Accountants for
ecommerce startups

UK ecommerce businesses now face one of the more complex VAT positions of any startup vertical, driven almost entirely by the post-Brexit regime. Sales to UK consumers carry standard UK VAT at 20 per cent (or 5 per cent or zero-rated by product category) and flow into the normal UK VAT return. Sales to EU consumers from UK stock can be filed through the One Stop Shop (OSS) register, which lets a UK business file a single quarterly return covering all EU consumer sales with the VAT applied at the destination member state rate rather than registering in each country separately. Sales of goods from outside the EU to EU consumers with a consignment value under €150 can use the Import One Stop Shop (IOSS) scheme, which also simplifies the VAT collection but applies destination-country rates on each consignment.

Online marketplaces add a further layer through the Online Marketplace (OMP) deemed supplier rules. For imports into the UK of consignments with a value of £135 or less sold through an OMP (Amazon, eBay, Etsy, and similar), the OMP is deemed to be the supplier for VAT purposes and collects VAT on the sale. The seller’s supply to the OMP is zero-rated. For a UK ecommerce startup selling imported low-value goods through Amazon UK, this changes who remits the VAT and requires clear accounting separation between OMP-handled sales (where VAT is not the seller’s responsibility to remit) and direct DTC sales (where it is).

The practical accounting requirement is reconciliation at each platform. Amazon Seller Central reports, Shopify tax reports, eBay transaction reports, and TikTok Shop settlement reports each have their own structure and conventions, and the VAT return needs to be built from reconciled platform data rather than aggregated sales totals. Monthly reconciliation between platform VAT reports, the general ledger VAT accounts, and the actual bank settlement is the control that prevents quarterly VAT returns being filed from a rolling set of platform exports that do not tie back to a single source of truth.

Stock for an ecommerce startup is valued at cost, and the cost that matters for gross margin reporting is landed cost rather than the invoice price from the supplier. Landed cost comprises the wholesale unit cost, plus inbound freight allocated per unit, plus customs duty and import VAT (where import VAT is not recoverable, which is uncommon but possible), plus broker and clearing fees, plus any inbound inspection or testing cost. Landed cost calculated properly is typically 10 to 30 per cent higher than the wholesale unit cost for a UK brand importing from Asia, and the gap directly reduces the reported gross margin. Ecommerce startups that report gross margin based on wholesale cost alone typically find at the first year-end that the actual gross margin is three to eight percentage points lower than the dashboards suggested.

The cost flow assumption (FIFO or weighted average) needs to be set at the first accounting period and applied consistently. Under FRS 102 both are permitted, with FIFO typically preferred for inventory that is physically identifiable by lot or batch (apparel with seasonal cuts, cosmetics with expiry dates, electronics with serial numbers) and weighted average typically applied for fungible inventory (commodities, raw materials, generic components). The choice affects reported cost of sales and gross margin in periods of changing wholesale prices, and the consistency requirement means a change in method later requires restatement.

Multi-location stock adds a layer. Amazon FBA holds inventory across multiple fulfilment centres, often on two or three continents under FBA Europe and FBA North America. Third-party logistics (3PL) warehouses hold their own sub-allocations. Own-warehouse inventory sits separately. Each location has its own inventory balance, and the monthly inventory reconciliation needs to tie the location-level quantities to the general ledger balance and to the Amazon Seller Central inventory reports. Reconciling differences is where shrinkage, damage, unrecorded transfers, and FBA inbound discrepancies show up, and accountants in our network build the reconciliation process at the first FBA shipment rather than at the first year-end stock take.

Ecommerce return rates vary substantially by product category. Apparel and footwear routinely see 30 to 40 per cent return rates in the UK market, consumer electronics 15 to 25 per cent, home goods 10 to 15 per cent, and consumables under 5 per cent. Under IFRS 15, the return right is a form of variable consideration, and the transaction price for revenue recognition needs to exclude the portion expected to be returned. This means that recognised revenue in the period of sale is already net of an estimated return provision, and the corresponding returned inventory is recognised as an asset at expected realisable value rather than at full cost.

The practical calculation is that a startup selling £1 million of apparel in a month with a 35 per cent expected return rate recognises £650,000 of revenue (with a liability for expected refunds and an asset for expected returned inventory), not £1 million of revenue and a reversing entry when returns come through in the following months. The provision is estimated by category and channel based on historic data where available, with a monthly true-up to actual returns. Provisions that are set up incorrectly (blended rate across categories, or recognised at the point of return rather than at the point of sale) produce month-to-month revenue that moves with the return timing rather than the sales performance, which obscures the actual commercial trajectory for both management and investors.

Chargebacks and fraud losses are separately provisioned, typically at a smaller percentage of gross sales (0.1 to 1 per cent depending on channel and category), recognised as a deduction from revenue under IFRS 15 where the chargeback represents a reversal of the sale. Refund fraud and friendly fraud at scale on Amazon and Shopify is a specific exposure that the provision needs to cover if the historic data supports it.

