Conversational social robots can use backchanneling to provide real-time feedback, convey an understanding, demonstrate attention and maintain a flowing dialog. However, backchanneling implementations are often project- and robot-specific, and we do not yet have standardized backchanneling toolkits to enable robot dialog designers to easily configure a robot's back-channel behaviour. This highlights a need for high-level toolkits to enable dialog designers to engage with backchanneling and customization, allowing rapid exploration of backchanneling as a part of dialog design. To engage with this problem, we surveyed recent works on backchanneling social robots, performed an initial analysis to describe the range of backchannel techniques, and outlined preliminary criteria for high-level backchanneling design.