Udesk's new feature: smart robots make a strong debut

The domestic enterprise-level SaaS customer service sector has been booming recently, with many major events. Udesk has been relatively active recently. In the customer service sector, Udesk has developed faster than its peers, and its professionalism has always been recognized by the industry. Its good customer reputation has also established a good public image for the company. The "2015 Customer Service Industry Data Analysis Report" just released two days ago has not yet dissipated, and today it was announced that the intelligent customer service robot has been officially launched, showing the style of a big boss in the customer service industry. The launch of the intelligent customer service robot also proves that Udesk has always been at the forefront of technology, and has brought a new customer service concept of "multi-channel integration + intelligent robots", reinterpreting intelligent customer service and opening a new chapter in the domestic enterprise-level SaaS customer service sector.

[[156932]]

According to the introduction, the algorithm of the customer service robot that Udesk just launched is based on natural language understanding semantic retrieval, multi-channel knowledge service and large-scale knowledge base construction. In terms of functions, it has eight functional modules: humanized Q&A, intelligent learning, detailed user management, personalized configuration, all-round real-time monitoring, in-depth data analysis and mining, multi-channel management system and customized expansion interface. The intelligent Q&A robot can solve more than 85% of common problems. In addition, the accuracy rate of Q&A matching and semantic understanding is as high as 97%.

Yao Guang, product director of Udesk, expressed his views on the robot that has just been launched. In terms of improving the customer service industry in the future, he said that robots rely on standardized interfaces and data to obtain services. The customer service industry has a large number of repetitive questions every day, and the answers to these questions are organized into a standard database, which is a knowledge base. The knowledge base is presented in the form of documents, which is the traditional help document. The interactive method is to view the classified directory, but no one likes to read documents, just like everyone doesn't like to read instructions. The same is true for customer consultation issues. By providing users with answers in the form of natural language dialogue, robots are more easily accepted by users, thereby improving the problem-solving rate.

In terms of the applicable scenarios of robots, he gave a more detailed analysis. He said: "After we understand the principles of intelligent robots, we understand that intelligent robots will not magically know the answers and help users solve problems. Instead, people are needed to fill the robot's knowledge base and continuously adjust the robot according to user feedback data. Both robots and manual customer service have costs, but the customer experience of both is relatively better than that of manual customer service. Therefore, choosing an intelligent robot is a balance between cost and customer experience. First of all, for particularly high-value customers, it is not recommended to use intelligent robots, and using manual customer service will have a better conversion rate. Companies should carefully analyze the customer's consulting scenarios based on actual conditions. We have an e-commerce partner who uses manual customer service to improve the sales success rate of non-standard products. For standard products, most of the customer's problems are related to logistics, and robots are used to connect to logistics. The service effect after the flow interface is very close to manual service, but the customer service cost has decreased. In general companies, after-sales service companies have limited resources to invest, and manual service always has to wait in line. We once had a partner test our product, first using an intelligent customer service robot to provide consulting services to customers, and most common problems can be solved by the robot. Afterwards, we conducted a survey on the customer satisfaction of this company and found that with the same number of customer service, the rate of immediate problem solving increased by 200% after the use of intelligent robot pre-service. After the customer service workload decreased, more time was invested in complex problems, and the problem solving rate increased by 15%. In short, intelligent robots are not perfect, but after reasonable use, companies can provide customers with better services at the same cost. "

Intelligent robots are very popular recently, and will become more and more popular. The basis for their popularity is that relevant technologies in the industry have indeed made breakthroughs, such as big data and machine learning. The industry leaders have also released some products one after another, including Google Now, Apple Siri, Microsoft recently built Cortana into Windows 10, and Baidu in China also placed a Xiaodu robot next to the search box. In the future, we expect robots to play a big role in the customer service field, and even use robots to increase sales. For example, when chatting with robots, users can be divided into after-sales support and pre-sales consultation through intent classification, and pre-sales consultation can be promptly transferred to manual customer service to improve conversion rate. The key lies in the "intent classification" technology.