If you run an Amazon or Shopify business, your bank feed is lying to you – not on purpose, but by omission. The deposit that lands in your account each week or two isn’t your revenue. It’s a net settlement: gross sales minus refunds, discounts, marketplace and payment-processing fees, shipping adjustments, ad spend, and sometimes a reserve Amazon is holding back for later. If your books post that single deposit number as “sales,” you’re not just misstating revenue – you’re erasing the fee line, the refund line, and the inventory cost line that tell you whether a given product is actually profitable.
Consider an illustrative Shopify-based seller doing roughly $4.2M in annual revenue across a core product line and a handful of Amazon FBA SKUs (a composite example, not a real client). When their bookkeeper switched from recording bank deposits as sales to reconciling full settlement details – gross sales, fees, refunds, and COGS booked separately – reported gross margin moved by roughly nine points. That’s the difference between a product line that looks healthy and one quietly funding a loss. The swing is illustrative of the type of distortion net-deposit bookkeeping can create, not a universal benchmark – your own numbers will differ.
This article walks through why net deposits distort your numbers, how Amazon and Shopify settlement data actually works, how to separate gross sales from fees from COGS, and what a monthly close process that gets this right looks like. The central idea to hold onto: a marketplace deposit is a cash-flow event, not a complete income statement.
In a traditional retail or service business, a sale and a cash receipt are close to the same thing: you invoice, the customer pays, the bank shows the full amount. Marketplace and payment-platform selling breaks that link. Between the click of “buy” and funds hitting your account, several things happen that never show up in the deposit total: the platform nets out commission and processing fees, nets out refunds or chargebacks, may net out advertising spend billed against the account, and may hold back a reserve. What lands in your bank account is what’s left over – not what you sold.
Booking that net number directly as revenue understates gross sales, hides the fee expense entirely, and makes it hard to track return rates by product since refunds are netted out before you see them. It also breaks any attempt to compute a real gross margin, since the deposit already has an unknown mix of COGS-adjacent and non-COGS items baked into it.
Here’s an illustrative example using round numbers. Suppose a seller generates $100,000 in gross sales during a settlement period. Refunds and discounts reduce that to $92,000. Marketplace commission, payment processing, and fulfillment fees take another $10,000. Shipping label costs and minor settlement adjustments account for $2,000 more. The seller’s actual bank deposit for the period is $80,000.
A bookkeeper who records the $80,000 deposit as “Shopify sales” has understated gross revenue by $20,000, recorded zero fee expense (which belongs on the P&L as an operating cost, not a silent haircut to revenue), and has no visibility into the $8,000 of refunds that may be concentrated in one or two SKUs worth investigating. Reconciling correctly means booking the full $100,000 as gross sales, then separately posting refunds, fees, and shipping adjustments – landing at the same $80,000 cash figure, but with every component visible.
A settlement period should flow through your books in this order:
Running parallel to this, on the balance sheet and COGS side:
Amazon issues settlement reports roughly every two weeks (cadence depends on your account), and each one bundles several transaction types. Order-related items (“ItemPrice” entries) capture product revenue, shipping charges to the customer, and any gift-wrap or tax collected. Refunds appear as their own transaction type and represent cash returned to a buyer – not the same thing as a physical return, which is a separate operational event your inventory system needs to track independently. Fees – referral commission, FBA fulfillment fees, and other service charges – are deducted within the same settlement.
Two mechanics are worth understanding specifically because they distort the deposit-to-revenue relationship further:
These mechanics are described here at a general, operational level based on how Amazon settlement reporting is structured; specifics can change, and reserve and fee policies should be confirmed against your current Seller Central account and Amazon’s own seller documentation.
Shopify Payments (or a connected third-party processor) batches transactions over a payout period and deposits the net amount, typically daily or every few days depending on your plan and processor. The payout figure nets out the payment-processing fee, refunds processed since the last payout, and chargebacks or disputes. If you also run app-based subscriptions or loyalty charges through the same account, those need to be separated out rather than lumped into “sales.”
