“Always keep the conversation going!” This motto is no longer just for the countless stars, stars, influencers or local politicians who inform us on various channels almost continuously about the news of their lives. It also applies to all service-oriented companies that value customer proximity and want to perfect their journey.
The old motto “Click, shop and forget” has been a thing of the past. On the one hand, because companies always want to keep their customers up to date with product news, current offers and promotions. On the other hand, because existing and potential new customers also demand this communication.
In the last three decades, the range of communication options available has expanded dramatically. Traditional options, such as personal contacts, letters, phone calls and faxes, have been complemented by emails, chats or social networking publications. Sometimes the channels even mix in a wild and unpredictable mix.
But how do companies react to new requirements in the area of customer communication? In addition to the simple offer of connecting through a special channel, are there other technologies and methods to make contact and communication as profitable as possible for both the customer and the company? Fortunately, this question can be answered with a decisive “yes! However, it must be borne in mind that each of the numerous communication channels must be considered individually in order to be able to develop customized solutions for each one.
Synchronous and asynchronous channels
Basically, communication channels can be divided into two classes: synchronous and asynchronous. Customers who contact through synchronous channels generally expect an immediate reaction: calls must be answered immediately. In the case of chat requests, via web chat or the messaging application via a smartphone, a quick, at best, immediate response is required. And, of course, the customer expects to receive a fast and satisfactory service even in personal contact, e.g. by visiting a branch.
Asynchronous channels include letters, faxes, but also e-mails or social networking publications. Of course, reactions are also expected here, but they don’t have to be immediate. On the other hand, some delays do not seem to be tolerated, but are usually even expected. In the classic letter request, this waiting time is probably higher (sometimes weeks pass until the response is received). For emails, the accepted duration is significantly shorter with only a few days, and in some social network channels an even faster reaction is expected in a matter of hours.
Bots as helpers in the synchronous dialogue with the customer
The transition between synchronous and asynchronous channels is particularly smooth at the threshold between chat and social networks. Next, the focus is on synchronous channels. Here, companies face the challenge of meeting high service and customer requirements. You must have enough competent staff to be able to respond to the customer enquiries not only properly, but also as quickly as possible.
Numerous companies are meeting this challenge by providing their customers with an artificially intelligent point of contact in certain situations. So-called bots can make phone calls and chat in the customer dialogue.
Basically, the dialogue in both channels, language and text, is based on an almost identical principle. Only in the pre-processing of the input, as well as in the preparation of the output, there are big differences: while the voice bot in the input may have to do with murmurs of calls or dialects, the chatbot must be able to handle spelling errors and symbolic input (emojis). When delivering the speech, Bot must make sure again that this is done very quickly, as in the phone call even the shortest breaks are perceived as unpleasant. The chatbot, on the other hand, should be able to signal and transmit messages symbolically, if desired. As a result, each of these two variants of human-machine communication has its own particularities.
Always the same steps
First, the question arises of how chat communication really works. The conversation is usually initiated by the customer who sends a request to the company. The contact is linked to the expectation of receiving a quick response. By the way, at this point it doesn’t matter if the contact person responding is a human or a machine. An independent contact person will accept the chat request, greet the customer in the interest of the company, and ask a specific question about their concerns. Depending on the customer’s response, a dialogue takes place in which their concerns are more precisely defined and all framework data for meeting this concern must be clarified.
Dialogues always follow the same steps. While the chatbot reveals information and asks more questions, the client provides answers and, if necessary, makes selections. In addition, the chatbot extracts supplementary information from back-end systems and can make decisions based on it, which are important for dialogue, sometimes even in a basic way. For example, when a customer asks about the whereabouts of a long-awaited order, the bot will ask for the order number and verify the delivery status. If there is also a tracking number, this can also be transmitted so that the customer can independently inform himself about the progress of his order.
How does the bot work?
To discover what technology allows chatbots to use intelligent dialogue, it is essential to look behind the façade. Again, basically two variants can be distinguished. Or, the course of the entire dialog is manually designed so that the chatbot ultimately only needs to work through a dialog script and follow a strict logic of if-then. Or rely on special automatic learning algorithms to provide appropriate answers based on the broadest possible information base.
Artificial intelligence (AI) is often used in this case, although the keyword applies to both variants of the automated dialog. Both types have their advantages and disadvantages. In the case of the programmed variant, bot operators face the challenge of anticipating all eventualities in the dialogue processes and planning appropriate reactions. At the same time, the choice of words and the logical course of events should not be based on the company’s own practices, but should be conceptualized from the perspective of a customer who lacks extensive knowledge of internal processes and contexts.
Anyone who relies on self-learning bots faces a completely different challenge. In this case, a sufficiently large amount of data must be prepared and prepared, from which the chatbot can nourish your knowledge. For example, existing employee-led chat dialogues could be collected and analyzed. In post-processing, it must be observed manually whether the individual steps of the dialogue should be classified as correct or incorrect.
From the data generated in the process, the algorithm can learn in the next step how it should behave in similar situations in the future. If such a chatbot is practically used, there is a possibility that he, like the famous lying baron of Münchhausen, will remove his hair from the swamp. It means The dialogues he manages are saved and employees evaluate the individual dialog steps to increase the database and improve the reliability of the bot.
However, once the chatbot has hardened to always deal with certain situations incorrectly, learned misconduct, unlike the programmed version, cannot be easily avoided. Instead, the database needs to be augmented with examples of the desired behavior pattern until the bot changes its learned notion of “right” and “wrong”.
Optimization the communication with the customer but…
How can chatbots be used meaningfully in communication with the customer? And how should your service be dosed? Regardless of the variant chosen, the question arises of how to deal with misunderstandings and where the limits of the bot are in the communication. So when does it get to the point where the chatbot can no longer solve a problem independently, but has to deliver it to a human colleague?
While robots do not know after hours and are generally used 24 hours a day, 7 days a week, human services employees are often not available 24 hours a day. It is therefore necessary to decide whether the chatbot will only be used if transfer to a human colleague is possible or if the dialogue is interrupted outside working hours if necessary. In the latter case, the bot could refer to service times or initiate the agreement of a personal appointment with an employee. The decisive factor for this decision is the extent of self-service that the chatbot can cover. Provided that a reasonably complete portfolio of frequently used services is achieved, 24/7 operation seems to make sense.
To avoid misunderstandings, ambiguities and ambiguities in the dialogue with the chatbot, there is an option to refrain from free text entry and instead provide the customer with a limited selection of responses. In this way, the bot’s reaction options, and therefore the effort for the operator, can be limited by means of a linear dialogue. The selection options can be displayed in different formats. In addition, they can be enriched to visualize the various possibilities using graphical information.
In short, automating customer dialogue is a very complex field of application, where it is important to analyze and weigh the type of requests for which a particular approach can be implemented and used cost-effectively. Bots are one of several ways in which companies can efficiently manage one of the most difficult tasks in the business environment, communication with the customer. After careful analysis and taking into account the peculiarities described above, bots can become a crucial factor in optimizing communication with the customer. Their integration in the dialogue with the customer leads to significant savings of two resources: time and money.
Jürgen Haas is a senior consultant at IP Dynamics, founded in 2005.