History of Marketplaces: From Stalls to AI

Marketplaces began as physical places where people gathered to exchange goods and services. They later became classified pages, websites and mobile apps; the emerging next step is a marketplace that behaves more like an assistant, interpreting what someone wants to accomplish rather than merely directing them towards a category.

This development is not a simple replacement of old models by new ones. Each era preserves useful elements of the previous one: proximity from the physical market, structure from classifieds, scale from the web and convenience from mobile apps. The history also helps explain why marketplaces are evolving towards more conversational and context-aware experiences.

The iMarketplace is a category we are proposing for this latest stage. It is not yet an industry standard term. As we define it here, it describes a native-AI, conversational and multi-universe platform organised around user intentions.

The physical marketplace

A shared place for exchange

For much of history, a marketplace was literally a place: a square, hall, street, fair or covered market where buyers and sellers met. Its value came from concentration. Bringing many participants together made it easier to compare goods, negotiate terms and learn who could be trusted.

The market also carried information that was difficult to formalise. A buyer could inspect a second-hand bike, ask why it was being sold and judge its condition. A household could speak directly to a local craftsperson about a repair. Reputation travelled through repeated encounters and community knowledge.

Physical markets nevertheless depended heavily on time and place. A buyer generally needed to attend when sellers were present, while a seller could reach only the people willing or able to travel there. Available supply was also difficult to know in advance.

The foundations that remained

Several enduring marketplace principles were already present:

a meeting point for supply and demand;
some method of organising offers;
information about price, condition and availability;
trust between strangers;
a way to complete or arrange an exchange.

Every later marketplace model changed how these functions were delivered, but did not remove the need for them.

Classified advertisements separate discovery from place

Printed classified advertisements turned the marketplace into an information system. Instead of standing beside an item or offering a service in person, someone could publish a short description under a heading such as property, employment, vehicles or household goods.

This was a major improvement in reach and persistence. A person looking for a private tutor or weekend rental could review several offers without visiting multiple physical markets. Sellers could remain discoverable for as long as the advertisement appeared.

The organising principle was the category. Categories made large numbers of unrelated advertisements manageable, but they also required the reader to translate a real-life need into the publisher’s structure. Someone moving house might have to consult property, furniture, transport, storage and services separately. The publication recognised five categories, even though the reader experienced one move.

That distinction between human intention and catalogue structure remains central to understanding intent instead of keywords.

The web creates marketplaces at scale

Searchable, persistent catalogues

Online marketplaces removed many of the physical and publishing constraints of classifieds. Listings could be searched, updated and illustrated with photographs. Buyers and sellers could communicate across much greater distances, while account systems, transaction tools and reputation mechanisms made exchanges between strangers more practical.

Different platforms developed genuine strengths around particular forms of exchange. eBay built a broad online-selling environment with auction and fixed-price formats, established transaction tooling and access to a wide audience. Its evolution is explored in how eBay approaches online selling. Leboncoin became closely associated with local classifieds in France, while specialist services developed strong habits and communities around particular categories.

These platforms were largely organised as digital catalogues because that was an effective structure for the technology and user expectations of their design era. Search boxes and filters improved access to the catalogue, but users still usually needed to select the right category and phrase their request in terms the system recognised. This is one of the structural characteristics of platforms designed before AI, rather than an absence of useful technology or expertise.

Trust becomes a platform function

As marketplaces expanded beyond established local relationships, trust increasingly had to be supported by the platform. Profiles, ratings, messaging, payment options, delivery tools and moderation helped people assess unfamiliar counterparties.

These systems did not eliminate risk, but they made remote exchange easier to manage. They also established expectations that modern marketplaces should do more than publish advertisements: they should help participants communicate, evaluate an offer and complete the transaction.

Mobile apps make the marketplace continuous

Smartphones placed marketplaces in people’s pockets. Taking photographs, publishing a listing, receiving a message and checking nearby offers became possible within a single device. Notifications made exchanges more immediate, while location services improved local discovery.

Facebook Marketplace benefits from its connection to an established social platform and familiar user accounts, making local selling accessible to a broad audience. Its model is described in how Facebook Marketplace approaches local selling. Vinted, meanwhile, developed a focused experience and strong user habits around second-hand fashion, as covered in how Vinted approaches second-hand fashion.

These focused and established platforms offer scale, liquidity, familiarity and mature tooling. Their structural limits arise partly from that focus: an app designed around fashion resale, local classifieds or broad online selling is usually optimised for that primary activity. Adding AI to such a system can improve search or listing creation, but does not necessarily reorganise the entire experience around intention.

Terms, fees and available features change, so readers comparing platforms should check each service directly.

From apps to assistants

The marketplace becomes conversational

Recent AI systems can work with ordinary language rather than relying entirely on exact keywords and fixed filters. This makes a different starting point possible: the user can describe the outcome they want.

Consider the request: “I need a medium suitcase within five kilometres, preferably available this evening, because I am flying tomorrow.” A conventional catalogue may ask the user to choose luggage, size, condition, distance and collection options separately. A conversational marketplace can extract those constraints from the sentence, ask whether the user wants to buy or borrow, and then search accordingly. The mechanics of this approach are explained in conversational search.

