Poor product discovery experiences cost UK online retailers dear
A majority of e-commerce websites fail to deliver results when processing complex search queries, which account for 17% of all searches, causing £8 million of lost revenue, according to research from Klevu.
Its research into product discovery methods revealed that there are usability and accessibility issues with the search experience on 80% of UK sites.
The main ones surround the ability to process complex or natural language phrases such as those involving price range or product attributes, or handle misspellings.
In the most critical cases, 26% of e-commerce websites couldn’t process simple misspellings and were unable to return relevant results and 30% showed zero results when faced with a query such as ‘mens jacket under £100’.
According to Klevu, this signifies that a vast majority of retailers are unprepared for the future of product discovery, including voice search, bots, and other zero interface interactions.
In addition, data from the Klevu platform shows that 17% of all searches are considered complex, using three or more words.
The data is based on an average number of sessions per month, average order value online in the UK, typical search-led sessions are around 20% of all web traffic, and conversion rate from search is often 5%, bringing potential lost revenue per month to £671k and £8 million annually.
These results are from the Klevu E-commerce Discovery Index which includes a review of 50 major UK retailers over 40 criteria including mobile search, voice search and natural language processing, personalisation, and product recommendations.
Made.com
Made.com has been using Klevu's product discovery technology since 2018 and has found a 10% increase in sessions with search and a 20% reduction in search refinements, meaning more shoppers are finding products from the first query.
Spencer Wong, Head of Digital Experience at Made.com, says: "I was really looking for a solution where we’re not having to build our own search engine.”
“Something that I could get off-the shelf that would be able to provide visual suggestions as well as text-based suggestions. That's what Klevu has given us and is able to do."
