Amazon seller accounting is the discipline of converting Amazon's settlement reports, FBA fees, reserves, reimbursements, and Marketplace Facilitator sales tax into clean books that match the bank. For Shopify brands who expanded to Amazon, the hard part isn't the software — it's keeping COGS, revenue, and fees split correctly across two channels selling overlapping SKUs.
Most guides on this topic walk you through connecting QuickBooks to Seller Central and call it done. That's the easy 20%. The 80% that breaks books at month-end is the stuff competitors gloss over: settlement period cutoffs that straddle month-end, reserve balances Amazon holds for 7 to 14 days, FBA reimbursements that look like revenue but aren't, and sales tax Amazon collects on your behalf that still has to land in the right GL bucket.
This guide covers how the 100+ Shopify brands Ottit closes books for actually handle Amazon accounting when Amazon is a second channel. We'll get into chart of accounts setup, the journal entry structure for a settlement, the COGS allocation problem, and the reconciliation cadence that keeps things from compounding into a quarterly mess.
What Makes Amazon Seller Accounting Different From Shopify Bookkeeping?
Amazon seller accounting differs from Shopify bookkeeping in four core ways: deposits are net of dozens of fee types, sales tax is collected by Amazon under Marketplace Facilitator laws, inventory is held and shipped by FBA warehouses, and settlements run on a 14-day cycle that rarely aligns with month-end. Each difference creates a specific reconciliation challenge.
On Shopify, a payout is mostly clean: gross sales minus Shopify Payments processing fees, deposited every 1-3 days. Amazon dumps a settlement every two weeks that bundles 30+ line item types — referral fees, FBA fulfillment fees, storage fees, long-term storage fees, removal fees, advertising charges, refunds, chargebacks, returns processing, lightning deal fees, coupon redemption fees, and more.
If a brand posts the net deposit as 'Amazon Sales,' the P&L is wrong by definition. Gross revenue is understated, every fee category is invisible, and there's no way to see what's actually eating margin. We see this pattern constantly when brands come to us from a generalist bookkeeper.
| Dimension | Shopify | Amazon |
|---|---|---|
| Payout cadence | 1-3 days (Shopify Payments) | Every 14 days (settlement) |
| Sales tax handling | Store collects and remits (or via tool) | Amazon collects and remits (MTF) |
| Fee complexity | Processing fee + app fees | 30+ line item types per settlement |
| Inventory custody | Brand or 3PL (e.g., ShipBob) | FBA warehouses |
| Reserve balances | Rare on established stores | Common, especially for new sellers |
| Refund timing | Same-day or near-real-time | Can lag a settlement period |
Reserves are the silent killer. Amazon withholds a portion of each settlement to cover potential refunds and chargebacks, and that reserve balance can move several thousand dollars between settlements. If books treat the reserve change as cash, the bank reconciliation breaks. If books ignore it, the asset side of the balance sheet is wrong.
The Amazon settlement is not a transaction. It's a summary of hundreds of transactions, and accounting it as one line is the most common reason multi-channel books are wrong.
Actionable takeaway: before connecting any tool, list every fee type and reserve movement that appears on a recent Amazon settlement and confirm there's a GL account for each. If 'Other Amazon Fees' is doing heavy lifting, the chart of accounts isn't ready.
How Should a Shopify Brand Set Up the Chart of Accounts for Amazon?
A multi-channel chart of accounts uses parent accounts for revenue, COGS, and fees, with sub-accounts or class/location tracking to split Shopify and Amazon. Most Shopify brands selling on Amazon use 4-6 dedicated Amazon revenue and contra-revenue accounts and 8-12 Amazon-specific expense accounts to capture the major fee categories without over-engineering.
In QuickBooks Online, the cleanest setup uses Classes (or Locations) to tag every transaction by channel. In Xero, Tracking Categories do the same job. This way, the chart stays flat — one 'Product Revenue' account — but every report can be sliced by channel. The A2X documentation for Shopify accounting and the equivalent A2X Amazon mapping both support this pattern out of the box (the A2X documentation for Shopify accounting){nofollow}.
