research
Two Percent of Routes Can Loosen a Whole City
A six-month Google study in ten U.S. cities found that small, coordinated route changes produced measurable network-wide traffic gains.
Summary
A six-month Google study in ten U.S. cities found that small, coordinated route changes produced measurable network-wide traffic gains.
Google Research modified routing recommendations around recurring bottlenecks while keeping alternative routes similar in travel time and road type. Fewer than two percent of observed trips changed, yet the study reports median speed gains of about two percent on targeted segments and smaller positive effects across all affected roads, with estimated fuel-use reductions. The experiment alternated treatment and control days over six months in ten cities. It does not establish that every city or navigation system will respond alike, but it shows how optimizing a small set of individual trips can improve a shared network rather than merely moving one driver faster.
Why it matters
A six-month Google study in ten U.S. cities found that small, coordinated route changes produced measurable network-wide traffic gains.
Limits and context
- It does not establish that every city or navigation system will respond alike, but it shows how optimizing a small set of individual trips can improve a shared network rather than merely moving one driver faster.
Key claims
A six-month Google study in ten U.S. cities found that small, coordinated route changes produced measurable network-wide traffic gains.
Qualification: It does not establish that every city or navigation system will respond alike, but it shows how optimizing a small set of individual trips can improve a shared network rather than merely moving one driver faster.
Evidence: source-2026-07-11-014
Sources
- Google Research: The power of collaboration in traffic routingGoogle Research · secondary reporting
Corrections
: Correction, July 11, 2026: An earlier automated version repeated five ticker briefs and nine dispatches from the July 10 edition. Those items were removed and replaced with previously unpublished editorial stories; the two front-page stories and three genuinely new dispatches were preserved.