- should cut ETH spend particularly in times of high traffic - want to add batching but still an open question as to implementation - One idea: - Internally track moving avg of rate of dispatch (updater turns on indexing messages) - If rate above threshold, introduce added pause for batching