Be inspired by examples and think about how you can use chatbots to attract new customers and better serve existing customers.
Author: Patrick Knab
Companies that today communicate with their customers exclusively via traditional websites have already been overtaken by innovative companies that rely on mobile apps. The same fate could befall those meetings that will not have a presence on the usual chat platforms such as WhatsApp and Facebook Messenger in the next few years.
Messenger apps such as Facebook Messenger and WhatsApp are very popular. Young people in particular, but also their parents and grandparents, are increasingly using these platforms. Especially well suited are these
Messenger Apps for communication on mobile devices. They can be seamlessly integrated into the usual communication patterns for text messages. Today, if a company wants to stay in touch with customers on the messenger channels, this is associated with substantial personnel expenditure. Chat bots, on the other hand, offer the opportunity to achieve a great effect with comparatively little effort.
Chatbots are available in the form of simple question/answer dialogues, up to full wizards integrated into backend systems. Chat bots do not have to be limited to text only, but can use all possibilities of modern devices. While the user input is primarily text, the answers can be supplemented with pictures or even videos.
The simplest chatbot solution is the simple question/answer dialog variant. Typical FAQs can thus be made more interesting in an uncomplicated way and offer the customer a new, refreshing experience. The more a chatbot is integrated into other systems, the more extensive its use becomes. A chatbot, which is available e.g. for registered users, can draw on past interactions and provide better tailored answers using information from the customer profile. It can also use this data to present further information or make recommendations to the customer using a document management system or other company data.
When designing a chat bot, you should focus on a task that is as specific as possible. The more specific the task, the better the user experience typically is and the deeper the programming effort. Although great progress has been made recently in the field of artificial intelligence, one cannot expect miracles from the services offered. First and foremost, services such as IBM Watson Conversation or MS Azure offer powerful tools for text recognition. IBM Watson, for example, requires only a few example sentences as training so that the chatbot can recognize the intention of the user. The more examples the system is fed with, the better it becomes. With a clever design of the bot, you can also let the users themselves help. Therefore, it is best to start the construction of a chatbot with an initial training set and then let users train the system further. Typically you start with focus groups, e.g. your own employees or selected customers.
A central point for success is the use of real training data. The training data must come from the business or from the customers. The training of the systems should not be left to IT alone.
Chat offers the possibility to enrich the customer experience with comparatively modest effort. Many existing processes can use chatbots to open up new customer segments and improve the experience for existing customers. Bots never get tired, even if they are in use all day. This makes them an ideal complement to existing channels and enables them to provide customer support 24 hours a day.
Two selected examples of existing solutions