Amazon Settlement Report: how to read each column vs the bank deposit

An Amazon settlement report is the only document that explains why a bank deposit is smaller than the sales you think Amazon owes you. Sellers often download the file, stare at thousands of rows, and still cannot see why total paid does not equal product sales. The report is not broken. It is a full ledger of one settlement period: every sale, refund, fee, advertising charge, reimbursement and reserve movement that Amazon netted into a single disbursement. Reading it by column and summarizing by type is how you turn that noise into a deposit that matches QuickBooks.

Amazon settlement report: get the right file first

Before you interpret columns, confirm you downloaded the settlement that belongs to the bank line you are trying to explain.

In Seller Central, open Payments and use either the statements view or Reports → Payments → Date Range Reports. Pick the settlement period whose total paid equals the Amazon deposit in your bank feed to the cent. Write down three anchors:

• Settlement ID (also called settlement group ID on some exports)
• Settlement start and end dates
• Total paid (the disbursement amount)

If total paid does not equal the bank deposit, stop interpreting columns. Check whether the deposit belongs to a different settlement, whether Amazon split one settlement across two transfers, whether a transfer failed and was re-sent, or whether the money landed in another bank account. No amount of column analysis will reconcile a deposit that belongs to a different settlement ID.

Prefer the detailed settlement transaction report for that settlement, not an order report, sales dashboard export or Business Report. Order-level sales reports show what customers paid Amazon for products. The settlement report shows what Amazon paid you after every adjustment. Those are different questions.

Major columns and what they mean vs the deposit

Settlement exports vary slightly by marketplace and report version, but most detailed files share a core set of columns. Knowing what each one is for saves hours of confused filtering.

Identity and timing columns

• settlement-id / settlement-group-id: the period this row belongs to. Filter to one ID when you reconcile one deposit.
• transaction-date / posted-date: when Amazon recorded the event. Refunds and fees can post on dates inside the settlement window even when the original order was earlier.
• order-id / order-item-code: useful for tracing a single refund or fee back to an order; not required for a summarized bookkeeping entry.
• SKU / description: product context. Helpful for exception hunting, not for the deposit total.

Money columns

• amount-type / amount-description (or equivalent): the bucket label Amazon uses for the line — Principal, Product charges, Commission, FBAPerUnitFulfillmentFee, Refund commission, Shipping, ShippingChargeback, LightningDealFee, Cost of Advertising, FBA Inventory Fee, Subscription Fee, WarehouseDamage, CompensatedClawback, Reserve, and many others. This is the column you group on when summarizing.
• amount / currency: the signed money movement. Sales and credits are typically positive; fees, refunds of principal and holds are typically negative (sign conventions can flip by report layout — always check that your pivot's grand total equals total paid).
• quantity: units for fulfillment or storage lines. Useful for fee reasonableness checks; it does not change how you match the deposit.

Metadata that trips people up

• transaction-type: Order, Refund, ServiceFee, FBA Inventory Fee, Adjustment, Transfer, and similar. Combining transaction-type with amount-description gives cleaner pivots than either column alone.
• marketplace-name: critical if you sell in more than one Amazon country. Reconcile each marketplace to its own deposit and currency.
• fulfillment-id: AFN vs MFN can explain why FBA fee lines appear for some orders and not others.

The deposit only cares about one number: the sum of every amount line in the settlement (excluding the Transfer/disbursement row itself, depending on export). That sum is total paid. Every other column exists to explain how you got there.

Summarize the Amazon settlement report by type

Raw rows are for forensics. Bookkeeping needs a short list of buckets whose net equals total paid. Build a pivot table (or equivalent group-by) on transaction type and amount description, then roll descriptions into a handful of accounts you actually use in QuickBooks:

1. Product sales and related income — Principal / Product charges, shipping credits, gift wrap credits and similar positive product lines.
2. Refunds — refunded principal and related refund shipping or gift wrap (often split across several amount-descriptions).
3. Selling fees — referral / commission, refund commission (which may partially offset), selling fees on orders.
4. FBA and logistics fees — per-unit fulfillment, weight-based, removal, return processing, inbound transportation and related.
5. Advertising — Cost of Advertising and sponsored ads charged against the balance.
6. Other fees and subscriptions — monthly subscription, storage, long-term storage, Lightning deals, coupon fees, service fees.
7. Reimbursements and adjustments — lost or damaged inventory credits, clawbacks, other positive adjustments.
8. Reserve movements — amounts held or released that change cash without being a sale or fee.

