Research and practice in Human-Robot-Interaction often focus on improving the human-likeness of the robots. Developers often place their focus on how to make robots more human-like, aesthetically (e.g., facial features), functionally (e.g., finger movements), or both. However, these types of improvements can be costly. Through a series of studies, in the first essay, I find that customers may not perceive humanoid robots to be superior to other self-service technology products (SSTs) such as kiosks. Additionally, customers expectation of service performance by humanoid robots is guided by their past experiences with SSTs. In the second essay, I propose that to enhance the evaluation of service robots, providers will need to increase the perceived liveliness of the robots. To assess this prediction, in addition to traditional methods (e.g., surveys), I employ a multi-modal approach that measures perceptions, behavioral reactions, and neuro-physiological responses to a lively humanoid robot.