With the advent of ChatGPT, generative AI — and what the technology could mean for businesses — has generated more headlines and predictions than possible to keep up with. But for every company jumping into the AI deep end, there are many more waiting to see how it’s best used.
Businesses are already using AI in a variety of CX functions — from contact-center automation to customer-service analytics. Generative AI is making a home in sales platforms, said Ejieme Eromosele, GM of EMEA at Quiq, a conversational CX and business messaging platform. Companies are building platforms to make it easier for customers to shop, especially if there are multiple variations of a product or an experience to sell.
Eromosele points to TUI, a U.K.-based travel booking company, which recently launched a generative AI experience in its app to help consumers curate a vacation, as an example.
Instead of overwhelming consumers with choices — and possibly decision fatigue — the customer is “guided by this AI so that they can reach an outcome that they're going to feel more confident in,” Eromosele said. The hypothesis is that a customer is going to be more likely to convert.
But businesses are still wary of bringing the technology directly to consumers. Experts who spoke to CX Dive say that companies are instead choosing to bring the technology to workers who work with customers.
Take a call center agent, who’s overworked and getting screamed at by people on the phone all day, Eromosele said. Generative AI can make their lives easier — and in turn benefit the customer’s experience on the call.
Companies have already begun using AI to provide a call agent information on what would be a relevant offer based on existing information the company has on the customer. Advancements in generative AI takes that one step further, offering conversational scripts and answers.
“Now imagine that you have a generative AI that can whisper to an agent,” Judy Weader, a principal analyst at Forrester, said. Generative AI can tell agents how to answer the question and provide them with the sources of information should they need to dig deeper into the topic. This not only speeds up resolution but also educates the agent for future calls.
Already, there are signs that generative AI can boost productivity. Agents who used conversational scripts created by AI increased their productivity by 14% on average, according to a study by Stanford University and MIT that looked at the number of issues resolved per hour.
The increase in productivity, however, was not consistent across skill levels. Less-skilled workers benefited more — with a “34% increase in the number of issues they are able to resolve per hour,” the study found.
But the effects were negligible — and in fact may decrease the quality of some conversations — with more skilled, experienced workers. Ultimately, the study concluded “AI assistance improves customer sentiment, increases employee retention, and may lead to worker learning.”
Generative AI can summarize calls too, so agents don’t have to spend the time manually writing summaries themselves.
For example, a user could ask an AI to pull together the top three takeaways of a call and then analyze call summaries to deduce pain points in the customer experience. Gartner also points to its use in closing the loop on customer issues, to “generate emails and texts to customers instantly and efficiently.”
As with any new technology, there are hiccups. Chatbots invent information at least 3% of the time — and as high as 27% — even in simple situations, according to research by the start-up Vectara. These errors, also known as “hallucinations,” are one of the reasons why Pete Jacques, a principal analyst at Forrester, advocates for companies to move slowly with the adoption of generative AI.