When every bucket is totaled, the sum must equal total paid. If it does not, your filter dropped rows, double-counted the Transfer line, or mixed two settlement IDs. Fix the filter before you post anything to QuickBooks.

An illustrative example with round numbers:

In this example, product sales alone were 10,000 and the bank received 4,700. The settlement report is where the 5,300 difference lives — not in a mystery "Amazon shorted me" line.

Catch missing reimbursements and fee lines

Sellers who only scrape one column often understate either income adjustments or expenses. A common pattern on seller forums is recognizing that the same dollars can land in different P&L places depending on which description you map. As one Amazon seller put it: "Same dollars, different place in the P&L if your sheet only scrapes one column."

Watch for these easy-to-miss groups when you build the summary:

• Inventory reimbursements under Adjustment or WarehouseDamage-style descriptions. They increase total paid and often sit outside Order/Refund transaction types. Filtering only to Order rows will hide them.
• Refund commission lines that partially return referral fees when a customer refunds. Omitting them overstates fee expense for the period.
• Advertising charged to the settlement versus advertising paid by credit card. If ads hit the settlement, they must appear in your summary or total paid will not reconcile.
• Storage and long-term storage billed on a different cadence than fulfillment fees. They still belong in the settlement that shows them.
• Reserve holds and releases. A hold reduces total paid now; a later release increases a future settlement. Track them in a balance-sheet reserve or receivable account so they do not vanish into "miscellaneous."
• Peak or surcharge fee descriptions that Amazon adds under new amount-description labels. If your mapping sheet only knows last year's labels, new fee names fall into an "other" bucket — or worse, get filtered out.

A practical check: sort the pivot by absolute amount and scan the largest lines. Then scan transaction types you rarely use. Anything non-zero that you cannot name should be researched in Seller Central help or the transaction detail before you force it into "other fees."

From report summary to QuickBooks posting

Once the buckets add to total paid, you are ready to post. Do not leave the work at the spreadsheet. The operational handoff is:

1. Confirm total paid equals the bank deposit for that settlement ID.
2. Confirm no duplicate sales already exist in QuickBooks for the same orders (apps that push per-order sales receipts will double-count if you also post a settlement summary).
3. Post one balanced journal entry through a marketplace clearing account: debit clearing for total paid, credit sales, debit refunds and fees and reserves as needed.
4. Match the bank deposit to clearing so clearing for that settlement returns to zero.

That posting sequence is covered in detail in the reconcile Amazon settlements in QuickBooks guide. Use this article for reading and summarizing the report; use that guide for the journal and bank match.

In journal-entry form, the example summary looks like this:

Account names are examples; map each bucket to the chart of accounts your accountant has approved. Notice reimbursements are credited (they increase what you were paid) and still sit inside a balanced entry whose clearing debit equals the bank deposit.

When the report still does not explain the bank

If your buckets add to the report's total paid, but total paid still does not equal the bank line, the issue is outside the column logic:

• Timing. Settlements often disburse a few business days after the period ends. Match by settlement ID and amount, not by assuming same-day posting.
• Split disbursements. Rare, but one settlement can produce more than one bank transfer.
• Currency and multi-marketplace mix. Each marketplace settles separately. Do not blend EU and US rows into one pivot when the bank deposit is from one store only.
• Previously recorded fees. Credit-card or invoice-based Amazon fee entries already in QuickBooks are a books problem, not a report problem. Do not invent a second fee line from the settlement if the charge never hit the settlement balance.

Resist plugging an unexplained remainder into a catch-all expense. A true rounding difference of a cent or two is different from an unexplained tens or hundreds of dollars, which almost always means a missing amount-description in the pivot.

Keeping settlement reading repeatable

Reading one Amazon settlement report carefully teaches the pattern. The cost is repetition: download, filter to one settlement ID, pivot, map new fee labels, check reimbursements, confirm total paid, then post. Sellers with two Amazon settlements a month already do this twenty-four times a year before adding Walmart or eBay.

ClearPayout automates the middle of that loop. It reads each Amazon settlement, maps amount-description buckets to your QuickBooks accounts once, and posts one balanced entry per settlement with the settlement ID in the memo so the deposit matches clearing without a manual pivot each time. See what gets posted on the Amazon to QuickBooks integration page, and compare plans on the pricing page. If you are evaluating other settlement tools, the A2X comparison explains how the approaches differ.

Whether you automate or stay manual, the discipline is the same: treat the Amazon settlement report as the source of truth for the deposit, summarize every bucket until the math equals total paid, then post that summary — do not try to match the bank to raw sales alone.