Bookkeeping for E-commerce Sellers: Inventory, Fees & Sales Tax

Selling online looks like it should be the easiest kind of business to keep books for. The platform tracks every order, the money lands in your bank automatically, and there's a dashboard for everything. In practice, e-commerce is one of the messiest businesses to keep clean books for — because the number that hits your bank account is almost never your real revenue, your costs are buried inside a dozen fee lines, and your inventory and sales tax obligations sprawl across every state you ship to. Here's how e-commerce bookkeeping actually works, and how to keep numbers you can trust.

Why e-commerce books get messy so fast

A traditional service business has a simple money trail: send an invoice, get paid, categorize the deposit. An e-commerce seller has none of that simplicity. Between the buyer clicking "checkout" and the money reaching your bank, the sale passes through a checkout system, a payment processor, and a platform payout engine — each of which takes a cut, holds a reserve, or nets something out along the way.

Four things make e-commerce uniquely hard to reconcile:

• Payment processors sit in the middle. Stripe, PayPal, Shopify Payments, Amazon, and Square all collect the full sale, subtract their fees, hold back refunds and reserves, and deposit a rounded-off "net" amount days later. The deposit doesn't match any single order.
• You're usually multi-channel. One brand often sells on Shopify, Amazon, Etsy, eBay, Walmart, and wholesale — each with its own fee structure, payout schedule, and report format. The books have to add up across all of them without double-counting.
• Deposits are not revenue. This is the single biggest mistake we see. A $9,400 Shopify payout might represent $12,000 in gross sales, minus $1,100 in sales tax you're holding for the states, minus $360 in processing fees, minus $600 in refunds, minus a $540 reserve. Booking the $9,400 as "sales" understates revenue and hides every cost.
• Inventory ties up real cash that never shows on the P&L when you buy it. The $40,000 you spent on product this quarter isn't an expense yet — it's an asset sitting in a warehouse until it sells. Get this wrong and your profit is pure fiction.

None of this is exotic. It's just that the accounting has to undo what the platforms have bundled together. Do that well and e-commerce books are clean and decision-ready. Skip it and you have a QuickBooks file full of bank deposits labeled "Shopify" with no idea what's actually happening. If you want the industry-specific version of this playbook, our e-commerce & retail bookkeeping page covers how we set these files up.

The cardinal rule: a payout is not your revenue

Everything in e-commerce bookkeeping flows from one principle. Record sales at gross, then account for every deduction separately. The deposit that hits your bank is the net of a long list of activity, and each item on that list needs its own home in the books.

Take a single Shopify payout of $9,400. Reconstructed properly, it might break down like this:

• $12,000 — gross product sales (this is your revenue)
• $900 — shipping charged to customers (revenue, or a shipping-income account)
• +$1,100 — sales tax collected (a liability you owe the states — never revenue)
• −$360 — payment processing fees (an expense)
• −$600 — refunds issued (a contra-revenue reduction)
• −$540 — reserve held back by the processor (a receivable / clearing item, not a loss)
• −$3,100 — the tax remitted, netting, and timing that reconcile it to the $9,400 that actually landed

Notice that the $12,000 in real revenue is higher than the $9,400 deposit, and there are three or four expense and liability accounts that would be completely invisible if you just booked the deposit. That gap — between what sold and what landed — is where e-commerce profit hides. Recording only the net payout doesn't just understate revenue; it erases the fees, refunds, and tax liabilities that you need to see to run and price the business.

Platform fees and payment fees: where the margin leaks

E-commerce is a thin-margin game, and the fees are where thin margins quietly get thinner. They deserve their own accounts so you can actually see them — not lumped into a vague "bank charges" line.

The fees fall into a few buckets:

• Payment processing fees. Roughly 2.9% + 30¢ per transaction is the familiar benchmark for card processing, though your effective rate varies by processor, card type, and volume. On a 15%-margin product, a 3% processing fee is eating a fifth of your profit — worth tracking to the penny.
• Marketplace / referral fees. Amazon referral fees commonly run around 8–15% of the item price by category; Etsy charges listing plus transaction and processing fees; eBay takes final-value fees. These are cost of doing business on that channel and belong in their own expense account per channel.
• Fulfillment fees. If you use Fulfillment by Amazon (FBA) or a third-party logistics (3PL) provider, pick, pack, ship, and storage fees are real costs — often the difference between a channel being profitable or not.
• Subscription and app fees. Your Shopify plan, apps, email tools, and ad spend. Small individually, meaningful in aggregate.

The reason this matters beyond tidiness: fees are how you compare channels. When Amazon's referral plus FBA fees are itemized separately from Shopify's processing fees, you can finally answer the question every multi-channel seller should be asking — which channel actually makes me money after everyone takes their cut? That analysis is impossible if every fee is buried in the net deposit.

Inventory and COGS: the part DIY sellers get wrong

If there's one place e-commerce books go off the rails, it's inventory. Get this wrong and every profit number you look at is fiction.

