A critical analysis of seller circumstances, marketplace and payment data, reconciliation, UK tax and accounting expectations, incumbent tools and the evidence still needed before significant investment.
Official guidance and platform documentation converge on the same structural issue: several economically different events are compressed into one bank deposit, while key costs and evidence live elsewhere.
The money that arrives in your bank from a marketplace is not your sales figure, and it is not your profit. The platform takes its fees out before it pays you, so the deposit is already a net number.
The platform also cannot know three things that matter: what the item cost you, what you spent away from the platform, and whether the thing you sold was your own old coat or stock you bought to sell. Those facts only exist with you.
Everything else in this report follows from that one gap.
Example One sale can honestly be described by three different numbers: £125 from the customer, £85 into your bank, £51 of profit.
Marketplace interfaces optimise listing, conversion, payment and fulfilment. They do not necessarily produce one accounting-ready record that combines the economic event, cash settlement, inventory cost, external evidence and the seller’s legal circumstances.
Selling takes minutes. Being able to prove what happened takes records, and the records live in different places.
The platform holds the order. The bank holds the cash. The receipt for what you paid is in your email, your pocket, or nowhere. Nobody joins them up for you.
Use the sliders below to see it happen. Change any number and watch the bank figure and the profit figure move apart.
eBay separates gross amounts, expenses, refunds and earnings; Etsy distinguishes sales, payment-account activity and deposits; Shopify states that its payout reconciliation report is not a revenue statement; TikTok Shop distinguishes order details, statements, reports, payments and reserves.[20][22][31][34]
A visible balance or bank credit feels definitive because it is concrete and spendable. Yet it usually represents a settlement calculation: sales and possibly buyer-paid delivery, less fees, refunds, labels, advertising, taxes, holds or other adjustments. It can reconcile cash without explaining business performance.
HMRC’s self-employed record guidance includes receipts for stock and goods, bank statements, sales invoices and other proof. A marketplace normally cannot know the original cost of a charity-shop purchase, an external courier receipt, a home-use apportionment or whether an item was originally bought for personal use.[5][50]
E-commerce year-end checklists commonly ask for channel sales reports, payment-processor statements, bank statements, stock records, expense evidence and explanations of unusual adjustments. Where these are missing or do not reconcile, the adviser must query the seller or qualify assumptions.[46][47]
| Issue | Legal or regulatory significance | Accountant request / good bookkeeping | Management information |
|---|---|---|---|
| Gross sales versus net payouts |
Legally significant Tax and accounts are not normally prepared by treating a net settlement as turnover. The applicable basis, VAT treatment and legal form determine the entries.[5][7] |
Order or transaction reports, settlement statements and bank credits are commonly reconciled so omissions and timing differences are visible. | Channel revenue, take rate and cash conversion are useful, but no law requires a seller to monitor those exact KPIs. |
| Fees, refunds, postage, advertising, discounts and adjustments |
Often legally significant They can change income, expenses, VAT and period allocation; the correct treatment depends on the facts. |
Advisers commonly ask for component detail rather than one net number. Refund reasons and unmatched adjustments may require seller explanation. | Contribution margin, return rate, advertising efficiency and postage recovery are management choices, not tax-return fields. |
| Different reports, platforms and payment providers |
Mechanism, not a separate rule The law does not require one standard export format. It requires adequate records and applicable digital links; fragmented formats make compliance harder rather than creating a new tax duty.[8][15] |
Normalisation, deduplication and payout-to-bank matching are good bookkeeping and a common working-paper need. | Cross-channel profitability is useful for decisions, but usually exceeds the minimum needed for a return. |
| Personal items mixed with trading activity |
Legally significant Disposing of personal possessions and trading are not the same. Platform reporting does not decide the answer, and some valuable personal possessions can raise Capital Gains Tax questions.[1][69] |
A seller explanation and consistent classification are commonly needed because the platform cannot know intention or ownership history. | Optional unless the seller wants a combined household and business view. |
| Stock cost, expenses and receipts |
Legally significant Claims and accounts need supporting business records. Stock records are especially important for companies, traditional accounting and VAT margin schemes.[5][7][62] |
Receipt capture and a documented missing-evidence policy reduce year-end queries; an adviser may decline unsupported claims. | Item-level cost and margin can be commercially valuable even where cash-basis tax records do not require accrual stock matching. |
| Bank reconciliation |
Usually supporting practice There is no universal rule forcing every sole trader to perform a formal monthly bank reconciliation in a prescribed format. However, retained bank records and accurate, complete figures are required; companies must record money received and spent.[5][7] |
A core control used to test completeness, duplicates, reserves, transfers and payout timing. | Provides cash visibility and can expose failed payouts or unclaimed refunds. |
| Spreadsheets, guesswork and repeated queries |
Not inherently non-compliant A spreadsheet can be lawful, including in an MTD-compatible bridging workflow, if records and digital links meet the rules. Guessing without evidence is the risk, not the file format.[15] |
Manual sheets are often serviceable at low volume but become fragile when schemas, periods and mixed activity multiply. | Automation may save time, but that benefit and willingness to pay remain unmeasured in this market. |
Move any slider and watch the bank figure and the profit figure move apart. This is arithmetic, not tax advice: it ignores VAT, the accounting basis you use, and every other business cost.
Starting figures match the worked example in this report.
| Line | Amount |
|---|
If refunded-adjusted receipts are £110, item cost is £35 and another £1 of postage was bought outside the platform, an illustrative pre-other-expense result is £51: £110 less £13 fees, £4 advertising, £7 total postage and £35 stock cost. The £2 reserve is still an asset/timing item, not automatically an expense.
Classification: arithmetic example, not a claim about any named platform and not tax advice. VAT, accounting basis, stock treatment and other expenses could change the result.
Order, payment, settlement, fee, ad and fulfilment exports use different columns, identifiers, signs, date conventions and file windows. Combining them requires mappings, not simple addition.
A seller may use marketplace-managed payments, PayPal, Shopify Payments, card processors and direct bank transfers. The bank shows cash endpoints, while provider balances, reserves and fees explain the journey.
Personal possessions and goods bought or made for profit can share an account and a payout. The marketplace generally records an order, not the seller’s original intention or the business-use facts HMRC considers.[1][4]
A sale record does not prove the purchase price of stock, materials, packaging or external postage. Missing receipts turn a calculable margin into an estimate and can prevent a clear audit trail.
Manual exports are downloaded repeatedly, copied into different tabs and reclassified. Overlaps can duplicate transactions; missing date ranges create gaps; formulas and signs change without an audit log.
A deposit may aggregate many orders, net later refunds, include holds released from earlier periods, and omit charges billed elsewhere. A match by amount alone may be impossible or misleading.
When twelve months are reconstructed at once, the seller is less able to remember why a cost was incurred, whether an item was personal, or which report version superseded another.
Advisers cannot derive missing seller facts from a bank statement. They commonly return questions about unexplained deposits, stock, mixed-use costs, missing reports, refunds and unsupported differences.[46][47]
A year-end total may be numerically plausible yet impossible to trace back to a platform row, payout, bank entry and supporting document. That weakens correction, review and explanation.
These categories overlap. “Side business” describes scale or working pattern; “sole trader”, “partnership” and “limited company” describe legal or tax form. Transaction count alone does not decide trading status.
Two people can use the same app and be in completely different positions.
Clearing out your own wardrobe is normally not a business, however many things you sell. Buying things in order to sell them at a profit normally is a business, even if it is small and even if you have a full-time job.
Nobody decides this by counting your sales. It is decided by the facts: what you bought, why you bought it, and what you did with it.
| Seller circumstance | Typical platforms and activity | Records generated | Records likely to be needed | When activity may be trading | Common tools and workarounds | Accountant involvement | Likely difficulty |
|---|---|---|---|---|---|---|---|
|
Casual seller Facts vary |
Occasional eBay, Vinted, Depop or local-platform sales; often a handful of transactions, but “casual” can also include small profit-seeking activity. | Listing/order history, messages, balance history, postage labels, payout and bank entries. | Enough evidence to show what was sold, whether it was personal, dates, original purpose/cost where relevant and any profit-seeking purchases. | Where items are bought or made with a view to profit, or the pattern and surrounding facts indicate a trade. No single badge is decisive.[1][4] | Platform history, screenshots, email receipts, bank app or a basic sheet. | Usually none unless gross trading income, a high-value disposal or another tax issue creates uncertainty. | Low if activity is genuinely personal and evidenced; higher if labels such as “casual” are used instead of recording intention and facts. |
| Personal-item disposer | Wardrobe, furniture, electronics, collections or household clear-outs; commonly Vinted, eBay and Depop. | Sale/order records and payout history; rarely a complete original-purchase record for older possessions. | Evidence of original personal ownership and, for valuable assets, acquisition and disposal details. Bank trail if mixed with business cash. | Usually not trading merely because possessions are sold. A personal possession sold for £6,000 or more can raise a separate Capital Gains Tax question, subject to the asset and available exemptions such as those for cars and wasting assets.[69][1] | Platform history, photos, old receipts where available, a simple disposal log. | Normally limited; advisable for valuable chattels, repeated ambiguous activity or mixed personal/business accounts. | Original cost may be unavailable; personal items can be indistinguishable from resale stock in a payout. |
| Reseller | Repeated sourcing and resale on eBay, Vinted Pro, Depop, Amazon or multiple channels; tens to thousands of orders may be operationally plausible. | Orders, returns, fees, payout statements, promotions, labels, stock purchases and supplier receipts. | Complete sales and expense records, payout-to-bank reconciliation, per-item or pooled stock evidence, returns/write-offs and closing stock if traditional accounting applies.[5][50] | Buying items with the intention of reselling at a profit is a strong indicator of trading.[1] | Sheets, SKU labels, seller dashboards, Vinta, Veeqo, sellerboard, connectors and a general ledger. | Often annual at lower complexity; periodic bookkeeping becomes more common with volume, VAT, staff, borrowing or multiple channels. | Item-cost matching, bundle allocations, unsold stock, returns, mixed personal items, multiple schemas and gross-to-net reconciliation. |
| Handmade-product business | Etsy, Shopify, TikTok Shop, Amazon Handmade and fairs; low-to-high order counts with variants or custom work. | Orders, discounts, marketplace fees, advertising, payment settlements, material invoices, production records and postage. | Sales and payout records plus materials, packaging, work in progress/finished goods where relevant, direct production costs and off-platform sales. | Making goods for sale with a profit-seeking commercial pattern will generally point towards trading; facts still govern.[1][4] | Etsy/Shopify exports, spreadsheets, Craftybase, inventory tools and accounting software. | Often needed as materials, stock, VAT, fairs and several payment methods accumulate. | Bill-of-materials cost, wastage, labour as management information, custom deposits, multi-channel duplication and sales outside marketplaces. |
| Side business | Part-time resale or making around employment, caring or study; often one or two channels before expansion. | The same business records as a full-time operation, often mixed with a personal bank account and household purchases. | Complete business-income and expense evidence, separation from employment/personal activity and information for the whole tax return. | “Side” does not change the trading test. Gross trading income across relevant activities is aggregated when considering the £1,000 trading allowance.[3] | Personal bank account, spreadsheet, receipt folder, platform reports; sometimes entry-level ledger software. | Commonly annual Self Assessment support; more frequent if MTD for Income Tax applies or records deteriorate between year ends. | Time scarcity, mixed accounts, missed receipts, duplicate annual clean-up and interactions with employment and other income. |
| Sole trader | Any platform and scale; the individual owns the business and may operate several distinct trades. | Platform and bank records plus business purchases, expenses, assets, tax correspondence and, where applicable, VAT/PAYE records. | All sales/income, all expenses and supporting proof; additional debtors, creditors, stock, bank balances, capital introduced and drawings under traditional accounting.[5] | The individual is carrying on a trade rather than merely disposing of personal property. Self Assessment is generally required where annual gross trading income exceeds £1,000, subject to other filing reasons.[3] | Sheets or general ledger; marketplace connector at greater scale; bridging or MTD-compatible software if in scope. | Optional but common. The seller remains legally responsible for accurate records and return information even when an agent acts. | Owner and business are not separate legal persons, so private/business separation, drawings and mixed-use costs require discipline. |
| Partnership | Two or more people carrying on the marketplace business together, across any channel or volume. | Partnership books, platform accounts, bank, stock, expenses, partner capital/current accounts and allocation information. | Partnership income and expense records, a partnership return and figures for each partner’s individual return. A nominated partner manages records and the partnership return.[9] | The partnership itself carries on the trade; individual partners report their shares. | General ledger, sheets at small scale, accountant working papers; channel connectors where supported. | Usually higher because a partnership return must align with each partner’s return. | Partner-funded purchases, drawings, profit shares, ownership of marketplace accounts and reconciling partnership totals to individual filings. |
| Limited company | Any marketplace, often more formal operations, employees, wholesale or higher volumes; the company is legally separate. | Company sales, costs, assets, liabilities, stock, payroll, VAT, bank and director transactions. | Adequate company and accounting records, annual statutory accounts and Company Tax Return, with stocktaking and goods-bought/sold support where relevant.[7][8] | The company’s activity is assessed as its business; the owner’s personal sales must not be treated as company activity merely because the same person controls both. | Business bank account, full accounting ledger, inventory/connector apps and accountant or finance team. | Common and often continuous, because accounts, Corporation Tax, payroll, VAT and Companies House obligations can interact. | Separate-entity discipline, director transactions, stock valuation, cut-off, VAT, payroll, retained reserves and platform accounts opened in the wrong name. |
Pick the description closest to you. This is general information drawn from the sources in this report, not advice about your situation.
