AI recommendations in retail: customers expect hyper-personalization

Almost one in two people now order something online at least once a week. Online shopping is popular, but is undergoing radical change. «Customers want more than functioning technology and a large selection of products,» says Marius Browarczyk, Head of Retail at Lufthansa Industry Solutions. «The keyword is hyper-personalization.»

According to a recent survey by Lufthansa Industry Solutions of almost 1,100 people in Germany, 81 percent use the product recommendations displayed on the platforms. However, only one in ten online customers is really enthusiastic about it. The respondents hope that the increased use of artificial intelligence will help.

Customers are satisfied, but demanding

89% of Germans are very or at least somewhat satisfied with their online shopping experience. However, the issue of advertising is a cause for concern: 44% of respondents complain that they receive too much of it, and 39% even feel that they are downright «persecuted». Many would prefer real advice instead of advertising, but this is often in short supply online.

There is also an increased need for advice when buying online, particularly for electronics, travel, medicines and DIY products. For electronics, 36% of customers would like advice, 24% for travel or tickets and 22% for medicines and healthcare products.

Chatbots not yet established

Advice in online stores is often provided by chatbots, but these have not yet become established. Only five percent of respondents use them very frequently, with a further 15 percent using them fairly frequently. More than a third (36%) never use chatbots, 44% rarely. However, they have potential, as 60 percent of users are somewhat or even very satisfied with them.

Recommendation functions are more popular. They are used frequently or very frequently by 38 percent of respondents. Only 19 percent do not attach any importance to product recommendations. In terms of satisfaction rates, the recommendation systems perform slightly better than chatbots: 10% are «very satisfied» and 62% are «somewhat satisfied».

67 percent see advantages through AI

Overall, 67% of respondents see benefits from the use of AI in online shopping. For example, 36% expect it to save them time and effort, and 27% assume that they will receive more suitable offers thanks to AI. A further 31% expect to find the desired products more quickly, 26% hope for more accurate information and 24% for optimal recommendations.

However, there are clear differences between the age groups. Among the younger respondents up to the age of 29, only 17% are skeptical, while in the 60+ age group, this figure is almost half (47%). Generation Z and the generation following it, Generation Alpha, are particularly open to AI-supported offers - they are the financially strong customers of tomorrow.

Transparency and data protection are key

However, 93% of online customers want to decide for themselves which of their data is used. And for 89 percent, it is important to know when and what AI is used for when shopping online. Eight out of ten respondents also want to understand why certain offers are presented to them.

«These figures show that consumers only accept AI if it delivers real added value and is comprehensible,» emphasizes Tobias Kämpf, Product Owner AI in Retail at Lufthansa Industry Solutions. Almost 90 percent of those surveyed would prefer to be able to reach a real person when needed.

Data quality as a success factor

«Retail faces many challenges. A clean data structure is crucial, without which there can be no successful AI applications,» says Marius Browarczyk. With clean data, retailers can improve the customer experience, work more efficiently with suppliers and secure further steps in digitalization.

Tobias Kämpf adds: «The main prerequisite for a successful AI recommendation system is reliable data. Customer data, article master data and interaction data are the basis on which the system recognizes semantic correlations and generates suitable recommendations. Without clear product descriptions or sufficient history, the desired precision would be lost.»

Shape: all-in-one solution for retailers

With Shape (Semantic Hybrid Algorithm for Personalized Engagement), Lufthansa Industry Solutions has developed an all-in-one solution that provides a recommendation engine and a digital sales assistant in the form of a language model chatbot. The associated product search is not based on keywords or tags, but on the semantic understanding of the item details and the customer's query.

Shape combines the control of personalized product recommendations and customer advice in one tool. It is a white label solution that remains invisible to the end customer. It therefore does not replace the web store, but extends its functionalities.

The new white paper from Lufthansa Industry Solutions shows how AI-supported recommendation systems help retailers to create real added value with personalized recommendations. It can be downloaded free of charge.

More information: www.lufthansa-industry-solutions.com