The difference is not that filters cease to be useful. Filters remain efficient when someone knows exactly what category and attributes to select. Intent-based search adds another route for needs that are incomplete, contextual or spread across several categories.

One intention can contain several needs

A weekend away may involve accommodation, a driver to the airport, pet care, an event and perhaps the rental of outdoor equipment. In a conventional app landscape, these are commonly handled through separate specialist services. A multi-universe platform instead treats them as related parts of one intention.

This is the purpose of connecting different domains, or universes, within a single experience. Multi-universe design changes the user journey because the system can retain context between goods, services, housing, mobility, missions, events, animals and skills. The aim is not simply to place unrelated features beside one another, but to understand how they contribute to the same outcome.

Native AI rather than an attached chatbot

An AI marketplace is not automatically an iMarketplace. A chatbot added to a conventional listing database may provide useful support, but the underlying marketplace can remain centred on categories.

In an iMarketplace, AI is part of the architecture from the beginning. It can support interpretation of intentions, clarifying dialogue, semantic matching, listing creation, moderation and recommendations across universes. The interface may therefore behave as an AI assistant marketplace rather than merely a search bar followed by a results grid.

The marketplace eras compared

| Era | Main organising principle | Typical interaction | Principal strength | Common constraint | |---|---|---|---|---| | Physical market | Place and time | Visit, inspect and negotiate | Direct contact and local trust | Limited reach and availability | | Printed classifieds | Published category | Read and contact the advertiser | Persistent, wider discovery | Sparse information and manual coordination | | Web marketplace | Searchable catalogue | Search, filter and message | Scale, comparison and transaction tools | Users must navigate platform structure | | Mobile marketplace | App, category and location | Photograph, swipe, message and transact | Convenience and continuous access | Needs may remain divided between apps | | iMarketplace concept | Intention and context | Describe, clarify and act | Connected, individualised journeys | Depends on capable AI, sufficient supply and user trust |

The last row describes an emerging model, not a completed industry transition. Catalogue-based, specialist and conversational services may coexist because they suit different tasks and user preferences.

WEVONE as an early illustration

WEVONE takes a different approach by designing one app around several universes. Available today are Tutus for second-hand goods and fashion, Nest for housing and space rental, Mission for services and local gigs, Events, and Pilote for transport and parcel delivery. More universes are planned.

Mia, its built-in AI, supports conversational search in natural language, including a request such as “a black T-shirt with an eagle on it”. It also assists with listings through photo analysis, title and price suggestions, moderates listings and photographs, and provides cross-universe recommendations. Local discovery is map-based and filtered by the area currently being viewed, supporting familiar searches such as “find near me”.

For example, someone wanting to sell second-hand clothes can receive help preparing the listings, while another user looking for a second-hand bike can describe size, budget and collection area conversationally. A request for a driver to the airport belongs to mobility, but it could also form part of a broader weekend journey. This connected logic is why universes can work better together.

WEVONE has been public since 2026 and has a few hundred registered members as of August 2026. It is far smaller than Vinted, Leboncoin, eBay or Facebook Marketplace, so it cannot offer their scale or liquidity. It is one concrete illustration of the proposed iMarketplace idea, not evidence that the category has already prevailed. Its “Earn from every action” positioning covers buying, selling, renting, booking and earning, but income is never guaranteed and depends on demand, location, condition and pricing.

Conclusion

The history of marketplaces is a history of reducing the distance between a need and someone able to meet it. Physical markets concentrated people in one place; classifieds made offers persistent; websites added search and scale; mobile apps made participation immediate and continuous.

Assistants introduce a further possibility: organising the marketplace around intention rather than requiring the individual to navigate categories first. The proposed iMarketplace combines native AI, conversation, interaction, individualisation and multiple universes so that several everyday needs can form one coherent journey. It may suit users who prefer to explain what they need, while established catalogue and specialist platforms will remain useful where scale, habit or category expertise matters most.

FAQ

What was the earliest form of marketplace?

The earliest marketplaces were physical meeting places where buyers and sellers exchanged goods, labour and information directly.

How did classifieds change marketplaces?

Classifieds separated discovery from physical attendance. They made offers persistent and easier to scan, but organised them primarily through publisher-defined categories.

Why were marketplace apps important?

Apps made listing, messaging, local discovery and transaction management continuously accessible through a smartphone. They also reduced the effort required to photograph and sell second-hand items.

What is an AI marketplace assistant?

It is an AI system that helps users express needs, refine requests, discover relevant offers or create listings. Its role can range from a support feature to the main marketplace interface.

What makes an iMarketplace different from a conventional marketplace?

As defined here, an iMarketplace is centred on intentions, built around native AI, conversational, individualised and able to connect several universes within one journey. The term is a proposed category, not yet an industry standard.

Will assistants replace search bars and filters?

Not necessarily. Search bars and filters remain fast and familiar for precise tasks. Conversation is particularly useful when a request is contextual, uncertain or involves several connected needs.

A Short History of Marketplaces: From Stalls to AI Assistants