The difficult part is not one calculation. It is preserving enough context from the original event for the seller, adviser and filing process to agree what a number represents.
Think of it as a relay race where each runner drops something.
The platform hands on a total instead of the orders. The payout hands on cash instead of the sales. The bank hands on one number instead of the detail. By the time an accountant sees it, most of the story has gone.
Two batons never get picked up at all: the evidence of what you paid, and your own confirmation that the year is complete. No software can carry those for you.
Read it left to right in three stages. Stage 1 is what the systems record for you. Stage 2 is what a person still has to do. Stage 3 is what reaches HMRC. The two boxes no software can fill are marked in red.
Your selling activity creates a marketplace order record. That becomes a payout statement, and then a single net deposit in your bank. Your own evidence of what you paid, and why you owned the item, arrives on a separate route and only if you kept it. Those strands are then reconciled, categorised, reviewed by you and worked into an accountant’s papers, before a decision about which filing applies: a Self Assessment tax return, a Making Tax Digital quarterly update if you are in scope, or another filing such as VAT, a partnership return or company accounts. Box 5, your own evidence, and box 8, your own review, cannot be produced by any software, because only you hold the facts.
| Stage | Data created | Primary responsibility | Common errors | Evidence retained | Information lost or transformed | Why reconciliation becomes difficult |
|---|---|---|---|---|---|---|
| 1 · Seller activity | Listing, SKU, description, sale intention, acquisition, dispatch and return facts. | Seller for truthful listing and business records; platform for captured system events. | No SKU; reused listing; wrong account; personal item presented like stock; stock purchase not logged. | Original purchase receipt, listing, photos, messages, supplier invoice, proof of dispatch. | Original intention, condition, bundle allocation and private-use history are rarely encoded in finance exports. | Later systems see only an order ID and amount, not why the seller owned the item. |
| 2 · Marketplace order record | Order date, item, buyer price, discount, tax, postage, status and platform identifiers. | Platform creates system record; seller checks completeness and downloads it. | Cancelled orders included; gross/net columns confused; local date/time misread; overlapping exports duplicated. | Native order export and account statement, preserved unchanged. | Platform-specific terms replace common accounting concepts; some components may be aggregated. | Different platforms use incompatible signs, headers, identifiers and download windows.[19][22][27] |
| 3 · Fees, refunds and adjustments | Commission, payment fees, ad charges, labels, fee credits, disputes, reserves and other movements. | Platform calculates; seller must preserve and understand the account activity. | All deductions called “fees”; refunds dated to original order rather than event; reserves expensed; fee VAT ignored. | Transaction-level account activity, fee invoices, credit notes and refund records. | Several line types may be netted into an “earnings” or balance total. | A single order can have later adjustments and multiple settlement periods.[20][23][34] |
| 4 · Payout statement | Payout ID, period, opening/closing balance, component transactions, net amount and status. | Marketplace/payment provider produces; seller verifies period and destination. | Treating payout as sales; ignoring failed/reversed payouts; omitting balance spend or reserve release. | Settlement report, balance statement and payout export. | Many orders become one cash total; order date becomes settlement date. | Settlement windows cross month, quarter and tax-year boundaries; not every provider uses a common payout ID. |
| 5 · Bank transaction | Posting date, value, currency, bank reference and counterparty. | Bank records cash; seller identifies all relevant accounts; bookkeeper matches. | Grossing up by guess; duplicate feed/import; personal transfer categorised as sale; fees absent from bank. | Full statements and feed history, including personal accounts used for business.[5] | The bank normally retains only the net deposit and a short reference. | Date can lag payout; batch totals obscure orders; foreign exchange and withheld balances create differences. |
| 6 · Stock and expenses | Supplier, amount, item/material, business purpose, VAT, payment source, receipt and stock movement. | Seller captures source evidence; adviser decides accounting/tax treatment where engaged. | Missing cash purchases; whole bundle assigned to first item; personal spend claimed; postage duplicated. | Receipts, invoices, mileage/logs, courier evidence, stock count and valuation basis. | Bank description rarely shows item, business purpose, VAT or allocation to sale. | Costs may be paid months before sale, outside the platform and from another account. |
| 7 · Reconciliation | Links between order activity, balance movement, payout and bank deposit; residual difference. | Seller/bookkeeper prepares; accountant reviews in line with engagement. | Forcing a zero by adding an unexplained adjustment; omitting opening/closing balance; matching amount but wrong payout. | Reconciliation schedule, source row references and explanation of timing/unmatched items. | Correctly explained differences become classifications; undocumented plugs destroy provenance. | Missing source files, no stable identifiers, fee credits, reserves, multi-currency and partial refunds. |
| 8 · Categorisation | Ledger categories, tax categories, personal/trade decisions, VAT codes and stock treatment. | Seller supplies facts; bookkeeper applies agreed rules; accountant/tax adviser handles professional judgement. | Personal sale treated as turnover; reserve treated as expense; stock purchase and COGS conflated; disallowable costs included. | Decision log, mapping rule, seller explanation and supporting document. | Detailed platform lines become category totals; rationale can disappear if not recorded. | Similar cash movements can require different treatment because of intent, legal form, basis or VAT. |
| 9 · Seller review | Completeness statement, corrections, confirmations and responses to queries. | Seller; an agent cannot know unreported channels or personal facts without disclosure. | Approving totals without source review; forgetting dormant accounts; guessing old costs; leaving queries until deadline. | Signed/dated responses, list of accounts and unresolved-item schedule. | Memory substitutes for contemporaneous evidence when records are late. | Time pressure encourages broad assumptions rather than item-level answers. |
| 10 · Working papers | Trial balance, reconciliations, stock schedule, adjustments, tax computation and cross-references. | Accountant or bookkeeper within the engagement; seller remains responsible for complete information. | Platform summary does not agree to bank; missing year-end stock; unsupported journals; prior-year timing not carried forward. | Source pack, queries/responses, journals and review sign-off. | Operational data is transformed into accounts and tax categories. | Every unsupported total generates a query or a documented limitation. |
| 11 · Submission | Quarterly category totals, annual Self Assessment return, VAT return, partnership return or company accounts/return. | Taxpayer/company and authorised agent according to the filing regime. | Quarterly update mistaken for final tax return; platform calendar year used as UK tax year; one entity’s activity filed under another. | Submission receipt, filed return/accounts, computation and underlying retained records. | HMRC normally receives totals, not every receipt or transaction line for an MTD update.[12] | A successful electronic submission proves transmission, not completeness or correct upstream classification. |
The legal duty depends on the facts, legal form, turnover, VAT status and tax year. Advisers often request more structured information than the minimum law states because they must turn retained evidence into reviewable accounts and returns.
Three different jobs get muddled together, so it helps to name them.
Keeping records means being able to show what happened. Sending a quarterly update means sending running totals to HMRC if the new digital rules apply to you. Filing a tax return means working out what you actually owe. They are not the same job, and doing one does not do the others.
The digital rules started on 6 April 2026 for the largest group, and the first quarterly deadline is 7 August 2026. You are not signed up automatically, even if the rules apply to you.
HMRC points to buying or making goods for sale at a profit and uses the wider “badges of trade”, including profit motive, repeated/systematic transactions, asset type, modifications and relationship to an existing trade. The conclusion comes from the overall facts, not a marketplace label or transaction count.[1][4]
The trading allowance is up to £1,000 of annual gross trading income across relevant activities. Above £1,000, a person will generally need to tell HMRC/register for Self Assessment; where partial relief is elected, the allowance replaces actual expense deductions. Partnership income is excluded from the allowance.[3]
From 1 January 2024, in-scope digital platforms collect, verify and report seller information annually. For goods sellers, a platform need not report a seller only where both fewer than 30 sales and less than €2,000 are met. This is a reporting exclusion, not a tax-free allowance or trading test. A professional-body summary makes the same point: the regime changed what platforms report, not the underlying tax rules.[2][52]
Self-employed people keep sales, income, expense and supporting records, normally for at least five years after the 31 January filing deadline. Companies keep adequate company/accounting records, generally for six years from the financial year end.[5][6][7]
VAT registration is generally compulsory when UK taxable turnover exceeds £90,000 over the previous 12 months or is expected to exceed it in the next 30 days. Taxable turnover is not payout or profit. VAT-registered sellers need output/input VAT records and valid evidence; eligible second-hand sellers may need advice on the optional margin scheme.[16][17][18]
Trade status, chattels, mixed business/personal costs, cash versus traditional accounting, stock valuation, VAT margin schemes, overseas stock/sales, imports, foreign currency, company/director transactions and corrections to earlier returns cannot safely be decided from marketplace exports alone.
Qualifying income means gross self-employment and property income added together, before expenses. It is not profit, not a payout total and not one platform's figure.[54]
| Qualifying income | Start date | Tested on | Status |
|---|---|---|---|
| Over £50,000 | 6 April 2026 | Tested on the 2024/25 return | Live now |
| Over £30,000 | 6 April 2027 | Tested on the 2025/26 return | Announced |
| Over £20,000 | 6 April 2028 | Tested on the 2026/27 return | Announced |
| £20,000 or less | Not applicable | Not applicable | Automatically exempt |
This means turnover before taking off any expenses, added together across your self-employment and property income. It is not your profit, not your bank payouts and not one platform’s figure. Employment wages, dividends and a partner’s share of partnership profit do not count.
