How AI Is Creating New Marketplace Behaviours

Artificial intelligence creates new marketplace behaviours by allowing people to describe a complete need in one sentence. Instead of deciding which category to open, which filters to apply and which keywords a seller might have used, a person can begin with what they are trying to achieve.

This changes more than the search box. It makes clarification, connected requests and context-aware assistance practical parts of an everyday marketplace journey.

The result is an emerging model in which people may ask for “a small suitcase within five kilometres that I can collect tonight” or “a driver to the airport tomorrow morning, with room for two large bags”. The platform’s task is to understand the intention, identify missing details and look across the relevant possibilities.

From category navigation to expressing an intention

Traditional marketplaces generally ask users to translate a real-world need into the platform’s structure. Someone looking for a second-hand bike might select a sporting-goods category, choose a bicycle type, set a price range and enter a location. This is familiar and efficient when the user knows exactly what filters matter.

Established marketplaces have substantial strengths here, including scale, liquidity, trusted habits, specialist tools and large catalogues. Their structural limits often reflect the design era and category focus in which they developed, rather than a failure of execution. A fashion resale platform, local classifieds service or auction marketplace may be exceptionally effective within its chosen model.

AI introduces another starting point: the user states the intended outcome. Understanding intent instead of keywords means interpreting details such as urgency, distance, condition, purpose and practical constraints together.

For example:

“I need a reliable second-hand bike for a six-kilometre commute, preferably nearby, and I will need someone to check it before I buy.”

This sentence may contain several connected needs: finding the bike, limiting the search area and possibly locating a local craftsperson or knowledgeable service provider. A conventional catalogue can still return bicycles, but an intent-aware system can recognise that the request is not solely about an object.

The new behaviours enabled by one-sentence requests

People describe situations, not products

A keyword search usually names an item: “suitcase”, “tutor” or “garden help”. Conversational use adds the circumstances around it.

A traveller might say, “I need a cabin suitcase five kilometres away because I leave tomorrow morning.” Distance, size and urgency all influence what counts as a useful result. If the traveller is unsure about airline dimensions, the system can ask a clarifying question rather than forcing an early selection from filters.

This behaviour is explored more fully in Conversational search: how it actually works. The important shift is that incomplete language is no longer necessarily an error. It can be the beginning of a dialogue.

Requests become iterative

People rarely express every relevant detail at once. They remember constraints as they see options:

“It needs to be available this evening.”
“Actually, collection by bicycle is too difficult.”
“Could someone deliver it?”
“Show me a cheaper alternative if delivery is included.”

A conversational marketplace can retain these details during the exchange. This differs from repeatedly resetting filters because the conversation carries the developing context. The distinction between a catalogue and this form of guided discovery is examined in Catalogue or conversation?.

The AI does not need to pretend that every request has a perfect answer. A useful assistant should say when supply is limited, ask the user to widen the area or explain that a requirement cannot currently be matched.

Several needs become one journey

Everyday intentions often cross commercial boundaries. Planning a weekend can involve accommodation, transport, an event, pet care and the temporary rental of equipment. Moving home can involve a van, packing boxes, gardening help, furniture and someone to assemble it.

A multi-universe platform is designed to treat these as parts of one situation. Why universes work better together explains why connecting different forms of supply can be more useful than merely placing unrelated services under one brand.

This is not the same as a super-app assembled from separate features, nor is it an aggregator that forwards the same query to multiple websites. The defining behaviour is continuity of intent: information supplied for one part of the journey can inform the next relevant step.

Users move more easily between roles

Category-led platforms often establish a clear role for the user: buyer, seller, host, passenger or service provider. AI-assisted participation can make those roles more fluid.

Someone searching for a private tutor may also have clothes to sell second-hand and gardening skills they could offer locally. A person renting equipment for a weekend event might later decide to list equipment of their own. One account can therefore support buying, selling, renting, booking and earning, although income is never guaranteed and depends on demand, location, condition and pricing.

Reducing the effort required to create supply is part of this change. Photo analysis, draft titles and price suggestions can help a first-time seller prepare a listing, as described in AI pricing and listing assistance. The user must still review the information and remain responsible for an accurate description.

“Find near me” becomes contextual

Local discovery traditionally begins with a postcode or radius. Intent-based search can consider location alongside the task itself.

A request for “someone to water my plants this afternoon” has a narrow practical radius. A search for a rare second-hand phone may justify travelling farther. The appropriate meaning of “near me” changes with urgency, rarity, transport and the value of the transaction.

