San Francisco pilots e‑bike deliveries for app and small‑business couriers; early data shows lower miles and mixed earnings effects

San Francisco Environment staff and program partners updated the commission April 28 on an e‑bike delivery pilot that provides no‑cost e‑bikes, accessories and safety training to eligible delivery workers and small organizations. The pilot is funded by the California Energy Commission and the U.S. Department of Energy (DOE) and is administered with partners including GRID Alternatives, the San Francisco Bicycle Coalition and Sports Basement.

"We are collecting participant delivery data to quantify the benefits and barriers to this type of mode shift," said Anna Charuto, the departmentclean transportation specialist. The pilot initially focused on app‑based food delivery (Uber Eats, DoorDash) and enrolled cohorts separated into e‑bike riders (pilot) and car drivers (control). The program offers e‑bike recipients full ownership after a one‑month study period.

Key findings and changes: Presenters said they have analyzed roughly 1,000 e‑bike deliveries and 2,500 car deliveries from the first cohorts. GRID Alternatives reported the two initial e‑bike cohorts collectively eliminated about 2,296 vehicle miles recorded for delivery trips, and that e‑bikes emit a fraction of the carbon dioxide equivalent emissions of cars on those trips. Staff said e‑bike riders reported fewer hours worked and fewer deliveries per shift, but also lower vehicle maintenance and operating costs.

Presenters and commissioners emphasized that differences in earnings appear linked to app platform behavior the project team could not fully measure. "We weren't able to track data on the time that someone was spent actually waiting for a delivery to be sent to them from the app," staff said, noting unpaid wait time is a statutory factor under California law. Program staff said they attempted to request backend data from transportation network companies (TNCs) but were denied and that the team will instead pursue recruitment‑based strategies and other data collection methods.

Program adjustments for the DOE expansion: Based on lessons learned the team said they will reduce the data collection period from four months to one month to reduce attrition, change bike models to improve range and cargo capacity, and expand to small businesses and community‑based organizations that want cargo e‑bikes (shared or organization‑based use). Staff said they will run additional cohorts under the DOE grant and finalize recruitment for small‑business pilots. Maya Chaffee (GRID Alternatives) said participants reported increased confidence in biking and interest in using e‑bikes for nonwork trips after the study period.

Risks and next steps: Staff noted the DOE grant covers three years and the program is in its second year; they said they had not received direction to stop work but acknowledged federal funding changes are a risk. Commissioners asked whether platform metrics such as ratings or idle time could explain earnings differences; staff said they did not collect driver ratings but had tracked some cancellation rates and planned to consider additional metrics in the DOE cohorts. The team also said participant feedback on unsafe intersections was collected via mapping during cohort meetups and shared with SFMTA.

No commission vote was required; staff said they will launch two more app‑based cohorts and finalize small‑business pilots in the coming months.