This compares one number against HMRC’s published timetable. It is not a decision about your circumstances, it does not check the exemptions and deferrals that apply to some people, and it is not advice.
| Question | Current position | Records implication | Important boundary |
|---|---|---|---|
| Who is in scope? | Individuals registered for Self Assessment with qualifying gross income from self-employment and property above the relevant threshold. Qualifying income is tested using the previous tax return and is before expenses.[10][14] | Marketplace turnover contributes where it is self-employment income; unrelated personal disposals do not become trading income merely because a platform reports them. | Employment, dividends and an individual partner’s partnership profit share do not count as qualifying income for this test.[14] |
| Mandatory dates | Over £50,000 qualifying income: from 6 April 2026, based on 2024/25. Over £30,000: from 6 April 2027, based on 2025/26. Over £20,000: from 6 April 2028, based on 2026/27.[10] | A seller inside a band needs compatible software and digital records from the start of the relevant tax year. | These are qualifying-income thresholds, not profit, platform payout, VAT or per-marketplace thresholds. |
| Structures | Sole traders can be in scope. A timetable for partnerships has not yet been set. Limited companies are outside MTD for Income Tax and follow Corporation Tax/company-account regimes.[14][8] | Entity and account ownership must be known before records are grouped. | A side business can be in scope if the person’s qualifying income crosses the band; not every marketplace seller is in scope. |
| Digital records | The general rule is to keep digital amount, date and category records in compatible software and retain supporting evidence. Two important easements qualify that statement: an eligible retailer may elect to record one daily gross-takings total, explicitly including third-party online-platform sales; and a business below the VAT-registration threshold may record income and expenses in broader categories.[11][85] | Item-level source data may still be needed to prove accuracy, reconcile payouts, answer queries and manage stock, even where MTD itself permits a daily total. | MTD does not make an expense allowable, determine whether a sale is personal, or invariably mandate one digital line per marketplace order. |
| Digital links | Where more than one digital product is used, data must move by digital link. Linked spreadsheet cells, emailing a spreadsheet, CSV or XML import, API transfer and other automated links are permitted. Manual retyping and copy or paste between products are not, and HMRC frames that prohibition around records that have been included in a quarterly update.[11] | A spreadsheet can remain part of the record if it connects to compatible submission software without manual transfer. | A bridge can transmit compliant totals; it cannot repair incomplete or misclassified source records. |
| Quarterly updates | Software totals income and expenses for each business and sends cumulative summaries. No accounting or tax adjustments are required before the update, and HMRC does not receive each invoice or receipt line.[12] | Standard-period deadlines for 2026/27 are 7 August 2026, 7 November 2026, 7 February 2027 and 7 May 2027. No penalty points arise for missing a quarterly update in 2026/27.[55] | An update is not a tax return and does not normally show the final tax liability. |
| Year end and final submission | After the tax year, the business records are adjusted as needed; the person adds or checks other income and gains, reviews the calculation and submits the annual tax return through compatible software by 31 January.[13] | Final records need year-end adjustments, unresolved-query closure and consistency with other return sections. | Some software/API material uses “final declaration”; current GOV.UK user guidance presents this as completing and submitting the tax return. It remains distinct from quarterly updates. |
| Agent process | Where an agent submits, current guidance says the agent shares the return with the client and obtains written confirmation before submission.[13] | The seller review remains a real control, not a passive hand-off. | Agent involvement does not transfer the underlying duty to provide complete and accurate facts. |
| Layer | What it covers | Marketplace-seller examples | What it does not mean |
|---|---|---|---|
| Legal / regulatory | Statutory records, retention, returns, accounts, tax payment, VAT and applicable digital-record duties. | All business sales and expenses; supporting receipts; company stocktaking evidence; and MTD digital records where in scope, subject to HMRC’s daily gross-takings and category easements.[5][7][11][85] | HMRC does not prescribe one universal seller spreadsheet, require every seller to use MTD for Income Tax, or require every eligible retailer to digitise each order separately. |
| Common accountant request | Inputs and reconciliations needed to prepare defensible working papers efficiently. | All platform exports, processor statements, bank statements, stock count, receipts, VAT invoices, unusual-adjustment explanations and payout reconciliations.[46][47] | A firm’s preferred pack format is not automatically a statutory form. Requirements vary with engagement and materiality. |
| Good bookkeeping | Controls that make records complete, accurate, timely and traceable. | Separate accounts, regular downloads, immutable originals, duplicate/gap checks, payout-to-bank matching, documented mappings and visible exceptions. Professional training material frames these as source-document and reconciliation disciplines rather than optional tidiness.[45] | Good process cannot replace missing evidence or decide disputed tax treatment. |
| Management information | Information used to run the activity, not necessarily required on a tax return. | Per-item margin, sell-through, channel profitability, ad return, return rate, aged stock and contribution after fulfilment. | A useful dashboard is not proof that the ledger, VAT return or taxable profit is correct. |
| Professional advice | Fact-sensitive accounting, tax and legal conclusions. | Trade status, VAT registration/margin scheme, stock valuation, capital treatment, cross-border supplies, partnership shares and prior-period corrections. | Software categorisation or a platform tax report should not be treated as personalised advice. |
Platform names, seller IDs, opening/closing dates, payment providers, bank accounts, currencies and confirmation that the period is complete.
Order/sales reports, fee and refund activity, statements, payout/settlement reports, processor fees, chargebacks, reserves and VAT invoices where available.
Full bank/credit-card statements, stock invoices, courier/postage receipts, packaging, software, advertising and other operating expenses.
Since the 2024/25 tax year the cash basis is the default way a sole trader or partnership works out trading profit, with traditional accruals accounting available by election. The old turnover limits and the £500 interest restriction were removed. Under the cash basis, goods bought for resale are deducted when paid for, with no opening or closing stock adjustment.[59][50]
Buy heavily in March and the whole cost lands in that year even though nothing has sold. Spend the next year selling the same stock and the profit looks inflated. The figures are correct but the year-on-year picture stops resembling margin, which is why a stock-heavy seller may be better served by electing traditional accounting.[50]
A points-based regime applies: one point per missed quarterly update or return deadline, a £200 penalty at four points, and £200 for each further miss. HMRC has confirmed no penalties for missing a quarterly update deadline in 2026/27. That grace does not cover the annual return or late payment, where charges run at 3% at day 15, a further 3% at day 30, then 10% a year from day 31.[55]
Qualifying income of £20,000 or less is automatically exempt. Some groups are deferred to April 2027, including people who claimed averaging relief, qualifying care relief, trust or estate income, or the remittance basis. Others are permanently outside, including anyone without a National Insurance number before the tax year starts.[56]
Government has announced that the Self Assessment reporting threshold for trading income will rise from £1,000 to £3,000 gross, with a simpler online route for those who still owe tax. No commencement date is set. Read it carefully: the trading allowance stays at £1,000, so income between the two figures can still be taxable, just reported differently.[70]
Companies House identity verification became mandatory on 18 November 2025 with a twelve-month transition, so existing directors and people with significant control must verify during 2026. It is separate from anything tax-related and it applies to the person, not the company.[71]
VAT can make record gaps materially expensive, but no evidence reviewed establishes that it is the “most expensive” error for marketplace sellers or how frequently it occurs. This section therefore explains the mechanism without estimating prevalence.
VAT is a separate question from income tax.
A UK-established seller selling its own goods already in the UK generally remains responsible for its VAT position. Turnover is not the net payout. Different rules can apply to overseas sellers, imported consignments and some business-to-business transactions, so seller location and the movement of goods matter.
Eligible second-hand sellers may use a margin scheme, but only with the required stock book and purchase/sale evidence. The benefit depends on the margin; it is not one fixed multiplier.
Since January 2021 an online marketplace accounts for the VAT in two situations: goods of any value already in the UK at the point of sale where the seller is an overseas business, and goods imported in a consignment not exceeding £135. The £135 test applies to the whole consignment, not each item. If the buyer supplies a VAT number, liability moves back to them.[60]
A UK-established seller selling its own goods that are already in the UK is generally outside those marketplace deemed-supplier rules. It monitors taxable turnover against the £90,000 registration threshold and, once registered, accounts under the VAT rules applicable to its supplies.[60][16]
If a seller assumes that platform-collected VAT on some lines settles every VAT obligation, or monitors only net payouts, taxable turnover can be understated. The size and consequence depend on the seller’s actual supplies, registration status and evidence; professional advice is appropriate near the threshold or across borders.
| What it is | Where it belongs | The trap |
|---|---|---|
| The price the customer paid | Your sales figure, and output VAT on it. | Using the payout instead. This is the central error. |
| Marketplace commission | An expense you buy, with input VAT recoverable if UK VAT was charged on it. | Netting it against sales instead of recording both sides. |
| Payment processing | An expense. | Treating it as a reduction in the sale price. |
| Advertising and promoted listings | An expense, with input VAT if UK VAT was charged. | Missing it entirely when it is billed separately rather than deducted. |
| Postage bought on the platform | An expense. Note that universal-service postage is largely VAT-exempt, so there may be no input VAT to reclaim. | Assuming there is VAT to recover on every postage cost. |
| Refunds | A reduction of output VAT in the period they occur. | Dating them to the original sale. |
| The payout itself | A bank receipt. Nothing else. | Treating it as a sales figure. |
For eligible second-hand goods, VAT is charged on the difference between purchase and selling price rather than the whole selling price. For a business buying from private individuals there is normally no input VAT on the purchase to reclaim, so the scheme can materially reduce VAT compared with the normal method; the saving depends on the item’s actual margin.[62]
HMRC requires purchase and sales invoices plus a stock book recording each item’s purchase and sale details, including dates, prices, identifying information, margin and VAT. If buying from a private individual, the dealer may need to make the purchase invoice; VAT is not separately shown on a margin-scheme sales invoice.[62]
The scheme is optional and limited to eligible goods and acquisition routes. If an item is not eligible, or the records do not support the scheme, normal VAT treatment can apply. Global Accounting can be relevant to certain low-value goods, but eligibility and record rules should be checked rather than inferred from a platform category.[62]
Inadequate evidence can put margin-scheme treatment at risk and lead to VAT being calculated under the normal rules. The financial increase is case-specific: it depends on purchase price, selling price, VAT rate and recoverability. The previous version’s universal “roughly sixfold” claim was wrong and has been withdrawn.[62]
Above £135 a consignment meets normal import VAT and duty at the frontier. Postponed VAT accounting lets a VAT-registered business declare and reclaim import VAT on the same return, with no application needed. The catch is documentary: you must download the monthly postponed import VAT statement, and statements are only available for six months.[63]
For VAT, amounts must be expressed in sterling on a consistent, evidenced basis, and you keep the original foreign-currency document as well as the sterling figure. Platform, processor and bank conversion figures can differ, so the retained rate and method matter.[64]
Marketplace collection, import VAT, export evidence, the EU Import One Stop Shop and any registration or intermediary requirement depend on consignment value, customer status, fulfilment route and platform arrangement. The report found no basis for the previous universal claim about “most small sellers”; obtain case-specific advice before choosing a treatment.
Native reports explain the platform ledger; profit tools explain operations; connectors translate settlements into a general ledger; accounting software holds the books; bridges transmit selected totals; professionals apply judgement. The intended scope of each category matters when judging a gap.
There are good tools here, and none of them is bad at its job. The point is that each one does a slice.
The marketplace tells you what happened on the marketplace. A connector moves those figures into accounting software. Accounting software keeps the books. A bridge sends totals to HMRC. An accountant applies judgement.
Nobody sells the whole chain, because a large part of the chain is you.
