Ecommerce Accounting for Shopify and Amazon Sellers

A service business gets paid and the money in the bank is close to the revenue number. An ecommerce seller almost never sees that. Amazon, Shopify Payments, and most payment processors pay out a net figure on a delay, after subtracting fees, refunds, chargebacks, and any sales tax the platform collected and remitted for you. The deposit that lands in your account on a Tuesday might represent two weeks of orders, minus a dozen different deduction categories, none of which are broken out unless you go find the settlement report behind it.

Book that deposit as one line called "Amazon income" and you've thrown away the information you actually need: how much you sold, how much you paid in marketplace fees, how much came back as refunds, and how much sales tax was never yours to keep in the first place. Do this for a full year and your P&L understates both revenue and expenses by the same amount, which happens to cancel out at the bottom line and hide the problem completely, until you need real numbers for a loan application, a valuation, or your own pricing decisions and discover the books don't actually show any of it.

The fix is settlement accounting: pulling apart each payout into its components before it hits the general ledger, so revenue, cost of goods sold, marketplace fees, refunds, and tax collected on your behalf each land in their own account, in the period the sale actually happened rather than the period the cash arrived.

Both tools sit between your sales channels (Amazon, Shopify, Etsy, Walmart Marketplace, and others) and your accounting software. They pull the settlement or payout report from each channel, break it into its underlying categories, gross sales, discounts, refunds, shipping charged to customers, marketplace or referral fees, FBA fulfillment fees, advertising spend deducted at source, and sales tax collected, and then post one summarized journal entry per settlement period into QuickBooks Online or Xero.

The practical difference from doing it manually: instead of importing hundreds or thousands of individual order lines, you get one clean entry per payout that ties exactly to the bank deposit, with every component visible and auditable back to the source report. Both tools also handle the accrual timing question. A sale can happen in one month and settle (get paid out) in the next; A2X and Link My Books can accrue the sale into the period it occurred, so your P&L reflects when the business actually made the sale rather than when the bank happened to receive the cash.

Which one fits depends mostly on channel mix and existing stack. A2X has broader multi-channel and multi-currency coverage and deeper inventory-costing features; Link My Books tends to be priced lower for a single-channel Amazon or Shopify seller and integrates directly with some expense-management tools. Neither replaces bookkeeping. They replace the manual, error-prone step of turning a settlement report into something your books can use, so the bookkeeping that follows is working from correct numbers instead of one lump deposit.

If you buy physical stock and hold it before selling it, the cost of that stock is not an expense the day you pay your supplier. It's an asset (inventory) until the unit actually sells, at which point its cost moves to cost of goods sold (COGS) on the P&L, matched to the same period as the revenue from that sale. Expense the whole purchase order the day you pay for it instead, and your P&L will show a loss in the month you restocked and an inflated profit in the months you sell through it, which is not what actually happened to the business.

Here's a worked example. A seller buys 1,000 units at a landed cost (unit cost plus freight and duty) of $8.00 each, for $8,000 total, added to inventory as an asset. That month, 300 of those units sell at $25 each, for $7,500 in revenue. COGS for the month is 300 units times $8.00, or $2,400, leaving a gross profit of $5,100 for the month, a 68% gross margin. The remaining 700 units, worth $5,600, stay on the balance sheet as inventory until they sell in a later period.

Landed cost matters more than most sellers assume. Freight, duty, and any inspection or prep fees are part of the unit's cost, not a separate expense line, because they were required to get that unit into sellable condition. Leaving them out understates COGS and overstates margin on every unit sold. Most ecommerce accounting tools and inventory-management add-ons (like Cin7, Skubana, or QuickBooks' built-in inventory feature) support weighted-average or FIFO costing; either is workable, but pick one and apply it consistently rather than switching methods between periods.

Since the Supreme Court's 2018 decision in South Dakota v. Wayfair, states can require you to collect and remit sales tax based on sales volume alone, with no physical presence in the state required. The original benchmark most states adopted was $100,000 in sales or 200 transactions into that state in a year; several states have since dropped the transaction-count leg and kept only the dollar threshold. These thresholds, and the look-back period used to measure them, differ by state and change over time, so check the current figure on the specific state's department of revenue site before assuming last year's number still applies.

