Best practice for chatbots using practical examples
Posted: Mon Apr 21, 2025 5:11 am
Chatbots are becoming increasingly popular in messenger marketing. Even though the technology is still in its infancy and some bots are driving users crazy, more and more companies are trying out this strategy. In 2017, for example, 100,000 chatbots were developed for the Facebook Messenger market – one year later, the number had tripled.
But that's no surprise! With Google Assistant, Alexa, Cortana, and Siri, the big giants ghana phone number data like Google, Amazon, Microsoft, and Apple are showing us how it can be done and the immense potential behind artificial intelligence, voice assistants, chatbots, and the like.
But how can you prevent your chatbot from suffering from teething problems? What are the worst chatbot cases? And while we're at it: What are the best and most creative ones?
Read all about this and how you can ultimately create your chatbot optimally in this blog article.
The worst use cases for chatbots
#Case1: Not finding any products as a product finder is not a brilliant achievement.
Source: Screenshots from the conversation with Dr. Oetker in Messenger
Dr. Oetker's chatbot, which is designed to find products, isn't performing well. When asked to search for tiramisu, crumble cake, or yeast cake, the bot always gives the same response: Dr. Oetker doesn't know these products.
#Case2: Advisory chatbots should be helpful. But sometimes that's not possible.
Source: Screenshots from the conversation with Kayak in Messenger
Even though the user has entered search queries for various locations, Kayak doesn't provide any useful answers about suitable flight connections or convenient travel times. Advisory chatbots are supposed to help users find the right product, not complicate their search with incorrect results. Because let's be honest: The user has just wasted their time.
But that's no surprise! With Google Assistant, Alexa, Cortana, and Siri, the big giants ghana phone number data like Google, Amazon, Microsoft, and Apple are showing us how it can be done and the immense potential behind artificial intelligence, voice assistants, chatbots, and the like.
But how can you prevent your chatbot from suffering from teething problems? What are the worst chatbot cases? And while we're at it: What are the best and most creative ones?
Read all about this and how you can ultimately create your chatbot optimally in this blog article.
The worst use cases for chatbots
#Case1: Not finding any products as a product finder is not a brilliant achievement.
Source: Screenshots from the conversation with Dr. Oetker in Messenger
Dr. Oetker's chatbot, which is designed to find products, isn't performing well. When asked to search for tiramisu, crumble cake, or yeast cake, the bot always gives the same response: Dr. Oetker doesn't know these products.
#Case2: Advisory chatbots should be helpful. But sometimes that's not possible.
Source: Screenshots from the conversation with Kayak in Messenger
Even though the user has entered search queries for various locations, Kayak doesn't provide any useful answers about suitable flight connections or convenient travel times. Advisory chatbots are supposed to help users find the right product, not complicate their search with incorrect results. Because let's be honest: The user has just wasted their time.