Visii uses AI-driven image and contextual user behaviour analysis to give your customers a more relevant visual search experience.

Who uses this technology?

General usage

More than 60 percent of Millennials and Generation Z want to use visual search to find and buy products. Overall, 50 percent of online shoppers say they have made a purchase based on product images.

Visii solves this need with a basket of AI-driven visual search features that produce highly relevant search results and product recommendations. This has proven to increase sales and user engagement, as well as reduce churn and product returns.

For AI and deep learning algorithms to work properly though, they need to be fueled with the right data. Visii combines the look and feel of products (texture, colour, shape), user behaviour and context (browsing or purchase history, time of day), and product trends (reviews, popularity) to refine its algorithm.

Solutions include:

Visii Similar – generates related product suggestions based on the images in search results or product pages.

Visii Explore – powered by deep-learning AI which refines search results based on previous product selections. Makes it easier to browse huge collections of products for more relevant results. Can be combined with existing search filters.

Visii Lens – reverse image search that allows customers to search for products by uploading an image. Visii will produce a product list as close as possible to the uploaded image.

Visii Assistant – intuitive chatbot that can deliver helpful suggestions at different stages of the buyer funnel. One of its benefits is to resolve out-of-stock scenarios with alternatives as they occur.

Visii can also be used for more relevant retargeting campaigns on third-party platforms and automating the curation of best-seller or seasonal collections.


Visii tends to be used by highly-visual niche applications such as art marketplaces.

Marketplace benefits

CobbleWeb used Visi to create an intuitive image search feature for art marketplace, Affordable Art Fair. It allows novice art collectors to browse visually by clicking on images they like. It then uses machine learning and custom AI to provide users with more personalised search results and recommendations. 

We customised AAF’s search algorithm further with standard display conditions, such as number of reviews, stock levels, best sellers, and IP locations. The results were quite spectacular. Conversion rates jumped by 200% and revenue per user by 611%.

We're here to help!

Would you like to know more about the suitability of a particular technology for your marketplace? Get in touch for a free audit and recommendations.