All 27 rows shown. Filtering and search need JavaScript; without it every row remains visible.
| Provider / category | Target and platforms | Data in / out | Sales, deductions and payouts | Bank reconciliation | Stock / COGS and personal activity | Accountant / HMRC hand-off | Pricing / model | Strengths, limitations and adoption path |
|---|---|---|---|---|---|---|---|---|
| eBay Seller Hub & reportsMarketplace-native | Private and business eBay sellers; eBay activity only. | Orders, transactions, payouts, statements and fee invoices, downloaded as CSV. eBay documents CSV only; the files open in spreadsheet software but no XLSX export is published.[19] | Transaction/earnings reports expose gross amounts, fees, fee credits, refunds, labels and payouts. Earnings reports exclude postage labels bought with methods other than eBay funds. eBay describes the report as earnings "at any given point in time", so re-running the same order period later can return different figures once refunds land.[20][19][20] | Payout IDs/dates and bank references support matching, but the seller or ledger completes the bank reconciliation. | No original stock cost or off-platform evidence. Private/business account distinction exists, but the report cannot decide tax status. | Useful source files for a seller or accountant; no UK tax return or MTD for Income Tax submission. | Seller Hub/reporting included with account; private selling is generally fee-free in the UK except for stated categories and options, while business fees vary by category and service. From 12 February 2026 the per-order business fee rose from 30p to 40p on orders above £10, with several category rates revised.[21][74][21] | Solves: authoritative eBay transaction and payout detail. Boundary: deliberately eBay-only; external stock, expenses, banks and accounting treatment remain elsewhere. Adopt for source truth; outgrow when cross-channel control or a ledger is needed. |
| Etsy Shop Manager & Payment accountMarketplace-native | Etsy makers, vintage and craft-supply sellers; Etsy only. | CSV downloads for order items, orders, Etsy Payments sales/deposits and monthly statements. | Shows sale and fee activity, refunds, advertising and postage labels. Deposit calculation reflects available sales less account charges and timing; Etsy states deposits are not net profit.[22][23] | Deposit records support bank matching; external payment activity and costs still require other sources. | No complete materials/BOM or external stock-cost evidence; no personal/trade conclusion. | Strong source export; no direct UK accounts, Self Assessment or MTD filing. | Reporting included; seller charges are transaction-based and include listing, transaction, payment, advertising and optional-service fees depending on circumstances.[24] | Solves: Etsy-native order and payment-account evidence. Boundary: separate exports and timing must still be combined with materials, bank and other channels. Adopt as source record; outgrow with multi-channel or production-cost complexity. |
| Vinted Balance & transaction historyMarketplace-native | Private wardrobe clear-outs and Vinted Pro businesses; Vinted only. | Account order and balance history, withdrawals, and the DAC7 reporting workflow.[26] The public first-party material reviewed did not document a standard transaction-level finance export for private or Pro sellers. | Private sellers ordinarily pay no seller fee; buyer-paid protection and optional seller services affect the wider transaction. Funds move from pending to available balance and then withdrawal.[25] | Balance/withdrawal history can be compared with bank entries; public documentation reviewed does not show a full payout-reconciliation workflow. | Platform cannot prove original item ownership/cost or determine personal disposal versus trade. Commercial sellers are directed to Vinted Pro.[25] | Account history can support records; no direct UK filing. | Core private selling has no seller fee; optional visibility features and Pro/commercial terms vary. | Solves: first-party transaction and balance evidence. Open verification gap: Pro invoices and DAC7 copies are not transaction-level sales exports, but absence from reviewed public documentation is not proof that no logged-in export, data-access route or partner integration exists. Test with current seller accounts before treating this as a product opportunity.[26][76] |
| Depop sales downloadMarketplace-native | Fashion, vintage and creative sellers; Depop only. | Sales CSV for up to three months per download, with buyer, item, quantity, prices, fees, shipping and taxes. | Export contains payment, marketplace, selling and boosting-fee fields. UK selling fees were removed on 20 March 2025 and a buyer-paid fee was introduced from 15 April 2025, so the "Selling" column now reads zero on UK sales while payment processing and optional boosting at 8% carry the real cost. All fees reverse on refund.[27][28] | CSV assists matching; three-month download windows increase file-overlap/gap risk. | No acquisition-cost evidence, inventory valuation or personal/trade determination. | Useful CSV input for a sheet/accountant; no direct HMRC submission. | No UK selling fee since 20 March 2025. Payment processing still applies, and boosting is optional at 8% of the sale price excluding taxes and postage.[28][75][28] | Solves: accessible transaction-level sales export. Boundary: payout, bank, stock and off-platform evidence still need assembly, and any fee mapping built before March 2025 will now mis-state UK cost of sales. Adopt early; outgrow with volume, long history or multiple channels. |
| Amazon Seller Central reportsMarketplace-native | Individual and Professional Amazon sellers; FBM/FBA and multiple Amazon marketplaces. | Orders, payments/settlements, fees, tax, inventory, fulfilment and advertising reports; APIs on eligible plans. | Settlement reports break down order amounts, refunds, fees and other transactions. The XML and original flat-file settlement reports are withdrawn on 11 November 2026, and integrations must move to the Flat File V2 report, which identifies each line by amount type and description.[30][77] FBA, storage, refunds, ads and cross-border activity add report families and accounting dimensions.[29][30] | Settlement total can match deposit, but seller/connector/ledger must post components and resolve reserve or currency differences. | FBA reports add inventory movement; acquisition/manufacturing cost still comes from the seller. No personal/trade split. | Rich adviser source data; no direct UK annual return or MTD for Income Tax filing from Seller Central. | Individual plan £0.75 per item; Professional £25/month, both excluding VAT, plus referral fees generally between 8% and 15% and optional fulfilment, storage and advertising charges. Referral and FBA fees were reduced across several UK categories on 5 January 2026, so older fee tables are stale.[29][78][29] | Solves: detailed first-party operational and settlement data. Boundary: report complexity is proportional to Amazon’s operational scope. Often paired with a specialist connector/accountant rather than replaced. |
| Shopify analytics & Shopify PaymentsCommerce-platform-native | Direct-to-consumer merchants; Shopify store, with many payment gateways and channels. | Orders, finance analytics and Shopify Payments payout transaction CSV; apps/API. | Payout detail includes charges, refunds, fees, adjustments and reserves. Shopify explicitly says payout reconciliation is not a revenue statement and explains order-versus-payout timing and third-party exclusions.[31][32] | Detailed balance and payout reporting supports reconciliation for Shopify Payments; Shopify billing fees, third-party processors and payment methods such as Afterpay are excluded, and the report carries a three-day data delay.[31] | Product cost/inventory can exist in Shopify, but completeness and accounting valuation depend on setup. Personal/trade separation is outside intended scope. | Exports/apps feed accountants and ledgers; no native UK Self Assessment filing. | Basic from £19/month billed annually (£25 monthly), then Grow £49, Advanced £259 and Plus from £1,800, plus card and transaction charges and app costs. The middle tier is now named Grow rather than Shopify.[33][33] | Solves: integrated order, payments and operational store data. Boundary: Shopify billing, third-party gateways, marketplaces and tax records can remain separate. Adoption is strong for store operations; accounting integration grows with complexity. |
| TikTok Shop Finance reportsMarketplace-native | TikTok Shop UK sellers, including affiliate-led and fulfilment activity. | Excel finance report with six tabs: order details, statements, reports, payments, reserve details and fee explanations. The reserve sheet is a recent addition. Settlement is stated as net sales less shipping less fees plus adjustments.[34] | Settlement amount bridges net sales, shipping, fees and adjustments. Statement timing follows settlement rather than original order, and affiliate, advertising, logistics and reserves can add components.[34] | Payment/report tabs support deposit matching; settlement timing and reserve periods require roll-forward logic. | No original stock-purchase proof or personal/trade conclusion. | Strong source workbook for bookkeeping; no native UK annual/MTD filing. | Finance reporting is included with shop operation; commissions, fulfilment, advertising and affiliate charges depend on category/service. | Solves: broad first-party finance workbook. Boundary: fast-changing fee types and settlement rules need version-aware interpretation; other channels and source costs remain external. |
| Facebook Marketplace / Commerce ManagerMarketplace-native / local selling | Local personal sellers and businesses using Meta commerce tools; the workflow differs materially by checkout and payment route. | Commerce Manager documented transaction, payout and reconciliation reports for supported commerce accounts. Meta discontinued its own Shops payment processing, although historic balance, payout and financial reports remain available to affected merchants.[86] | Commerce reports can show orders, refunds, fees and payouts where Meta processed the transaction. Ordinary local Marketplace deals completed by cash, bank transfer or an external provider have no single Meta settlement ledger. | Possible for Commerce Manager payouts; external-payment and local cash sales must be matched from the chosen rail. | No original stock cost, receipt evidence or personal/trade determination. | Reports can support a seller or accountant; no native UK accounts or MTD filing. | Listing/payment terms vary with route; historic reporting access is not a promise of a current integrated checkout. | Solves: finance history for supported Meta commerce flows. Boundary: “Facebook Marketplace” is not one uniform payments product. The unresolved gap is fragmented local/external settlement, not the absence of every Meta report. |
| PayPal & Stripe reportingPayment-provider-native | Sellers receiving wallet, card or direct-store payments across channels. | PayPal provides downloadable activity, statements and financial summaries. Stripe provides balance and payout-reconciliation reports, including automatic and manual payout views.[87][88] | Provider reports explain charges, refunds, fees, disputes, balance movements and payouts within that processor; they do not necessarily carry the marketplace’s full order, delivery, advertising or tax context. | Strong for provider-balance-to-bank reconciliation when identifiers and dates are retained. | No acquisition cost, inventory ownership or personal/trade classification. | Useful source files and integrations for ledgers/accountants; no standalone UK annual-return conclusion from payment data. | Reporting is part of the provider account; payment fees and advanced products vary by transaction and contract. | Solves: the money-rail layer. Boundary: one provider may aggregate several storefronts and cannot reconstruct facts that never entered its ledger. |
| sellerboardProfit & inventory | Amazon sellers needing near-real-time profitability and operational analytics. | Amazon sales, returns, advertising, FBA/FBM fees, inventory and user-entered costs; dashboards/exports. | Attributes more than 100 Amazon fee types, refunds, ad spend and fulfilment costs to profit; estimates payouts. | Profit and payout views assist investigation but are not a full bank-ledger reconciliation. | COGS and stock are supported, but cost accuracy depends on seller inputs. No personal/trade classification. | Reports can inform bookkeeping/accountants; no UK MTD or tax-return filing. | Subscription; Standard listed at US$19/month monthly or US$15/month billed annually for the stated order allowance.[40] | Solves: Amazon operational profit and fee visibility. Boundary: deliberately Amazon-centred management information rather than statutory books or evidence repository. Adopt for Amazon margins; pair with accounting as compliance needs grow. |
| StocksmithMaker inventory & costingRenamed from Craftybase | Handmade and small-batch manufacturers. Published integrations: Shopify, Amazon, WooCommerce, Faire, Etsy, Square, Wix and Squarespace. eBay is not supported, and Amazon is gated to higher tiers. | Materials, recipes/bills of materials, production, orders and expenses; channel imports and accounting exports/integration. | Tracks order revenue/fees and production costs; not primarily a payout-to-bank settlement engine. | Accounting hand-off rather than native bank reconciliation is the emphasis. | Strong materials, work-in-progress, finished stock and COGS logic; still depends on complete purchase and production inputs. No personal/trade split. | QuickBooks Online integration and reports, with Xero listed as coming soon. UK filing is outside scope and some tax language is not UK-oriented. | Rebranded from Craftybase during the period since v1. Advertised plans now start at US$83/month billed annually (Indie), with a Studio tier at US$41 and a Pro tier "from US$20" mentioned only in the FAQ. Quoting US$20 as the price would misrepresent the current positioning.[41][79] | Solves: maker-specific cost genealogy that marketplaces omit. Boundary: production and inventory by design, not UK tax status or settlement reconciliation. The rebrand and the higher entry price both post-date v1, which is a reminder that this layer of the market moves. |
| VeeqoMulti-channel operations | Retailers managing Amazon, eBay, Etsy, Shopify and other published channels. | Orders, shipping, inventory, revenue, marketplace fees, ad spend and user-supplied product costs. | Unified profit analytics and fulfilment/order management across supported channels. | Not positioned as a statutory bank-reconciliation ledger. | Inventory sync and COGS/profit analytics are strengths; personal/trading judgement and receipt evidence are outside scope. | Exports/integrations support operations and downstream accounting; no direct UK MTD for Income Tax return. | No longer free for inventory. Shipping features are free, funded by commission on labels bought through Veeqo; inventory management starts at US$19/month by order volume, and a High Volume tier starts at US$350/month. Pricing is in US dollars and the pre-negotiated carrier set is US-centric.[42][80] | Solves: operational multi-channel order, shipping and stock consistency. Boundary: the free tier is now shipping only, so the v1 description of free inventory management no longer holds. An operations layer is not the legal books or the tax computation either way. Adopt for fulfilment/inventory; add finance controls as obligations deepen. |
