36Kr learned that Yi chat technology, an AI customer service saaS service provider focusing on online customer acquisition scenarios, has recently completed a B round of financing of 100 million yuan. It is reported that this round of financing is led by CYTS Hongqi Fund, and the A round of investors Mingyu Venture Capital follows.
According to public information, Yiliao Technology was established in 2014, with online customer service SaaS software as the main product in the early days, after 2018, Yiliao Technology began to enter the AI customer service robot market, and launched a series of intelligent customer service solutions for online customer acquisition scenarios.
Online customer acquisition AI customer service robot is a typical AI + MarTech (marketing technology) product, in the eyes of practitioners, this is a landing scenario with considerable commercial value - the target enterprise has clear evaluation indicators, the product value is quantifiable, the product effect is obvious, and the awareness of payment is strong...
According to the "China Intelligent Customer Service Industry Research Report" released by whale research institute, the current stock market size of China's customer service software is 10-15 billion yuan, and the future AI intelligent market space is expected to reach 50-80 billion yuan. Among them, the market space of AI customer service robots is 20-30 billion yuan.
A few days ago, 36Kr chatted with Liu Xincheng, CEO of Easy Chat Technology, trying to sort out three problems:
1. What kind of track is intelligent customer service? Why is it so sought after by capital?
2. In the intelligent customer service track, what are the characteristics of the online customer acquisition scenario? Why has it become a separate track with great commercial value in the entire intelligent customer service market?
3. For online customer acquisition track intelligent customer service vendors, what are the opportunities to stand out in this track?
NLP technology is mature, and intelligent customer service has entered the AI era
In 1950, the British mathematician Turing, who was later known as the "father of artificial intelligence", published the classic paper "Computer And Intelligence", and for the first time proposed the classic question of "can a machine think". Now, 70 years later, the question is clearly answered — machines have learned to think to some extent, and even compete with humans in certain positions, and customer service is one of them.
Intelligent customer service has become a major scenario for the landing of artificial intelligence, which is divided into two main areas: voice intelligent customer service and online text intelligent customer service. Among them, online text intelligent customer service can be divided into after-sales service intelligent customer service, e-commerce shopping guide intelligent customer service, and online customer acquisition intelligent customer service according to application scenarios. According to Yiliao Technology, it is the first manufacturer to launch an online customer acquisition intelligent customer service robot. At the same time, as the head brand of the vertical track, it has attracted considerable attention from capital.
In recent years, the head brand of the intelligent customer service vertical track has been highly sought after by capital. Especially since 2020, the amount of financing in the intelligent customer service track has increased significantly. IT Orange data shows that in 2019, the total financing amount in the field of intelligent customer service was 1.443 billion yuan, while by 2020, this figure will grow to 3.326 billion yuan, achieving more than 1 times the growth.
The reasons why intelligent customer service is sought after by capital are not complicated. Customer service has always been a manpower-heavy industry, and the emergence of AI intelligent customer service technology has brought new solutions to the customer service industry. Early smart customer service products could trigger fixed replies based on specific word rules, for example, whenever a consumer mentions the word "return," bot agents automatically send a return request process. This simple intelligent customer service product has been able to help after-sales service customer service answer some simple and repetitive questions, reducing the workload of labor with a small amount.
In recent years, the maturity of AI technologies such as NLP (natural language processing) has brought new development opportunities to the intelligent customer service industry. Different from the rules-based intelligent customer service software, the NLP-based AI intelligent customer service can understand the natural language of human beings, and generate reply words through natural semantics, "can think, can speak", so that intelligent customer service can enter the AI era.
Online customer acquisition, AI customer service landing is an excellent field
The maturity of AI technology has enabled intelligent customer service to rapidly grow into a new market of nearly 100 billion.
In the entire intelligent customer service market, Easy Chat Technology chooses to enter the segment of online customer acquisition. In Liu Xincheng's view, this market is the most commercially valuable area in the entire intelligent customer service track, and the customer's demand for products is clear, the input-output ratio is controllable, and there are relatively clear return indicators.
Different from the after-sales service AI customer service that consumers are familiar with, the online customer service for customer acquisition focuses on improving the online "traffic-service-conversion" chain of enterprises. For example, a stomatological hospital put the keyword "orthodontics" on the search engine, and after searching for the keyword, potential customers will enter the landing page built by the hospital, and it is difficult for potential customers to complete the purchase closed loop just by displaying information on the landing page. At this time, if there is a customer service that can follow up on the purchase needs of consumers and take the initiative to introduce relevant products and services to consumers, it can significantly increase the traffic conversion rate.
