Smart search for eCommerce

What you don't find, you won't buy. Indulge your customers with robust tolerance for typos and synonyms. Guide your customers to the best results, no matter how they search and in what language.

Business case

The longer you spend searching for a product, the lower your chance to make a purchase. In order to help your customers find exactly what they are looking for, you need to eliminate full-text results, route your users directly to relevant categories or products and understand users' search behaviour to provide them with the best possible results. You need to navigate your users towards products they want to buy, and while doing that, you might as well capture current search trends.

Solution

Our solution supports mobile and voice search, is developed friendly, and A/B tested. It can be extensively personalised and based on users' behaviour to understand and detect categories, brands, or properties. It also decodes the semantic of requests based on BERT (Bidirectional Encoder Representations from Transformers), and offers trends leaning on search history.

Benefits

Directing customers directly into the relevant categories instead of a full-text search helps them find what they are looking for faster; thus, it increases loyalty and the chance of purchasing.
Offering trends based on search history makes the navigation faster and more relevant by "making a conversation" with the customers.

Reference

Anastasi Korjenevskaya, Search Product Manager at Heureka Czech Republic

"The search query understanding solution that we developed together with DataSentics helps our customers to find what they want to find very quickly. The solution understands a search query and shows only the relevant product category to the customers. If also other entities are recognised, those are directly filtered in the category as well (e.g. a brand). For example, when customers search for “HDR monitor Samsung” they instantly get monitors with relevant filters checked and can apply whatever additional filters that are available for this category. This AI solution covers approx. 50% of all queries and in total decreases the leave rate after searching by 12.5%, improves the median click position by 25% and increases CTR from category to product detail by 10 %."