Because Shopify and Amazon settle on different schedules with different fee and reserve behavior, a multi-channel seller’s bank feed is really several interleaved settlement streams. Reconciling channel by channel – not treating the combined bank balance as one undifferentiated sales number – is what makes channel-level profitability analysis possible at all.
The standard fix for the deposit-vs-revenue problem is a clearing account (sometimes called an undeposited funds or marketplace suspense account) on the balance sheet for each sales channel. Gross sales, refunds, and fees post to the clearing account as they occur, based on the settlement report detail – not the bank feed. When the deposit hits, it’s matched against the accumulated balance. If the two don’t tie out, the difference is a real reconciling item (a reserve, a timing lag, an unrecorded fee) rather than a mystery variance buried in “sales.”
This is the mechanism most eCommerce-specialized bookkeeping setups are built around, including settlement-automation platforms like A2X: they pull settlement detail from Amazon and Shopify, break it into component categories, and post summarized journal entries to the clearing account and P&L so gross sales, fees, and refunds appear separately instead of collapsing into one deposit line.
Settlement Line Item | Accounting Treatment |
|---|---|
Gross product sales (ItemPrice / line-item sales) | Revenue – Gross Sales |
Shipping charged to customer | Revenue – Shipping Income |
Refunds / returns (cash) | Contra-Revenue – Refunds & Allowances |
Marketplace commission / referral fee | Operating Expense – Marketplace Fees |
Payment processing fee (Shopify Payments, etc.) | Operating Expense – Payment Processing Fees |
FBA fulfillment fee | Operating Expense – Fulfillment Fees (or COGS, per policy) |
Advertising spend billed through platform | Operating Expense – Advertising |
Reserve withheld | Balance Sheet – Marketplace Receivable |
Reserve released | Clears Marketplace Receivable |
Inventory shipped / COGS recognized | COGS – Product Cost |
Chargebacks / disputes | Contra-Revenue or Bad Debt Expense, per policy |
Net settlement deposit | Cash – clears Clearing Account balance |
Whether fulfillment fees sit in COGS or operating expense is a policy choice – discuss it with your CPA and apply it consistently, since it affects how gross margin compares across channels.
Cost of goods sold is the cost directly attributable to the units you sold in a given period – not everything you spent on the product line. For a physical-product eCommerce business, COGS typically includes the unit product cost, inbound freight and duties to get inventory into a sellable location, and directly related packaging. It generally excludes marketing, platform fees, warehousing overhead not tied to specific units, and general administrative cost – those are operating expenses, tracked separately so gross margin reflects product economics specifically.
A common error is costing inventory at the vendor unit price alone. The more accurate figure is landed cost – everything it takes to get one unit into inventory and ready to sell. For an illustrative SKU with a $20 vendor unit cost, landed cost might also include $2 of inbound freight, $1 of duties, and $1 of allocated inspection or prep-center handling, bringing the true landed cost to $24 per unit – 20% higher than the vendor invoice alone would suggest. These figures are illustrative only; your actual landed-cost components depend on your suppliers, freight lanes, and fulfillment setup.
Using vendor cost instead of landed cost systematically overstates gross margin, sometimes by several points, and the gap tends to be worst on imported or freight-heavy SKUs – exactly the products where an accurate margin read matters most for pricing and ad-spend decisions.
Landed cost is only useful if it’s tracked at the SKU level rather than as a blended average across the whole catalog. A blended average smooths over the fact that a slow-moving, high-freight-cost SKU can look profitable next to a fast-moving, low-cost bestseller when the two are averaged together – and product-level pricing or ad decisions made on the blended number will be wrong for both products. SKU-level costing also matters for physical inventory count reconciliation, shrinkage tracking, and identifying which products are actually driving contribution to overhead.
