Skydio’s Adam Bry: How a US Drone Startup Is Outselling DJI

Skydio CEO Adam Bry says the largest US drone maker will outsell DJI to American civilian customers this year, after Chinese sanctions, a dock bet with no market and a decade of building. Speaking with host Ti Morse on the Relentless podcast, recorded inside Skydio’s factory, Bry walks through how the company rebuilt its supply chain after Beijing seized its battery supplier, why it bet everything on autonomous docks in 2021, and why it refuses to grow headcount as fast as its revenue.

TLDW

Skydio sells to public safety agencies, utilities, construction firms and the military, and by Bry’s count it now beats DJI on US professional sales because the market shifted from manually flown drones to dock-based autonomous systems. The company was sanctioned by China two years ago and is spending about a billion dollars with US suppliers to reach zero China dependence. Bry explains the company’s big irreversible bets (computer vision autonomy, rolling shutter cameras, docks, public safety as the beachhead), how he decides where to spend his time, and why a lean team that will scale “no faster” than its people can carry is the company’s identity.

Thoughts

Bry’s strongest claim is about headcount, not drones. After raising easy money in 2020 and 2021, Skydio overhired, lost its scrappy habits, and then had to get lean when the dock business took longer than planned and capital dried up. Headcount stayed flat or slightly down for several years while output grew exponentially, and now that the company could raise as much as it wants, Bry says it will not. His benchmark is blunt: public hardware and software comps run two to four times Skydio’s headcount on products he considers an order of magnitude simpler. The line about LinkedIn is the sharpest part of the interview. When a hardware company doubles its team year after year, he says he knows “with 100% certainty” it is not operating well inside. That is a useful filter for anyone evaluating the current wave of richly funded robotics startups, and it connects to his warning elsewhere that many of them are underinvesting in the unglamorous work of reliable, end to end systems that customers deploy at scale.

His test for a bet-the-company decision is also worth stealing. Bry says he cannot convince investors or recruits of something he does not believe in his bones, so when he catches himself explaining a plan without that conviction, he treats it as a signal to change course. Docks passed that test in 2021 when there were no customers, no budgets and no market, and Skydio was burning heavily while its only revenue came from standalone drones. He could not hold two ideas at once: that Skydio would be very valuable, and that it would not invest in docks. A spreadsheet in 2022 showing today’s business, he says, would not have been believed. This is the honest version of “conviction”: not a feeling, but a logical inconsistency you refuse to live with.

The rolling shutter story is the best illustration of what a one-way door looks like in hardware. For Skydio 2, the company needed better 3D sensing at lower cost, so in 2018 it switched from global shutter cameras, the standard for computer vision, to the cheap rolling shutter sensors used in phones, where each row is exposed at a slightly different moment. Almost nobody did computer vision on rolling shutter because the math gets explosive. Skydio built the cameras into the drone before it had solved the algorithms, and there was no going back. It worked, and the result was better than the old system. Software companies can ship and patch. Hardware companies sometimes have to commit to the silicon and trust the team to make the math catch up, which is what Bry and Hadrian’s Chris Power mean by rolling a hard six.

The China sanctions get the headlines, but Bry’s more lasting point is that the hardest manufacturing failures happen upstream. Inside Skydio’s own factory, the team can throw people and brainpower at any problem. The trouble comes when a supplier quietly changes a process or a material in a way that seems harmless to them but couples to something on Skydio’s line or, worse, in a customer’s hands. Nothing changed on your side, yet the product behaves differently. His answer is to go deeper into the supply chain and exert more control over every component, which is also why Skydio’s billion dollar commitment to US suppliers reads less like patriotism and more like risk management. Being sanctioned forced the move, but the lesson applies to any hardware company with long supply lines.

One small operating habit stands out. Skydio’s senior leadership meeting opens with roughly half an hour of videos of products failing in customers’ hands over the previous two weeks. It is deliberately painful, and the point is not to fix everything in the room but to make sure there is a line of work aimed at every failure. Combined with Bry’s cheerful view that a drone getting shot, thrown out a window or attacked with a baseball bat is good news because the violence was not aimed at an officer, it paints a company that treats the field as the source of truth rather than the lab.

