Many retailers have started to plow full steam ahead on artificial intelligence. But consumers may not be ready for the next step.
New data from marketing automation platform Omnisend shows that 66 percent of consumers would spurn allowing AI to purchase items for them, even if it resulted in paying a lower price.
That type of transaction would likely require the intervention of an AI agent, which are designed to autonomously complete tasks after receiving guidance from a human user. For instance, a consumer might ask an agent to “Buy me a formal, short-strap red leather purse for less than $200,” and the agent would set about its task without the user.
Greg Zakowicz, senior e-commerce expert at Omnisend, said that, while it surprised him to see one-third of consumers would consider such an interaction, e-commerce is still a ways from successfully being able to offer that option to consumers.
“We are so far away from that, it’s not even funny,” he said. “It’s a tall task to ask consumers to put your credit card in, and let [AI] buy stuff for you.”
Four in 10 consumers reported facing struggles with the accuracy and usability of AI interactions on retailers’ sites, resulting in them abandoning a purchase. As such, almost half of consumers suggested that retailers focus on upping the ante on customer service, rather than on AI agents. Zackowicz said many retailers would do well to listen to that request.
“Customer service is a valuable asset to your company. If you have great customer service, that is how you differentiate,” he said, noting that, if companies plan to use chatbots to streamline customer service interactions, they should do so cautiously and with a keen eye on solving real problems with greater levels of efficiency.
Still, consumer sentiment around AI isn’t all doom and gloom. Some AI systems put into place by retailers, like those designed and trained to tackle personalization and discovery, have started to show promise for retailers.
Omnisend’s data shows that 38 percent of consumers said they’ve found product recommendations the most useful AI-powered feature retailers offer today, while three in 10 consumers said product discovery tools supercharged by AI were most beneficial. Zakowicz said he expects to see both functions improve, but noted that today, the AI-enabled chatbot and search functions working to show consumers relevant products don’t fully work to consumers’ advantage.
“Right now, according to this study, consumers are finding [AI] most useful in the things that have existed for a while—product recommendations, product discovery,” he said. “Make sure it’s actually improving [experience]. If it’s not improving, hit pause. Do something else.”
One surefire way to improve many AI systems is by training them with valuable, clean data—especially datasets that give some insight into the mind of the consumer. But today, consumers still lack a sense of trust in the security of their data when it’s being used by AI systems. The survey showed that 58 percent of consumers said they have concerns over how AI handles their data, and nearly three in 10 consumers don’t trust any company to appropriately handle their data.
Still, Zakowicz said, consumers will continue to hand over their data—namely because it’s the status quo.
“I think people, like me, go, ‘I don’t trust anyone; I certainly don’t trust AI to do it.’ But this is the kind of world we live in, and it’s just the cost of doing business,” he said. “I think people just become desensitized to it and trust you’ll do your best with [that data].”
However, once a brand or retailer has access to consumer information, it’s imperative for companies to show consumers they plan to use data responsibly, to build better customer experiences as they work to develop more complete strategies around emerging technologies, Zakowicz advised.
“Customers don’t expect you to be perfect, but they expect you to be good enough. And I think if you can build that trust and be good enough, it gives your room to kind of massage and maneuver those things as you move through,” he said.