The use of algorithms to mine big data for media preferences presents a transformation in the structure of the public sphere that is amplifying the tyranny of the majority. Whereas previous scholarship has lamented the fragmentation of the public sphere caused by the use of big data to inform audience analysis and media production, I argue here that fragmentation itself is not an implicitly bad thing for public debate, as fragmentation can encourage participation from the otherwise disempowered. Instead, I suggest that the use of big data to inform media production causes problems in the public sphere not because it fragments public debate, but because it somewhat paradoxically recentres public engagement around the complementary interests of the broad majority and profitability. The problem for public engagement is not that there are no overarching or all-encompassing media structures anymore but rather that these systems are informed by algorithms that promote a particularly populist ‘profitable and normal’ media experience.