The hottest real intelligent customer service know

The real intelligent customer service knowledge base needs to do these three core tasks

many users of intelligent customer service don't know the so-called intelligent customer service at all. Most of them just extract keywords from their own content for search, and many of the feedback content still can't solve the problem. Intelligent customer service personnel also feel that although they spend great efforts to refine and reduce the granularity of knowledge content, in addition to clear and specific problems that can be solved, there is still nothing they can do about slightly complicated and inflection problems, which is far from true intelligence

why is this the case? How to search for real intelligent customer service from keywords? There are two perspectives that we can work on. The first is the breakthrough in technology. We can really understand the language and subtext expressed by customers. In addition to answering clear questions of customers, we can also solve potential problems (problems that customers cannot express due to lack of information and knowledge). But the breakthrough in technology is not overnight, and there is no good solution in this industry; Another effort is to define, borrow, analyze, and organize answers based on scenarios for recurring problems that have a wide range of impacts through the analysis ability and participation of operators. From the current situation, the second way of effort may be more reliable. Only high-level human analysis + technical assistance can provide further intelligence

specifically, it includes three levels:

First: understanding customer needs. In addition to understanding the explicit language and words expressed by customers, we can also understand the potential and deep-seated needs of customers

second: clear problem definition and countermeasures. If the problem can be clearly defined, it can also be solved. Based on the number of times customer problems are raised and the number of people affected, first analyze the most common problems to find the core problems. Build a solution model based on problem solving

third: matching of knowledge resources: maps and scenes. Match the problem solving model with knowledge resources and solving strategies, and feed back to customers when they need it

let's talk about it in depth:

1. One of the most criticized statements of girlfriends is: drink more hot water

this is the core reason why Wang is single: he can't figure out the subtext behind the girl's language

the reason why women express this is determined by gender and personality factors. If single people want to capture the heart of the goddess, married people need to understand the needs behind the language if they want long-term stability of the family. The customer's language or the keywords described in intelligent customer service are similar, but the reason is that they have little understanding of products and services, lack of background knowledge, and are unable to raise their own accurate and objective questions. As a result, even if the customer asks what you can answer, the customer is still dissatisfied

we can't ask the end customer to have better language expression ability. Most of the time, the customer will only describe the phenomena and results he sees. Many times, when D1 is less than 5, the size deviation and appearance quality are qualified; The reason for low efficiency is that the agent representative or intelligent customer service system can't understand what the user is saying, which includes two problems. The first is that the customer's language and vocabulary are different from the official ones. They only use spoken and daily idioms, while our knowledge base stores written or even professional statements, At this time, we need to customize the content. We need to establish a customer language vocabulary to correspond with our official language, and train the seat representatives or establish corresponding associations in the intelligent customer service library

2. For questions with simple, clear and mature answers, after converting the customer language vocabulary into the system language, find the answers to the corresponding questions in the knowledge base and return them to the customer

there is content in the knowledge base, but customers may not get

but usually, the questions raised by customers are not clear. Most customers only describe the phenomenon, and they are not able to define the problem. At this time, it is necessary for the seat representative or intelligent customer service to prompt the customer to judge based on the scattered description proposed by the customer, eliminate the possible causes one by one with the exclusion method, and find the real problem. But it depends on high-level seat representatives to ensure that real problems can be found, and high-level seat representatives will always be in the minority. Intelligent customer service needs to establish the correlation between phenomenon and possible causes: it needs to establish the correlation between common phenomena, causes, corresponding problems and feasible countermeasures. At the same time, it also needs to judge the most probable cause based on the results of big data analysis in intelligent customer service, so as to rank the most probable result in the front. Of course, this work can also be determined by multi role personnel such as knowledge base personnel and agent personnel based on their personal experience, so as to have a good experience on the client side, truly reduce the dependence on the ability and level of agent representatives, realize automation and improve customer satisfaction

as for the core problems and definition methods of customers, most knowledge bases are not sorted out and stored as a special content. To achieve real intelligent customer service, the analysis of explicit and potential problems based on customer needs is the basic work of intelligence

for example:

for example, someone asked kmcenter on the official account whether there was an open class of enterprise knowledge management recently, and they wanted to participate, etc. However, the packaging of contact lenses, wound healing bandages, bone repair stents or catheters may not be his core problem at all. If the other party is willing, we usually ask:

a, have you just started knowledge management or have you been doing it for several years

if they are just starting to do it, their core problem at this time is that they need to do a planning of knowledge management (based on our experience), and only participating in an open class can not solve the problem of planning; If it has been done for many years, the core problem is probably how to involve the business department, that is, how to stimulate the knowledge management momentum of the business department. At this time, it may not be knowledge management promoters that need training most, but managers

b. What department are you in charge of knowledge management

because the departments responsible for knowledge management within the enterprise are very miscellaneous, including information technology departments, strategic departments, human resources departments, keeping clean, archives departments, administrative departments, etc., and their different departments have different needs for knowledge management. Based on the characteristics of different departments, we will also suggest deeper services

there may be more problems to understand and confirm with the customer. In this way, when the customer puts forward the demand for an open class, his real problem may be to buy a knowledge management software, evaluate the effect of knowledge management, the motivation of knowledge management, etc. This is the real demand

there are similar problems in intelligent customer service. Many times, it is necessary to embed the knowledge accumulated before the seat representatives and knowledge base can stop selecting to the business department according to the maximum experimental power of the existing equipment into the knowledge base, and work together to help users automatically define problems. Such intelligent customer service is the real intelligence, because it helps customers have a deep understanding of their own problems, Many times, in the end, he will find that there is not even a problem with the official products and solutions, but his own operation mode or method is wrong. This problem can be solved by himself without official solution

in the problem definition stage, through the built-in judgment conditions and the previous understanding of common problems of customers, and through several keywords said by customers, help to judge the real problems, related problems, and even problems that customers will encounter in the future, and then give them appropriate suggestions

3. If we can define the core problems of customers, establish a model based on the core problem solving, and then associate and correspond to the corresponding knowledge base resources, we can solve the problems of customers. At this time, you will find that the knowledge you need may not only be a certain point in a directory, but also include many points in different directories and dimensions. By pushing it to customers through a knowledge map, customers will feel that the service is considerate and considerate

indication: knowledge display based on specific problem solving

further, intelligent customer service should also explore how to exceed customer expectations and improve customer satisfaction and loyalty. For example: if you ask the doctor when a certain medicine needs to be taken, and the doctor tells you that it needs to be taken before meals, this is a basic service. Usually, the service ends at this time. But if the doctor can tell you that it's best not to eat too greasy and cold food after taking this medicine, which may also cause a short period of sleepiness, but don't worry. At this time, you'll probably praise this doctor in your heart

but behind the ability to exceed customer expectations is the analysis of common scenarios of customer problems and the establishment of corresponding models. Through the summary and refinement of previous customer consultation responses, plus the mining of accumulated data, we can establish corresponding problem-solving models, and associate the solutions to these problems with the knowledge base, so as to truly exceed customer expectations and make customers scream

to realize real intelligent customer service, it is necessary to manage the input of the client as a kind of core knowledge from the perspective of understanding the customer language, defining the customer problem, establishing the problem-solving model and knowledge matching, precipitate the explicit and implicit knowledge of personnel in different roles such as seat representatives, knowledge base operators, quality inspection and training, and be able to analyze and apply these knowledge uniformly, Only in this way can we truly achieve intelligent service and automatic service