What 300+ bookkeeping projects taught me about ecommerce accounting

When I took my first bookkeeping gig on Fiverr in 2015, I had no idea I'd end up with over 300 projects under my belt across four marketplaces. What surprised me most isn't how different ecommerce businesses are from each other — it's how similar their problems are. The same five or six issues come up over and over, regardless of whether the client sells supplements on Amazon US, jewellery on Etsy UK, or runs an AirBnB portfolio in Lisbon. This piece is about those patterns — what generalist accountants get wrong about ecommerce, and what every cloud bookkeeper learns the hard way.

The pattern that comes up first, every single time

When a new ecommerce client books a clean-up project with us, there's a script we could almost write in advance. Within the first hour of looking at their books, we find one of these:

• Gross sales recorded as net (after marketplace fees), making revenue look smaller than it is
• Net sales recorded as gross (with fees ignored entirely), making revenue look larger and gross margin look healthier than reality
• Bundled marketplace payouts reconciled as a single revenue line, with no breakdown of fees, refunds, or chargebacks
• Inventory expensed at the point of purchase rather than at the point of sale, making the P&L wildly volatile

I'd estimate over 80% of the new ecommerce clients we onboard have at least one of these issues. Not because they're careless — most are smart, hard-working business owners. The reason is that ecommerce accounting is genuinely different from the standard "high-street business" model that most accountants are trained on, and the difference isn't obvious until you've worked through it.

Why generalist accountants struggle with ecommerce

The traditional accountant's mental model is: cash comes in via bank transfers, cash goes out for purchases and payroll, the difference is profit. This works perfectly for a plumber or a consultancy.

It breaks immediately for an ecommerce seller. The cash that hits your bank from Amazon isn't your revenue — it's a net payout after Amazon has already deducted referral fees, FBA fees, refunds, and storage costs. The cash that hits from Stripe isn't your revenue either — it's net of processing fees and any refunds. The accountant who treats either as "revenue" gets numbers that look reasonable but tell the wrong story.

To do ecommerce bookkeeping properly, the accountant has to understand the platform's settlement structure — what's in the payout, what isn't, what's a fee versus what's a refund versus what's a reserve. Most accountants don't, because their other clients don't sell on Amazon or Stripe, and they've never had reason to learn.

The five lessons that come up the most

1. Inventory is where most P&Ls go wrong

If you treat inventory as an immediate expense when you buy it, your P&L will look terrible the month you place a big purchase order, and amazing the months when you don't. Neither matches what's actually happening in the business. Inventory is an asset until it's sold. Cost of goods sold is what you spent on the items that actually went out the door this month.

This sounds obvious to a trained accountant. To a self-taught seller running their own books, it's the single most common mistake we see. And it's not just an accounting nicety — it's the difference between thinking you lost money in March (you didn't, you just bought stock) and thinking you crushed it in April (you didn't, you just sold last month's stock).

2. The cleaning-up cost is always more than the doing-it-right cost

One client came to us with three years of Etsy bookkeeping done as monthly summary entries from a notebook. Cleaning it up took weeks, several thousand pounds, and an HMRC voluntary disclosure. Setting it up properly from year one would have cost a few hundred pounds total.

This pattern repeats. We've never met a seller who said "I wish I'd waited longer to set up proper bookkeeping." We've met dozens who said "I wish I'd done this two years ago." If you're running an ecommerce business and your books aren't clean, the question isn't whether you'll fix them — it's whether you'll pay for that fix on your terms now, or HMRC's terms later.

3. Multi-currency is harder than people expect

Sellers expanding internationally tend to underestimate currency complexity. They open a US Amazon store, see USD landing in Wise, transfer it to GBP in batches, and assume their bookkeeping software will figure out the rest. It won't.

Each currency needs its own clearing account, gains and losses on conversion need their own categorisation, and the accounting platform needs multi-currency enabled (which is a different subscription tier in Xero — Standard, not Starter). Get this right and you'll have clean reporting forever. Get it wrong and untangling it after the fact is genuinely painful.

4. Real-time bookkeeping isn't a luxury — it's a decision-making tool

The accountants who do annual books for ecommerce sellers are giving them a history lesson. By the time the year-end accounts arrive, the seller has already made eleven months of decisions blind. The whole point of cloud accounting is to give you reliable monthly data while it's still useful.

Sellers who reconcile monthly know whether their gross margin is improving or deteriorating. They know when ad spend is creeping above ROI thresholds. They catch fraud-pattern chargebacks when there are five, not when there are fifty. Annual books on ecommerce isn't really bookkeeping — it's tax preparation.

5. The right software stack changes everything

The single biggest lever for ecommerce bookkeeping quality isn't the bookkeeper — it's the software stack. A competent bookkeeper with Xero plus A2X plus a multi-currency bank produces dramatically better books than a brilliant one with QuickBooks Self-Employed and a spreadsheet.

This is genuine good news for ecommerce sellers. The infrastructure exists, the integrations work, and the cost is modest (£50-£100/month all-in). The hard part is choosing the right combination for your specific situation and setting it up properly the first time.

What I'd tell my 2015 self

If I could go back and give 2015-me one piece of advice about specialising in ecommerce bookkeeping, it would be: spend more time on the setup, less time on the monthly grind.

The clients whose books are easiest to maintain ten years later are the ones where we spent extra hours up front getting the chart of accounts exactly right, mapping the integrations carefully, and documenting the workflow. The ones whose books are still painful are the ones where we rushed setup to start invoicing faster.

This applies to ecommerce sellers too. If you're starting an Amazon, Shopify, Etsy, or AirBnB business in 2026 and you're going to take it seriously, spend a day getting the books set up properly before you've made a single sale. It's the highest-ROI day of bookkeeping work you'll ever do.

Why I keep doing this work

Ecommerce bookkeeping isn't glamorous. Nobody throws parties when the gross margin reconciliation matches to the penny. But there's a particular satisfaction in opening a new client's books, finding three years of accumulated mess, and slowly turning it into something the owner can actually trust to make decisions with.

Most of our long-term clients started as clean-up projects. Once a business owner sees what good bookkeeping actually looks like — clean monthly reports, accurate gross margins, defensible tax positions — they don't go back. The clean-up turns into an ongoing relationship, and the relationship usually outlasts the platform they're selling on.

Three hundred projects in, that's still the most rewarding part of the work. Not the technical wins. The moment a client looks at a properly-prepared P&L for the first time and says "oh — that's what's actually happening?"

Md. Nasir Uddin FCA

Fellow Chartered Accountant. First & only Silver Xero Champion from Bangladesh. Ex-KPMG. 300+ small business bookkeeping projects since 2015 across Amazon, Shopify, Etsy, AirBnB, and beyond.

More about us →