The introduction of a volumetric-pricing has been a critical intervention in household water conservation programs. However, it is typically impossible to observe how individual households respond to introducing a price signal as metering and billing start simultaneously. We report results from a near-ideal quasi field experiment wherein we measured daily water use both before and after the introduction of volumetric pricing (n = 59,563). As expected, the introduction of volumetric pricing (that replaced the previous fixed fee regime) resulted in an overall reduction in aggregate water use. However, the aggregate conservation effect of volumetric pricing (5%, p<.01) masks how nearly half of households increased water use. Further, a large share of households consuming above the median in the fixed price regime further increased water use. Using daily household-level water use data at three metered points for each household, we also detected that more than a third of all households increased water usage after the water bill was delivered. Triangulating our findings using multiple methods, including fixed effects panel and intervention time series, we also uncover key drivers of price response, such as water curtailment habits and a prior non-price intervention. The heterogeneous price response framework that we develop here reveals a large potential to design water conservation programs that combine price and non-price interventions to seal leaky water conservation buckets.