The resurgence of the chatbot has been facilitated by recent developments in machine learning. Recently, chatbot technology has been more widely available on the web and on mobile devices. According to recent statistics, 57% of businesses have already used or plan to execute a chatbot in the close to future. As a result, chatbots will be crucial in a future powered by AI. The goal of chatbots, which can be text-based computer systems, is to mimic human communication skills. In numerous industries, including customer service, e-coaching, and entertainment, they are spreading like wildfire. The effectiveness of these virtual agents cannot yet be evaluated in a consistent manner. As an alternative, a variety of individuals and groups have developed their guiding principles, either exclusively for their chatbot project or with some inspiration from other groups. In this paper, we explore the trends and evaluation techniques currently in use for chatbots. We investigate chatbot evaluation techniques and rank them in accordance with the ISO 9214 usability characteristics of efficiency, effectiveness and satisfaction. After that, we examine the techniques applied in the literature between 1970 and 2022 and contrast the outcomes. We find a clear trend toward examining the efficacy of chatbots in several recent studies, which is likely due to the growing popularity of task-based chatbots, which are increasingly being deployed in many business situations.