ChatBots are currently the new black in customer communication and in the effectiveness of the Digital Customer Journey. But what exactly are ChatBots, what does it need, and where should you use it?
First of all, it should be noted that the customer communicates more and more with Messenger services. And he also expects this from the brands that interest him. Since early 2015, more messages have been sent on services such as Facebook Messenger, WhatsApp, WeChat, or Kik than via Facebook, Twitter and Instagram – with a strong upward trend
New customer expectation: speed!
This means new customer expectations in terms of speed and relevance of communication. According to social media, this poses the risk of companies being overburdened again, as communication has to take place much more in real time than is already the case today. This is difficult to achieve in terms of personnel and organisation. One possible solution is the use of a friendly ChatBot.
A chatbot is software that automates communication with people over the Internet. As virtual assistants, they answer questions asked in normal language in a personal and human way. They thus imitate a human interlocutor, but are a program, i.e. a pure human-machine interaction takes place. This is cheaper for the company and faster, easier for the customer to use and better accessible than most of today’s communication channels.
Not all ChatBots are based on AI
Some of these programs use AI, some are simply based on preconceived scripts based on rules or decision trees. The now famous Taco-Bot of the American fast feed chain Taco Bell, where you can order and deliver your tacos to the office via a chatbot within the collaboration app Slack, for example, is a good example of a rule-based, simple but effective and personal ChatBot. AI-based systems include Amazon Alexa, Microsoft Cortana or Apple Siri, which – like IBM with Watson software integration – support ChatBot development with the provision of AI functions.
Zum Teil nutzen diese Programme AI, zum Teil basieren sie einfach nur auf vorgefaßten Skripts, die auf Regeln oder Entscheidungsbäumen basieren. Der mittlerweile berühmte Taco-Bot (https://www.tacobell.com/feed/tacobot
) der amerikanischen Fast-Feed-Kette Taco Bell, bei der man z.B. innerhalb der Collaboration-App Slack über einen Chatbot seine Tacos bestellen und ins Büro liefern kann, ist ein gutes Beispiel für einen regelbasierten, einfachen, aber effektiven und persönlichen ChatBot. Als AI-basierte Systeme wären z.B. Amazon Alexa, Microsoft Cortana oder Apple Siri zu nennen, die – ebenso wie IBM mit der Watson Software-Integration – die ChatBot Entwicklung mit der Bereitstellung von AI-Funktionen unterstützen.
Here are a few examples that outline the range of uses of ChatBots:
Two aspects are important when using chatbots: the customer should not notice that he is communicating with a machine (even if he knows it – as in the case of Siri or Alexa, or because the provider reveals it). And the answers that the chatbot offers must be correct and relevant. This requires a very complex and expensive content management. Questions and answers must be available and constantly updated, which can also be partly ensured by self-learning algorithms. At this point, the job description of a “content engineer” is already being developed, who is not only responsible for the creation and provision of the content, but also for how the content is algorithmically and thus automatically further developed and how its relevance and also its accuracy are ensured. Regulations must also ensure, for example, that country-specific legal requirements are complied with or that the customer’s privacy is protected.
How to decide
The decision to use chatbots is based on various factors and is not trivial. It basically starts with the customer and asks who the target group is, what requirements the target group has of the brand, which messenger services are preferred in the target group and what additional brand loyalty the customer gets with the use of a ChatBot. This results in clear requirements for the underlying technology, the expected communication behavior of the ChatBot, the interfaces offered and the content on which the ChatBot must be based. Marketing, IT and external service providers must work very closely together to ensure that a ChatBot project does not fail as early as the requirements definition phase.
What to expect
In the long term, more and more ChatBots will become established in the areas of communication and services. Standardized development and integration technologies are increasingly available, the challenges now lie less on the technical side than in enabling the integration of existing data and creating the right user experience. The most successful ChatBots are those to which the customer voluntarily and gladly returns and which create consistent added value for him.