Marketplace facilitator laws add a second layer that ecommerce sellers specifically need to understand. In most states, when you sell through Amazon, Etsy, Walmart Marketplace, or a similar platform, the platform itself is legally required to calculate, collect, and remit sales tax on that sale, not you. That's a real operational relief for marketplace-only sellers. It does not, however, extend to sales made through your own website (a Shopify store you run directly, for example): those are your own nexus and your own filing obligation, tracked separately from marketplace sales even though both might show up as revenue in the same bank account.

The practical implication for a multi-channel seller: track nexus and taxable sales by channel, not just in total. A seller doing $150,000 through Amazon and $40,000 through their own Shopify store has marketplace-collected tax handled by Amazon in most states, but still needs to track whether that $40,000 in direct sales, combined with total economic activity in a given state, has crossed that state's threshold and created a filing obligation on the direct-channel revenue.

A return or refund is not simply revenue disappearing. On Amazon and Shopify both, a refund typically comes with its own fee treatment: some platforms return the original transaction fee, some keep a portion of it, and returned inventory that comes back damaged or unsellable needs to be written off separately from inventory that goes back into stock in resellable condition. Book every refund as a straight reduction to the sales account and you lose the ability to see how much refund activity is actually costing you in fees and written-off stock, on top of the lost sale itself.

Chargebacks (a customer disputing a charge through their card issuer rather than requesting a refund through the platform) are a different category again, usually carrying their own fee even when the seller wins the dispute, and often settled weeks after the original sale. These need their own account too, separate from ordinary refunds, so a pattern of chargebacks on a specific product or payment method is visible rather than buried inside a general "returns" line.

The settlement-accounting tools mentioned earlier generally break refunds and chargebacks out automatically as part of the payout breakdown, which is one of the clearer arguments for using one rather than posting settlement deposits by hand. Without that separation, a business can look profitable on paper while a return rate creeping up on one product line quietly erodes margin somewhere the P&L never shows it directly.

Every channel deducts a different mix of fees before you ever see the cash, and lumping them together hides which channel is actually more expensive to sell through. A reasonable chart of accounts for a multi-channel seller separates at least: marketplace referral or commission fees (Amazon's category-based referral fee, Etsy's transaction fee), fulfillment fees (Amazon FBA per-unit and storage fees, or your own 3PL's pick-and-pack charges), payment processing fees (Shopify Payments, Stripe, PayPal), advertising spend deducted at source (Amazon PPC billed against seller balance rather than a separate invoice), and any subscription or platform fees charged flat regardless of sales volume.

Why this matters beyond bookkeeping neatness: a seller comparing an Amazon listing against the same product sold direct on Shopify needs fee data broken out by channel to know which one actually nets more per unit after fees, not just which one has higher gross sales. A product that looks like the better seller on Amazon by revenue can lose that comparison once referral fees, FBA fulfillment, and ad spend deducted at source are all counted against it.

Most of this breakout happens automatically through A2X or Link My Books if the mapping is set up correctly when the integration is first configured; the manual alternative is opening every settlement report and re-categorizing each fee line by hand, which is exactly the kind of repetitive work that tends to get skipped once volume grows past a handful of orders a day.

Most ecommerce sellers run more than one channel within a year or two of starting: Amazon plus Shopify is the common pairing, sometimes with Etsy, Walmart Marketplace, or a wholesale channel added on top. Each one pays out on a different schedule and reports settlement data in a different format, which is the actual reason reconciliation gets harder as channels are added, not the volume of orders itself.

Payout timing differs meaningfully by channel, written here as comparison rows instead of a table:

• Amazon: settlement periods roughly every two weeks, with a lag between sale and deposit that can run one to three weeks depending on account standing and reserve holds
• Shopify Payments: payouts typically land two to three business days after the sale, faster than Amazon but still not same-day
• Stripe and PayPal (used for a direct Shopify or standalone checkout): payouts usually settle in one to two business days, with rolling reserves possible on newer or higher-risk accounts
• Etsy: payouts on a set schedule (often daily or weekly depending on account settings), with fees deducted from the same payout

Because each channel's payout timing and reporting format differ, reconciling by simply matching bank deposits to a single "sales" total per channel per month tends to drift out of balance quietly, usually noticed only when the bank balance and the books diverge by an amount nobody can immediately explain. Reconciling each channel's payout against its own settlement report, ideally through an integration that automates the breakdown, and then reconciling the resulting bank deposits against the books, catches discrepancies while the settlement report is still available to check against, rather than months later.

