PM OS

Kozmo.com: quick commerce twenty years too early, with no delivery fee and no minimum order

Kozmo delivered DVDs, snacks, and convenience items within an hour, with no delivery fee and no minimum order, using couriers on bikes and scooters. Customers loved it; the cost of carrying a single candy bar across a city made sure it could never pay for itself.

  • Primary cause: Broken unit economics
  • Scaled too early
  • Ran out of runway
Founded
1998
Ended
2001
Headquarters
United States
Sector
Commerce & Delivery
Capital raised
~$250M

What happened

Shut down in April 2001 after a planned IPO was withdrawn.

What Kozmo.com was

Founded in New York in 1998, Kozmo promised to bring small items — movies, ice cream, magazines, snacks — to your door within an hour. Its couriers became a familiar sight in Manhattan, and the service expanded to around ten US cities.

It raised roughly $250 million, including money from Amazon, and signed a large marketing partnership with Starbucks that put return boxes in coffee shops. A planned IPO in 2000 was pulled as markets fell, and Kozmo closed in April 2001.

Why it failed

  1. 1

    No fee, no minimum, no margin

    Delivering one low-value item by courier costs far more than the margin on that item. Kozmo removed the two levers that make delivery viable — fees and minimum basket sizes — to drive adoption, and couldn't put them back without breaking the habit it had created.

  2. 2

    More cities before more density

    Delivery economics improve with density: more orders per courier per hour in a small area. Expanding into new cities spread the operation thin before any single market had enough order density to cover courier costs.

  3. 3

    A plan that needed the IPO

    Like many of its peers, Kozmo needed a successful listing to fund its way to scale. When that window closed in 2000, there was no alternative path.

The product lesson

The quick-commerce boom of 2020–2022 relearned Kozmo's lesson at a much larger scale: speed is easy to sell and expensive to deliver. The numbers that decide survival are basket size, orders per rider-hour, and how much of the delivery cost customers will actually pay. Answer them in one dense neighbourhood before launching a second city.

Early warning sign to watch for

The average basket is smaller than the fully loaded cost of one delivery, and the plan relies on scale to fix it.

Compiled from public reporting, court records, and company statements. Figures are approximate. Where legal proceedings are mentioned, convictions are described as convictions and charges as allegations. Corrections welcome at hello@productmanagement.space.

Failed for similar reasons

PM OS