Former enterprise sales director at Skydio on selling autonomy to energy & government buyers

We spoke with a former enterprise sales director at Skydio, the U.S.-based drone manufacturer known for its advanced obstacle-avoidance and autonomy systems, to break down how enterprise drone adoption happens in practice—from oil fields and railroad lines to law enforcement and retail experiments. He walked us through what makes Skydio competitive against DJI, how BVLOS (beyond visual line of sight) regulations have slowed scale deployments, and why drones sell best when they displace a manual inspection process with a clear ROI.

Can you tell me a bit about your background in the drone industry?

I started my career in telecom and software, then jumped into custom software development. I was involved in designing and developing the first 50 native iPhone apps for ExxonMobil working with Apple. After starting an oil and gas software company that faced challenges when the market collapsed in 2014, I spent four years at Salesforce managing energy, utilities, and manufacturing clients.

In November 2020, I joined Skydio. It was an opportunity to take what I'd done in the software space and complement it with truly innovative hardware that nobody else was doing in the world at that time. We were competing with China on a daily basis. I would probably still be at Skydio had I not joined when I did, because after I joined we faced supply chain issues that delayed deliveries to customers like BNSF, Exxon, and the Department of Defense.

Since leaving Skydio, I've done more consulting work around drones. I got deeply involved in the drone ecosystem because you need your Part 107 certification to sell effectively. Our motto was "you fly, they buy"—instead of doing a traditional software demo, we'd get out to places like Columbus, Ohio and fly around these cell phone towers with the right decision-makers present. I was once able to get a Skydio drone up to 1600 feet, these things are powerful.

What was the client mix like at Skydio, and did you see that shift over time?

You're going to see three main segments. First is energy and manufacturing—anything that's blue collar. Down here in Texas we know the difference between the oil and gas supply chain and railroad transportation. But some of these drone sellers think that everything with a hard hat belongs in one bucket, but that’s not really the case—it’s much more nuanced than that. But anyway: think chemical, refineries, construction, inspection. If you're inspecting fixed assets like cell phone towers or railroad lines where you either put a guy up on a 200-foot tower or fly a drone, that's a major use case.

Number two is government. Adam [Bry], Skydio's founder, developed the first-fixed wing mode, the R2, named after Star Wars of course, at MIT and got some black-box government funding. When I was at Skydio, it was primarily federal and state and local government. The Department of Defense was buying, we had several three-letter government agency customers, think CIA and FBI, but we were also seeing major uses with first responders. To sell into law enforcement, we inked a major relationship with Axon [Enterprise, fka as TASER International], the company that makes all the body cameras for law enforcement. So that’s how we got in the door with first responders at the state and local level.

The third bucket was what I would call innovation. For instance, with Walmart retail, we were doing, I would say it was more skunk-works than anything else. Walmart has been trying to crack that code to deliver via drones for so long, but I think they’re not going to be able to do that because of concerns around BVLOS or beyond visual line of sight flights. People don’t want that out in cities and the suburbs, so I think it’s kind of a tough market barrier. We also had customers like King Ranch and large construction firms like David Weekley Homes using drones for inspection. We also inked a deal with the NFL—any drone you see them use in a football game that’s not on wires, they’re Skydios.

Hollywood started picking up on that too, drones for filming. But we never wanted to be a Hollywood drone company, it wasn’t a great fit in terms of willingness to pay for our features. There are systems that are more geared to that use case. So, mostly Hollywood moved to other drones.

What was the role of software vs. hardware in sales?

About three-fourths into my tenure, 3D scan was developed. Skydio Cloud and 3D scan were the two big software products, and as you can imagine, both led to higher multiples on those deals, in the sense of much higher ACVs per year. Skydio Cloud was meant somewhat to compete with Pix4D or Bentley Systems, some of your post-processing players. And 3D Scan was basically enabling a 3D rendering of whatever your asset is.

What made Skydio stand out in the market?

The number one reason Skydio is the best, is obstacle avoidance. I can put a bird up in the air and fly it at 27 miles an hour straight toward a brick wall or even thin wires, and it will completely stop. As a pilot, this takes away your apprehension about crashing the craft. You can still do it, trust me. But for a normal guy inspecting a rig, it’s almost foolproof.