Here's the account structure the typical $1M-$10M Shopify brand uses once Amazon is added:
- Revenue accounts: Product Revenue (with Class = Amazon or Shopify), Shipping Revenue, Gift Card Revenue
- Contra-revenue: Refunds, Promotional Discounts, Coupon Redemptions
- Amazon fee accounts: Referral Fees, FBA Fulfillment Fees, FBA Storage Fees, FBA Long-Term Storage Fees, Amazon Advertising, Removal & Disposal Fees, Subscription Fees (Pro Seller)
- Amazon balance sheet accounts: Amazon Clearing (asset), Amazon Reserve (asset), Sales Tax Payable - Amazon MTF (liability, often pass-through)
- FBA inventory: separate Inventory - FBA account from Inventory - 3PL or Inventory - Warehouse
The Amazon Clearing account is the workhorse. Every settlement gets posted in full to clearing, then the net deposit clears it when it hits the bank. The Amazon Reserve account holds the running balance Amazon is withholding. At any point, Clearing should be near zero after the deposit lands, and Reserve should match what Seller Central shows.
Separating FBA inventory from non-FBA inventory matters more than it sounds. A 5,000-unit transfer to FBA is not a sale — it's a location change. But if inventory is one bucket, that movement is invisible, and you can't tell whether the units sitting in FBA match what Amazon's reports say. Brands that run Shopify COGS accounting properly already have this discipline; adding Amazon just extends it.
If 'Other Amazon Fees' on the P&L is bigger than 'Referral Fees,' the chart of accounts isn't catching what it should — and margin analysis is fiction.
Actionable takeaway: a typical setup has roughly 15-20 Amazon-specific GL lines (revenue, contra, fee, balance sheet) plus class/location tracking on every transaction. Anything less and channel reporting won't hold up at $5M+.
How Do You Reconcile an Amazon FBA Settlement Report?
Reconciling an FBA settlement means posting one summary journal entry per settlement period that captures gross sales, refunds, every fee category, advertising, reserve movement, and Marketplace Facilitator sales tax — with the offset to the Amazon Clearing account. The bank deposit then clears against Amazon Clearing, leaving a near-zero balance if everything was captured correctly.
Here's a realistic example. Say a Shopify brand also selling on Amazon closes a 14-day settlement period with the following activity:
The journal entry that captures this looks like the one below. Note that Marketplace Facilitator tax flows in and right back out — Amazon collects and remits, so it's a wash for the seller, but tracking it through the books matters for state reporting (more on that below).
When the $43,655 deposit hits the operating bank account, it clears the Amazon Clearing debit to zero. The Amazon Reserve account now sits at $4,500 (or whatever the cumulative reserve balance is), and that should match the 'Account Level Reserve' figure visible in Seller Central.
Most teams don't write these entries by hand. Middleware does it. Ottit's stance on the middleware layer is documented in our guide on setting up A2X with QuickBooks, and the logic translates cleanly to Amazon settlements. The point is: whoever does the entry, the structure above is what correct looks like.
Actionable takeaway: pull the most recent Amazon settlement and compare the journal entry that posted to the structure above. If categories are missing or rolled into 'Amazon Fees,' the books need a remap before the next close.
How Should Settlement Period Cutoffs Be Handled at Month-End?
Settlement periods rarely end on the last day of the month, so a portion of each cross-month settlement belongs to the prior month. The standard practice is to accrue the in-period portion of an open settlement at month-end, posting an estimated entry to the right month and reversing it when the actual settlement closes. Without this, revenue and fees swing month to month based on calendar quirks.
Example: a settlement runs from April 22 to May 5. April 30 is month-end. Eight of the 14 days of activity belong to April, six belong to May. If the entire settlement posts to May when it closes, April revenue is understated by roughly the eight-day share, and May is overstated by the same.
The accrual approach pulls the date-stamped activity from Seller Central reports (the 'Date Range Reports' in Seller Central give daily-level data), splits it at month-end, and books the in-month portion via journal entry to the right accounts. The offset goes to an 'Accrued Amazon Activity' account on the balance sheet. When the full settlement closes, the accrual reverses and the actual settlement journal posts in full.