Key Takeaways

DJI can still sell in the US. The main restriction, in effect since earlier this year, bars new models, so Skydio competes with DJI head to head and expects to beat it on US civilian sales this year.
Skydio has essentially no consumer business. Its customers are public safety agencies, energy utilities, construction companies, departments of transportation and the military.
The market moved from manually flown drones to dock-based autonomous drones that live in a networked charging base and are operated remotely. Bry calls that shift the real basis of competition.
Skydio’s connected customers, excluding the military, fly about 2 million flights a year, roughly 40,000 a week, and the number is doubling year over year. The factory will triple output this year.
China sanctioned Skydio and Bry personally two years ago and shut down its battery supplier the same day. Skydio rationed to one battery per drone, then stood up new cells and packs, and gained a couple of minutes of flight time in the process.
Skydio plans to spend about a billion dollars with US suppliers over five years to build a domestic supply base, which Bry still describes as fragile.
The first product, the R1, was a “technology marvel but commercial failure.” It took four years to ship, and three to four more years after pivoting to enterprise in 2020 to get the product ingredients right.
Battery swapping docks add little for real customers. Usage data on flight patterns and time between flights shows the gain is small and the mechanism costs reliability.
The F10 fixed wing drone is launched and caught by a robot arm, flies for 2 hours at up to 100 mph, and mounts its camera gimbal on the wingtip so the wing never blocks the view during turns.
Skydio tracks payback per product. The R10 indoor drone paid back its development cost within a short time on the market, while some research bets are expected to take five years or more.
Skydio writes model policies for police drone programs and frames public safety drones as targeted response, not blanket surveillance. Bry says he does not want to help build an always-on surveillance world.
Vertical integration is relative. Skydio has a world class camera team but does not grind its own lenses or make its own chips, because chips for AR, VR and other robots already fit its needs.
Bry estimates Skydio is about 1% penetrated even in its most mature markets, so the company must keep inventing while running a large-scale core business.

Chapters

0:06 Outselling DJI in the US

Bry corrects a common misconception: DJI still sells in the US and only new models are blocked. Skydio expects more US civilian sales than DJI this year, driven by professional customers and the move to dock-based autonomy. He credits a customer preference for US suppliers, but says the real story is a product shift. DJI has copied Skydio’s fisheye, vision-centric design and its docks, which he treats as normal competition.

4:29 The Ten Year Hardware Baking Period

Skydio started almost 13 years ago and took four years to ship the R1, a consumer drone with great autonomy and little else. The company pivoted to enterprise and government in 2020, spent three to four years getting the products right, and has grown exponentially in the last couple of years. Customers now fly about 40,000 flights a week.

7:02 Scaling Production and Getting Sanctioned by China

Skydio built drones in house from day one, against Silicon Valley wisdom that said avoid hardware or outsource it all to China. When Beijing sanctioned the company, it seized the battery supplier’s equipment. Bry chose to treat it as burning the boats, rationed batteries, and rebuilt the supply chain with US and allied vendors.

11:33 Product Tradeoffs and the F10 Fixed Wing

Product decisions blend usage data with conviction about what could exist. Battery swapping loses on reliability for little gain. The F10 came from conversations with rural sheriffs and puts the gimbal on the wingtip, a choice that looks aerodynamically odd but delivers the “magic camera floating in the sky” experience.

16:52 Maximization Functions, Payback and the Dock Bet

Bry describes a good organization as a maximization function that expresses the best of each person, and explains why he keeps asking why in technical meetings. Cheap decisions get tried, expensive ones get careful upfront aiming. He tracks payback per product and recalls the 2021 dock bet, made with no market, heavy losses and wavering investors.

23:59 Government Sales, Politics and Privacy

Selling to government means taxpayer money and political accountability, which Bry calls healthy even when officials decide against drones. The real customer of a police drone program is the community. Skydio builds transparency into its products, writes model policies with agencies, and draws a line at blanket surveillance.

31:02 Designing for Scale and the Entropy Problem

A new low-volume product like the R10 gets different tradeoffs than the successor to the X10 flagship, which must launch with better reliability and manufacturability. Bry calls manufacturing a fight against entropy, and says the worst problems come from small supplier changes that ripple downstream.

35:30 Drones That Get Destroyed and a Hardware Bubble

R10 drones get thrown, clubbed and shot during tactical calls, and Skydio replaces them. Bry describes how skeptical employees became champions of public safety work after seeing it up close, and the staff meeting that opens with failure videos. He warns that hype-era robotics companies are underestimating what reliable deployment takes.

42:38 Rolling a Hard Six

Bry lists the irreversible bets: vision-based autonomy in 2014, rolling shutter cameras before the algorithms existed, docks, and public safety as the beachhead for dock infrastructure. He explains how conviction carries a bet, and how the company survived the R1 flop and the lean years before Dock for X10 scaled.

49:10 Vertical Integration, Wartime Mode and VC Dollars

Control over components can come from building them or from deep supplier partnerships, and Skydio does both. Bry loves operating with his back against the wall and says competition from Anthropic and OpenAI has been good for Google. He closes on staying lean even when capital is available, and scaling as fast as great teams allow but no faster.

Notable Quotes

“We will do more sales in the US to civilian customers than DJI this year.”
Adam Bry, on competing with DJI in the US market

“There’s nothing like burning the boats to motivate you to like to solve some hard important problems.”
Adam Bry, on being sanctioned by China

“I don’t want to live in that world. I don’t want to help build that world.”
Adam Bry, on blanket drone surveillance

“I actually love it when our drones are victims of violence.”
Adam Bry, on R10 drones attacked during police calls

“I couldn’t simultaneously explain to somebody why Skydio is going to be a super valuable company and we’re not investing in Dock.”
Adam Bry, on the bet-the-company decision to build docks

“I think the rise of Anthropic and OpenAI has been great for Google.”
Adam Bry, on competition as a forcing function for focus

“We’ll scale that as quickly as we can, but no faster.”
Adam Bry, on refusing to double headcount every year

Watch the full conversation with Adam Bry here.