(AP) — As the holiday season approaches, many shoppers are seeking inspiration for gifts. Artificial intelligence chatbots may offer a solution. However, while they can assist, they are not infallible and often require human oversight.
Many consumers are expected to turn to online platforms in search of Cyber Monday deals, where they will encounter chatbots designed for improved customer interaction.
Several retailers have begun utilizing advanced generative AI technologies. This innovation allows customers to engage with chatbots through everyday language, asking questions like, ‘What’s the best wireless speaker?’
Retailers aim for these chatbots, often termed shopping assistants, to serve as virtual advisors, helping identify and compare products, a shift from earlier versions that mainly managed order tracking and returns.
Amazon has launched a generative AI-powered shopping assistant named Rufus, designed to answer questions about products such as coffee makers or suggestions for children’s birthday gifts.
This year, select Walmart shoppers will also experiment with a comparable chatbot, currently being tested in areas like toys and electronics.
Perplexity AI recently introduced a new feature allowing users to ask for specific product recommendations, like ‘What’s the best women’s leather boots?’ with results that are non-sponsored, as mentioned by the company.
‘It has been adopted at pretty incredible scale,’ remarked Mike Mallazzo, an analyst at Future Commerce.
In late 2022, as ChatGPT gained popularity, many retailers recognized chatbots’ potential, fueling interest in the generative AI technologies that secure user interaction.
Companies like Victoria’s Secret, IKEA, and Instacart have also started exploring chatbot integrations powered by OpenAI.
Although online retailers have long used purchasing histories to create tailored recommendations, Rufus’s ability to converse and ask follow-up questions enhances the shopping experience.
However, chatbots do have limitations and can produce inaccurate results. For example, Juozas Kaziukenas, founder of Marketplace Pulse, noted that when Rufus was asked about gaming TV recommendations, its suggestions included unrelated products.
During a trial, Rufus provided holiday gift ideas for a brother, ranging from T-shirts to multifunctional knives. Although it offered some tailored suggestions after further questions, it failed to compare prices accurately.
According to Amazon’s Rajiv Mehta, the company is progressively enhancing Rufus by allowing it to engage customers in meaningful dialogue, thus improving deal searches and personalizing recommendations.
Shop AI, launched by Shopify, employs a similar approach but struggles with precise product suggestions.
Moving forward, true transformation in the shopping experience hinges on developing deeply personalized tools that remember customer preferences and purchase histories, according to a report by McKinsey & Company.