A map remains useful, but it becomes one expression of context rather than the entire discovery method. Users may also search the area currently visible on the map, which is useful when planning somewhere other than their present location.

A comparison of emerging behaviours

| Everyday task | Category-led behaviour | AI-assisted, intent-led behaviour | |---|---|---| | Find a suitcase | Choose luggage, size, price and radius filters | Describe size, deadline, distance and collection needs in one request | | Book a driver | Enter route and time in a dedicated transport flow | Explain the journey, passengers, luggage and flexibility conversationally | | Plan a weekend | Visit separate accommodation, mobility and event services | State the overall plan and explore connected needs while preserving context | | Sell clothes | Photograph each item and write listings manually | Receive draft descriptions, titles or pricing guidance for review | | Find gardening help | Select a service category and contact providers | Describe the garden, task, timing and tools, then answer clarifying questions | | Locate a tutor | Filter by subject and level | Add learning goal, schedule, location and preferred teaching style naturally |

The two approaches can coexist. Filters provide precision and predictability, while dialogue is useful when the request is complex, uncertain or crosses categories. Semantic search versus filters considers how these methods can complement rather than simply replace one another.

Why native AI matters

A chatbot placed over an existing catalogue may make a search interface more conversational, but it does not automatically change the underlying platform. If the data, matching logic and user journeys remain isolated by category, the assistant may only translate sentences into conventional filters.

An AI marketplace designed around intention requires artificial intelligence within the architecture from the outset. It must be able to interpret natural language, request clarification, work across different universes and connect recommendations to the user’s stated circumstances. This distinction is covered in Native AI versus added AI.

Such systems also require safeguards. AI can misunderstand an ambiguous sentence, infer an irrelevant preference or produce an unsuitable suggestion. Users need clear control, understandable recommendations, accurate listings, moderation and ordinary ways to correct the assistant. Conversation should reduce effort without concealing uncertainty.

The iMarketplace as a proposed framework

An iMarketplace is a category we are proposing and documenting publicly; it is not yet an industry standard term. As defined here, the four letters represented by the “i” are Intelligence, Intention, Interaction and Individualisation.

Its central properties are native AI, conversational discovery, organisation around intentions and a multi-universe structure spanning areas such as goods, services, housing, mobility, missions, events, animals and skills. Most importantly, it should be able to connect several everyday needs within one coherent journey.

WEVONE is one concrete illustration of this idea, not proof that the category has prevailed. As of August 2026, it is a young platform, public since 2026, with a few hundred registered members. It is far smaller than Vinted, Leboncoin, eBay or Facebook Marketplace, whose scale and established user habits can provide much broader supply and demand.

Available today in WEVONE 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. Mia, its built-in AI, supports natural-language search, 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, helps moderate listings and photos, and can make cross-universe recommendations. Local discovery is map-based and filtered by the area being viewed. Additional universes and more connected journeys are planned, while some capabilities continue to be tested before wider use.

Conclusion

The most significant new use created by marketplace AI is not voice input or a more attractive search bar. It is the ability to begin with a situation: what someone needs, where, when, why and under which constraints.

Once that becomes normal, people can refine requests through dialogue, combine related needs and move more easily between buying, selling, renting, booking and earning. An iMarketplace takes a different approach from established category-led platforms by designing the experience around these intentions. It may suit users whose everyday needs do not fit neatly into a single catalogue.

FAQ

What is a one-sentence marketplace request?

It is an ordinary-language description of an outcome, such as “I need a driver to the airport tomorrow at 6 am with room for two suitcases”. The platform interprets the details rather than requiring the user to construct separate filters first.

Does conversational search replace filters?

Not necessarily. Filters remain useful for exact comparisons, while conversation helps with ambiguity, context and connected needs. A well-designed service may offer both.

What makes an AI assistant marketplace different from a chatbot?

The difference lies in the underlying architecture. An assistant marketplace is designed to understand intent and work across platform data and journeys; a chatbot may simply sit above an unchanged catalogue.

Can an iMarketplace handle several needs at once?

That is one of its defining aims. A weekend request could connect accommodation, transport and events, although the available results still depend on local supply and the platform’s current capabilities.

Is iMarketplace an established technology category?

No. It is a concept WEVONE is proposing and defining publicly. It should not be treated as a recognised analyst category or an industry-wide standard.

Does AI guarantee a good match or extra income?

No. AI can reduce search and listing effort, but matches depend on supply, demand, location, availability, condition and pricing. Recommendations also need user review.

New Everyday Uses Emerging With Artificial Intelligence