Study highlights retailers’ lag behind consumer shopping behaviour

Retailers have been slow to update their portals and apps to keep pace with new consumer behaviours, such as multimodal search, AI shopping assistants, and even real-time stock tracking.

These are the findings of an international study unveiled by Google on 17 September at the NRF Europe trade show in Paris. Titled “Garage Retail”, the study examined 23 assessment criteria recorded in spring across 56 department stores, fashion and beauty retailers, as well as home furnishings and food specialists.

The study found, for example, that only 6 out of 56 retail portals return results when a word is misspelt, and that only 36 out of 56 sites return results for semantic searches (such as “dry skin” or “oversized look”).

Speaking to NRF Europe attendees on Wednesday, Mark Steel, director of industry and retail strategy at Google Cloud, did not hide his surprise.

“Only one of the sites we observed offers multimodal search (combining image and text, editor’s note),” said the executive, noting that Google itself is seeing growing adoption of this feature among internet users.

There is still some way to go on personalisation, too, with only 18 of the portals surveyed remembering previous searches to better target recommendations. By contrast, nine portals, including two in the fashion/beauty segment, are said to offer a high degree of personalisation. And only nine portals provide a real-time view of stock.

On product pages, the study notes that only 23 of the sites examined (including eight in fashion/beauty) fail to provide detailed information about the product. This represents a potential avenue for AI providers looking to generate product listings, suggests Mark Steel. The study also reveals that only four of the 40 sites offering customer reviews use AI to produce summaries of the comments.

When it comes to payment, 34 out of 56 sites offer five or more payment methods, and 27 indicate same-day or next-day delivery options directly from search. By contrast, only 18 offer a wide range of delivery options. A further 13 adapt their language according to detected emotion.

Chatbots are also covered. Of the 29 portals that use them (including 14 in the fashion/beauty sector), 22 support natural-language interactions, avoiding robotic exchanges. However, only one of the chatbots observed offers a personalised shopping assistant function.

“Consumers are clearly in favour of this innovation,” said Google Cloud. “In fact, 61% of consumers say they expect more personalised services thanks to AI”.