Honey bee signals are primarily studied through natural observation combined with manipulations of the colony or environment, not direct manipulation of the signal stimulus or receivers. Consequently, we know little about which signal aspects are necessary to reproduce behavioral responses. Here, we focus on the shaking signal, wherein a worker grabs onto another bee and vibrates. All castes receive shaking signals, but individual responses depend on context, and the signal may be multi-modal (mechanical, odor, sound, etc.). We designed a tool to mimic the shaking signal. We tested whether a purely mechanical stimulus elicited the same behavioral response as a natural shaking signal, teasing apart the effects of signal and receiver characteristics. We found that both workers and drones increased their movement after being artificially shaken, and that shaken drones were more likely to engage in feeding and grooming than a sham control. These behavioral changes support the idea that the shaking signal serves to generally increase worker activity, but also serves to activate male reproductives (drones). With this tool, we show that vibration itself is responsible for eliciting much of the shaking signal's behavioral response, in one of the few examples of direct playback in social insects.