Second, Skydio is the only drone companies where all the hardware components are made in the US. This became important in 2020-2021 when concerns about Chinese-made drones like DJI emerged. With DJI drones, your aircraft is controlled by GPS, by Chinese satellites, raising security concerns. For government customers especially, this was a deal-breaker—they had to use American-made. When we went in for sales and people didn’t want Chinese components, the room would basically empty except for Skydio. It was like that scene in Forrest Gump where they’re the only shrimp boat left in the water.

Third is autonomy. We had eight cameras that weren't even for recording—they were just designed to avoid obstacles and provide situational awareness. With machine learning and AI, this led to world-class autonomy for autonomous missions, which unlocks capabilities like drone-in-a-box, where a drone can sit in a remote location, wake up on schedule, fly missions, return to dock, charge, and upload data, maybe using 5G or LTE.

Was “drone in a box” or drone nest/docking technology part of the product mix while you were there?

Well, it was on the product roadmap. I didn’t have a dock but all my solutions engineers did. It was a prototype at the time. It didn’t have all the glitz or polish of a finalized product. They came out with it about six months after I left.

What were common objections that slowed down deals, and what helped close them?

It’s always the same thing: find about their requirements, if they have a greenlit requirements docs, then they have a budget. Then you find out what they are, match the pitch to those requirements. It’s Sales 101.

The biggest objection was beyond visual line of sight (BVLOS) concerns. Large companies are very risk adverse, especially in high-risk industries like oil and gas. At ExxonMobil they’d have signs in the bathroom that said, “Please adjust your stride to the surface you’re walking on.” This all revolves around liability—when you have autonomous drones flying autonomous missions, people look at that like Skynet. Again, nobody has really cracked the consumer retail use case, like ordering a pizza and having it delivered in 15 minutes, yet at scale, because there's too many opportunities for problems.

There are legitimate safety concerns—if a drone fell in the middle of traffic, you could have a massive pileup. It’s all about safety. But privacy is an issue too—these things have 4K or 8K cameras on them, and people don't necessarily want them flying around their backyards. Sometimes I think fixed wing drones, things that are gliding versus hovering and are kind of dumb hardware, might be better in these retail and delivery applications.

BNSF would have to go get BVLOS waivers from the FAA, and a good salesperson takes that as an opportunity to get on the same side of the table as a customer. But unless it’s a self contained property or company-owned land or facility, businesses themselves just did not feel comfortable with BVLOS.

Second, you also have people that are misinformed and want these moonshot capabilities or have these innovation ideas, and those deals could falter. So you would find out quickly whether a deal would have legs or not. For anything that feels skunkworks, which was the case with Walmart, you could see the budget for that disappear pretty quickly.

For example, I went to Bentonville Arkansas to sell drones to Walmart to do inventory bar scanning, they wanted to find out how much revenue leakage they had and whether things were in the right place. But the Walmarts where they wanted to use these drones were 24 hour stores, but at the time they were not 24 hours because of COVID. But that was an artificial condition that was going to go away. So it was just a skunk works thing, a special point-in-time innovation use case, but that was not going to scale.

It was much better when there was an existing process that you would be unseating—inspection, logistics etc. Then you can bring your business finance guy and show them the numbers. You can go out and fly with them, sell 20 to 30 drones, and then you just scale the amount of birds over time. It’s crawl, walk, run. It’s nice in that way, and different from software, because the customer sees the gains very concretely, and you can grow the contract in phases.

So how did you sell it, the sales pitch for closing deals? It all came down to efficiency and ROI. How many full-time employees do you have at this oil site in West Texas? How much gasoline and wrench time is it taking away from this worker to do manual inspections? How much trailer time is it taking for him to go back to the trailer and sit there to send that data to headquarters? Then customers could see the efficiency gains. So those deals would sell themselves when they already had something going.

Speaking to the FAA restrictions on unsupervised flights that require waivers for BVLOS, are those waivers becoming more routine because the technology is itself making drones safer? Is the loosening of regulations a consequence of where the technology is going?

