A polished online shop can take orders while quietly losing money. The product pages look convincing, checkout works and revenue appears to rise, yet stock errors, fulfilment labour, returns, payment deductions and support time consume the margin after the sale.
The problem is structural: a storefront is the customer-facing entrance to an operation, not the whole operation. Before buying an e-commerce platform—or adding another app—map what must happen from the moment a customer sees a product to the moment the order is reconciled and the return window has passed.
The six systems behind one online order
A profitable order usually depends on six connected systems. A platform may cover several of them, but the presence of a feature does not prove that the workflow is complete.
- Catalogue and pricing: product data, variants, discounts, delivery promises and the total price shown to the customer.
- Inventory: available stock, reservations, bundles, damaged goods and returned items awaiting inspection.
- Checkout and payment: authorisation, fraud checks, payment fees, taxes where applicable and failed-payment recovery.
- Fulfilment: picking, packing, postage, tracking, delivery exceptions and replacement shipments.
- Returns and support: cancellation requests, refunds, exchanges, customer messages and restocking decisions.
- Finance and reconciliation: matching orders to payouts, fees, refunds, disputes and cash actually received.
If information stops at any hand-off, someone compensates with a spreadsheet, inbox search or memory. That may be reasonable at low volume. It becomes expensive when the workaround is invisible and therefore never included in the cost of each order.
Start with contribution, not revenue
Revenue tells you what customers paid. Contribution margin tells you what remains after the variable costs required to win and fulfil those orders. A practical per-order calculation is:
Net order revenue − product cost − payment and platform fees − pick, pack and delivery cost − acquisition cost − expected returns and support cost = contribution per order
Illustrative example only: suppose a seller receives £45 after a discount. The product costs £15, packaging and delivery cost £6, payment and platform charges total £2, customer acquisition costs £8, and the seller allocates £5 for expected returns and support. Contribution is £9 per order—not £45.
| Illustrative order | Amount |
|---|---|
| Net order revenue | £45 |
| Product cost | − £15 |
| Packaging and delivery | − £6 |
| Payment and platform charges | − £2 |
| Customer acquisition | − £8 |
| Returns and support allowance | − £5 |
| Contribution before fixed overhead | £9 |
At 100 similar orders, revenue would be £4,500 while contribution before fixed overhead would be £900. Salaries, software subscriptions, storage, accounting, insurance and tax still have to be paid from that £900. For a fuller worksheet, see how to calculate contribution margin per order.
Inventory must have one operational truth
“In stock” is not a decorative label. It is a promise backed by a sequence of rules. A buyer evaluating a platform should ask when stock is reserved: when an item enters the basket, when payment is authorised or when an order is confirmed. Each choice trades conversion opportunity against overselling risk.
Then test awkward cases. If a bundle contains three components, does selling one bundle reduce all three component quantities? Can staff separate sellable returns from items awaiting inspection? What happens when the same product is offered on the website, a marketplace and at a physical event? If two systems can independently declare stock available, neither is a reliable source of truth.
Small sellers do not always need sophisticated inventory software. A controlled spreadsheet and a daily stock check can be adequate when there are few products, one sales channel and manageable order volume. The warning sign is not manual work itself; it is manual work that nobody owns, times or audits.
Returns are part of the product economics
A returns page is not the returns operation. The business still needs to receive the request, identify the order, issue instructions, inspect the item, decide whether it can be resold, update stock, refund the correct amount and reconcile the deduction from a later payout.
For UK consumer sales, the legal workflow also matters. GOV.UK’s distance-selling guidance sets out information businesses must provide, while its returns and refunds guidance explains that online customers can have cancellation rights even when an item is not faulty, with exceptions for certain goods. Faulty, misdescribed or unfit items trigger separate refund obligations. This article is operational guidance, not legal advice; check the rules that apply to your products and markets.
Model returns before selecting software. Use your own expected return rate, outbound and return postage, inspection time, lost packaging, write-down on opened stock and support minutes. A “free returns” promise can still be a sound commercial choice, but only when its full cost is visible in pricing and margin decisions.
Gross sales are not cash available
An order can be marked paid while the money is still awaiting settlement. Payouts may combine several orders and subtract fees, refunds or disputes. Currency conversion and reserves can add another timing difference. Treat the order ledger, payment-processor ledger and bank statement as three records that must reconcile—not as interchangeable versions of the same number.
A useful buying test is to choose one real historical payout and ask a supplier to demonstrate how every line would be matched to its orders, refunds and charges. A dashboard total is insufficient if a bookkeeper cannot trace the underlying entries.
Price presentation is an operational requirement
The price shown on a product page must survive the path to checkout. The Competition and Markets Authority’s price-transparency guidance, published on 18 November 2025, explains what businesses should include in pricing information, including mandatory fees, taxes and charges. Its guidance on unfair commercial practices notes that the relevant provisions of the Digital Markets, Competition and Consumers Act apply to commercial practices from 6 April 2025.
This affects platform selection. Test whether mandatory delivery or service charges can be included at the appropriate point, whether discounts have clear conditions, and whether the basket maintains a transparent running total. Do not rely on a theme or plug-in to make the policy decision for you.
Run a failure test before you buy
| Workflow | Evidence to request | Failure test |
|---|---|---|
| Inventory | Variant, bundle and channel rules | Two customers attempt to buy the last unit |
| Fulfilment | Pick list, label and tracking flow | A parcel is delayed or partially shipped |
| Returns | Cancellation-to-restock workflow | An opened item cannot be resold |
| Payments | Order-to-payout reconciliation | A refund lands in a later payout |
| Pricing | Complete customer-visible total | A mandatory charge varies by order |
| Data | Export, ownership and retention terms | You leave the platform next month |
Ask for a live demonstration using the awkward case, not the happy path. Record which steps are native, which need a paid add-on and which depend on staff. Include migration, integration, training, data-cleaning and exit costs alongside the advertised subscription.
Who should buy an integrated platform?
An integrated platform can suit a seller with several channels, meaningful order volume, recurring stock errors or expensive reconciliation work—especially when one team needs shared visibility. It can also be sensible when consumer-law or audit requirements are difficult to manage across disconnected tools.
It may be the wrong purchase for a new shop still validating demand, a creator selling a handful of low-return digital products, or a business whose main problem is weak product-market fit. Automation will not repair an uneconomic product, unclear positioning or insufficient demand.
Three realistic alternatives
- Keep the current stack: document ownership, add a weekly reconciliation and measure the hours spent on workarounds.
- Add one specialised tool: fix the highest-cost break—such as inventory synchronisation or returns—without replacing the storefront.
- Do nothing for now: set a trigger for reconsideration, such as a defined order volume, error rate or monthly labour cost.
The best option is the one with the lowest total cost for the reliability you actually need. Count subscriptions, transaction charges, implementation, staff time and failure costs over a realistic period—then compare that figure with the value of errors avoided or labour released.
A restrained next step
MOC Marketplace is part of the MaryChuks.com business ecosystem. If you want to observe how digital businesses and software assets are presented, you can browse the general MOC Marketplace gateway. Treat displayed claims as a starting point for due diligence, not as proof of profitability or operational performance; this link is for general discovery, not a recommendation of a particular listing.
The decision rule
Do not buy a better-looking storefront to solve an invisible operating problem. Map the order, calculate contribution, test failure cases and identify the single source of truth for stock, payments and returns. Only then can you tell whether you need a new platform, one targeted integration or simply a more disciplined process.
Discover more from Marychuks.com AI, Psychology, Business & CreativeVerse
Subscribe to get the latest posts sent to your email.