| VintaVinted-specific records | Vinted resellers and Pro sellers; single-platform specialism. | Provider states automatic order history, purchases, inventory, shipping, analytics and CSV reports. Its described connection uses a Chrome extension to transfer Vinted session tokens for server-side access; Vinta states it is not affiliated with or endorsed by Vinted. | Sales, orders, purchase tracking and per-item margin are advertised; public material is provider-authored rather than independent verification. | Bank-reconciliation functionality was not established from the public pages reviewed. | Item purchase/inventory tracking and bundle cost allocation are promoted; personal/trade status still depends on facts. | CSV can be given to an accountant; “tax-ready” reports are not the same as HMRC-compatible filing. | Provider states £20/month, or £49 one-off for lifetime access. No VAT treatment is stated.[43][81] | Solves: a Vinted-specific history and item-cost workflow. Due-diligence boundary: this is not an official API flow. Reliability, security, revocation, platform-authorisation and continuity should be tested directly; the public sources reviewed do not justify a definitive conclusion about contractual permissibility or account enforcement.[81] |
| Dext CommerceE-commerce data and bookkeeping | E-commerce businesses and accounting firms; published coverage spans major marketplaces, stores, payment providers and accounting platforms. | Connects commerce channels and consolidates financial activity for export or synchronisation to accounting software.[89] | Dext says it captures and categorises sales, fees, refunds and supplier costs, giving a consolidated view across connected channels. | Accounting integration supports reconciliation; exact coverage and exception treatment depend on the connected systems. | Supplier-cost capture broadens the source set, but seller-entered evidence and personal/trade judgement remain necessary. | Designed for bookkeeping and accountant workflows; the connected accounting platform, not Dext Commerce alone, provides the relevant HMRC filing functionality. | Commercial subscription/quote structure; confirm the package, channel limits and ledger costs directly. | Solves: more of the evidence-to-ledger chain than a settlement-only connector. Boundary: public product claims are broad and require a real-file trial; unsupported consumer channels and seller-only facts remain outside automated capture. |
| TaxomateE-commerce connector | Smaller online sellers using Amazon, eBay, Shopify, Etsy or Walmart with Xero, QuickBooks Online or Wave. | Imports platform activity and posts summaries to accounting software; plan limits vary by monthly order volume.[90] | Advertises automated sales, fees, refunds and tax categorisation plus COGS support. | Accounting postings are designed to support payout/deposit matching; exceptions still require review. | COGS is supported from supplied cost data; purchase proof and personal/trade classification are not created by the connector. | Ledger hand-off is core. HMRC filing depends on the connected UK-capable accounting stack; Wave is not itself evidence of UK MTD coverage. | Published pricing starts at US$12/month on annual billing and scales with orders and features.[90] | Solves: a lower-cost supported-channel connector option. Boundary: no published Vinted, Depop, TikTok Shop or Facebook Marketplace support in the reviewed pricing material. |
| Link My BooksE-commerce connector | E-commerce businesses/accountants using Xero or QuickBooks Online. Published channels: Amazon, eBay, Etsy, Shopify, TikTok Shop, Walmart US, WooCommerce and Square. | Connects channel accounts and posts settlement summaries/audit attachments into Xero/QBO; historical depth depends on plan. | Imports sales, refunds, fees and taxes; auto-posts settlements, supports payout reconciliation, VAT mappings and channel P&L.[35] | Designed to match marketplace payouts with Xero/QBO bank deposits. | COGS tracking is included but source costs must exist; no published Vinted/Depop support and no personal/trade adjudication. | Strong accountant-user model and ledger hand-off. It supports bookkeeping/VAT files, while the connected Xero or QuickBooks product can provide relevant UK filing functions. | Monthly subscription calculated from rolling monthly orders and channel count, excluding VAT. The public calculator returned £17 to £18/month for one channel at low volume, while the vendor's own July 2026 price-change notice gives a worked example at £36/month. Treat the entry price as unsettled rather than quoting a single figure.[35] | Solves: settlement breakdown, VAT mapping and bank matching for supported channels. Boundary: intentionally a connector into two ledgers, not a receipt archive or source of seller facts. Adopt when supported-channel volume makes manual journals inefficient. |
| A2XE-commerce connector | Sellers/accountants on Amazon, Shopify, eBay, Etsy, Walmart and PayPal; accounting integrations include Xero, QuickBooks, Sage and NetSuite in published material. | Imports settlement/payout data and posts categorised summaries; exports to connected ledger. | Maps sales, fees, refunds, taxes and other settlement lines for deposit matching; multi-channel plans available.[36] | Core proposition is reconciling deposits to settlement summaries in the accounting system. | Optional COGS uses costs entered/uploaded/synchronised by the user; inventory asset and COGS journals do not create missing purchase evidence. | Accountant-focused workflows and major ledger integrations; no direct personal/trade decision or UK Self Assessment/MTD filing. | Channel subscriptions start at US$29/month for the Amazon, Shopify, eBay, Etsy and PayPal connectors; Walmart starts at US$79. Plans rise with order volume, and multi-channel pricing is separate. Sage appears only in an integrations FAQ rather than on the product pages.[36] | Solves: mature settlement accounting and deposit reconciliation. Boundary: supported-channel and ledger dependency is deliberate; Vinted, Depop and evidence capture are outside the published core. Adopt for reliable settlement postings; pair with stock/evidence processes. |
| XeroBookkeeping & accounting | Sole traders and companies; broad small-business ledger with app ecosystem. | Bank feeds, invoices, bills, receipts, manual/CSV entries and connected apps; reports and accountant access. | Records correctly separated sales, fees, refunds and expenses once imported or posted; marketplace detail usually arrives via connector, summary or manual entry. | Strong bank rules, matching and reconciliation.[37] | Inventory capability depends on plan/setup/apps; personal transactions can be marked/reconciled but the software cannot infer trade status. | MTD for VAT and MTD for Income Tax functionality is advertised on eligible plans; accountant collaboration is native. | UK list prices run from £7/month (Simple) through Ignite £16, Grow £37 and Comprehensive £50 to Ultimate £65, excluding VAT. The tiers were renamed from Starter, Standard and Premium. MTD for Income Tax is included on every plan, a deep introductory discount is running, and Xero has published that prices rise from 1 September 2026.[37] | Solves: durable general ledger, bank control, reporting and accountant access. Boundary: general-purpose by design; raw marketplace payouts need mapping or a connector. Adopt when formal books and adviser collaboration justify setup. |
| QuickBooks OnlineBookkeeping & accounting | Sole traders and small companies; tiered cloud ledger. | Bank feeds, receipts, invoices, expenses, CSV/apps and accountant access. | Holds separated marketplace postings; connector or manual import is commonly needed for settlement detail. | Bank matching, categorisation rules and reconciliation are core functions. | Stock/inventory and COGS are available on relevant tiers; cost and personal/trading facts depend on input. | UK MTD for VAT/Income Tax features are advertised; direct accountant access and tax categories. | UK list prices, plus VAT: Sole Trader Plus £10, Simple Start £16, Essentials £38, Plus £56 and Advanced £123. Every tier is marked ready for MTD for Income Tax, and the £10 Sole Trader Plus tier is positioned as the entry product for it. A deep introductory discount is running.[38] | Solves: general books, cash matching, receipt capture and accountant hand-off. Boundary: not a universal marketplace data source. Adopt for complete bookkeeping; additional mapping effort depends on channel mix. |
| Sage AccountingBookkeeping & accounting | UK sole traders and companies, often with accountant relationships. | Bank feeds, sales/purchases, receipt capture on relevant plans, VAT and reports. | Records marketplace components after manual import, journals or integrations. | Bank reconciliation is core. | Stock/inventory is included on the Plus tier; accurate COGS still needs cost and movement data. No personal/trade inference. | MTD-ready VAT/accounting, accountant collaboration and broader Sage ecosystem. | List prices of £20 Start, £43 Standard and £59 Plus per month excluding VAT after a six-month introductory discount. Stock and inventory management is on the Plus tier only. Sage also publishes a separate free Sole Trader and Landlord product that is HMRC-recognised for MTD for Income Tax.[39] | Solves: UK ledger, banking, VAT and adviser workflow. Boundary: channel-native detail is an integration/import problem outside the core ledger. Adopt for established bookkeeping; specialist connectors can reduce settlement work. |
| 123 SheetsMTD bridging | Sole traders/landlords retaining spreadsheets and needing MTD for Income Tax transmission. | Imports/links selected spreadsheet totals and submits to HMRC; optional spreadsheet templates. | Depends on the seller’s spreadsheet totals; does not itself ingest or reconcile marketplace settlements. | No marketplace-to-bank reconciliation. | No native item stock/COGS or personal/trade decision unless built in the source spreadsheet. | Direct bridge to HMRC for quarterly information within published scope; annual affairs may require other compatible software/functionality. | Published price £19.75 per taxpayer per year excluding VAT, which the page states includes a 50% first-year discount, so the standing rate is roughly double. The price covers multiple self-employments and property lets per taxpayer, and an optional bookkeeping spreadsheet is extra. HMRC's finder lists it for quarterly updates and the tax return.[44][82] | Solves: low-cost digital transmission while retaining a spreadsheet. Boundary: a bridge deliberately transmits records; it does not make marketplace records complete. Adopt for a sound existing sheet; reject if source records are the actual problem. |
| MTD SheetsMTD bridging | Spreadsheet users seeking pay-per-submission MTD for Income Tax bridging. | Published support for Excel, CSV, OpenDocument and Numbers-derived files; sends selected totals. | No marketplace ingestion or settlement breakdown. | No native bank reconciliation. | Relies on source spreadsheet for stock, costs and ownership. | Quarterly HMRC transmission only. The vendor states year-end submissions are coming, and HMRC's own software finder independently lists the product for quarterly updates with no tax return, which corroborates the gap.[82] | First submission free, then £2.99 per credit, falling to £0.49 at volume. The published pricing is framed around VAT returns and no separate Income Tax price is stated.[49] | Solves: inexpensive quarterly transmission. Boundary: filing-layer specialism; not bookkeeping or year-end completion. Adopt only when source records and final-return route are separately controlled. |
| ZipSaleUK crosslisting & inventory | UK resellers across eBay, Etsy, Depop, Vinted, Shopify, WooCommerce and others. | Crosslisting, inventory, auto-delisting and item records across connected channels. | Item-level sale records across channels; not a settlement-accounting engine. | Not positioned as a bank-reconciliation ledger. | Item purchase and inventory tracking are core; personal versus trading status still depends on the seller's facts. | Records can be handed to an accountant; no HMRC filing. | From £15/month plus VAT, with per-item credits from £0.18 plus VAT.[83] | Solves: one of the few tools that reaches Vinted alongside the other UK resale channels, with UK pricing and VAT stated. Boundary: listing and inventory operations, not books, VAT or filing. |
| Free and low-cost MTD toolsMy Tax Digital, Pandle, Sage Sole Trader Free | Sole traders and landlords needing HMRC-recognised digital records and submission. | Manual entry, bank feeds on some tiers, bridging from spreadsheets. | No marketplace ingestion. Settlement detail must arrive by import or manual entry. | Bank feeds on paid or specific tiers; no marketplace matching. | Depends entirely on what the user records. | HMRC-recognised for quarterly updates, and for the tax return in some cases. | My Tax Digital is free with no paid tier; Pandle has a free tier with Pro at £5/month plus VAT; Sage publishes a free Sole Trader and Landlord product.[82][84] | Solves: the cost objection to MTD compliance almost entirely. Boundary: none of them connects to a marketplace, so the seller still assembles the figures. Named here because a report that only lists paid ledgers overstates the cost of getting compliant. |
| Excel, Google Sheets & templatesDIY records | Casual sellers through small businesses; any platform if the user designs/imports the model. | Manual entry, CSV import, formulas, pivots and linked sheets; exports to CSV/PDF or bridging software. | Can represent every category and payout if designed correctly; automation and schema maintenance are user-owned. | Possible through formulas and schedules, but matching logic and exception control must be built. | Highly flexible for stock, COGS and personal flags; quality depends on design, evidence and disciplined use. | Universally portable to accountants; can satisfy MTD digital-link rules when linked to compatible software without copy/paste.[11][15] | Google Sheets has a free consumer route; Excel may be owned or subscribed to; templates range from free to paid. | Solves: flexible, transparent, low-entry-cost records. Boundary: error controls, versioning, imports and maintenance are not automatic. Adopt at low complexity; reject or outgrow when duplicated work and review risk exceed the saving. |
| Bookkeepers, accountants & managed servicesProfessional service | Sellers needing recurring books, year-end accounts/returns, VAT or complex advice. | Receive native reports, bank feeds, receipts and explanations; produce reconciliations, ledgers, working papers, accounts and returns within engagement. | Can resolve complex settlements and apply accounting/tax treatment, but cannot reconstruct undisclosed activity or evidence that never existed. | Can perform/review full bank and control-account reconciliation. | Can advise on COGS, stock valuation, mixed use and trade status; seller facts and source proof remain essential. | Strongest route to reviewed accounts and authorised filing; service scope, qualifications and responsibilities vary. | Hourly, fixed-fee, monthly package or transaction-volume pricing; quotes vary materially with clean-up and complexity. | Solves: judgement, review, filing and exception handling. Boundary: human expertise does not remove data collection and can become expensive when inputs are incomplete. Adopt for risk/complexity; clean records reduce repeated queries. |
| Manual seller workflowNo dedicated system | Low-volume sellers or anyone deferring record preparation until year end. | Screenshots, emails, downloaded reports, notes and bank statements assembled ad hoc. | Can be accurate at very low volume if complete and carefully checked; often nets payouts or omits components. | Visual/manual matching, frequently without a documented reconciliation schedule. | Relies on memory for purchase cost and personal/trading distinction; receipts may be scattered. | Handover is usually a folder or unexplained bank total; adviser must rebuild before filing. | No direct software fee; potentially substantial seller time and professional clean-up cost. | Solves: avoids setup cost for genuinely simple circumstances. Boundary: weak scalability, repeatability and traceability. Acceptable for simple, evidenced cases; rapidly fragile with refunds, volume, channels or stock. |
HMRC's finder listed 122 recognised Making Tax Digital for Income Tax products on 1 August 2026. A name-based review found visible landlord, banking, agriculture and trades verticals, but no product explicitly branded for marketplace sellers.[82] Product names do not establish actual capability or demand, and a seller can use a general ledger reached through a connector.