Tools like A2X handle this split automatically when configured for accrual-basis reporting, which is one reason multi-channel brands at scale move past simpler payout-only workflows. The mechanics are the same as the cutoff logic in our Xero Shopify reconciliation playbook — just applied to Amazon settlements instead of Shopify payouts.
For brands on cash-basis books, the cutoff matters less, since cash hits when it hits. But cash-basis Amazon accounting falls apart fast above $1-2M in revenue because inventory and fee timing distort the P&L too much. Most brands at that scale shift to accrual, and once they do, settlement cutoffs become non-negotiable. This is also a revenue recognition under ASC 606 requirement for any brand pursuing audit-ready books, per FASB ASC 606 (Revenue from Contracts with Customers).
Actionable takeaway: at the next month-end, identify any open Amazon settlement that crosses the cutoff date. If books are on accrual basis, an entry needs to allocate the in-month portion. If books are still cash basis above $2M revenue, it's worth a conversation with a CPA about transitioning.
How Do You Allocate COGS When the Same SKU Sells on Shopify and Amazon?
COGS is allocated per unit sold using the same costing method (FIFO, LIFO, or weighted average) across both channels — the cost layer doesn't change based on where the unit ships from. What changes is the channel tag on the COGS journal entry, so the income statement can show Shopify margin separately from Amazon margin. Inventory in transit between a 3PL and FBA stays on the balance sheet, not in COGS.
Here's the scenario we see most often. A brand sources 10,000 units of SKU A at $4.20 landed cost. They split inventory: 6,000 units go to ShipBob for Shopify fulfillment, 4,000 units transfer to FBA for Amazon. Both channels sell the SKU at $24.99. Over a month, Shopify sells 1,800 units and Amazon sells 1,200 units.
The COGS calculation per channel:
The journal entries are channel-tagged but the underlying cost is the same. This is straightforward — until inventory moves between channels mid-month. If 1,000 units transfer from ShipBob to FBA on the 15th, that's an inventory location change, not a COGS event. The entry is DR Inventory - FBA, CR Inventory - 3PL, both at $4.20 × 1,000 = $4,200. No P&L impact.
Where it gets messy: shipping costs to move inventory from a 3PL to FBA. Those costs are typically capitalized into inventory (added to landed cost) for the units transferred, which means the FBA cost layer for those units is now slightly higher than the original $4.20. Most brands under $10M skip the per-unit recalc and expense the transfer freight to a 'Inventory Transfer Costs' account, which is technically a small departure from strict GAAP but immaterial in practice. A CPA should weigh in on materiality thresholds for any brand approaching audit.
The deeper mechanics of FIFO vs weighted average for multi-location inventory are covered in our Shopify inventory accounting methods guide. The core point: pick a method, apply it consistently across both channels, and don't let channel be the costing variable.
Channel doesn't change cost. Channel changes attribution. Once a brand internalizes that, multi-channel COGS stops being scary.
Actionable takeaway: confirm that the costing method (FIFO, LIFO, or weighted average) is identical across Shopify and Amazon channels in whatever inventory system tracks landed cost. If Amazon COGS uses 'estimated' margin while Shopify uses actual landed cost, the P&L is comparing apples to grapefruit.
How Are FBA Reimbursements and Reserves Accounted For?
FBA reimbursements for lost or damaged inventory are generally recorded as a reduction of inventory loss or a recovery, not as new revenue. Reserve balances Amazon withholds are recorded as a current asset on the balance sheet, with movements between settlements posted as part of the settlement journal entry. Both line items hit the books regularly and need their own GL accounts to stay clean.
FBA reimbursements come in three flavors: (1) inventory lost in the warehouse, (2) inventory damaged by Amazon, (3) customer returns Amazon failed to return to inventory. Amazon reimburses at a rate they calculate based on recent sale price, often less than full retail, sometimes more than landed cost.
If a brand wrote off 50 units of inventory at $4.20 cost ($210 total) when Amazon flagged them lost, and Amazon later reimburses $410, the entries are:
The split between cost recovery and other income matters because the $210 portion is offsetting an earlier write-off (no net income effect over time), while the $200 above-cost portion is genuinely income. Lumping both into 'Other Income' inflates margin and obscures the inventory loss pattern.
