazon, Shopify, and other channels automatically. The operational overhead of adding Shopify as a channel is small compared to the margin improvement.

Option 4: Negotiate COGS

At 2,000 units/month (24,000/year), this seller has significant purchasing power. A 10% reduction in manufacturing cost, from $6.50 to $5.85, saves $0.65/unit, flipping the product to breakeven. Suppliers who have been with you for years often have room to negotiate, especially if you offer longer-term commitments or faster payment terms.

Option 5: Discontinue

The hardest option. Walking away from $50K/month in revenue feels suicidal. But $50K/month in revenue that costs $50,800/month to generate is not revenue: it is a subsidy program for Amazon, your ad platform, and your manufacturer. The capital, time, and mental energy going into this product could fund 2-3 smaller products with healthy margins.

The Exercise That Changes Everything

Rank your products by revenue. Take the top 20. Run the full unit economics analysis on each one. I guarantee at least 2-3 of your top 20 are either losing money or making less than $1/unit.

Then rank your products by per-unit profit. Your #1 revenue product might be #18 on the profitability list. Your #15 revenue product might be #1 in per-unit profit.

That profitability ranking is where your growth strategy should come from. Double down on the products that make you money. Fix or cut the ones that do not. Revenue without profit is just activity. And activity without profit is the most common reason ecommerce businesses fail while looking successful from the outside.