A shop's money, in columns
A handmade shop is a real business run on the economics of a hobby, and the bank statement is where that tension shows. Marketplace deposits arrive net of a stack of fees. Materials go out constantly in small amounts. Postage, packaging, ads, the studio, the printer, the market stall — all of it in a personal-looking account that also buys the groceries.
And it is printed in a PDF no spreadsheet can read. FlowParse converts the statement with AI — every transaction with its date, description, signed amount and running balance — and validates it against the balance the bank printed, so nothing is missing before you start.
What comes out is what a shop's books actually need: a categorised table of what the marketplace paid, what making and shipping the work cost, and what was genuinely left — from any bank, with no template to configure.
The deposit is not the sale
This is the fact that governs marketplace accounting. A customer pays one amount. Your shop receives a very different, smaller amount, days later, bundled with other orders, after the marketplace has deducted transaction fees, listing fees, payment processing, offsite ads, and any refund that happened to fall in the same window.
So the deposit is a net residue of many things, and totalling deposits and calling it revenue is wrong in every direction at once. It understates turnover — the figure tax thresholds are measured against — and it silently deletes your fee costs, which are deductible and are usually the largest cost after materials.
Converting the statement gives you the banked side exactly: every deposit, dated, as a real number. Bridged against the marketplace's own payment report, gross sales become provable and the fee stack becomes an expense with a figure attached instead of a vague sense that the marketplace takes a lot.
Why a marketplace shop is not an own-store business
If you run your own store on your own domain, money comes from a payment processor whose deposit you can trace back to specific orders, you set your own shipping, and your fees are a processing percentage. That is a different job, covered on the ecommerce bank statement page.
A marketplace shop is structurally different. You do not have a processor relationship; you have a marketplace that pays you a residue. Fees come from several sources at once, ads can be charged whether or not you asked for them, and the marketplace may collect and remit sales tax on your behalf so that it never appears as income at all.
Same engine, different question. An own-store seller asks which orders a payout covers. A marketplace seller asks what was taken out before the payout arrived — and whether the shop is actually profitable once it is all counted.
What FlowParse pulls from a seller's statement
Extraction is by meaning rather than by template, so it works with whichever bank your shop deposits into: every transaction's date and value date, the full description including the marketplace's deposit reference, the signed amount, the running balance, and the counterparty where named.
References matter because that is what identifies a deposit as belonging to a particular payout period — the link back to the marketplace's report. Descriptions that wrap across lines are joined back into a single field rather than truncated.
Every amount exports as a typed number with the correct sign, so a year of deposits and a year of materials each total in one formula.
How to convert a seller bank statement
Upload the statements
Drop in one or many PDF statements from any bank — digital or scanned.
Let AI extract them
Every transaction is read with its date, description, signed amount and balance.
Check the balance
Opening balance plus transactions is checked against the closing balance the bank printed.
Export and categorise
Download clean Excel or CSV — or an accounting-ready file for QuickBooks or Xero.
The honest boundary: we do the bank side
FlowParse does not connect to Etsy, Folksy, Not On The High Street or any other marketplace. No integration, no API, no login to your shop.
What we convert is the statement your bank produced. The marketplace's payment account and monthly statement are the other half, and you download those from your shop dashboard. The value is in having both as clean data, because that comparison is the only way the fee stack becomes visible.
We also do not issue tax forms, do not file returns, and do not determine your sales-tax or VAT position. We produce accurate, validated, categorised data — the raw material, not the conclusion.
The fees you agreed to and the ones you did not
Marketplace fees are not one fee. There is a listing fee, a transaction fee, a payment processing fee, and — on Etsy specifically — an offsite ads fee charged on sales that came through the marketplace's own advertising, which sellers past a revenue threshold cannot opt out of.
That last one is the one that causes genuine anger, because it is a percentage of the whole order value, it can be charged on a sale you would have made anyway, and it is deducted before you ever see the money. Whatever you think of it, it is a real cost and it is deductible — and only if you know what it was.
Structured deposits are what make the total findable. Deposits against the marketplace's reported gross, month by month, and the difference is your all-in cost of selling there. Most sellers, seeing that number for the first time, reprice.
Materials, stock and the cost you never counted
Handmade sellers buy materials the way people buy groceries: often, in small amounts, from many suppliers, whenever something runs out. Silver wire, fabric, yarn, blanks, resin, glaze — dozens of small transactions a month that never feel like inventory purchasing.
Which is why cost of materials per item is the number handmade sellers are most likely to have wrong, and it is usually wrong in the direction of optimism. Items get priced from a remembered cost, and the remembered cost is out of date.
As a structured category, materials become an actual monthly total against actual sales in the same month. That ratio — materials as a share of banked revenue — is the fastest honest read on whether the pricing works, and it takes about a minute once the data is in columns.
Postage, packaging and the shipping you undercharged
Shipping is where handmade shops lose money most quietly. The customer pays a shipping amount set months ago; the courier charges what it charges today; and the packaging — the box, the tissue, the ribbon, the card, the mailer — is a cost nobody includes because it feels like presentation rather than cost of sale.
Free shipping makes it worse by hiding the cost inside the item price, where it can be forgotten entirely and then eroded by a carrier price rise nobody noticed.
Structured rows total postage and packaging per month against the orders they shipped. If postage is rising as a share of revenue, that is not a feeling — it is a line, and it is usually the cue to change the shipping settings rather than work harder.
