{"id":352,"date":"2019-01-14T18:31:37","date_gmt":"2019-01-14T17:31:37","guid":{"rendered":"https:\/\/eddi.labs.ai\/?p=352"},"modified":"2020-12-11T15:49:01","modified_gmt":"2020-12-11T14:49:01","slug":"why-a-great-technology-is-not-enough-to-build-great-chatbots","status":"publish","type":"post","link":"https:\/\/eddi.labs.ai\/de\/why-a-great-technology-is-not-enough-to-build-great-chatbots\/","title":{"rendered":"Why a great technology is not enough to build great chatbots?"},"content":{"rendered":"
Thanks to advancements in natural language processing technology, big tech companies such as Microsoft, IBM, among others\u2013have made conversational interface technology (chatbots) available via their respective platforms. These proprietary platforms are offered with a specific license model as great technology only.<\/span><\/p>\n However real life projects proof everyday, that this is not enough and that chatbot developers are looking for knowledge transfer together with a great technology. Most of the projects turn out to be not successful because the focus was only on the technology.<\/span><\/p>\n Ask yourself whether this makes sense and is fully achievable. Nothing is worse than creating wrong expectations about the capabilities of the chatbot. This will disappoint the users and the chatbot is at the end of the day not used at all.<\/span><\/p>\n \u201cUnderpromise and Overdeliver\u201d is in this context highly important. Set the tone right and you will have success with your chatbot. Once the chatbot is accepted by the user, you can always gradually extend the capabilities. Believe me – this will come as a positive surprise to the users.<\/span><\/p>\n Bevor getting into the technology the scoping of the conversation flow should be done. This starts with the users of the chatbot.<\/span><\/p>\n Who are they? What questions will they ask? What are the possible answers?<\/span><\/p>\n When you have defined this work yourself backwards from the answers and define the conversation flow. Start with a bot who takes control over the conversation. This can be achieved with a well defined use case, where the user knows exactly what to expect and what not. Don\u2019t try to create a bot which can answer everything.<\/span><\/p>\n<\/h3>\n
Defining the Chatbot Use Case right from the very beginning<\/span><\/h2>\n
<\/h3>\n
Focus independent from the technology you intend to use on the definition of the chatbot use case at the very beginning. <\/span><\/h4>\n
<\/h4>\n
What do you want to achieve with the chatbot?<\/span><\/h4>\n
<\/h3>\n
Conversation Design is evenly important than using the Right Technology<\/span><\/h2>\n
<\/h3>\n
Chatbot Technology and Knowledge Transfer<\/span><\/h2>\n