The Deployment of Chatbots within Customer Experience Management

由于持续的数字化和互动渠道的多元化，公司和客户之间的接触点数量每年增加20%以上。因此，顾客之旅，即顾客从购买前阶段，通过购买阶段到购买后阶段的路径，变得更加复杂。由于顾客体验直接影响到企业的评价和顾客的购买意愿，因此企业有兴趣在各自的接触点设计最佳的顾客体验。人工智能领域的快速发展现在允许公司使用数字交互工具。这种交互工具，特别是聊天机器人，在客户旅程中的客户体验管理方面具有巨大的潜力。聊天机器人是人工智能的一种形式，可以与客户进行一对一的互动。迄今为止，关于在客户体验管理中有效部署聊天机器人的研究还很缺乏。本项目的总体目标是通过基于社会反应理论、恐怖谷概念和媒体丰富性理论的实证研究来弥补这一研究差距。为此，我们计划调查客户如何看待聊天机器人。在这样做的过程中，我们特别考虑了使聊天机器人看起来或多或少像人类的特征。我们假设这些特征会影响客户对聊天机器人的感知，进而影响聊天机器人的有效性，如客户感知的服务质量和购买意愿。此外，我们的目标是揭示基于特定阶段的客户需求，聊天机器人在客户旅程阶段之间的有效性差异程度。为了获得关于在客户体验管理中部署聊天机器人的初步见解，我们首先对已经使用聊天机器人的公司进行了专家访谈。然后，我们通过对包含客户旅程所有三个阶段的消费者进行三个实验室实验，在这个定性数据基础上进行构建。最后的现场实验旨在调查在实践中部署聊天机器人的有效性，并验证实验室实验结果。该项目对研究和实践具有重要意义。特别是恐怖谷的概念被转移到一个全新的应用环境中，被更新，并以这种方式刺激了卓有成效的未来研究。根据项目结果并遵循基于消费者的战略方法，我们可以为公司提供关于在整个客户旅程中部署聊天机器人的基本建议。

Caused by the ongoing digitalization and the plurality of interaction channels, the amount of touchpoints between companies and customers increases by more than 20 percent each year. As a consequence, the customer journey, that is the customer path from the pre-purchase phase, via the purchase phase through to the post-purchase phase, gets more complex. As customer experiences directly influence company evaluations and customers' willingness to buy, companies are interested in designing the customer experiences at the respective touchpoints in an optimal way. Rapid developments in the area of artificial intelligence now allow companies to employ digital interaction tools. Such interaction tools, particularly chatbots, bear great potential for the management of customer experiences along the customer journey. Chatbots are a form of artificial intelligence and enable one-to-one interactions with customers. To date there is a paucity of research on the effective deployment of chatbots in customer experience management.The overarching aim of this project is to close this research gap by conducting empirical studies based on the Social Response Theory, the Uncanny Valley concept, and the Media Richness Theory. To this end, we plan to investigate how customers perceive chatbots. In doing so, we especially consider characteristics that make chatbots appear more or less human-like. We postulate that these characteristics affect customers' perception of the chatbots which in turn affects chatbot effectiveness like customers' perceived service quality and willingness to buy. Additionally, we aim to uncover to what extent the effectiveness of chatbots differs between the customer journey phases based on phase-specific customer requirements.To gain first insights regarding the deployment of chatbots within customer experience management, we initially conduct expert interviews with companies that already use chatbots. We then build upon this qualitative data basis by conducting three laboratory experiments with consumers that comprise all three phases of the customer journey. A final field experiment aims at investigating the effectiveness of the deployment of chatbots in practice and at validating the laboratory experiment results.The project bears valuable implications for research and practice. Particularly the Uncanny Valley concept is transferred into an entirely new application context, is updated, and, in this way, spurring fruitful future research. Based on the project results and following the consumer-based strategy approach, we can derive essential recommendations for companies regarding the deployment of chatbots along the entire customer journey.