Form 1099-K is issued by payment settlement entities, including Amazon, Shopify Payments, Stripe, and PayPal, reporting gross payment volume processed through the platform. It is not something you file for yourself; the platform sends it to you (and to the IRS) based on your account's activity. The 1099-K reporting threshold has moved more than once in recent years, so check the current figure directly at irs.gov before relying on any number quoted elsewhere, including here.

The number on a 1099-K is gross payment volume, not net income and not taxable income by itself. It will not match your P&L's revenue line, because it includes amounts that later got refunded, sales tax collected and remitted on your behalf, and shipping charged to customers, none of which are your income even though they passed through the platform. Reconciling your 1099-K totals against your books each year, rather than assuming they should already match, avoids a mismatch that can otherwise draw an unnecessary IRS notice.

Beyond 1099-K, most ecommerce sellers with employees or contractors still owe the standard set of filings covered elsewhere (1099-NEC for contractors, payroll filings for employees), plus, if operating as an LLC or corporation with foreign ownership or foreign warehousing arrangements, additional forms that a credentialed signer should review case by case. None of this is filed by a bookkeeping team directly; preparation is handled here and the actual filing goes through a credentialed signer or your registered filing partner.

A chart of accounts built for a service business, a handful of revenue lines and generic expense categories, breaks down fast for an ecommerce seller with inventory and multiple channels. A workable structure separates revenue by channel (Amazon sales, Shopify sales, Etsy sales, wholesale) so channel-level performance is visible without pulling a separate report every time, and separates cost of goods sold by its actual components (product cost, freight-in and duty, fulfillment fees) rather than lumping everything into one COGS line.

On the expense side, marketplace and payment fees deserve their own category group entirely separate from "cost of goods sold," since they scale with sales volume the same way COGS does but represent a different kind of cost (getting paid and getting found, not making the product). Inventory itself sits on the balance sheet as a current asset, tracked either in aggregate or by SKU depending on volume and whether a dedicated inventory tool is in use alongside the accounting software.

Getting this structure right before transaction volume grows saves a rebuild later. A seller who starts with three generic income and expense categories and later wants to see margin by channel and by product usually can't reconstruct that history accurately; the old transactions were never coded with enough detail to split apart after the fact.

The mistakes that show up most often in ecommerce books trace back to a short list: booking the net payout as gross revenue and losing the fee and refund detail behind it; expensing inventory purchases immediately instead of carrying them as an asset until sold; treating sales tax collected on your behalf as your own income instead of a liability you're holding for the state; missing that marketplace-collected tax and direct-website tax are tracked separately; and letting settlement reconciliation lapse for a few months, which turns a ten-minute monthly check into a multi-day cleanup project once it's caught up on.

A business running a single channel with low order volume and no inventory (a dropshipping or print-on-demand model, for instance) can often manage on straightforward cash-basis bookkeeping without a dedicated settlement tool. Once inventory, multiple channels, or meaningful sales tax exposure enter the picture, the case for A2X or Link My Books integration, paired with inventory accounting done properly, gets considerably stronger, because the manual alternative scales badly with order volume in a way service-business bookkeeping does not.

Signs it's worth bringing in outside help specifically for ecommerce accounting: settlement reports are more than a month behind being reconciled, nobody can currently say what gross margin looks like by channel or by product, a state sales tax notice has arrived and nobody is sure whether nexus existed before it did, or inventory on the books hasn't matched a physical count in longer than anyone wants to admit. Our team sets up the channel integrations, structures the chart of accounts for multi-channel and inventory accounting, and reconciles settlements monthly, so the numbers behind pricing and growth decisions are the real ones.