| Vertical | Recognised products |
|---|---|
| Landlord and property | 12 |
| Banking and fintech | 5 |
| Agriculture | 3 |
| Construction and trades | 2 |
| Marketplace and e-commerce sellers | 0 |
Read this carefully. It shows an absence in one list on one date, not that sellers are unserved: a seller can use any general product, and the connector tools feed ledgers perfectly well. It is evidence about specialisation, not about adequacy, and this report proposes nothing on the strength of it.
Best authority for what happened inside one platform. Weakest by design on costs and evidence that occurred elsewhere.
Best at operational margins, SKUs, materials, fulfilment or channel performance. They do not necessarily create statutory books or file UK returns.
Best at translating supported settlements into ledger entries that match bank deposits. They assume supported channels, a destination ledger and correct configuration.
Best at durable books, bank control, accountant access and filing integrations. General ledgers require a clean feed or mapping for marketplace detail.
Best at moving already-prepared digital totals from a sheet to HMRC. Their narrow scope is a feature, not a failure; they do not repair underlying records.
Professionals add judgement and review; spreadsheets add flexibility and visibility. Both depend heavily on the quality and timeliness of seller-supplied source evidence.
Section 04 set out the stages and section 06 set out the categories. This section joins them. It is synthesis of the sourced material above, not a new evidence claim, and it describes intended scope rather than product quality.
This section is the argument in one picture.
Line up every stage of the job against every kind of tool. Nine of the eleven stages have somebody designed to cover them. Two do not.
Those two are the evidence of what you paid for things, and your own final check that nothing is missing. They are unowned because they are yours.
| Category | 1 Activity & intention | 2 Order record | 3 Fees & refunds | 4 Payout | 5 Bank | 6 Stock & expenses | 7 Reconciliation | 8 Categorisation | 9 Seller review | 10 Working papers | 11 Submission | What the seller must still supply |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Native platform reportsOne platform each | Partial System events only | Core | Core | Core | Out of scope | Out of scope | Partial Within the platform | Out of scope | Out of scope | Out of scope | Out of scope | Everything that happened off the platform, plus the intention and ownership behind each item. |
| Profit & inventory toolsOperational analytics | Partial Item tags | Core Supported channels | Core | Partial Often estimated | Out of scope | Core From entered costs | Partial Investigation, not control | Partial Management categories | Out of scope | Out of scope | Out of scope | Accurate purchase costs, receipts, unsupported channels and the statutory books. |
| Marketplace connectorsSettlement to ledger | Out of scope | Partial Summarised | Core | Core | Core Deposit matching | Partial Optional COGS | Core | Partial Mapped rules | Out of scope | Out of scope | Out of scope | A destination ledger, correct mappings, source costs and any unsupported channel. |
| Accounting softwareGeneral ledger | Out of scope | Partial Once imported | Partial Once imported | Partial Once imported | Core | Core If entered | Core | Core | Partial Approval workflow | Partial With adviser access | Core Where advertised | A clean marketplace feed or mapping, and the facts behind every personal or mixed transaction. |
| MTD bridging softwareTransmission layer | Out of scope | Out of scope | Out of scope | Out of scope | Out of scope | Out of scope | Out of scope | Partial Inherited from the sheet | Out of scope | Out of scope | Core Within published scope | A complete, correctly classified source spreadsheet and a route for the annual return. |
| Spreadsheets & templatesDIY records | Depends | Depends | Depends | Depends | Depends | Depends | Depends | Depends | Depends | Partial Feeds the adviser | Partial Via a digital link | Design, discipline, version control and the evidence behind every row. |
| Bookkeepers & accountantsProfessional service | Out of scope Cannot know intention | Partial If supplied | Partial If supplied | Partial If supplied | Core | Partial If evidenced | Core | Core | Out of scope They ask, the seller answers | Core | Core | Complete disclosure of every channel, account and personal fact, plus source documents. |
Counting only the stages a category treats as core scope. Nine of the eleven stages have at least one owner. Stage 1 and stage 9 have none, and both of them are you.
| Lifecycle stage | Categories owning it |
|---|---|
| 1 Activity | 0 |
| 2 Order | 2 |
| 3 Fees | 3 |
| 4 Payout | 2 |
| 5 Bank | 3 |
| 6 Stock | 2 |
| 7 Reconcile | 3 |
| 8 Categorise | 2 |
| 9 Review | 0 |
| 10 Papers | 1 |
| 11 Submit | 3 |
The same eleven stages, read across each category. Nothing here reaches from end to end, and nothing is trying to.
| Category | Core | Partial | Depends | Out of scope |
|---|---|---|---|---|
| Native platform reports | 3 | 1 | 0 | 7 |
| Profit and inventory tools | 3 | 4 | 0 | 4 |
| Marketplace connectors | 4 | 3 | 0 | 4 |
| Accounting software | 5 | 5 | 0 | 1 |
| MTD bridging software | 1 | 1 | 0 | 9 |
| Spreadsheets and templates | 0 | 2 | 9 | 0 |
| Bookkeepers and accountants | 5 | 4 | 0 | 2 |
| Circumstance | Plausible combination | What the combination does not answer | What usually breaks first | Advice trigger |
|---|---|---|---|---|
| Personal-item disposer | Platform history retained, plus proof of original ownership where an item is valuable. | Whether any item was in fact bought to resell, and the original cost of older possessions. | Nothing, unless a high-value item or a profit-seeking purchase is mixed into the same account. | A single disposal above the chattels threshold, or repeated sourcing.[1] |
| Casual seller who has started buying to sell | Platform exports and a simple sheet with an item-cost column. | Whether the trading allowance applies, and whether the pattern has already become a trade. | Stock cost, because the sourcing habit predates any decision to keep records. | Gross trading income near or above £1,000, or unclear trade status.[3][4] |
| Single-channel reseller | Native reports, a designed spreadsheet, an annual accountant. | Payout-to-bank differences across period ends, and closing stock if traditional accounting applies. | Reconciliation at year end, once refunds and reserves cross the boundary. | First VAT question, first bad reconciliation, first year with unsold stock.[50] |
| Multi-channel reseller | Native reports plus a connector into a ledger, with an inventory or profit tool alongside. | Channels the connector does not support, cash purchases, and whether personal items entered the trading account. | The unsupported channel, which quietly reverts to manual work and breaks the audit trail. | VAT registration, margin scheme eligibility, cross-border activity or stock valuation.[16][18] |
| Handmade maker | Channel exports plus a materials and production tool, feeding a ledger. | Wastage, work in progress and the boundary between a hobby and a trade in the early years. | Bill-of-materials accuracy, because materials are bought in bulk and used across many items. | Stock valuation, fairs and off-platform sales, or a first employee.[5] |
| Partnership | A shared ledger with defined partner capital and current accounts. | How partnership totals reconcile to each partner's individual return. | Partner-funded purchases and drawings recorded as business costs. | Any change in profit share, a new partner, or a marketplace account held in one partner's name.[9] |
| Limited company | Business bank account, full ledger, connector and a continuing accountant relationship. | Whether every transaction belongs to the company rather than the director personally. | Separate-entity discipline, especially where a personal marketplace account predates the company. | Director transactions, stock valuation, cut-off, payroll or a platform account in the wrong name.[7][8] |
The residual problem is a coordination problem across commercial systems, seller knowledge, documentary evidence, accounting rules and tax filing. Automation can reduce transformations; it cannot manufacture missing facts.
The problem survives because no single party sees the whole thing.
The platform knows its own orders. The bank knows the cash. The shop you bought stock from knows what you paid. You know why you bought it. An accountant arrives last and can only work with what reaches them.