Marketplace fees are the single largest line of cost for an Amazon-led ecommerce business after cost of goods, and the fee structure varies by channel in ways that change the comparative economics of each channel. Amazon FBA typically carries three cost layers: a referral fee (a percentage of the sale price, ranging from 5 to 15 per cent depending on category), FBA fulfilment fees (a per-unit fee based on size and weight), and FBA storage fees (a monthly fee per cubic metre, with additional long-term storage fees after six and twelve months). Amazon Vendor Central has a different cost structure based on wholesale pricing and co-op marketing contributions. Shopify direct DTC carries payment processing (typically 1.5 to 2.5 per cent plus fixed fee), app subscriptions, and variable direct marketing costs but no platform commission. eBay carries a final value fee and insertion fees. TikTok Shop and Etsy carry their own fee structures.

The channel margin calculation needs to strip fees and channel-specific costs from revenue before comparing channel profitability. An SKU that shows a 55 per cent gross margin at blended level may carry a 40 per cent margin on Amazon FBA (after referral and FBA fees) and a 62 per cent margin on Shopify direct DTC (after payment processing but before marketing spend). Channel-level margin reporting is the single most valuable output of monthly management accounts for an ecommerce startup, because strategic decisions (price changes, promotion spend, channel expansion or contraction, and advertising budget allocation) all depend on it. Accountants in our network build the channel-margin reporting from the first month of trading rather than assembling it for the first investor update.

R&D tax credits for ecommerce startups are narrower than founders often expect under the merged scheme. Standard ecommerce platform work does not qualify: Shopify theme development and customisation, standard app integration, typical Amazon listing optimisation, product photography and content, and routine operational software do not meet the qualifying R&D bar. First-time ecommerce claims built around those activities are routinely rejected on enquiry under the post-2023 HMRC position.

The qualifying work that does exist on ecommerce products is specific. Novel product development involving material science, chemistry, biotechnology, or engineering can qualify where the work resolves scientific or technological uncertainty not deducible from existing approaches. Proprietary inventory optimisation, demand forecasting, or dynamic pricing systems that go beyond commercially available tools can qualify where the work meets the technological uncertainty bar. Specialist packaging or logistics technology (for example, novel cold-chain solutions, new recyclable materials, new fulfilment automation) can qualify where it addresses genuine scientific or technological uncertainty. For the majority of ecommerce startups, the honest answer is that qualifying R&D is modest or absent, and the SEIS or EIS reliefs are the more important tax planning focus.

For consumer product businesses genuinely developing new formulations, materials, or proprietary technology, specialist documentation matters. The technical narrative needs to identify the specific scientific or technological uncertainty, the existing approaches that did not resolve it, and the work undertaken to advance the state of knowledge in the field. Cost allocation between qualifying R&D and routine commercial development is a significant exercise in an ecommerce context where most of the engineering and creative spend is routine. Accountants in our network who work with ecommerce startups will often recommend against a marginal R&D claim rather than pursue one that invites enquiry without strong evidence.

Ecommerce working capital is dominated by inventory. A growing ecommerce startup typically holds three to six months of forward sales as stock (split across FBA inbound, FBA on-hand, 3PL, and own warehouse), which ties up substantial cash compared to a pure SaaS or services business. Supplier payment terms on import purchases are typically 30 to 90 days, sometimes with letter of credit or factored arrangements, and the terms vary by supplier and by the negotiating position as the startup scales. Amazon settlement cycles run on a two-week basis for most sellers, meaning revenue earned on Amazon today settles to the bank account approximately two weeks later, with further delays on new seller accounts or during trust-building holds. Shopify direct DTC typically settles daily via the payment processor, which is operationally simpler but produces different cash flow timing.

The working capital calculation for an ecommerce business is effectively: inventory value plus receivables from marketplaces minus payables to suppliers, and the number is almost always larger than founders expect. A brand doing £5 million in revenue at 50 per cent gross margin typically has £400,000 to £800,000 of working capital tied up, which scales directly with growth. Cash flow forecasting needs to model the working capital expansion as a use of cash, not treat it as neutral, and the 13-week rolling cash flow model needs specific ecommerce features: inventory reorder schedule, Amazon settlement timing, supplier payment schedule, and channel-specific refund lag.

Ecommerce exits carry specific diligence topics that do not appear in SaaS, tech, or fintech exits. Amazon account dependency is the first diligence item for any Amazon-led business. An acquirer reviewing a startup with 80 per cent Amazon FBA revenue will discount the valuation for the platform risk unless there is a clear plan for channel diversification, and the discount is typically material. The accounting function supports the diversification narrative by reporting channel mix, channel growth rates, and channel margin trajectory over time, not just the current-period blend. Brand IP ownership and trademark registration across jurisdictions (UK, EU, US, Canada, Australia as a minimum for a UK brand with international aspiration) is a standard diligence item, and incomplete trademark coverage reduces the acquirer confidence in the brand as an ongoing asset.

Supplier relationships and concentration are the third diligence area. An ecommerce business with 70 per cent of gross margin coming from a single supplier or factory carries single-point-of-failure risk that affects valuation. Supplier agreements, minimum order commitments, exclusivity arrangements, and IP assignment from supplier-designed elements all surface at diligence. Inventory reconciliation through the acquisition process (closing stock take, third-party inventory audit) is standard, and the accuracy of the ongoing inventory records through the growth period affects how smoothly the acquisition diligence runs. Accountants in our network who work with ecommerce startups treat the supplier and inventory records as exit-readiness items from first major shipment, not as something to clean up at the point of interest from an acquirer.