Inventory is an asset, not an expense

When you buy $40,000 of product, you have not spent $40,000 in expenses. You've converted $40,000 of cash into $40,000 of inventory — an asset on your balance sheet. It only becomes an expense (Cost of Goods Sold, or COGS) when the product actually sells. This is the accrual-matching principle in action, which is exactly why serious inventory businesses can't run on pure cash-basis books; we walk through that tradeoff in our guide to cash vs. accrual accounting.

The DIY mistake is expensing inventory purchases the moment the money leaves the bank. That makes the month you restock look terrible and the months you sell down look artificially fat. Your P&L becomes a chart of when you happened to buy product, not whether you're making money. Done right, COGS lands in the same period as the revenue from selling that product — so gross margin means something.

Choosing a COGS method

There are three common ways to value inventory as it sells:

• FIFO (First In, First Out). Assumes the oldest stock sells first. The most common and intuitive method, and it usually mirrors how physical goods actually move.
• Weighted average cost. Blends all units into one rolling average cost. Simpler when you buy the same SKU at fluctuating prices across many purchase orders.
• Specific identification. Tracks the exact cost of each individual item. Practical only for low-volume, high-value goods (think furniture or jewelry), not for thousands of identical units.

Most small e-commerce sellers land on FIFO or weighted average. What matters more than the choice is picking one and applying it consistently — switching methods mid-stream distorts your margins and complicates your tax return.

What actually goes into unit cost (landed cost)

Your true cost per unit is more than the price on the supplier invoice. Landed cost includes the product price plus freight/shipping to you, import duties and tariffs, customs brokerage, and any inbound handling. A $6 item that costs $2 to ship in and $0.80 in duty has a landed cost of $8.80 — nearly 50% more than the sticker price. Sellers who price off the sticker cost instead of the landed cost routinely think they're profitable when they're barely breaking even. Building landed cost into your inventory valuation is one of the highest-leverage things you can do for pricing.

Refunds, returns, and chargebacks

Physical-goods sellers live with returns, and each type needs consistent handling:

• Refunds and returns reduce revenue (contra-revenue), not increase expenses. If the item comes back in sellable condition, it also flows back into inventory — the COGS from the original sale reverses.
• Chargebacks are disputed transactions the bank pulls back, often with a dispute fee on top. They reduce revenue and add a fee expense, and they frequently show up on a later payout than the original sale — which is exactly why the deposit never ties to a single order.
• Restocking and disposal. Damaged returns that can't be resold are inventory write-offs, not just lost revenue.

Handled loosely, refunds and chargebacks quietly overstate your revenue and understate your true return rate. Handled properly, your net sales reflect reality and you can actually measure how much returns cost you.

Multi-channel reconciliation

The moment you sell on more than one platform, reconciliation becomes the real work. Each channel has its own payout schedule (Amazon pays roughly every two weeks; Shopify on a rolling daily-to-weekly basis; Etsy on its own cadence), its own fee structure, and its own report layout. Your job is to make sure that:

• Every channel's gross sales, fees, refunds, and tax collected are recorded — once, and only once.
• Each bank deposit ties back to a specific payout report, so the money in your account is fully explained.
• Reserves and in-transit funds (money the platform is holding) are tracked as clearing accounts, not lost in the gaps.

A useful mental model is a clearing account per channel. Sales, fees, and refunds post into the clearing account from the platform's settlement report; the bank deposit posts out of it. When a channel's clearing account nets to zero after the payout clears, you know that channel is fully reconciled. When it doesn't, you've found a missing fee, an untracked reserve, or a timing difference — before it snowballs into a year-end mystery.

Sales tax: nexus and marketplace-facilitator rules

Sales tax is the part of e-commerce that keeps owners up at night, and for good reason — the rules changed dramatically after 2018 and now reach into states you've never set foot in. This is a summary; for the full walkthrough of where and when you owe, see our companion guide on sales tax nexus. (Tax rules change often — the figures below are current as of 2026; always confirm each state's current thresholds before you rely on them.)

Economic nexus after Wayfair

Before 2018, you only owed sales tax in states where you had a physical presence. The Supreme Court's South Dakota v. Wayfair decision changed that, letting states tax sellers based on economic activity alone. The result is "economic nexus": cross a state's sales threshold and you're obligated to register, collect, and remit there — even with zero physical footprint.

As of 2026, the landscape looks like this:

• Of the roughly 45 jurisdictions that enforce economic nexus, most set the trigger at $100,000 in annual in-state sales — the single most common threshold.
• A handful sit higher: California, Texas, and New York use $500,000; Alabama and Mississippi use $250,000.
• The old "200 transactions" secondary test is fading fast. More than a dozen states have dropped it in favor of a sales-dollar-only standard — Illinois eliminated it effective January 1, 2026, Utah in mid-2025, and Alaska in early 2025, with Kentucky removing it August 1, 2026. Roughly half of jurisdictions are now sales-only, while the rest still run a transaction count.
• A couple of states use "AND" tests: New York requires $500,000 and 100 transactions; Connecticut requires $100,000 and 200 transactions.

