Some companies have tried to make things better for their customers by removing the human factor from the equation, and they are doing this by fielding artificial intelligence robots (AI robots), known in industry parlance simply as “chatbots”.
However, the effectiveness of chatbots has been inconsistent at best.
The chatbot either guides the caller to someone who can deal with his problems or simply tries to do the problem-solving by itself. However, this is not exactly foolproof. Sometimes it causes a lot more frustration than brings relief to the customer.
The AI robot went into popular use in many countries in 2015, according to Chatbot Malaysia CEO Carliff Rizal Carleel. CIMB introduced the technology to field calls in 2019, making it the first business in Malaysia to have a computer talk to its customers.
But for all the hype, and despite the many years it has been in use, the AI robot is still “not that effective” in serving the needs of customers, Carleel told FMT Business.
The idea of having computers interact with humans began with the Turing Test. In 1950, British computer scientist Alan Turing developed a test to determine if robots could pass off as human, a test that came to bear his name.
A man way ahead of his time, Turing predicted that by 2000, machines would be able to fool 30% of human judges.
He was too optimistic. It would take another 14 years before that was achieved. In 2014, a chatbot named Eugene Goostman successfully fooled 33% of judges into believing that he was human. But that seems to be the limit of the AI robot’s capability … at least for now.
As an irate user once pointed out, it is one thing to get a computer-generated voice to guide you through a specific service, it is another to get the desired response.
This is especially challenging for the older generation, according to psychologist Dr Gerard Louis. He pointed out that while those born in the digital age have become very comfortable with interacting with AI robots, their parents are not.
“To the young the responses are instant and mostly appropriate, but for older people like me, it can be irritating sometimes,” Gerard, who is CEO of Help Education Services, told FMT Business.
It is for this reason that getting the buy-in from users was difficult initially. Carleel said that when he set up Chatbot Malaysia in 2017 to sell the idea of using computers to field calls, there were many challenges.
“It was not easy to get people to understand the concept of having a machine interact directly with customers,” he said.
“We had to put in a lot of effort to educate people about the advantages of using the chatbot.”
While it may be unwise to go entirely digital in customer service, there are benefits that businesses will undoubtedly gain.
According to DefinedCrowd, an AI marketplace, businesses could shave off up to 30% in customer service costs by turning to chatbots. Companies could see a return on their investments in just three years.
In a 2020 projection, DefinedCrowd said that by 2022, businesses would be able to cut about US$8 billion (RM34 billion) a year in customer support costs.
“For every second chatbots can cut average call centre handling time, there will be as much as US$1 million (RM4.2 million) in savings in annual customer service costs,” its CEO, Daniela Braga, said.
A survey on how artificial intelligence can revive retail through chatbot-based interaction shows that global sales are expected to rise from US$7.3 billion in 2019 to US$112 billion in 2023.
The study, by UK-based digital technology market research outfit Juniper Research, also shows that the number of chatbot interactions will rise from 2.6 billion to 22 billion in the same period.
The automation of customer interactions through chatbots is expected to save retailers up US$7 billion annually, according to the same research.
For better or worse, the artificial intelligence robots – call them callbots or chatbots – are here to stay.