Ecommerce Bookkeeping: The Seller's Guide to Books That Match Reality

Ecommerce bookkeeping differs from standard bookkeeping in three structural ways: marketplace and processor deposits are settlements (sales minus a dozen fee types), not revenue, and must be broken apart; inventory and COGS accounting determines whether your margins are real or fictional; and sales tax spans dozens of states through economic nexus rules. The standard 2026 solution is a settlement connector (A2X or similar) feeding QuickBooks Online or Xero, inventory tracked outside the accounting file, and a bookkeeper who has seen an Amazon settlement report before. Sellers who record deposits as sales are understating revenue by 15–30% and have no idea what their margins are.

That last sentence describes a remarkable share of DIY ecommerce books, so let's start there.

Why is "deposits as revenue" the cardinal sin of ecommerce books?

Because the deposit is what's left after the platform takes its cut. A single Amazon settlement nets together: gross product sales, shipping and gift-wrap income, referral fees, FBA fulfillment fees, storage fees, advertising spend, refunds, reimbursements, chargebacks, and sales tax the marketplace collected and remitted on your behalf. Shopify payouts do the same with processing fees, refunds, and disputes; every platform settles on its own calendar, and settlements routinely straddle month-end.

Record the deposit as "sales" and every one of those components collapses into one wrong number:

• Revenue is understated by the full fee stack — often 15–30% of gross
• Fees are invisible, so you can't see Amazon fee creep or rising ad costs eating your margin
• Refunds vanish instead of being tracked as a rate you manage
• Sales tax is misstated, since marketplace-collected tax isn't your money at all
• Month-end cutoff is wrong whenever a settlement period crosses months

Proper ecommerce bookkeeping decomposes each settlement into gross sales, each fee category, refunds, and tax — usually via an automation tool (next section) — and then reconciles the resulting entry to the actual bank deposit to the penny. That reconciliation is the integrity check: it proves the books capture everything the platform did.

How should you handle inventory and COGS?

Inventory is the second place ecommerce books quietly go wrong. The rule: money spent on inventory is not an expense when you buy it — it's an asset that becomes cost of goods sold when the unit sells. Expensing purchases on the buy date makes heavy-buying months look catastrophic and sell-through months look absurdly profitable; neither number is usable for decisions, and both misstate taxes.

What good practice looks like for a growing seller in 2026:

• Landed cost per SKU, not just supplier price: product cost + inbound freight + duties/tariffs + prep. For import-heavy brands, freight and duties can add 15–40% to unit cost; excluding them overstates margin by exactly that much.
• Purchases hit an Inventory asset account. A monthly COGS entry moves the cost of units sold to the P&L: beginning inventory + purchases − ending inventory = COGS.
• Inventory tracked outside the accounting file for multi-channel sellers — QuickBooks and Xero native inventory struggle beyond simple catalogs; tools like Finale, Cin7, or Katana own quantities and costs and feed summary entries to the books.
• Periodic counts and write-downs. Shrinkage, unsellable returns, and dead stock are real costs; books that never adjust for them drift optimistic.

Accrual-style books (at least for revenue and inventory) are effectively mandatory once inventory matters. Cash-basis ecommerce books produce gyrating fake margins — and lenders and aggregators evaluating your business will rebuild accrual numbers anyway.

What do you actually need to know about sales tax nexus?

Since South Dakota v. Wayfair (2018), physical presence isn't required: economic nexus means enough sales into a state (commonly $100,000/year; a few states use different thresholds or transaction counts) obligates you to register, collect, and remit there. The 2026 landscape in brief:

• Marketplace facilitator laws mean Amazon, Walmart, Etsy, and other marketplaces collect and remit for marketplace sales in essentially every sales-tax state. If you sell only via marketplaces, your remittance burden is small — though registration obligations can still exist, and the books must record marketplace-collected tax correctly (not as your revenue or your liability).
• Direct channels (Shopify, your own site) are on you. You track where you've crossed thresholds, register, collect, and file.
• Physical nexus still counts — and FBA inventory placed in Amazon's warehouses across many states remains a genuinely gray area that deserves professional judgment rather than a blog answer.