You also still have physical nexus anywhere you have an office, an employee, or inventory. That last one bites Amazon FBA sellers in particular: when Amazon stores your goods in a fulfillment center in another state, that stored inventory can create physical nexus there. Because thresholds and rules genuinely differ state by state, the safe posture is to track your sales by state and check each state's current rules rather than assume one number applies everywhere.

Marketplace-facilitator laws

Here's the piece that saves marketplace sellers real headaches: every state with a sales tax has now passed a marketplace-facilitator law. These laws shift the obligation to collect and remit sales tax onto the marketplace itself. So when you sell through Amazon, Etsy, eBay, or Walmart, the platform generally calculates, collects, and remits the sales tax on those transactions for you. You don't file for tax you never touched on those orders.

But — and this is where sellers get tripped up — marketplace-facilitator coverage does not make sales tax go away entirely:

• Your marketplace sales usually still count toward your own nexus. In most states, the gross revenue flowing through your Amazon or Etsy storefront counts toward your economic-nexus threshold — not just the platform's. Cross the line and you can be required to register even if the platform is doing the collecting.
• Direct sales are all yours. Sales through your own Shopify store or website are not marketplace sales. Once you have nexus in a state, you're responsible for collecting and remitting on those direct orders yourself.
• You may owe "zero" or informational returns. Once registered in a state, some states still expect a periodic return even when the marketplace remitted everything — reporting the marketplace-covered sales and showing $0 due. Miss those filings and you can rack up penalties for tax you never actually owed.

The practical upshot: a pure marketplace seller has a much lighter sales-tax burden than a Shopify-first seller, but "the platform handles it" is not a complete answer. You still need to monitor where you've crossed nexus, register where required, and keep your collected-tax liability clean in the books.

The tool stack that makes this manageable

The good news is you don't have to reconstruct payouts by hand. A mature e-commerce accounting stack has three layers:

• The ledger: QuickBooks Online or Xero. Your accounting system of record, where the P&L, balance sheet, and inventory live. Everything else feeds into it.
• The connector: A2X (and tools like it). This is the piece most DIY sellers are missing. A2X-style connectors pull each platform's settlement/payout data and post a clean summary journal entry into QuickBooks or Xero — splitting out gross sales, fees, refunds, shipping, and sales tax, and matching the entry to the exact bank deposit. It turns a $9,400 mystery deposit into a fully itemized, reconcilable entry automatically. If you sell on Amazon, Shopify, or Etsy at any volume, a connector like this is the difference between clean books and a nightmare.
• Sales-tax automation: Avalara, TaxJar, or similar. These track your sales by state against each state's nexus threshold, tell you where you've triggered an obligation, and can file returns. Worth adding as soon as you sell across many states.

Layer an inventory or operations tool on top if you need real-time stock counts across channels, and you have a system that scales. The key insight: the connector layer is what lets the ledger stay accurate without someone manually decoding every payout. Set up well, month-end becomes a review, not a reconstruction.

A monthly close routine for e-commerce

Pulling it together, a healthy monthly close for an online seller looks like this:

• Import every channel's settlement data (via your connector) and confirm gross sales, fees, and refunds posted correctly.
• Reconcile each bank deposit to a payout, so every dollar in the account is explained and each channel's clearing account nets to zero.
• Update inventory and record COGS for what actually sold, so gross margin is real.
• Reconcile the sales-tax-payable liability — what you collected vs. what's been remitted (by you or the marketplace) — and check whether you've crossed nexus anywhere new.
• Review the P&L by channel to see true profitability after all fees.

Do that every month and you're never more than 30 days from the truth — which is what makes pricing, ad-spend, and inventory decisions possible. If your books have drifted and you're staring at months of unreconciled Shopify deposits, that's a catch-up bookkeeping project before it's a monthly routine.

How Tides handles e-commerce books

E-commerce is one of the industries we specialize in, precisely because it's the kind of bookkeeping that punishes shortcuts. When we take on an online seller, we set up the connector layer so every payout posts cleanly, build a chart of accounts that separates fees, refunds, shipping, and sales-tax liability by channel, and get inventory and COGS onto the right basis so your margins are real. Every month, our monthly bookkeeping service reconciles each channel, updates COGS, and reviews your sales-tax exposure — and our financial reporting shows you profitability by channel so you can see which storefronts actually earn their keep.

We work with sellers on Shopify, Amazon, Etsy, eBay, and multi-channel combinations of all of them, all over the country — see the full range of businesses on our who we serve page. The goal is simple: books you can trust, a sales-tax position you can defend, and margin numbers that reflect what's really happening after every platform takes its cut.

The bottom line

E-commerce bookkeeping isn't hard because online selling is complicated — it's hard because the platforms bundle everything together and the accounting has to take it apart. Record sales at gross and account for every fee, refund, and tax liability separately. Treat inventory as the asset it is and let COGS land when product sells. Reconcile each channel's payouts to the penny. And stay on top of sales-tax nexus and marketplace-facilitator rules across every state you reach. Get those four things right — usually with a connector tool doing the heavy lifting — and your books stop being a source of anxiety and start being the thing that tells you exactly where your money is made.