Two routes to the same customer
Selling on a marketplace means listing products inside a store that someone else operates: the marketplace sets the page template, runs the search, handles payment and often fulfilment, and charges a referral fee on each sale plus, commonly, fees for fulfilment, advertising and subscription. The seller controls the listing content, the price and the stock, and little else. Selling through an own store means operating the whole thing: the site, the acquisition, the checkout, the delivery promise and the customer service. The seller controls everything and owns the customer data, and pays for that control by having to generate every visit.
The economic difference is where the cost sits. On a marketplace, acquisition cost is folded into the fee and paid only when a sale happens; on the own store, acquisition is paid up front as advertising, content and search work, and the cost per new customer (CACCost to acquire one new customer.) has to be measured and managed. The strategic difference is who the customer belongs to. A marketplace buyer has a relationship with the marketplace, which holds the address, the history and the recommendation; an own-store buyer can be emailed, segmented and brought back, which is where lifetime value (LTVTotal value a customer brings over time.) is built. The acquisition hub places both routes among the other sources of store traffic; this page compares them directly. The rules that apply on either route, from pricing and review claims to consent for marketing messages, are gathered in the e-commerce industry guide. Selling to retailers who buy the stock and resell it is a different route again, compared with the own store in the wholesale versus direct-to-consumer comparison.
Compare table
Marketplaces vs Your own store
| Row label | Marketplaces | Your own store |
|---|---|---|
| Reach and demand | Shoppers already present and searching; category demand available from the first day | Every visit must be earned through ads, search, social or email |
| Trust at first purchase | Borrowed from the marketplace's delivery and returns promise; ratings visible | Must be built with reviews, guarantees, clear policies and a professional site |
| Fees and margin | Referral fee on each sale, often plus fulfilment, advertising and subscription fees; margin per order is lower | No platform fee beyond payment processing and software; margin per order is higher but acquisition is paid separately |
| Customer data | Limited; the marketplace holds the relationship and restricts contact | Full: email, order history, behaviour, consent for marketing |
| Retention and repeat | Repeat orders go to whichever seller ranks when the buyer returns | Repeat purchase can be driven with flows, loyalty and service, and measured by cohort |
| Brand and pricing control | Template page, marketplace rules on content, pressure to match the lowest price on the page | Own design, own story, own pricing and promotions |
| Competition on the page | Direct competitors and resellers appear beside the listing, sometimes below the buy button | None on the page; competition happens before the visit |
| Advertising | The marketplace's own ads, bought inside its auction, with reporting limited to the marketplace | Shopping, catalogue, search, social and email, measured in the store's own analytics |
| Operations | Fulfilment and returns can be outsourced to the marketplace; its rules and penalties apply | Fulfilment, returns and service are the seller's to build or contract |
| Speed to first sale | Fast: a listing can sell within days if the demand exists | Slow: a store needs traffic, trust and a working checkout before it sells |
| Measurement | Sales and marketplace ad metrics; little visibility of the customer journey | Full funnel: sessions, conversion, order value (AOVAverage revenue per order.), repeat and margin by channel |
| Dependency risk | Policy changes, fee increases, suspension and ranking changes are outside the seller's control | Platform and ad-cost risk, but no single party can switch the store off |
Where marketplaces win
Marketplaces win when the demand exists and the seller cannot yet reach it. A new brand in a category people already buy on marketplaces will sell there before it sells anywhere else, because the shoppers are present, the trust is borrowed and the first sale needs no advertising budget. They win for products bought on price and specification, where the buyer compares a row of near-identical listings and the brand matters less than the rating and the delivery date. They win for sellers without the capacity or the capital to build fulfilment, because the marketplace's logistics can be rented. And they win in categories where the marketplace is, in effect, the search engine: if a large share of product searches start on the marketplace, an absent brand is invisible to those buyers.
They also win as a research tool. Marketplace sales and reviews show which products, variants and price points move, which is information a store launch needs and which paid search would cost money to gather. And they win, within limits, for reach into markets the seller cannot serve directly: cross-border selling with the marketplace handling tax, currency and delivery is far simpler than opening a store in each country. The price in all these cases is the fee, the competition on the page and the customer the seller never meets; what the marketplace gives in reach it takes in margin and data. In the terms of the five forces model, a marketplace that gathers the buyers and sets the fees holds both buyer and supplier power over the sellers on it, which makes heavy dependence on one a strategic question rather than only a channel choice.
