Drawing: Dai Yao (Image generated with AI assistance Data source: Social Survey Center of China Youth Daily)

From e-commerce platforms to food delivery, from travel navigation to government services... Nowadays, with the rapid development of artificial intelligence technology, intelligent customer service has appeared on major platforms. Recently, a survey conducted by the Social Survey Center of China Youth Daily in collaboration with Wenjuan. com, involving 1334 participants, showed that 62.9% of respondents felt that the use of intelligent customer service is more common now. 70.4% of respondents expect intelligent customer service to improve their ability to solve complex problems.

Sometimes it seems like there's no other choice. "Qi Min, a post-90s employee working in Sichuan, found that when encountering problems nowadays, it's usually the intelligent customer service who answers first, and only transfers to manual customer service if they can't solve them." Although some simple problems can be handled by intelligent customer service, it's usually the manual customer service who can solve them.

With the development of artificial intelligence technology in recent years, Zhang Xinxin, a post-95s working in Beijing, has found that many customer service jobs have been replaced by artificial intelligence. "Nowadays, AI is becoming increasingly intelligent and can solve more and more diverse problems. But she also believes that there is a consideration for businesses to "reduce costs and increase efficiency" behind this, by using intelligent customer service to reduce some communication costs.

However, some merchants now set up the entrance for manual customer service very discreetly, clicking on many platform customer service buttons, and basically connecting to intelligent customer service. Zhang Xinxin is very troubled by this situation, and she thinks that merchants should make the distinction between manual customer service and intelligent customer service more obvious. "Sometimes when you want to find manual customer service, the operation process is really too complicated.

A survey shows that although over half (50.2%) of respondents tend to choose manual customer service when using customer service functions, 62.9% of respondents feel that intelligent customer service is more common now.

Qi Min recently wanted to sell the idle goods she purchased in 2020, but couldn't find the invoice. At first, she wanted to solve the problem through the intelligent customer service on the product's official website, but before the issue was resolved, she switched to the next intelligent customer service and said, 'I have to describe the problem again.'. It wasn't until the end that the problem was resolved through manual customer service, "I think it's more convenient for manual customer service in terms of after-sales service or complaint and rights protection.

A few months ago, when Zhang Xinxin was riding a shared bike, her clothes were rolled into the wheels and damaged. She found that the platform where she purchased the clothes provided relevant after-sales support, so she tried to apply for a claim, but she only found the intelligent customer service entrance on the page. Finally, after countless applications to switch to manual labor, she was able to communicate with the manual customer service. "In the end, it was the manual customer service who helped me solve the problem through layers of feedback.

When encountering some simple problems, Yang Li, a post-95s from Shandong, tends to use intelligent customer service, such as "checking logistics or understanding product details". She thinks that many e-commerce platforms have trained their intelligent customer service quite well nowadays.

Qi Min also believes that in simple situations such as information inquiry or operation guidance, intelligent customer service can mostly solve problems. "Sometimes when applying for a refund, it can also be quickly resolved through intelligent customer service.

This survey investigated the use of human customer service and intelligent customer service by respondents in different scenarios. Data shows that more respondents choose manual customer service for after-sales service, complaint protection, and emergency situations, while more respondents choose intelligent customer service for information inquiry and operation guidance.

Specifically, when faced with account theft or abnormal deductions, 70.6% of respondents prefer to choose manual customer service; When it comes to complaints and rights protection, 74.0% of respondents tend to use manual customer service; 66.1% of respondents tend to choose manual customer service when it comes to refunds, exchanges, and returns.

When searching for price, logistics, and other information, 42.2% of respondents prefer intelligent customer service, 30.1% choose manual customer service, and 27.7% of respondents indicate that both options are acceptable; When requiring guidance on account registration and other operations, 42.5% of respondents would choose intelligent customer service, 32.2% would choose manual customer service, and 25.3% of respondents would agree.

Yang Li believes that the "reading comprehension" ability of intelligent customer service needs to be improved. "Everyone encounters different problems and situations, and some intelligent customer service representatives only capture the keywords and automatically generate replies, which are not actually the problems you described. She hopes that with the development of artificial intelligence technology, intelligent customer service can help users solve more complex and personalized problems in the future.

Zhang Xinxin hopes to simplify the process of transferring to manual customer service by clearly distinguishing between manual customer service and intelligent customer service entrances, so that users can easily and quickly make choices. In her opinion, human customer service and intelligent customer service need to cooperate with each other. "Both have their own advantages, and if they can be better used together, they can help everyone solve problems faster.

What aspects do you expect intelligent customer service to improve in? Data shows that 70.4% of respondents expect to improve their ability to solve complex problems, 62.5% of respondents expect to improve the accuracy of problem understanding, 56.9% of respondents expect to increase humanized and emotional communication, 56.6% of respondents expect to simplify the process of transferring manual answers, and 43.0% of respondents hope to reduce template based and mechanized answers.

Among the respondents, those born in 2005 accounted for 2.8%, those born in 2000 accounted for 23.7%, those born in 1995 accounted for 23.3%, those born in the 1990s accounted for 26.1%, those born in 1985 accounted for 19.3%, and other age groups accounted for 4.8%.

(Yang Li and Zhang Xinxin are pseudonyms in the article)

Reporter Dai Yao from China Youth Daily and China Youth Net Source: China Youth Daily

Version 08, July 9th, 2026