Inventory Location | Units on Hand |
|---|---|
Amazon FBA warehouses | – |
Third-party logistics (3PL) warehouse | – |
Owned/leased warehouse | – |
In transit (inbound to any location) | – |
Customer returns pending inspection | – |
Total inventory on hand | – |
This table is a template, not a filled-in example – the point is that inventory sitting in FBA, a 3PL, or in transit is still your asset and still needs to be reconciled to a total, even though you don’t have eyes on the physical shelf. Sellers who only track “what’s in my warehouse” routinely underestimate total inventory and, by extension, understate COGS timing when units in FBA sell.
Separate from how you present margin to yourself internally, there’s a real tax-accounting question: does your business have to maintain a full inventory-and-accrual method for tax purposes, or can it qualify for the small-business taxpayer exception? Under Internal Revenue Code §471(c), a taxpayer that meets the §448(c) gross-receipts test – average annual gross receipts over the prior three tax years at or below the inflation-adjusted threshold – can treat inventory as non-incidental materials and supplies, or use a method that conforms to its financial (book) accounting treatment of inventories, rather than the general accrual/inventory rules. For the 2025 tax year that threshold was $31 million; the IRS has since released the inflation adjustment for 2026, setting the threshold at $32 million (Revenue Procedure 2025-32, Section 4.30). The same gross-receipts test also governs eligibility for the cash method of accounting and the small-business exception from the §263A uniform capitalization (UNICAP) rules, which otherwise require capitalizing certain indirect costs into inventory.
A few points matter for accuracy here. First, this is an aggregation test – related entities under common control are generally combined when measuring gross receipts, so a seller operating through multiple LLCs can’t necessarily stay under the threshold by splitting revenue across entities; this requires a facts-specific analysis. Second, qualifying is a tax-method election with real consequences (potentially requiring Form 3115 to change an existing method), not something to adopt informally. Third, the 2027 threshold had not yet been published as of this writing – do not assume the 2026 figure carries forward; verify the current-year amount against the applicable IRS revenue procedure. Given the complexity, this is a conversation for a CPA who can review your specific structure, not a self-service determination.
Even a business that qualifies for the §471(c) small-business exception on its tax return generally still benefits from tracking inventory and COGS accurately in its management books – the internal financial statements used to price products, evaluate channels, and make ad-spend and hiring decisions. Tax method and internal reporting answer different questions: the tax method determines how income is measured for the IRS; management accounting determines whether you actually understand your margin. A business can legitimately use a simplified method for tax purposes while still building out full landed-cost, SKU-level, channel-level reporting internally – in fact, that combination is common and appropriate for a lot of growing eCommerce sellers.
Line Item | Amount |
|---|---|
Gross sales (all channels) | |
Less: refunds & discounts | |
Net sales | |
Less: cost of goods sold (landed cost of units sold) | |
Gross profit | |
Less: marketplace & payment processing fees | |
Less: fulfillment fees (if not in COGS) | |
Less: advertising | |
Less: other operating expenses | |
Operating income |
This is a structural template – no figures are populated – but the ordering matters: fees and advertising sit below gross profit as operating expenses, not netted into revenue or COGS, so gross margin reflects product economics and operating margin reflects the full cost of running the channel.
Gross margin (net sales less COGS, divided by net sales) tells you about product economics in isolation. Contribution margin goes a step further and also nets out the variable costs of actually selling the unit – marketplace fees, payment processing, and often a reasonable per-unit allocation of advertising – leaving the amount each sale actually contributes toward fixed overhead and profit. Two products can carry identical gross margins and very different contribution margins once fee structure and ad spend per channel are factored in, which is why channel-level and SKU-level contribution analysis, not gross margin alone, should drive pricing and channel-mix decisions.
Product | Primary Channel | Gross Margin | Contribution Margin |
|---|---|---|---|
Product A | Amazon FBA | Higher (fulfillment-fee heavy) | Lower after fees |
Product B | Shopify (owned fulfillment) | Lower (no FBA fee) | Higher after fees |
Product C | Both channels | Mixed | Varies by channel mix |
This table is illustrative and directional, not a real product set – it exists to show how the same product can rank differently on gross margin versus contribution margin depending on channel-specific fee structure.
