Are retailers losing sales to automation? 39% of shoppers have abandoned their purchase due to frustrating AI chatbots

AI-driven customer service tools are designed to make online shopping more efficient, yet many are walking away from purchases frustrated. A recent survey by Omnisend has found that AI support in ecommerce is one of the biggest concerns for online shoppers – 40% cite lacking customer service as AI’s biggest drawback. 

Further 39% have even abandoned a purchase due to frustrating chatbot interactions. Regardless of mixed reviews, 88% of consumers had at least one conversation with a chatbot in the past year.

“Given high abandonment rates, AI chatbots might cost retailers more than they save,” says Paulius Milišauskas, the VP of Customer Experience at Omnisend. “However, most of the time, the problem lies in poor implementation.” 

“Retailers often deploy chatbots without a good understanding of their customers’ needs, resulting in frustration instead of actual problem solving,” Milišauskas adds. “AI falls short when we forget it’s supposed to enhance human support, not replace it.”

Half of customers are asking for better support, only a third feel understood by AI chatbots

As more businesses embrace AI tools in customer support, the gap between efficiency and customer satisfaction grows. When asked about AI features they’d like to see improved, 48% of respondents pointed to customer service quality, transparency on data usage came in at a distant second with 36%.

In total, 53% rated their experience with AI agents as average to extremely poor. When asked about their general attitudes towards AI in ecommerce, only 28% of shoppers expressed that AI consistently understands their needs.

“Retailers have to reconsider what efficiency means in customer support. Fast responses from AI chatbots may appear productive, but speed without accuracy only worsens customer dissatisfaction,” says Milišauskas.

“Our findings reveal that shoppers are sensitive to how effectively – not how quickly – their problems are understood and solved. So, businesses should rethink how they evaluate AI support, measuring success by the accuracy of problem solving and reduced repeat inquiries rather than resolution speed alone.”

How can retailers make AI work for them?

By 2025, the number of businesses using AI chatbots is expected to grow by 34%, and by 2030, AI may even be handling 80% of all customer interactions. What can businesses do to make AI work for them instead of against them? 

Milišauskas suggests viewing AI as a customer experience tool rather than merely an efficiency tool. “Most problems with AI chatbots arise when we focus too much on efficiency alone instead of how to use it for customer satisfaction.” 

“Having a hybrid customer support model is non-negotiable,” Milišauskas comments. Chatbots may be faster, but they also lack a nuanced understanding of human sentiment. “While AI excels at handling routine queries, customers are usually happier engaging with a human representative.” So, always leave the option to connect with a human agent – especially when it comes to more complex issues. 

Personalization is equally as important. “Our research has shown that AI does improve product recommendations and help customers easily find desired items.” To take full advantage of this, he suggests “continuously analyzing customer behavior and sentiment to tailor each customers’ shopping experience” to make it more relevant and reduce friction.