This is a field report on a six-month experiment in 10 US cities where Google modified Maps' routing algorithm to suggest alternative routes with similar travel times away from identified bottleneck segments. The intervention rerouted under 2% of trips system-wide but produced measurable results: ~2% speed increase on targeted segments, 0.5–1% reduction in fuel consumption there, and modest gains (0.35–0.5%) across the broader affected network. The authors claim this translates to thousands of tons of CO2e savings per city annually. The mechanism is straightforward—dispersing traffic from high-volume highways onto peripheral roads maintains better flow everywhere—and was tested using daily crossover design with hierarchical Bayesian modeling. One open question worth pressing: what happens if the intervention scales beyond the ~2% adoption rate tested here, and does congestion simply shift to the peripheral segments that absorbed diverted traffic?