Reserve balances are a balance sheet concept, not a P&L concept. When Amazon increases the reserve by $4,500 in a settlement period, that $4,500 still belongs to the seller — it's just held by Amazon. Booking it as 'Amazon Reserve' (asset) keeps the balance sheet honest. When the reserve releases in a future settlement, the entry reverses.
Many third-party tools exist to audit FBA reimbursements (GETIDA, Refully, Seller Investigators, others). Brands recovering meaningful dollars through these services should still book the recoveries with the same split logic — cost recovery vs. above-cost income — and net out the service's commission as an expense, not a reduction in the recovery.
Actionable takeaway: pull the last 90 days of FBA reimbursements from Seller Central. If they're all booked to a single 'Amazon Income' or 'Other Income' account, the brand is overstating margin and missing inventory shrinkage signals.
How Does Marketplace Facilitator Sales Tax Affect Amazon Bookkeeping?
Marketplace Facilitator (MTF) laws require Amazon to collect and remit sales tax on FBA orders in every US state with a sales tax. For accounting, this means the tax flows through the books as a pass-through: Amazon collects it, Amazon remits it, and the seller's books should show both sides of the transaction without recognizing tax revenue or tax expense. Sellers still need to report MTF sales on state returns where they have nexus.
The mistake we see most often: a brand books the gross-of-tax Amazon sales as revenue and never offsets the tax piece. Revenue is overstated by the sales tax amount, and there's no liability to reflect that the tax was collected. Even though Amazon handles remittance, the seller's books need to show the collect-and-remit cycle for accurate financial statements.
The cleaner approach: book the MTF tax as a credit to a 'Sales Tax Collected - MTF' liability and a debit to 'Sales Tax Payable - MTF' simultaneously when the settlement posts. Both sides clear within the same entry, leaving the books showing the gross sales tax collected (informational only) and zero net liability.
Where this matters most is at the state return level. Even though Amazon remits the tax, many states require the seller to report MTF gross sales as a deduction on the seller's state return. If the books don't track MTF sales separately from non-MTF sales, the state return preparer is reconstructing the number from Seller Central reports instead of from the GL — which slows down filings and creates audit risk.
Brands managing direct sales tax obligations on the Shopify side typically use a dedicated tool. For Shopify stores Ottit closes books for, Bookkeep is the partner we use for sales tax automation across the 100+ stores in our portfolio — it handles the data flow into the GL and the state-by-state reporting layer. Avalara is a common alternative for brands with complex multi-channel obligations. The deeper mechanics of multi-state Amazon nexus are covered in our Amazon FBA sales tax nexus guide and our broader Shopify sales tax automation guide.
Actionable takeaway: confirm the Amazon MTF sales tax appears as both a debit and a credit in the settlement journal, netting to zero P&L impact. If revenue is gross-of-tax with no offset, the books are overstating revenue and a CPA should review state filings.
What Reconciliation Cadence Keeps Multi-Channel Books Clean?
The industry standard for multi-channel sellers is reconciling Amazon every settlement period (every two weeks), with a full month-end close that ties settlement activity, reserves, FBA inventory counts, and bank deposits to source data in Seller Central. Quarterly close cycles consistently produce variances that are unrecoverable because the source data ages out and memory of edge cases fades.
The cadence we run for the 100+ Shopify brands Ottit closes books for, when Amazon is part of the mix:
- Every settlement (every 14 days): post the settlement journal entry, match the bank deposit to Amazon Clearing, confirm reserve balance ties to Seller Central, flag any variance over $50
- Weekly: review FBA reimbursements posted that week, confirm split between cost recovery and other income, reconcile any inventory adjustments
- Month-end (within 5 business days of close): accrue any open settlement crossing the cutoff, reconcile FBA inventory units to Seller Central reports, tie MTF tax pass-throughs to state filing data, review channel-level margin
- Quarter-end: full review of channel mix, cost layer accuracy across SKUs sold in both channels, reimbursement audit, sales tax filing reconciliation by state
- Year-end: inventory count reconciliation across FBA and 3PL locations, full COGS true-up, tax package preparation
Brands that try to compress this into a quarterly cycle almost always discover a settlement that didn't reconcile cleanly four months earlier — and by then, Seller Central reports have fewer details available, and the bank statements have been closed out. The fix becomes a plug entry, which is the wrong answer.