The practical stack: a nexus-tracking/filing tool (TaxJar, Avalara, or Shopify Tax on the collection side) plus a professional review annually or whenever you cross into new states. Getting registrations and back-liability cleanup right is tax specialist work; the bookkeeper's job is making sure collected tax sits in a liability account and filings reconcile to it.

What does the 2026 ecommerce accounting stack look like?

The consensus stack is boring and effective:

Two notes on the connector layer, since it's the ecommerce-specific piece. First, A2X-style tools post summarized journal entries per settlement, not thousands of individual orders — order-level detail belongs in your platform analytics, not your general ledger, and sellers who sync every order into QuickBooks end up with unusable, bloated files. Second, the connector is only as good as its account mapping; a specialist sets up fee categories once, correctly, and your P&L shows referral fees, fulfillment, storage, and advertising as separate visible lines forever after.

When should you hire an ecommerce specialist bookkeeper — and what does it cost?

DIY with a well-configured stack is defensible under roughly $250k/year in revenue on one or two channels. Past that, the combination of settlement accounting, inventory, and multi-state tax outgrows evenings-and-weekends attention fast.

Typical 2026 pricing for outsourced ecommerce bookkeeping: $400–$800/month for a single-channel seller in the low hundreds of thousands in revenue, and $800–$1,800/month for multi-channel brands with inventory complexity — above generic bookkeeping rates because settlement and COGS work is genuinely more involved. Books that are months behind need a catch-up engagement first; ecommerce backlogs price at the higher end because every month contains settlements to decompose.

Questions that quickly separate specialists from generalists:

1. How do you record marketplace settlements? (Anything resembling "from the bank deposits" ends the interview.)
2. Which connector do you use, and who owns the mapping?
3. How do you compute COGS — and does your process include inbound freight and duties in landed cost?
4. How do you handle month-end cutoff for settlements that straddle months?
5. How do you track my nexus exposure as I grow?
6. Can your monthly P&L show me margin by channel?

A generalist can learn all this — on your invoice, slowly. Specialists have the chart of accounts, the connector mappings, and the platform quirks already templated. Browse ecommerce bookkeeping specialists to compare firms that work with sellers like you, or start from the broader ecommerce industry hub if you also need tax or CFO-level help — inventory financing and channel-mix decisions at the $3M+ scale are classic fractional CFO territory.

FAQ

How much does ecommerce bookkeeping cost?

Typically $400–$800/month for a smaller single-channel seller and $800–$1,800/month for multi-channel brands with inventory, as of 2026. Settlement decomposition, COGS work, and channel reconciliation make it more expensive than generic small-business bookkeeping — and more valuable, since the alternative is not knowing your margins.

Do I need accrual accounting for my ecommerce business?

Once inventory is significant, effectively yes — at minimum accrual treatment of inventory and revenue. Cash-basis books make margins swing with purchasing timing instead of sell-through, which makes every downstream decision (pricing, ads, reorders) worse. Your tax return basis can differ; your accountant bridges the two.

Does QuickBooks alone work for Amazon sellers?

QuickBooks Online is a fine ledger, but on its own it will tempt you into deposits-as-revenue bookkeeping. Pair it with a settlement connector like A2X so each payout is decomposed into gross sales, fees, refunds, and tax, then reconciled to the deposit. That combination is the standard for a reason.

Do I owe sales tax in every state I ship to?

No — only where you have nexus (economic thresholds, commonly $100k/year in sales into the state, or physical presence). And for marketplace sales, facilitator laws make Amazon and similar platforms collect and remit for you in essentially all sales-tax states. Direct-site sales are your responsibility to track, register, and remit as you cross thresholds.