A communication channel alone offers no added value — it is the information that is delivered via these channels.
Therefore, the elementary question is actually: In which way can information be transmitted to the customer in the fastest and easiest way?
A live chat can of course also be an answer to this question — as long as the answers are quick and correct. The same applies to telephone service, support via Facebook or WhatsApp, e-mail contact, etc. Whereby the number of communication channels should be viewed critically. The rule here is: Quality over quantity, anyone who offers various communication channels but does not use them correctly does not make their customers happier at the end of the day. It is better to focus forces and resources on the channels that the customer actually wants to use.
The sub-optimal thing about the options mentioned above, such as live chat or e-mail, is that it usually only works as long as the corresponding employee is at their workplace and can answer within a very short time. After work, it is therefore the end. It's just a shame that many customers work themselves within normal operating hours and only want to access a live chat or similar in the evening.
An AI chatbot can solve this problem.
The fact that there is no real person behind this technology is often denounced, but in many cases this is an advantage. This means that the smart bot never leaves work and is available around the clock. A clear advantage, because that cannot be required of a service employee. At the same time, this does not mean that live chats with employees must be abolished or that a person cannot be activated (human takeover) if the user does not get anywhere with the chatbot. However, it makes sense to invest in a strategy to 1. relieve the live chat and 2. also to be available to customers when employees are offline.
A good solution could therefore be to create a AI chatbot to integrate into customer communication. The “magic formula” for this looks like this:
► The use of an AI chatbot makes customer communication easier because it is available around the clock and even on weekends.
► It is also important that recurring inquiries must be answered automatically. 80% of all inquiries are repetitive and therefore stop customer service employees unnecessarily. Such a high proportion of repetitive requests almost calls for automation.
► Employees shouldn't have to “deal with” repetitive inquiries. This can be frustrating in the long run and does not lead to any profitable results. It is better to use employees specifically, e.g. for more complex issues or emotional issues.
The thing about expectations
Yes, there was something... Customers expect that they will receive a solution to their problem within a short period of time. And that is also completely justified. Unfortunately, the reality is usually that the average waiting time is 84 hours. That's too much — 83 hours and 50 minutes too much, actually. Because: According to Hubspot, 90% of all customers think it's important to get an answer to their question within 10 minutes.
Thanks to modern technology, almost everyone is now familiar with it from their own everyday life — they are used to getting answers after a short time, whether on WhatsApp or via e-mail.
The ravages of time show that we are becoming more impatient. And companies should also be prepared for this.
What is the score of the User acceptance of Chabots?
Many companies ask themselves: “Do my customers even want to communicate with a type of robot? ”. The answer is: “Yes.” If the chatbot really helps a user and actually offers added value, then the customer is also happy to accept the chatbot as a communication channel.
It's like this: As soon as a customer notices that a chatbot is actually helping them with their request, they will of course accept it. It is therefore not primarily the communication channel that is important, but the content that is delivered to the customer and the question: “Was I able to really and above all quickly help the customer with this? ”. Other channels, such as telephone or live chat, can also meet these requirements, but unfortunately usually not around the clock and within a few seconds.
Where should a chatbot be played out?
In order to achieve the desired goal, i.e. an increase in customer satisfaction, you should carefully consider which channel a chatbot makes the most sense. There are essentially three options and one decision criterion. The three options are:
► Web site
► Facebook Messenger
► WhatsApp
The question leading to the decision is: “On which channel does the company receive the most customer inquiries? ”. Because the be-all and end-all is still to accommodate the customer as much as possible and to “meet” them where they are anyway. It should be made as easy as possible for the customer to get in touch.
How can AI improve customer satisfaction?
The advantages of a chatbot, such as “round the clock” service and immediate answers, have already been reported. However, it is also important to mention that not all chatbots are the same. Because the two letters “KI” can make a significant difference. AI stands for artificial intelligence — and all the “knowledge” and “learning” of the AI chatbot solution moin.ai is based on this very intelligence.
In order for moin.ai to learn new things, it essentially needs two mechanisms:
► Dreaming: Generating new topics and question objectives
The chatbot tries to make sense of everything — even topics that it doesn't know yet. In other words, if similar phrases pop up again and again, but the chatbot doesn't actually know, it makes a suggestion for a new question, as the topic is obviously relevant. In this way, the chatbot is opening up new topics step by step and at the same time adapts to the wishes and needs of users and their topics.
► Reassuring: Intelligent queries
It often happens that users formulate a question in a very long and complex way, so that a chatbot is not immediately aware of what the user actually means or wants. In this case, the chatbot acts as a person would if he doesn't understand what his counterpart is actually talking about — he asks a question. An example of this:
users: “I went shopping this morning and I always write down my passwords on a piece of paper. I accidentally took this note shopping with me and I guess I lost it there somehow. In any case, he is not in the apartment and I don't want to go back to the supermarket now, it's a 15-minute walk. What should I do now? ”
Chatbot: “Do you mean you want to reset your password? ”
If the user now answers in the affirmative, this goes back as a strengthening signal and the topic of passwords grows and is further sharpened. In just a few months, the chatbot is constantly developing a better language understanding than at the beginning and can answer more and more queries automatically.
⇒ Thanks to these two mechanisms, the AI chatbot solution is constantly learning and can answer ever more diverse questions in an ever higher level of detail.
And to answer the question asked at the beginning: In short, AI can make a customer happier when it comes to the chatbot, because thanks to AI and natural language processing (NLP) usually really understands and can therefore immediately offer suitable solutions.
Insight: An exciting side effect
The Firm Geberit observed an interesting side effect a few months after the chatbot was in use, which was not planned as such, but was still welcome because it solves a core problem. The hot topic at stake here is the “shower toilet” from Geberit. “We found that requestors actually much prefer to talk to a machine than to a real person. Of course, it is a taboo topic,” says Thomas Brückle, Head of Marketing at Geberit. The knowledge that not a potentially judgemental person but a neutral AI is “sitting” at the other end of the chat obviously lowered the hurdle to writing about the topic of intimate hygiene. Thomas Brückle: “We are taking a huge hurdle in communication with customers” — a great and unexpected result.
If, apart from this topic, you would like to know more about Geberit and the AI chatbot, you are welcome to Here is the continuing case study download.
Conclusion: Real added value for the customer
It turns out that an AI chatbot can be an asset in customer communication. It offers permanent availability, answers in seconds and also user acceptance of chatbots, which grows with every successful conversation. NLP and AI ensure the success of a conversation, as they sharpen, clarify and expand the understanding of the chatbot and the answers, so that a customer is usually satisfied with the answer given. And should it happen that a customer is dissatisfied with an answer and wants to talk to a person, can via “Human Takeover” always be passed on to a human employee.
