First 1000 field note · Updated September 18, 2026

How Uber got its first customers

Uber recruited early riders through the founders’ friends and startup demonstrations, then used credits and referrals to make the first ride easier to try. The offer became more valuable when the alternative was getting stranded.

By Ali Abouelatta · Original issue: March 21, 2021

What the earliest account establishes

In his December 22, 2010 founding account, Travis Kalanick described a January test with three cars in New York. He dated the San Francisco launch to May 31, 2010, and credited the founders’ invitations and Ryan Graves’ startup demos with getting things started.

By December, that account described thousands of active San Francisco riders. It does not provide a ledger of the first 1,000 customers or a channel-by-channel breakdown. Claims that one campaign delivered exactly 1,000 paying riders go beyond this source.

Give publicity a concrete next step

On September 13, 2010, Ryan Graves addressed people who had found Uber through TechCrunch. His welcome post offered $15 in credits and explained where to enter the code. It also encouraged weekday use, since the service was already busy on Thursday through Saturday nights.

The press story created attention; the follow-up reduced the price of trying a ride. A founder using this approach needs a path from the story to a completed first use. A registration on its own does not establish that the reader experienced the product.

A taxi strike made the offer more relevant

Uber’s June 17, 2011 taxi-strike announcement promoted 50% off a first San Francisco ride for the following Tuesday. It also promoted $10 off for existing riders who referred a friend. This was after the initial 2010 launch: evidence of an early tactic, not proof of how the first 1,000 riders arrived.

The offer matched a specific situation: the reader might need transport on a day when taxis were harder to find. The referral gave an existing rider a reason to bring a friend into that situation. The announcement publishes the offer, not its profitability or retention results.

The growth story includes a failed night

On December 11, 2010, Uber apologized for a dispatch failure. Drivers had been assigned to multiple riders. Troubleshooting took drivers offline, reducing supply further.

The company described growth as “sometimes 30% week over week.” It did not claim a sustained weekly rate or define the metric precisely enough to turn that sentence into a customer-growth series.

Recruiting people at their moment of greatest need makes a failed first experience more costly. A customer waiting for a car that never arrives has a different story to tell friends than a customer who gets home quickly.

Method and limits

This historical synthesis combines my 2021 First 1000 issue with contemporaneous Uber announcements linked below. The dates, offers, and reported growth come from those sources. The implications for acquisition are editorial analysis. No new user-level dataset or verified count of the first 1,000 customers is available here.

What I would keep

Choose a moment when the current alternative is visibly failing.
Connect attention to a first completed use.
Check service capacity before widening the promotion.
Track fulfilled transactions, unsuccessful requests, waiting time, and repeat use alongside signups.

The takeaway

For a marketplace, a growing registration chart can hide a worsening first experience. Check whether new customers receive the service they came for.

Sources and original research

Historical facts are attributed to these sources. An observed product change does not establish conversion lift.

Updated September 18, 2026. Based on original First 1000 reporting and the sources listed above.