Tooling supports the cadence but doesn't replace it. A2X, QuickBooks, and Xero each handle pieces of the workflow, but the human review cadence is what catches the edge cases — the one-off promotional rebate Amazon ran, the reserve increase tied to a chargeback spike, the FBA inventory adjustment that didn't tie. Per the QuickBooks Online help center and the Xero Central help center, bank rule automation can speed the matching step but can't validate the underlying data.
For brands evaluating who runs this cadence — in-house, fractional, or full-service — the criteria framework is in our ecommerce bookkeeper vetting guide. The short version: ask candidates how they handle the settlement cutoff, the reserve account, and FBA reimbursement splits. If the answers are vague, they haven't done multi-channel at scale.
Bi-weekly Amazon reconciliation isn't a luxury — it's the only cadence that keeps the books from compounding errors into a quarter-end disaster.
Actionable takeaway: audit the last 60 days of Amazon settlements against the bank statement. Every settlement should match to a deposit within 1-3 days, and the cumulative reserve balance in books should match Seller Central. Any unmatched item over 30 days old is a sign the cadence isn't holding.
What Tools Actually Fit Into Amazon Seller Accounting?
The tooling stack for Amazon seller accounting has three layers: a general ledger (QuickBooks Online or Xero for most Shopify brands; NetSuite at higher scale), a settlement summarization tool that translates Amazon data into journal entries, and supporting tools for sales tax, spend management, and multi-channel attribution. The GL is the system of record; everything else feeds it.
| Layer | Common Tools | What It Handles |
|---|---|---|
| General Ledger | QuickBooks Online, Xero, NetSuite | Chart of accounts, financial statements, system of record |
| Settlement summarization | A2X, Bookkeep, Synder | Translating Amazon settlements into journal entries |
| Sales tax | Bookkeep, Avalara | Multi-channel sales tax tracking and filing |
| FBA reimbursement audit | GETIDA, Refully, Seller Investigators | Recovering missed reimbursements |
| Spend management | Ramp, Brex | Corporate cards, ad spend tracking, expense automation |
| Multi-channel attribution | Triple Whale, Northbeam | Channel-level marketing ROI |
| ERP (at scale) | Cin7, NetSuite | Inventory, multi-channel, multi-location |
The choice between QuickBooks Online and Xero is mostly preference for sub-$10M brands; both handle Amazon plus Shopify with the right middleware and chart of accounts. NetSuite enters the picture above $10-20M when SKU count and channel complexity overwhelm small-business GLs. We covered this comparison in detail in our QBO vs Xero vs NetSuite guide.
For settlement summarization, A2X is the most-used tool in the category — it's also a competitor to firms like Ottit on the bookkeeping services side, so we evaluate it on technical merit. It handles Amazon settlement translation cleanly and supports both QuickBooks and Xero. A2X is widely deployed; Bookkeep covers similar ground for revenue recognition workflows and is the partner we use across our 100+ Shopify portfolio for the revenue recognition layer specifically. The Synder Shopify integration guide outlines a third option that some brands prefer for transaction-level (rather than summary-level) syncing.
On spend management, Ramp is the corporate card and AP tool we deploy most often for Shopify brands managing Amazon ad spend across multiple cards. The auto-categorization and integration with QuickBooks/Xero saves meaningful time at month close. Brex covers similar ground for venture-backed brands.
On attribution, Triple Whale has become the default for DTC brands wanting channel-level marketing ROI across Shopify and Amazon. It's not an accounting tool — it doesn't replace the GL — but it gives the marketing-finance translation layer that a P&L alone can't show.
The tools matter less than the chart of accounts and reconciliation cadence. We've seen brands run clean books on basic QBO with disciplined process, and we've seen brands with the most expensive stack imaginable produce garbage P&Ls because no one mapped the settlement properly. Software amplifies the underlying setup; it doesn't fix it.
Actionable takeaway: list the current tool stack and trace one Amazon settlement from Seller Central → middleware → GL → bank reconciliation. If any handoff doesn't have a clear owner, that's the failure point — not the tool itself.
