Fashion Aggregators and Marketplaces

Revenue: four fundamentally different meters

The clearest economic split is how each model bills. A retailer earns the margin between wholesale cost and retail price on goods it owns. A marketplace earns a commission — a percentage of each hosted transaction — plus, often, fees for advertising and fulfilment. An affiliate aggregator earns a referral fee paid by the seller when a click converts, and nothing at all when it does not. A product search engine or comparison site sits close to the aggregator, monetising through referrals, sponsored placement or licensing of its data.

These meters reward different behaviours. Retail rewards inventory selection and sell-through. Marketplaces reward transaction volume and seller supply. Aggregators and search engines reward traffic quality and conversion, because they are paid on the click, not the stock. A business that mixes models has to keep these incentives from pulling against each other.

The six models by economics

The pattern is that inventory-owning models earn more per sale but carry risk, while data- and traffic-owning models earn less per event but scale cheaply — if they can defend their position.

Data ownership decides leverage

Data ownership is the quiet determinant of long-term power. A marketplace owns the transaction data of an entire ecosystem, which lets it see demand before anyone else and lean on that knowledge. A retailer owns its customers directly. Aggregators and search engines, by contrast, often borrow their product data from the sellers they list, which is a weaker position — the underlying feed can be withdrawn or degraded.

The strongest data position for a pure discovery business is a proprietary, normalized index that it builds and maintains itself, described in canonical product records. That index cannot be trivially copied, because the value is in the reconciliation work — the normalization and matching — not in the raw listings. Owning clean data is how an asset-light business acquires something durable to defend.

Margin and the cost of breadth

Margin shape follows the model. Retail margins are the fattest per sale but come loaded with inventory, markdown and returns risk. Marketplace and aggregator margins are thinner per event but asset-light, which is why they scale so well when volume arrives. The trap for aggregators and search engines is that their thin per-click economics only work at large scale, and reaching that scale requires broad, fresh coverage — which is itself expensive to maintain.

This is the central tension of the aggregator model: breadth is both the product and the cost. Every additional retailer indexed adds discovery value but also data-cleaning, freshness and integration overhead. The economics work only if the traffic and conversion earned from that breadth outrun the cost of keeping it clean. Many discovery businesses fail not on demand but on the unglamorous unit economics of coverage.

Defensibility: what stops a copycat

Each model defends itself differently. Retailers defend on brand, buying skill and service. Marketplaces defend on network effects — more sellers attract more buyers, which attract more sellers — a moat that is very hard to breach once established. Aggregators defend on audience and traffic relationships, which are real but rentable and can erode if a channel changes its rules.

Search and discovery engines defend on something subtler: the quality and breadth of their normalized data and the relevance of their models. A rival can copy the interface in a weekend but cannot copy years of match quality and coverage overnight. In a market as fragmented as footwear — see sneaker retail fragmentation — that data moat is often the only durable advantage a discovery business has. Understanding which moat you are actually building is the difference between a defensible position and a feature waiting to be cloned.