While it’s just getting started being put to serious use, GenAI is already beginning to transform online retail customer service. Because GenAI can gather insights from unstructured data, structured data, customer chat histories and more — these models can turn all of that information into accessible knowledge that will streamline and enhance the experience and efficiency of customer communications before and after the sale. This is especially so with chatbots — which have historically overpromised and underdelivered in their customer service promises.

To uncover consumers’ views of AI and learn how their attitudes toward AI may impact the current holiday season, conversational AI provider LivePerson surveyed 1,000 consumers. According to the respondents, the essential elements of the online retail experience during the online holiday shopping season include quick response time to inquiries, accurate responses to their questions and concerns and quality product recommendations. Further, half of the respondents said they would use an AI-powered chatbot to get help with suggestions and answer other questions for general holiday shopping.

The investments in AI customer service efforts are occurring in parallel with investments in supply chain optimization, personalization, fraud reduction and more.

This may be occurring now because of the rapid improvement of chatbots and virtual assistants due to AI-enhanced data curation and dissemination during the past few years. “Initially, based on keyword matching, they frequently failed to meet consumer expectations. AI allows these tools to collate information such as customer history, consumer patterns, and customer loyalty to identify patterns and recommend solutions or options more suited to a particular customer,” Karl Cama, senior chief architect, office of the CTO at Red Hat.

“Instead of getting the default response based on a keyword, the consumer can get a personalized response based on loyalty and history. They can also communicate with customers in multiple languages to ease the language barrier. This makes the consumer feel valued and rewards loyalty, resulting in increased customer satisfaction and personal recommendations to peer groups,” Cama says.

Reshma Iyer, head of product marketing at search-as-a-service provider Algolia, agrees that AI can considerably drive customer service efficiencies. “Many inbound customer service queries are typically repeatable responses and can be easily retrieved and fielded by the system,” Iyer says. “In many cases, this interaction can also continue async. For queries with more complexity, human intervention or triaging can be set up wherein the conversation is transferred at a certain point to a customer support staff. Due to the volume of this type of queries, the impact is almost instantaneous in terms of freeing up agents,” Iyer says.

“AI in the form of a conversational assistant can result in a rich and powerful customer experience as the AI plays the role of a trusted “guide” taking the customer through a journey that showcases results that are most likely to resonate. This is a creative and deeply engaging approach to bringing a shopper closer to a set of products they are likely to purchase,” Iyer adds.

Additionally, Red Hat’s Cama adds call center chatbot technology that can route calls to a specific set of individuals based on the nature of the call and the questions being asked. “Such that the customer receives accurate information on the first call and walks away with enhanced customer experience. In these situations, the model must [continuously] improve based on ongoing customer feedback and issue resolutions. Machine learning must be fast and accurate with minimal delay for the deployment of a new machine learning model to the AI environment. This means the machine learning should be deployed to the AI as frequently as possible, maybe even daily or hourly,” says Cama.

Considerable AI Challenges Remain

Still, the use of GenAI in customer service isn’t without challenges. Notable security and regulatory concerns persist. There’s the data scientist talent shortage and the accuracy of the data being fed into the models. According to Anthony Deighton, data products general manager at data products provider Tamr: Chatbots, and virtual assistants are trained on large amounts of data, including customer conversations, product information and frequently asked questions. This data is then used to teach the chatbot or virtual assistant how to understand and respond to customer queries. Often, chatbots are used to handle customer queries that are simple or repetitive, freeing up time for human customer service reps to take more complex or challenging questions.

“However, if the data used to train the chatbot is dirty or inaccurate, the chatbot/virtual assistant will not be able to learn properly and cannot provide accurate or helpful responses. This can result in longer wait times for a human to intervene, customer frustration and decreased customer satisfaction,” Deighton says.

Cama agrees. “Done incorrectly or poorly, they quickly become a negative experience for consumers,” he says.

Of course, GenAI is having an impact on online retail beyond front-line chatbots. “AI is being used in back-office settings to automate disputes, manage chargebacks, and streamline document processing for retailers and financial sectors,” says Jason Bohrer, executive director of the U.S. Payments Forum.

Finally, the impact of GenAI on retail e-commerce is just getting started. “Shortly, we anticipate many of the related technologies will come together in a seamless conversational format, which will fundamentally alter the customer service. Whether using text, videos, images, or voices, the customer’s reason to contact will be understood quickly and automatically. Their current history and other details will be connected in context and automatically, and aided by generative AI, the virtual agent will be able to ascertain promptly the most appropriate path to follow with the specific situation on hand,” Algolia’s Iyer concludes.