It is not difficult to see that in the online customer acquisition intelligent customer service scenario, the conversion rate is the core indicator pursued by enterprises. Intelligent customer service constitutes the last link in the entire traffic conversion chain of enterprises, playing the role of professional shopping guides, consultants and businesses in traditional stores. An excellent online customer service team can obviously significantly improve the conversion rate and ROI of the enterprise, however, relying entirely on manual completion of the online customer acquisition conversion work is obviously unable to adapt to the business needs of the enterprise.
For example, during weekends, nights, and holidays, most of the company's manual customer service resources are limited and may not be able to meet the traffic conversion needs; in the online marketing customer acquisition scenario, the artificial customer service will inevitably face the problem of multiple concurrent multi-visitor consultation, which seriously reduces the conversion rate and wastes traffic; in addition, the online customer service needs to master a lot of professional knowledge, the training cycle is long, and the level of business ability between customer service is also uneven, and in the context of rising labor costs, the cost of building an excellent customer service team is extremely high.
More importantly, today, with the development of China's Internet industry entering the second half, the cost of enterprises to buy traffic has increased significantly compared with before, and the traffic is becoming more and more expensive, so enterprises naturally have to work the conversion rate. Compared with direct purchase of traffic, digital transformation methods that can effectively improve conversion rates are beginning to show cost-effective advantages.
AI customer service services for online customer acquisition scenarios came into being. Different from after-sales service customer service, online customer service belongs to the revenue department within the enterprise, not the cost department, so its commercial value is more significant. "Unlike after-sales service customer service, which pursues the problem resolution rate index, online customer service pursues the final conversion rate. Through a series of R&D investment and product iterations, we have achieved a conversion rate comparable to or even better than that of human customer service. The head customers of many industries have realized the unmanned reception of all traffic through our products, and the effect of cost reduction and efficiency increase can be imagined. Liu Xincheng said.
But this effect is not easy to achieve. For example, compared with after-sales service intelligent customer service, online customer acquisition robots must adopt a more proactive service strategy, actively guide customers to understand products and services, and ultimately promote the purchase of closed loops, rather than just answering customer questions in the form of one question and one answer. This means that online customer acquisition AI customer service manufacturers need to have relevant service experience and knowledge engineering accumulation in various industries, and they also need to improve the technical ability of multi-round conversation algorithms.
The conversion rate is king, how can the intelligent customer service provider win?
In order to improve the conversion rate of AI customer service, AI technology itself is obviously an unavoidable topic.
At the technical level, AI customer service must be both "able to understand" and "speak", and is a complex product that integrates NLU (natural language understanding) and NLG (natural language generation). In order to polish the product of online customer acquisition AI customer service, Easy Chat Technology has developed multi-round session guidance, knowledge base automatic construction and other functions, so that AI can quickly grasp the relevant knowledge of specific industries, and can carry out multiple rounds of smooth conversations with visitors, and actively guide visitors to complete consumption conversion.
In this process, intelligent customer service providers not only need to have excellent algorithm capabilities, but also need to be "fed" with a large amount of data. At this level, Easy Chat Technology obviously has unique advantages, before the launch of AI customer service products, Easy Chat has been working in the online customer service software industry for many years, with customer service corpus resources covering multiple industries, these data have also become the basis for Easy Chat to enter the AI market.
According to the official data of Easy Chat, as of now, Easy Chat has accumulated 3,000 customers, with 900 million+ labeling corpus, and 10.2 billion + conversation records. The total number of reception sessions of its AI customer service robot has reached 768 million. Through these data, based on deep learning technology, e-Chat extracted 80,000+ topical conversation scenarios and 170,000+ intent recognition feature types, and the conversation intent recognition rate can reach more than 98%.
At the same time, at the product level, Yi Chat's intelligent customer service products also have the ability to dock through all-traffic channels, which not only support 40+ head public domain media channels such as Baidu, Douyin, Today's Toutiao, Kuaishou, Guangdiantong, 360, Sogou, Lemon Aimei, etc., but also dock private domain traffic such as APP, WeChat, Mini Program, and public account, which can meet the diversified traffic docking needs of enterprises.
Excellent algorithm technology foundation, as well as massive data accumulation, can become a barrier to an AI customer service provider. On the one hand, the best algorithm and the richest data can polish the conversational ability and conversion ability of customer service robots, and higher conversion capabilities will inevitably get more enterprise customers, and more customers will further improve their data accumulation, and further feed back algorithm research and development, and eventually form a virtuous circle. In this process, the head of the AI customer service providers will further consolidate their own advantages, industry resources will continue to be concentrated in the head enterprises.
In addition to products and technologies, market and service capabilities are also the key to the success of intelligent customer service providers. In terms of services, in order to help enterprises use AI intelligent customer service tools, Yiliao Technology will equip enterprises with exclusive AI trainers to help enterprises complete the construction of professional knowledge libraries, and continuously optimize and iterate data models.