Tax the marketplace collected for you
In many places the marketplace is legally the collector — it charges the buyer sales tax or VAT and remits it itself, meaning that money never reaches your account and is not your income. In others, or on other channels, the obligation is entirely yours.
Getting this backwards is expensive in both directions: remitting tax the marketplace already paid, or failing to remit tax it did not. And the correct answer depends on your country, your buyer's country and the channel — it genuinely changes.
FlowParse does not determine any of that. What it gives you is the bank truth: if tax money arrived or left, it is a row, dated and validated. That evidence is what your accountant needs to work out which regime applies — a question worth asking properly rather than assuming.
The market stall, the fair and the second marketplace
Handmade businesses rarely stay on one channel. There is a craft fair with a card reader and some cash, a wholesale order to a shop that pays on invoice, a second marketplace, maybe an own-store site — each with a different fee structure and a different way of reaching the bank.
That is a genuinely difficult comparison to make by eye, and it matters: the fair that felt wonderful may have cost more in stall fees, travel and a lost weekend than it returned, while the boring wholesale order may be the most profitable thing you did all year.
Converted and categorised, each channel becomes a column — revenue net of its own costs, side by side. Sellers regularly discover their least favourite channel is their best one.
When the shop banks into your own account
Almost every handmade shop starts in a personal account, because it started as a hobby that unexpectedly worked. So deposits, materials, the weekly shop and the rent all live in one statement.
Untangling that manually means reading every line of a year. In a structured export it is a filter-and-tag exercise done once, with the rules reused for the rest of the year.
It also handles the genuinely mixed cases honestly — the craft supplies bought on a personal trip, the phone that photographs the work and everything else. Your accountant can only apportion those if the transactions are visible.
What a sale actually earns you
The reason to do any of this is one number: what is left after everything. Selling price, minus the fee stack, minus materials, minus postage and packaging, minus the ads you paid for and the ones you were charged for.
Sellers who have never calculated it are frequently shocked — not because the shop is failing, but because the margin is thinner than assumed and the busiest products are often the worst ones. Volume feels like success and can be the opposite.
That calculation is impossible from a PDF and straightforward from a structured export. Once it exists, pricing becomes a decision rather than a guess, and the decision usually is not to work more hours.
Materials now, deposit later
A handmade shop funds itself in the wrong order. You buy materials, you make the thing, you list it, someone buys it, and the marketplace pays you days later. Cash goes out well before it comes back, and every growth spurt makes the gap wider.
That is why a shop can be busy, profitable and completely out of money at the same time — most acutely in the run-up to Christmas, when stock must be bought before the season that pays for it.
A structured monthly series shows the rhythm honestly: money out for materials, money in from deposits, and the balance between them. It is what turns a seasonal panic into something you can plan around.
A year of statements in one pass
Shop bookkeeping happens once a year, usually in a rush, and usually after a season that left no time for admin.
Batch processing takes up to 100 statements at once and merges them into a single sheet, with duplicate detection for overlapping periods and a source-file reference on every row, so the whole year arrives as one sortable dataset.
Sellers who finally opened a business account halfway through the year get both it and the old personal account consolidated into the same table.
The card the materials go on
Most of a shop's spending is on a card — materials, packaging, postage labels, the ads. The current account shows the deposits and the big things; the card shows the actual cost of making the work.
Card statements convert the same way and merge into the same dataset; see credit card statement conversion. Without them, cost of goods is a guess, and profit per order is a guess built on a guess.
Straight into QuickBooks or Xero
If you or your accountant run accounting software, the converted statement goes straight in. FlowParse produces real bank-feed files — QBO, QFX and OFX — with a transaction ID per row so a re-import does not double-post, and a Xero-ready CSV with the columns Xero expects.
That matters most for the accounts a live feed does not reach: the personal account the shop started in, an older bank, the year before you registered.
Scanned statements too
The early years of a shop — the ones that establish how long it has traded — often exist only as scans or phone photographs of posted statements.
OCR runs first on those, then the AI structures the recognised text and flags low-confidence figures for a quick check, so an old scan becomes the same clean rows as a fresh download.
Numbers you can price on
Around 98% field-level accuracy on standard layouts, with every low-confidence figure highlighted in an editable preview before anything exports.
And the statement proves itself: opening balance plus every transaction must equal the closing balance the bank printed, and FlowParse checks that on every statement. A missing materials purchase does not just cost a deduction — it makes every item priced from that data slightly wrong. That is why validation matters here: the numbers feed decisions, not just a return.
Your account, your data
A seller's statement is usually a personal document too — your shop and your household in one file. Uploads run over TLS, processing is EU-hosted, the original PDF is deleted immediately after processing, and documents are never used to train AI models.
Nothing is retained once your export is produced.
Who this is for
Handmade and vintage sellers filing their own return, sellers who suspect the fee stack is eating the margin, anyone selling across a marketplace and a fair and a wholesale order who wants them compared honestly, and the accountants who serve small makers and would rather not receive a shoebox.
If you cannot currently say what a typical order actually earns you after fees, materials and postage, converting the statements is the step that produces the answer.
Convert your shop's bank statements
Upload a statement and get clean, categorised rows — deposits bridged back to gross, fees and materials visible, profit per order finally computable.