☐ 1. Pull complete settlement reports from every channel (Amazon, Shopify, and any others) for the full period, not just the bank deposits.
☐ 2. Reconcile gross sales per settlement report to the sales channel’s own reporting dashboard.
☐ 3. Post gross sales, refunds, fees, and shipping as separate line items to each channel’s clearing account.
☐ 4. Match cumulative clearing-account activity to actual bank deposits for the period; investigate any variance.
☐ 5. Update landed cost for any SKUs with new vendor pricing, freight, or duty changes during the period.
☐ 6. Recognize COGS against units shipped, using current landed cost by SKU.
☐ 7. Reconcile inventory on hand across FBA, 3PL, owned warehouse, and in-transit locations to a single total.
☐ 8. Review and true up any Amazon reserve balances held versus released during the period.
☐ 9. Review deferred or cross-period settlement transactions and confirm they’re recorded in the correct period.
☐ 10. Reconcile advertising spend billed through each platform to the platform’s own ad reporting.
☐ 11. Review gross margin and contribution margin by channel and by top SKUs for anomalies.
☐ 12. Close the period and prepare management-basis financials for review.
Answer honestly – each “no” is a place net deposits are likely hiding your real margin:
☐ Do your books show gross sales, refunds, and fees as separate line items for every channel?
☐ Is inventory costed at landed cost, by SKU, rather than vendor price or a blended average?
☐ Does inventory sitting in FBA, a 3PL, or in transit appear on your books as an asset you track?
☐ Do you reconcile Amazon reserve holds and releases separately from your regular settlement activity?
☐ Do you know your contribution margin – not just gross margin – by channel and by top SKU?
☐ Have you confirmed your current-year §448(c) gross-receipts threshold rather than assuming last year’s figure?
☐ If you operate multiple related entities, have you tested gross receipts on an aggregated basis?
☐ Do you close your books monthly, rather than reconstructing the year at tax time?
If several of these are “no,” your reported margin is probably not your real margin. That’s the gap a proper diagnostic closes.
NexusWorks works with Amazon and Shopify sellers to rebuild settlement reconciliation, landed-cost tracking, and channel-level margin reporting from the ground up, as part of a broader eCommerce accounting and advisory practice. That work typically starts with the underlying eCommerce financial systems a growing seller needs, runs through the day-to-day bookkeeping that keeps settlement and inventory data current, and extends into financial advisory and optimization once the numbers can be trusted. For sellers who need company-level financial leadership, that can include fractional CFO support, coordinated with NexusWorks’ tax filing and compliance work so the same clean numbers feed both pricing decisions and the tax return.
Because the deposit is a net settlement figure - gross sales minus refunds, fees, and any reserve or timing adjustments - not a revenue figure. Reconciling the full settlement report, not just the deposit, is what closes the gap.
Gross sales is the total value of orders before any deductions. The net deposit is what's left after refunds, marketplace and payment fees, and any reserve holds are subtracted. Both numbers matter, but only gross sales belong on the revenue line of your P&L.
It's a policy choice that should be made deliberately and applied consistently; many sellers treat core fulfillment fees as a COGS-adjacent cost while treating referral commission and advertising as operating expenses. Discuss the right treatment for your business with your CPA.
Landed cost is the vendor unit price plus freight, duties, and directly related handling to get a unit into sellable inventory. Using vendor price alone typically overstates gross margin, especially on imported or freight-heavy SKUs.
It lets a business that meets the §448(c) gross-receipts test treat inventory as non-incidental materials and supplies, or conform to its book treatment of inventory, instead of using the general tax inventory and accrual rules. The threshold is adjusted for inflation each year - $32 million for 2026 - and eligibility should be confirmed with a CPA.