Automation can move information faster. It cannot invent a receipt that was never kept, or a reason that was never written down.
| Structural cause | Why it exists | What incumbents solve | What remains unresolved | Who can resolve the residual |
|---|---|---|---|---|
| Fragmented data ownership | The platform owns order/settlement events; banks own cash; carriers own postage; suppliers own purchase documents; the seller knows purpose and ownership. | Connectors join supported platform settlements to ledgers; accounting software joins bank and books. | External purchases, cash costs, unsupported channels and personal facts do not arrive merely because an API is connected. | Seller supplies complete sources; software preserves links; adviser reviews treatment. |
| Inconsistent schemas | Each platform uses its own commercial model, field names, signs, tax representation, identifiers and export windows. | Native reports faithfully describe one platform; connectors normalise a defined set of formats. | Format changes, long-tail platforms, regional variants and seller-edited files require maintenance and exception handling. | Platform documentation, integration provider and reviewer. |
| Several clocks | Order, dispatch, refund, statement, payout and bank-posting dates serve different operational purposes. | Settlement tools reconcile batches; ledgers handle timing and clearing accounts. | Tax-year and quarter cut-off, late refunds, reserves and cross-currency settlement still require consistent policy and roll-forwards. | Bookkeeper/accountant with complete source data. |
| Insufficient seller knowledge | Platforms make commercial steps simple without teaching double entry, accounting basis, VAT or tax classification. | Wizards, dashboards and managed services simplify categories and postings. | A seller may not recognise that payout is net cash, a reserve is not a cost, or a reported personal disposal is not automatically taxable. | Plain guidance plus adviser where fact-sensitive. |
| Stock and receipt gaps | Goods may be sourced in cash, in bundles, long before sale, from personal collections or without itemised receipts. | Inventory tools allocate known costs and track movement; receipt apps preserve captured evidence. | A mathematically precise allocation cannot prove an undocumented purchase or recover an unknown original cost. | Seller at acquisition; adviser judges acceptable evidence/treatment later. |
| Personal and business mixing | The same phone, account, marketplace balance and bank may be used for both. | Ledgers can mark transactions personal; specialist trackers can add item tags. | Intent at acquisition and ownership are not reliably inferable from price, category or frequency alone.[1][4] | Seller supplies facts; adviser determines uncertain tax treatment. |
| Cost, complexity and setup | Accurate integration needs channel connections, chart mappings, VAT configuration, opening balances, stock costs and review. | Sheets minimise cash cost; connectors minimise repeated settlement work; services absorb setup effort. | The lowest-cost route shifts work to the seller; the lowest-effort route costs more. Very small sellers may rationally accept manual work. | Seller chooses based on actual complexity and risk; no universal optimum follows from the evidence. |
| Trust, privacy and access | Automatic tools may request marketplace tokens, APIs, bank feeds or years of commercial data. | Major ledgers/connectors publish security and data-processing controls; CSV workflows can reduce live access. | Sellers must assess provider trust, access durability, data retention and the risk of credentials or platform-policy changes. Public feature pages do not fully answer each user’s risk tolerance. | Seller/business owner with security, privacy and contractual due diligence. |
| Bookkeeping versus tax | A ledger records economic activity; MTD quarters transmit summaries; annual returns apply tax adjustments and combine wider circumstances. | Accounting software and bridges can support defined filing steps; accountants prepare/review final positions. | A successful import or quarterly submission does not validate trade status, expense allowability, VAT, stock or the annual return.[12][13] | Taxpayer and authorised adviser under the applicable regime. |
| Incomplete adviser hand-off | Accountants enter after source records have already fragmented or been discarded. | Checklists and working-paper processes identify missing files and differences. | Repeated queries persist until the seller supplies facts/evidence or the adviser documents an estimate, adjustment or limitation. Draft ICAEW MTD engagement material expressly places source-information duties on the client.[51] | Seller responds; adviser assesses, documents and files within scope. |
Authoritative for the platform’s orders, charges, balances and payouts. It may use calendar-year reporting, settlement dates or platform-specific definitions and cannot see external costs.
Usually starts from the bank. This is observable and useful for reconciliation, but net cash omits gross activity and can mix trading, personal, reserve and transfer movements.
Separates revenue, costs, assets, liabilities, personal movements and timing under an accounting basis. It requires evidence and mappings beyond the bank feed.
Applies entity-specific rules, allowances, disallowable amounts, VAT and other return information. It is downstream of good records, not supplied by a platform payout total.
The surrounding figures establish exposure to online selling and general compliance burden. They do not measure the records problem, affected customers or commercial demand.
How common is this? The honest answer is that nobody has counted it directly, and you should distrust anyone who quotes a precise figure.
HMRC receives millions of platform reports, but those are not distinct UK marketplace sellers. Whole-economy tax-gap and administration-cost data concern every type of taxpayer and business.
These figures are useful context only. They cannot show that a marketplace seller has bad records, feels pain, will buy a service or can be reached economically.
Reports covering calendar 2025 named nearly four million sellers and £55bn of income, up from under 1.5 million and £25.5bn a year earlier. These figures come from HMRC data released under a freedom of information request, not from a published statistical series.[66]
| Calendar year | Sellers reported | Income reported |
|---|---|---|
| 2024 | 1,466,171 | £25.5bn |
| 2025 | 3,988,892 | £55bn |
Three caveats that matter more than the number. It counts platform reports, not distinct people, so a seller active on three platforms appears three times. It mixes individuals with companies, partnerships, trusts and charities. Goods sellers who meet both exclusions · fewer than 30 sales and no more than the €2,000 consideration threshold · need not be reported under the rules, so the smallest sellers are not comprehensively represented. It is not a count of UK marketplace sellers and must not be used as one.[66][2]
HMRC puts the 2024/25 tax gap at 6.4%, or £59.2bn, and attributes 62% of it to small businesses. Split by behaviour, failure to take reasonable care and error together account for just over half, against 12% for evasion.[65]
| Behaviour | Share of the tax gap |
|---|---|
| Failure to take reasonable care | 35% |
| Error | 16% |
| Evasion | 12% |
| All other behaviours | 37% |
| Statement | Figure | Status | How far it can be pushed |
|---|---|---|---|
| Sellers named in platform reports for 2025 | 3,988,892 | Official, released under FOI | Evidence of reporting volume. Not a seller count, not a compliance measure.[66] |
| Income covered by those reports | £55bn | Official, released under FOI | Same caveats. Gross consideration after specified platform fees, on a calendar year.[66][2] |
| UK tax gap, 2024/25 | £59.2bn (6.4%) | Official statistic | Whole-economy. Small businesses are 62% of it.[65] |
| Share attributed to failure to take reasonable care and error | 51% | Official statistic | Whole-economy behavioural context only. It does not support a claim about marketplace-seller records, intent or demand.[65] |
| UK marketplace businesses HMRC believes are not meeting VAT obligations | “Tens of thousands” | Official estimate, no methodology published | Quote as HMRC's stated estimate in a consultation, with the cost given as hundreds of millions a year.[61] |
| Internet sales as a share of UK retail | ≈28% | Official statistic | Context for the size of the channel, nothing more.[72] |
| Annual cost of tax administration to business | £15.4bn | Official, described by the NAO as likely an understatement | Whole-economy compliance cost, not seller-specific.[73] |
| Small businesses reporting a worsening compliance burden | 57% | Survey, self-selecting sample | Directional only. The Tell ABAB survey is open and self-selecting, so respondents skew towards the engaged and the aggrieved.[67] |
| Proportion of marketplace sellers with incomplete or unreconciled records | Unknown | No source exists | Nothing. Treat any published figure claiming to measure this as unsupported. |
The mechanism and difficult segments can be identified from documents. Pain intensity, purchase urgency, reachability and retention require customer behaviour, not more desk research.
Do not build “accounting for every seller”. Most casual sellers do not need it, and many serious sellers already have workable native reports, a connector, a ledger and an adviser.
Test one hard hand-off for a defined business seller: unsupported channel data, stock-cost evidence gathered when goods are bought, or an exception pack an accountant will actually accept. Charge for the manual outcome before automating it.
| Opportunity area | Best initial customer | Incumbent pressure | Unresolved value | Commercial test | Assessment |
|---|---|---|---|---|---|
| Another supported-channel settlement connector | Amazon, eBay, Etsy or Shopify seller already using Xero or QuickBooks | High: Link My Books, A2X, Dext Commerce, Taxomate and ledger app ecosystems. | Only a material gain in price, accuracy, workflow or service would justify switching. | Win paid migrations from an incumbent on a defined superiority claim. | Deprioritise |
| Unsupported-channel reconciliation and evidence pack | Serious Vinted, Depop or Facebook/local seller using mixed payment routes | Medium: some specialist and cross-listing tools exist; official access and coverage vary. | Repeatable order/balance/payout/bank lineage, gap flags and durable source copies. | Reconcile real record packs manually; measure exceptions, time saved and paid renewal intent. | Validate first |
| Stock-cost and purchase-evidence capture | Stock-based reseller, especially second-hand or VAT margin-scheme candidate | Medium: inventory tools track cost, while receipt tools store evidence; the acquisition-time join is less consistently covered. | Item identity, seller/source, purchase proof, bundles and later sale match without year-end reconstruction. | Run an acquisition-to-sale pilot and obtain adviser confirmation that the output is usable. | Promising hypothesis |
| Accountant-ready exception and working-paper hand-off | Bookkeeping or accounting firm with several marketplace clients | Medium to high: connector and practice platforms already serve firms. | A cross-channel completeness statement, unresolved-item schedule and traceable seller responses. | Ask firms to use and pay for the pack across several live clients; measure query reduction. | Validate B2B route |
| Managed clean-up or ongoing bookkeeping service | High-complexity seller who values an outcome over software | High: bookkeepers, accountants and e-commerce specialists already sell outcomes. | Potentially urgent one-off pain and rich discovery data, but labour and liability are substantial. | Sell a tightly scoped fixed-fee clean-up after clarifying supervision, engagement and liability boundaries. | Test cautiously |
Official platform and HMRC documentation establishes the gross-to-net, evidence and reconciliation mechanism. Defined high-complexity segments have more failure points and higher potential consequences.
Report counts, tax-gap data and software-list gaps do not show purchase urgency, budget, acquisition cost or retention. No market-size or willingness-to-pay claim is made.
A provider recording, reviewing or reporting financial information for clients may fall within UK accountancy service provider supervision and anti-money-laundering obligations. Obtain professional advice before offering bookkeeping, accounts or tax services; a self-service software tool is a different risk profile.[92]
| Evidence needed | Minimum practical test | Decision signal |
|---|---|---|
| Repeated pain in the target segment | 15–25 structured seller interviews and 8–12 accountant/bookkeeper interviews, recruited beyond personal networks. | The same job, trigger and current workaround recur without leading respondents. |
| Problem visible in real data | With consent, inspect 10–15 anonymised record packs covering at least two periods and several relevant channels. | Exceptions are common enough and standard enough to solve repeatably. |
| Willingness to pay | Sell 5–10 manual, tightly scoped pilots or take deposits; do not rely on hypothetical price questions. | Customers pay from a real budget and at least some ask to repeat or continue. |
| Adviser acceptance | Have qualified advisers review reconciliations, source links and exception schedules before they are used. | Outputs reduce queries or preparation work without shifting hidden liability. |
| Durable access | Test current logged-in exports, APIs, consent, rate limits and historical depth for each chosen platform. | The workflow does not depend on fragile access or unverified absence claims. |
| Reachable economics | Test two acquisition routes and measure booked calls, conversion, fulfilment time and gross margin. | Acquisition and service effort leave credible room for a sustainable price. |
The evidence supports the existence and mechanism of the problem. It does not support a fabricated count of affected sellers, average hours lost, error rate or willingness to pay.
The problem is real, it is not the same for everyone, and it is worst where somebody is genuinely running a business through a net payout.
A clear personal-item disposal is normally a low-complexity records case. A stock-based trader across several apps, providers and fulfilment routes has a materially harder job · but a capable incumbent stack may already solve most of it.
Few sales, clear personal ownership and retained high-value evidence can make records simple. Platform reporting does not turn this into a trade.
One channel and disciplined receipts may be manageable in a sheet, but gross/net and stock-cost distinctions still matter.
Volume, refunds, ads, fulfilment, several providers, stock and VAT compound both matching work and the consequences of missing evidence.