Through the self-developed BOSS (Business & Operation Support System) operation system, EasyTalk Technology realizes the digital management of the whole business process. For example, the customer success team "foresees" customer problems through "customer health indicators" and diagnoses based on many dimensions such as traffic, conversation volume, conversion rate, etc., to help customers optimize the solution and improve the traffic conversion rate.
Liu Xincheng revealed that the current customers of Yiliao Technology have covered education, medical care, finance, home improvement, tax law, corporate services, management consulting, information services, technology software and other industries. In these areas, due to the accumulation of sufficiently rich industry databases, Easy Chat can complete the deployment of AI robots for customers within 1 day at the earliest.
In terms of market model, the intelligent customer service service of Easy Chat Technology is delivered in a standard non-customized SaaS model, and its payment model provides a more scientific and reasonable and more user-friendly per-session payment model for small and medium-sized enterprises in addition to subscription payment.
The investment of E-Chat Technology in the fields of technology, products, services and markets has also received direct returns. According to the investment bank due diligence data revealed by Yi Liao, in the three years from 2019 to 2021, Yi Chat has achieved nearly three times the revenue growth, with the renewal rate of M12 subscription customers exceeding 80%, and more than 95% of the revenue comes from AI robot products rather than traditional customer service software. In addition, Liu Xincheng revealed that at present, about half of the customers of Easy Chat are from referrals. As the pioneer and leader of the online customer acquisition intelligent customer service track, EChat has established its own customer reputation.
According to the official data of Easy Chat, online customer acquisition AI robots have effectively increased the traffic conversion rate by 5%-60% in different industries, and effectively improved manual efficiency by more than 50%. Taking Jinsong Dental, the head brand of Beijing Dental Chain Agency, as an example, after the launch of the Yiliao AI robot, its "conversation-lead" conversion rate increased from 32% to 43%, and the labor efficiency was doubled.
It is this kind of effect performance that makes customers choose the products of Easy Chat Technology. At the same time, and more and more customers will further enhance the comprehensive ability of easy to chat, as mentioned earlier, more customers to bring data and algorithm improvement, will push up the conversion effect of the product, and the effect will bring more customers, data, algorithms, customer trinity of positive cycle model, will make the industry show the stronger the stronger, winner-take-all Matthew effect.
The intelligent customer service market is full of smoke, where is the road to online customer acquisition?
In the future, the intelligent customer service head brand represented by Easy Chat Technology will continue to focus, with their accumulated experience in the customer service industry, and achieve a breakthrough at a single point such as online customer acquisition scenarios.
Liu Xincheng told 36Kr that in the future, Yi chat technology will still uphold the concept of "value first, achieve customers", adhere to the scene of online customer acquisition, continue to cultivate deeply, and repeatedly polish around this "super single product".
In terms of conversion rate indicators, the AI customer service of Easy Chat Technology has achieved the leveling or even surpassing of human customer service, "especially the model of human-machine collocation, which has a conversion rate of 5%-60% higher than that of the traditional manual model." Liu Xincheng said.
In the intelligent customer service industry, the working mode of human-machine collocation has become another major trend in the future. Kate Leggett, an analyst at Forrester, a market research institute, wrote that the integration of AI into the customer service center has become an inevitable part of the customer service industry - AI is not a complete replacement for human customer service employees, but can help improve the working ability and efficiency of human customer service.
For example, AI can simplify the indexing process of human customer service, AI-specific insights can help guide customer service conversations, and AI can also actively guide visitors to speak, and remind human customer service to communicate with customers after the visitor opens.
And Easy Chat Technology has long integrated this model into its own products. Nowadays, the intelligent customer service robot of Easy Chat not only realizes the robot independently from the manual duty in practical applications, but also supports a variety of human-machine collaboration modes - in the night, holidays and other unmanned periods, AI customer service can independently complete customer reception and conversion; in addition, Easy Chat also supports the human-machine collaboration mode, the system can assign the dialogue to the human or robot reception according to the different sources, conversation content, intention and other tags of the visitor, and share the workload of the human customer service in the human-robot collaboration mode. Solve the problem of decreased labor efficiency caused by high concurrency.
It is not difficult to see that in the future, AI customer service will be able to completely replace human customer service, but completely replacing human service is a gradual process, at present, the biggest significance of AI customer service is to liberate human customer service from tedious work and deal with higher-level work. Looking at the various scenes of AI landing, it is more like the best partner of human customer service.
![Poor communication, non-answering... The "unintelligent" intelligent customer service makes consumers upset[fig]](data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=)