How Competitive Intel Product Teams Use Passenger Ground Transportation Data

Because each class ships separately, platform-versus-cab share is arithmetic: divide HVFHV trips by yellow trips in the same month and zone, and you have a defensible disruption curve built from primary records rather than vendor estimates. Two cautions shape the alerting window - files land monthly, and the newest months are the slowest to settle, so baseline against a trailing three-month average before calling a shift.

Transit agencies move faster than their press offices, and their feeds double as demand telemetry for any mobility product competing for the same riders.

The National Transit Database is the peer benchmark underneath them: roughly 370,000 agency-by-mode-by-month rows across 834 NTD IDs since January 2002, refreshed weekly and carrying unlinked passenger trips, vehicle revenue miles, hours and peak vehicles in the commercial delivery terms. It is what tells you whether a rival's claimed ridership surge is execution or simply the market moving.

Entry signals surface in registries first, and three of them cover the ground transport map.

Every operational signal above needs a demand denominator, and four slower sources supply it.

American Community Survey - Journey to Work / Commuting Data gives the structural baseline: Table B08301 mode shares splitting workers into drove alone, carpool, transit, walk, bike and worked from home, plus county-to-county worker flow tables that size catchments precisely. Five-year estimates reach block groups - roughly 85,000 tract geographies - while one-year estimates cover populations of 65,000 and up; releases run annually since 2005, currently the 2024 1-year and 2020-2024 5-year files, all commercial delivery terms.

National Household Travel Survey (NHTS) microdata adds behaviour: the 2022 wave's public-use core carries 31,074 trip records with full survey weights in a csv.zip of about 4.5 MB, downloadable with no registration, and waves reach back to 1969 for mode-choice trendlines. It is static, so treat it as calibration rather than monitoring.

Ranked by how soon each step pays off, using only the sources named above:

1. Diff fares where they are legal to diff - query the Chicago TNP tables anonymously over SODA, comparing fare-per-mile and shared-trip mix by platform across the 2018-2022, 2023-2024 and 2025-plus tables.

1. Wire service-change alerts - poll the TfL Unified API within its 500 calls-per-minute allowance, diff BODS timetable registrations, and watch Mobility Database merges for new agency feeds.

1. Normalize against demand - benchmark every anomaly against NTD monthly peers, MTA and CTA daily boards, ACS commuting flows and Eurostat modal split before briefing anyone.

This cluster is the competitive-intelligence angle on the industry. For the dataset-by-dataset treatment, read the passenger-ground-transportation data guide; for the same evidence base as a ranked shortlist, open the passenger ground transportation data for competitive intel product teams page. The industry hub ties together the best-of, free-list and API pages, and the persona hub collects every vertical covered for this team.