No source counts sellers with incomplete or unreconciled records. Platform-report volumes and whole-economy tax-gap figures are contextual and do not establish incidence.
| Question | Verdict | Confidence and boundary |
|---|---|---|
| Is the problem real? | Yes. | High confidence in the records mechanism; not every seller experiences material pain. |
| How widespread and severe is it? | Widespread exposure; prevalence unknown; severity uneven. | High-complexity traders have more failure points, but UK incidence and hours lost are unmeasured. |
| Who experiences it most? | Likely stock-based multi-channel resellers, VAT/margin-sensitive sellers, makers with materials and entities with mixed accounts. | A reasoned segment hypothesis, not survey evidence. |
| Underserved or inconvenient? | Underserved in selected hand-offs; merely inconvenient or already solved elsewhere. | Unsupported channels, acquisition evidence and exceptions are credible gaps. Basic sales export or filing alone is crowded. |
| Are incumbents adequate for most users? | Adequate for many casual and supported-channel users. | Manual records can be enough at low complexity; connectors plus Xero, QuickBooks or Sage and an adviser can be strong at higher complexity. |
| Credible gaps worth commercial investigation? | Yes, narrowly. | Cross-channel exception lineage, unsupported payment flows, acquisition-time stock proof and adviser hand-off merit tests. |
| Enough evidence for discovery or paid validation? | Yes. | The mechanism is clear enough to recruit precisely and test a manual outcome. |
| Enough evidence to build a full business? | No. | Demand, pricing, acquisition, retention, access durability and service boundaries remain open. |
| Recommendation | VALIDATE FURTHER · NARROW AND TEST. | Do not reject the problem and do not fund a broad build. Stop or reframe if real-file paid pilots do not show recurring value. |
Every material legal, platform, feature and price claim links to a directly relevant source. The report intentionally does not infer market demand from the existence of sellers, MTD scope or competitor products.
This report only says things it can point at.
Where the law is involved, it links to HMRC. Where a platform or a product is described, it links to that company’s own page. Where something is our reading rather than a fact, it says so.
Where we do not know, the report says we do not know instead of guessing.
HMRC/GOV.UK for law and filing expectations; official platform help for data and fees; official software/pricing pages for incumbent capabilities.
AAT/ICAEW material and UK accountancy-firm checklists were used for common hand-over practice, clearly distinguished from legal requirements.
Fact is directly sourced. Practice describes adviser workflow. Interpretation is reasoned synthesis. Open is not established.
Provider feature and price statements are self-descriptions, not independent proof of accuracy, reliability, security, user adoption or accountant acceptance.
No market size, affected-seller count, pain frequency, time saving or demand estimate is presented. Qualitative transaction bands are operating examples only.
Desk research only: no seller interviews, accountant file testing, live account access, API test or independent security review. Dynamic product pages and platform features can change.
Every legal, platform and vendor claim carried over from v1 was re-checked against the primary source on 1 August 2026. Corrections are listed in section 14 rather than made silently.
Three items were on a clock at the time of writing: Amazon withdraws its older settlement reports on 11 November 2026, Xero has published a price rise from 1 September 2026, and the marketplace VAT consultation closes on 18 August 2026.
92 sources. Where the law is involved the report cites HMRC directly. Vendor pages are used only for what a company says about its own product, never as proof that the product works.
| Source type | Count | Share |
|---|---|---|
| HMRC and GOV.UK guidance | 39 | 42% |
| Marketplace documentation | 24 | 26% |
| Vendor product and pricing pages | 18 | 20% |
| Official statistics and scrutiny bodies | 4 | 4% |
| Professional bodies | 4 | 4% |
| Accountancy firm guidance | 3 | 3% |
| Open question | Why it matters | What would resolve it | Status |
|---|---|---|---|
| Does Vinted publish a standard finance export? | Vinted is a major UK route for both personal disposals and resale, and a missing export shapes the whole record-keeping route for those sellers. | A logged-in account review, or first-party UK help that names the file. | Not documented publicly; logged-in test required |
| How many UK sellers actually have incomplete or unreconciled records? | Without it, severity can be described but prevalence cannot. It is the single largest gap in this report. | Primary research with sellers and agents. The surrounding official evidence is now set out in section 10, but it does not measure this. | Still no direct measure |
| How do marketplace VAT rules change the seller's own records? | Where a marketplace accounts for VAT on certain supplies, the seller's turnover, VAT return and reconciliation all look different from a straightforward domestic sale. | A dedicated pass over HMRC VAT guidance for online marketplaces, plus platform VAT documentation. | Addressed in v3, section 06 |
| How do imports, overseas stock and foreign currency affect the chain? | Sourcing from abroad adds duty, import VAT, exchange differences and timing that none of the lifecycle stages above address. | A separate cross-border edition with its own sources. | Summarised in v3, section 06 |
| Do accountants accept specialist tool output without rework? | A tool that produces a report an adviser will not rely on has not removed the work, only moved it. | Interviews or file testing with practising UK accountants across several tools. | Not tested |
| How durable are unofficial platform connections? | Where a tool connects to a platform without a published partner API, access can end without notice and take the record history with it. | Provider disclosure, platform policy review, or observed behaviour over time. | Not tested |
| Do MTD-compatible products handle marketplace-shaped data well? | Compatibility with HMRC is not the same as handling settlements, reserves, refunds and multi-channel stock. | Hands-on testing of compatible products against real settlement files. | Partially addressed |
| What is Link My Books’ actual entry price? | A price that cannot be pinned down is a warning about the whole pricing layer of this market, which moves faster than any report can track. | A quote for a specific order volume, or a stable published rate card. | Vendor calculator and vendor price notice disagree |
| Does the absence of a marketplace-seller vertical in HMRC-recognised software reflect absent demand or an unserved need? | It is the difference between a gap worth filling and a gap that exists for good reason. This report deliberately does not answer it. | Demand research with sellers and agents, which this report does not attempt. | Open by design |
| Does the platform report reconcile to the seller's own figures in practice? | The report uses a calendar year and platform fee definitions, so a mismatch with the seller's UK tax-year records is expected rather than exceptional. | Comparison of received platform reports against prepared records for the same sellers. | Not tested |
Version 4.0 critically re-tested the report’s conclusions, corrected overclaims and added a commercial feasibility decision. Earlier revision history is retained and explicitly marked where a v3 conclusion has been superseded.
Recommendation changed to validate further. The report now separates a proven records mechanism from unproven prevalence, purchase urgency and willingness to pay. It supports narrow paid discovery, not a full build.
Added HMRC’s retailer election for one daily gross-takings digital record, including third-party online-platform sales, and the below-VAT-threshold category easement. The prior blanket transaction-level statement was incomplete.
Replaced withdrawn VAT Notice 718, removed the false universal “roughly sixfold” margin-scheme impact and narrowed cross-border claims that were not adequately sourced.
Whole-economy tax-gap data is now contextual only and is not presented as evidence of marketplace record failure. Platform-report volumes are not used as a distinct seller count or demand proxy.
The Vinted export conclusion is reopened: public first-party material did not document a standard finance export, but absence from public help is not proof that no logged-in or data-access route exists.
Added Facebook Marketplace/Commerce Manager, PayPal, Stripe, Dext Commerce and Taxomate. Corrected the claim that marketplace connectors and MTD-capable ledgers are separate markets with no overlap.
Added opportunity hypotheses, incumbent pressure, minimum paid-validation evidence and the accountancy-service-provider supervision boundary.
Independently re-verified. Four parallel research passes re-checked every legal, platform and vendor claim against the primary source. All ten of the core HMRC claims held. Several platform and product claims did not, and the corrections are listed below rather than made silently.
The cash basis is the default for sole traders and partnerships from 2024/25, with accruals accounting now an election. v1 and v2 implied stock and cost-of-goods matching was the norm. Under the cash basis, resale goods are deducted when paid for and there is no closing-stock adjustment, which makes a stock-heavy reseller’s profit lumpy.
Chattels wording corrected to HMRC’s “£6,000 or more”. Digital-link prohibition given its proper timing qualifier.
Depop removed UK selling fees on 20 March 2025 and introduced a buyer fee. v1 described the old model. Its CSV still carries a now-zero “Selling” column, so older mapping logic mis-states cost of sales.
eBay documents CSV only, not XLSX. eBay UK raised the per-order business fee on 12 February 2026. Amazon reduced several UK fees on 5 January 2026 and withdraws its older settlement reports on 11 November 2026. Shopify’s middle tier is now named Grow.
Craftybase is now Stocksmith, and its advertised entry price has moved from around US$20 to US$83. Veeqo is no longer free for inventory: the free tier is shipping only. Xero renamed its tiers and has published a price rise from 1 September 2026. 123 Sheets’ £19.75 is a first-year discounted rate.
Vinta reaches Vinted through a Chrome extension that copies session tokens, not an official API. That is now described as a durability, terms-of-service and security consideration rather than a footnote.
v3 stated: “Vinted publishes no finance export.” Open in v1 and v2, was checked across the UK help centre, the Pro guide and the DAC7 pages and described as answered definitively. Pro sellers get invoices for what Vinted sells them; over-threshold sellers get a DAC7 copy. Neither is a sales export. v4 withdraws the proof-of-absence conclusion pending logged-in verification.
VAT and marketplace selling, previously declared out of scope. Covers when a marketplace is the deemed supplier and when it is not, where each part of a sale belongs once registered, the second-hand margin scheme and its stock book, and a summary of cross-border consequences.
It also flags the live HMRC consultation, opened 23 June 2026 and closing 18 August 2026, on extending marketplace VAT liability to UK-established sellers.
How big is it. v1 and v2 said the problem was not quantified and stopped. That was too strong. The direct measure still does not exist, but the surrounding official evidence does, and each figure is now stated with its status and its limits.
HMRC’s recognised software list carries purpose-built products for landlords, farmers, tradespeople and banks, and none for marketplace sellers. Separately, every marketplace-connected tool stops short of HMRC and every HMRC-recognised tool starts at the bank. A seller crossing the digital threshold needs one from each set and bridges them personally.
This was stated as an observed absence on one date. v4 corrects it: connector-to-ledger ecosystems span the layers, even though unsupported sources and seller-only facts remain.
Every section now opens with a plain-language summary, switchable off from the header. Section titles were rewritten as questions a seller would actually ask.
A settlement calculator with live sliders, a timetable checker, and a circumstance picker. Nine figures: a hand-built SVG flow diagram and eight charts.
Mermaid has been removed entirely. All diagrams and charts are now inline SVG with no external library, so the document renders identically offline, in print and behind a strict content policy. Every figure carries a data table or written equivalent.
Two further incumbent rows: ZipSale, a UK tool that reaches Vinted, and the free or near-free HMRC-recognised filing products, because a report listing only paid ledgers overstates the cost of getting compliant.
Sources expanded from 52 to 84.
The v2 changes are kept below for the record.
Coverage matrix. Joins the eleven lifecycle stages from section 04 to the incumbent categories in section 06, and shows that stage 1 (intention and ownership) and stage 9 (seller confirmation) have no incumbent owner at all.
A second table sets out what is left over for the seller under a plausible tool combination, by circumstance. Both are labelled interpretation and neither is a recommendation.
Open questions register. The unresolved items scattered through version 1.0 are now a single table with, for each question, why it matters and what evidence would settle it.
Two subjects are named as deliberately out of scope rather than left silent: marketplace VAT treatment, and imports, overseas stock and foreign currency.
Four sources were listed in version 1.0 but never cited in the body. Sources [26], [45], [48] and [52] are now attached to the Vinted row, the good-bookkeeping layer, the definition of profit and the platform-reporting card respectively.
Column headers now carry an explicit scope. Every footnote link announces which source it points to. Filter buttons expose their pressed state and a live region reports how many rows are visible.
A jumped-to source is highlighted on arrival, and the diagram falls back to its written equivalent if the diagram library fails to load.
The top navigation is hidden below 1060px, which left narrow screens with no way to move between sections. A collapsible contents list now covers that case, and the navigation order matches the document order.
Printed source entries now show their full URLs, so a paper copy remains checkable. The web font loads without blocking the first render.
No source was added, removed or re-checked. The research cut-off remains 31 July 2026, and prices and product features should be re-verified before reliance.
No demand estimate, market size, seller count or willingness-to-pay figure has been introduced, and no solution is proposed.
