Methodology

A shared picture of earnings, with the context that matters.

Three numbers, clearly defined

Drivers counts linked contributors with processed Uber or Lyft screenshots, including anonymous browsers. Registration does not add another driver; linking browser histories to the same account combines them. Separate anonymous browsers cannot reliably be identified as one person. Screenshots counts processed Uber or Lyft submissions, including those without enough information for an hourly average.

Average earnings per hour is total eligible driver earnings divided by total online hours. This is gross platform-reported driver earnings, including tips when included in the reported total, before vehicle expenses and taxes. It is not take-home profit or the average of individual hourly rates.

What counts toward the average?

Readable daily and weekly USD earnings from Uber or Lyft with total earnings and positive online time. One eligible screenshot can start the average. A missing year does not prevent the calculation when earnings and hours are clear; missing dates remain missing. Fare breakdowns, driver profiles, monthly and yearly summaries are saved separately for future analysis. Booked time is never substituted for online time.

For each linked contributor and platform, the earliest accepted eligible report takes precedence. A later report with an overlapping period is excluded from the hourly average. If the earlier report is deleted, the remaining reports are reconsidered. Uber and Lyft online time may overlap across platforms; we cannot reconstruct concurrent driving from summary screenshots.

The average uses available earnings and hours, not the total number of uploaded images. Screenshots with the same visible period for a linked driver and platform are not counted twice. Without complete dates, some overlaps cannot be detected reliably; this is a screenshot sample, not audited unique work time. Anonymous browser sessions are not verified unique people. Exact file duplicates are blocked, but changed crops and separate browsers may evade deduplication.

Understand the sample

This is a voluntary, self-selected sample, not a representative estimate of all US drivers. Screenshots are not independently audited. Editing or generation indicators can exclude an image from calculations, but neither metadata nor AI can guarantee authenticity. AI extraction can make mistakes. Location and vehicle details are self-reported and saved when each report is uploaded.

Public city, vehicle and time filters will be introduced as the sample grows. Missing fields remain missing; we never invent statistics.

Last updated November 14, 2026.