Amazon said Wednesday it will expand its Prime Air drone delivery service to nearly 500 cities and towns across the United States this year, roughly six times the service’s current coverage, as the company moves its long-delayed drone program from pilot projects toward a scaled business. The expansion will focus on light packages, the small orders that make up a large share of Amazon’s daily volume.
The announcement is the strongest sign yet that Amazon believes the economics of drone delivery have finally turned. The company said the marginal cost of delivering a package by drone has fallen to a level that makes scaling the service practical, a claim it has not been able to make in the program’s decade-long history. For years, Prime Air was a showcase project, generating headlines and videos but few deliveries. Now Amazon is treating it as an operating business.
The program has been a long time coming. Jeff Bezos promised in 2013 that drone delivery would arrive within five years, a prediction that became a running joke inside and outside the company as the deadline came and went. Amazon spent the following decade building and rebuilding the technology, testing it in a handful of towns, and at one point restructuring the entire program after a round of layoffs. The service was suspended in some markets, redesigned and relaunched, and its current form is the fourth or fifth version of a vision that never quite died.
The numbers behind the expansion are straightforward. Amazon moves millions of packages a day, and the small, light orders that drones can carry represent a meaningful share of that volume. If even a fraction of those deliveries move from vans to drones, the cost savings compound: fewer drivers, less fuel, faster delivery times and, critically, more capacity in the trucking network during peak hours. The company has said the drones are not a replacement for its delivery network but a pressure valve for its smallest orders.
The technology has taken years to reach this point. Prime Air’s drones have gone through multiple redesigns, and the service has been tested in a handful of markets, with mixed results. Amazon suspended the program in some areas, restructured its team and changed its drone designs more than once. Each setback fed a narrative that drone delivery was a science experiment that would never pay for itself.
What has changed is the combination of hardware, regulation and software. The current generation of drones is quieter, more reliable and able to fly in worse weather. Amazon has won approval from federal regulators for expanded operations, including the certifications needed to fly beyond the pilot’s line of sight, and its flight-control software has improved enough to handle the complexity of suburban airspace. The remaining constraints are familiar ones: weather, local rules and the pace of regulatory approval in each new market.
Weather is the variable Amazon cannot engineer away. Drones fly poorly in high winds, heavy rain and extreme temperatures, and the cities on Amazon’s expansion list include some of the country’s most unpredictable climates. The company has said it will operate where conditions allow and pause where they do not, a concession that its rivals have made as well. The industry’s other big operators, including Alphabet’s Wing and the logistics company Zipline, have run into the same limits.
The competitive stakes are real. Drone delivery has been a talking point in the industry for years, and Amazon’s scale gives it an advantage its rivals lack: a delivery network, a customer base and the willingness to lose money on a project for a decade. If Prime Air reaches its target of hundreds of cities, it will force competitors to respond, and it will give regulators a working example of what scaled drone delivery looks like. Wing and Zipline have both expanded their own services, but neither operates at Amazon’s volume or with its logistics backbone.
For Amazon, the program is also a hedge. The company’s logistics operation has grown into one of the largest in the world, and its labor costs are enormous. Drones do not replace the drivers who deliver big packages, but they do absorb the small, cheap orders that currently consume the most time per package. Amazon has said it views the technology as one tool among many, not a replacement for its network.
The path from announcement to delivery will not be smooth. Every new city requires approvals, mapping and local engagement, and the company has been careful not to promise a specific date for full coverage. The ambition, however, is unmistakable: Amazon wants drone delivery to be ordinary, the way two-day shipping became ordinary. The 500-city expansion is the moment the company stops describing the future and starts shipping it.
