The Fashion Discovery Market

A market defined by fragmentation

The defining feature of fashion retail online is that no single shop carries everything. A popular sneaker model might be sold by the brand direct, a dozen national retailers, several pan-European marketplaces and a long tail of regional specialists — each with its own pricing, size availability and delivery terms. Business of Fashion and McKinsey have described the sector, in their recurring State of Fashion work, as structurally fragmented and under persistent margin pressure, and that fragmentation is the root cause of most discovery friction.

Fragmentation is not an accident to be engineered away. It reflects genuine differences in geography, brand agreements, warehousing and taste. What it creates, though, is a coordination cost: the shopper has to do the work of finding, comparing and verifying availability across shops that do not talk to one another. That coordination cost is the raw material of the discovery market.

Marketplaces, brands and the DTC swing

Two structural forces shape supply. The first is marketplace growth: aggregating third-party sellers into one destination remains one of the most reliable ways to scale assortment, and analysts at Euromonitor and Statista consistently point to marketplaces capturing a growing share of online fashion spend.

The second is the swing between direct-to-consumer and wholesale. Brands that pushed hard into DTC to own the customer relationship have, in many cases, rebalanced toward marketplaces and wholesale to move inventory. The result for shoppers is that the same product surfaces across many channels at different prices — good for choice, punishing for anyone trying to find the best available offer. We unpack this tension further under sneaker business trends.

The aggregator role

Fragmented supply invites aggregation. An aggregator does not hold stock; it collects offers from many merchants, normalises them into a comparable form, and presents them as a single searchable layer. This is the same structural role that price-comparison played in electronics and travel, applied to a category where the product data is far messier.

Where an aggregator earns its keep is in the parts shoppers cannot see: reconciling that a shoe listed three different ways across three shops is in fact the same product, keeping prices and sizes fresh, and filtering out dead or duplicated offers. Done well, it collapses the coordination cost of fragmentation into a single query. We treat this as a category of its own in online fashion aggregators and product discovery platforms.

Affiliate discovery and the commercial model

Much of the discovery layer is funded by affiliate economics: the discovery site sends a qualified shopper to a merchant and earns a share of the resulting sale. It is an old model, but it aligns incentives usefully — the discovery site only wins when it genuinely helps a purchase happen, which rewards relevance and freshness over volume of clicks.

The model has limits. Affiliate feeds are only as good as the merchant data behind them, commission structures can distort which offers get surfaced, and attribution is imperfect. Operators who treat affiliate purely as an arbitrage game tend to erode trust; those who treat it as a reason to invest in offer quality tend to build something durable. Retail Dive and Think with Google have both documented how thin, low-trust intermediaries lose out as shopper expectations rise.

Product data as the quiet moat

Underneath every discovery experience sits a catalogue, and the quality of that catalogue is the real differentiator. If prices are stale, sizes are wrong or the same shoe appears as five separate listings, no interface — however slick — can produce a trustworthy answer. This is unglamorous work: feed normalization, product matching and canonical records.

We argue throughout this library that data quality, not the front end, is where defensibility lives. Interfaces are copied quickly; a clean, fresh, well-matched catalogue is expensive to build and hard to replicate. Grand View Research and Future Market Insights frame the broader retail-analytics and product-data tooling space as a growth category.

AI shopping and visual search

The newest layer is conversational and visual. AI shopping assistants answer questions in natural language; visual search lets a shopper start from an image rather than a phrase. Both change the shape of discovery, moving it away from the keyword box toward something closer to asking a knowledgeable friend.

The opportunity is real, but so are the constraints. An assistant that recommends a shoe which is out of stock, or invents a price, destroys trust faster than it builds it. The practical frontier is therefore less about model capability and more about grounding: connecting the assistant to a fresh, accurate catalogue so its answers hold up. We cover this in depth under fashion tech trends and AI fashion discovery.

Where the opportunity sits

Put the pieces together and a pattern emerges. The commercial opportunity in fashion discovery is not in adding another storefront; supply is already abundant. It sits in the coordination layer — in making fragmented supply legible, fresh and trustworthy for a shopper who does not want to visit fifteen shops. Whoever owns that layer, and the clean product data beneath it, sits between demand and a fragmented market and is hard to disintermediate.

• Supply is not the constraint. Assortment is abundant; findability is scarce.
• Data quality compounds. Freshness and matching are slow to build and slow to copy.
• Interfaces commoditise. AI and visual search raise the bar but do not replace the catalogue beneath them.