With advancements in detection devices and cognitive technology, experience-based displays (exhibits) reacting to user behaviors are gaining popularity. These systems use sensors to recognize users’ upper arm and forearm movements to switch images or the environment in images; however, communicating exactly what user action is required for the display to respond is difficult. To have viewers (users) behave in the manner intended by experience-based displays incorporated in media art and museum exhibits, this study proposes methods for instructing behaviors through pictograms. This study produced still (frozen) and moving pictograms after elaborating on forms and compositions that specify desired behaviors and compared the correct answer rates and average time spent by users in their twenties to seventies to achieve the desired behaviors. For the assessment experiment targeting users in their twenties, the correct answer rates varied among different still images, whereas all moving images successfully navigated >88.5% of users to correctly behave. Pictograms were successful in making middle-aged and older users understand the desired behaviors; however, they were frequently required more time to absorb the instructions. This study introduces use cases incorporating these results into a real-world experience-based system.