Where the own store wins
The own store wins on margin, data and relationship, which together decide whether a brand is a business or a supplier. Every order keeps the fee the marketplace would have taken, which over a catalogue and a year is the difference between funding growth and funding the platform. Every buyer becomes a contact who can be welcomed, brought back and served, which makes the second order cheaper than the first and the customer worth more than one sale; the repeat purchase and loyalty guide covers how that value is built, and it cannot be built on a marketplace. Return on advertising (ROASRevenue generated per unit of ad spend.) is measured in the store's own analytics, against the store's own margin, by channel, so the budget can be moved to where it earns.
The own store also wins on brand. The page says what the brand wants it to say, in its own design, with its own story, bundles, guarantees and pricing, and no competitor sits beside the buy button. For products that are chosen on trust, taste or expertise rather than on price and specification, that is the whole sale. The store wins on control of risk, too: a fee change, a policy update or a suspension on a marketplace can remove a channel overnight, whereas a store's risks are spread across several ad platforms and search; the guide to reducing dependence on a single channel shows how to measure that concentration from order records and build an owned audience as insurance. And for premium products, the store protects price: a marketplace listing invites resellers and discounting that the brand cannot police. The cost is that every visit must be earned, through the channels in the shopping and catalogue ads guide and the e-commerce SEO hub, and that trust must be built rather than borrowed.
Running both without cannibalising
Most brands end up with both, and the problem is that they compete with each other for the same buyer. The buyer who saw the brand's catalogue ad and then searched the marketplace, where the price was lower or the delivery faster, bought there; the brand paid for the ad and the fee. Running both well means deciding what each is for and arranging the assortment, pricing and messaging so that the marketplace feeds the store rather than draining it.
The usual design has four parts. Assortment: the full range on the store, a deliberate subset on the marketplace, so that the marketplace introduces the brand and the store rewards exploration; exclusive bundles, new launches and personalised items stay on the store. Pricing: the same headline price everywhere, because marketplaces monitor price parity, with the store's advantage expressed in bundles, loyalty points, delivery thresholds and service rather than in a lower sticker. Migration: every marketplace order is an introduction, so packaging inserts, product registration, warranty activation and care content give the buyer a reason to visit the store and join its list, within the marketplace's rules on contact. The guide to packaging and order inserts covers what a card in the parcel should ask for and how to tell whether it brings buyers to the store. Measurement: branded search volume, store sessions and new-customer counts read alongside marketplace sales, so that the brand can see whether marketplace presence lifts the store or substitutes for it; the channel mix guide covers the general method, and the e-commerce marketing hub places both routes in the growth equation. The flows that greet a migrated buyer are covered in the email and SMS flows guide.
A candle brand splits its range
A candle brand sells on one large marketplace and through its own store. Marketplace orders average £24 with a 15% referral fee and a £3.50 fulfilment charge, leaving £13.70 after £3.30 of product cost; store orders average £38, with £2.10 in payment and software costs and £5.20 of product cost, leaving £30.70 before the £9 the store spends to acquire each new customer. The brand moves its three gift sets and all new scents to the store only, keeps the six core scents on the marketplace at the same price as the store, and adds an insert offering a free wick trimmer on registration. Over six months, 18% of marketplace buyers register, 40% of those place a store order within a year, and the store's share of total revenue rises from 35% to 52% while marketplace revenue stays flat rather than falling.
If the own-store route wins, the Shopify versus WooCommerce comparison helps you choose a platform by weighing cost, control, integrations and the effort needed to maintain it.
When judging marketplace advertising, the ROAS calculator walkthrough for marketplace sellers shows how to enter the margin left after fees and read the result as ACoSMarketplace advertising cost divided by the sales the ad console attributes to ads, shown as a percentage; the inverse of ROAS..
Once a brand has decided to list, the marketplace channel page for online sales sets the range, listing, review and bidding tactics, and a ceiling on the marketplace's share of sales.
Common mistakes
- Choosing one route on principle. The question is what each is for in this category, not which is better in general.
- Pricing the marketplace lower to win the buy box. The brand's own ads then send buyers to the cheaper, fee-bearing channel.
- Putting the whole range everywhere. With no reason to visit the store, the marketplace becomes the store.
- Counting marketplace revenue as if it were store revenue. After fees, fulfilment and advertising the margin is different, and comparisons must be made after them.
- Treating marketplace buyers as the brand's customers. They are the marketplace's until they register or buy on the store.
- Ignoring the dependency. One policy change or suspension can remove the channel; the store is the hedge.
- Never measuring the halo. Marketplace presence can raise branded search and store sales, or cannibalise them; only reading both together tells which.
