Traverz Conversational Recommendation

A paradigm change in Product Search

At the core of every e-commerce and marketplace interaction is the consumer’s product search journey. This is a critical element to achieving conversion and loyalty. Yet traditional search mechanisms are far too often leading to user confusion, or worse, frustration.

Traverz Conversational Recommendation is a paradigm change in product search, that lets consumers tell you what they really want – beyond on-off filters. It recognises that user preferences are fuzzy and developed iteratively through exposure, and opens an agent-based dialogue channel using natural language and voice. Importantly, it puts consumers back in control – thereby raising their purchase intent through greater comfort and confidence in their product selection.

Traditionally the product search journey has been served by filters and search bars, turning user input into a product list that is a subset of the available products (“Interactive Filtering”). Originally developed in response to the observation that filter and search-bar input can drive an SQL statement on a product database, this technology-driven User Experience is now over 20 years old.

More recently, Recommender Systems were added to order the user’s product list. This black-box technique uses data collected about ‘similar’ users to infer the likely product preferences of the current user, and the user themselves is unable to directly influence the results. This approach to personalisation often suffers from very poor signal-to-noise ratio, and leaves the user with the impression that it is the site that controls their product search results.

With the advent of voice recognition, Conversational Commerce has emerged as a voice-driven shopping technology. Whilst voice is a powerful communication mechanism, the mostly prescriptive product search process steps and absence of visual cues take it very far away from the traditional Interactive Filtering approach.

Traverz Conversational Recommendation re-imagines the online product search mechanism to deliver a best-in-class experience. At its core, the Traverz technology is based on Conversational Recommendation – the concept that we recommend products based on the preferences of the individual consumer understood through a conversational style of interaction.


The Traverz search journey is therefore a consumer-driven multi-step process, based around preferences (rather than binary filters) and involving the support of an intelligent virtual agent that provides shopping-assistance like support. Uniquely, Traverz has built this technology on top of the existing user-driven web-ui and utilised modern AI to drive the interaction with the user.

On sites powered by Traverz Conversational Recommendation technology, consumers are supported right from the start of their journey – from initial awareness and discovery guidance, through to narrowing the search and the final step of making a purchase.

The benefits are substantial: Higher top-of-funnel conversion, a natural upsell process, more confident consumers, and substantially greater loyalty. Product data issues are no longer a problem, and far greater insight is gained into consumer needs.