Yes, having obstacle avoidance is getting really good. But I think it’s more that the regulations are changing as the lithium ion batteries get more powerful, and as drone-in-a-box capabilities get refined. That’s what’s putting pressure on the regulations.

Because you can play hopscotch with drone-in-a-box and drone nest systems, you can have one drone land and another take off. But it totally defeats the purpose if you have to have a guy following it. So for the industry to take advantage of the technology, you have to have looser BVLOS restrictions.

How did customers respond to the software offerings like Skydio Cloud and 3D scan?

Buyers who were educated on SaaS knew what we were doing right away—they recognized this was our recurring revenue tail, how we’d really make the money. And they had negotiated contracts like that before. But that wasn’t the case for everyone.

With Skydio Cloud, it was somewhat commoditized because if you were a developer, you could just use Dropbox or Box. Sure, you wouldn't get the post-processing capabilities we offered. But people weren’t that excited, so then you’d come back to a religious argument, kind of your Android vs. Apple, where we were the more premium offering, offering white-glove service, whereas maybe they wanted a cheaper, modular solution.

But 3D scan was incredibly exciting because the product roadmap was shared with customers all the way. We got to have early iteration and feedback at the start of that journey. Our customers felt like they were already wearing our jersey by the time we were coming out with the final version. They had already inked the contracts around it because we were developing it adhering to their standards. That was more successful.

How did economics work around pricing hardware and software?

It depended on how they wanted to structure the contract. It was kind of a Wild West with contracts at the time. There were a lot of bespoke deals. That isn’t surprising since we were a relatively new company.

As a general rule, pricing was typically either per bird, which we didn’t really want to do, or per user—around $1,000 per user per year. That was low because of the early stage of the ecosystem and the limited feature set at that time. It was really asking customers to take a leap of faith with us. A16Z was our main backer, and they thought that it was good for us to show we had penetration with 3D Scan and Skydio Cloud, it was going to feed into positive traction in the market, even if we were kind of giving away the software effectively sometimes.

How did you get initial interest from enterprise buyers as a relatively new startup?

It's easy to answer that—if you want to do something in technology and sell directly to the enterprise, you have to prove something that hasn't been done. You need a Trojan horse capability.

Adam [founder and CEO Adam Bry] had to develop groundbreaking technology. The R2 had a six-foot wingspan and could fly through a parking garage at 35 miles an hour by itself without hitting anything. When you saw that at the time, you’d think it was a magic trick. Obstacle avoidance and autonomy were the two big differentiators.

The second hurdle with enterprise sales for a startup is the paperwork—indemnity and security concerns. Sometimes you have to do non-standard things like putting code in escrow to show you're willing to work with enterprise requirements. It can be a lot of work.

Who actually had the decision-making power in these enterprise conversations, who owned the buying decision?

In an enterprise situation, it was almost always the leader of the business line—whoever owned a P&L. You’d think you’d be selling to IT. That’s true when I was at Snowflake. But with drones, it’s not the IT buyer. With drones, you're selling to the actual folks in the field and those responsible for the P&L, which is great because they usually get a faster answer than going through IT.

What else are you seeing out there in terms of technology, maybe that’s more at the bleeding edge?

More recently, drone companies are looking at how to sell swarm capabilities, commercial drone swarms, and like I said, drone-in-a-box and nest technology—autonomous systems are getting big.

Percepto is an Israeli company that’s pushing into a lot of the autonomy stuff.

But what people really want is to have computing reach toward the edge, that’s what it comes down to. Anduril’s system Hivemind is basically all about what you can do with drones, AI, and edge computing, you can also look at Armada International.

But I think a lot of this is about first-mover advantage. Stuff like BVLOS will help but you have to be there with the technology and in the conversation first, that’s what really matters, and that helped Skydio along the way.

This transcript is for information purposes only and does not constitute advice of any type or trade recommendation and should not form the basis of any investment decision. Sacra accepts no liability for the transcript or for any errors, omissions or inaccuracies in respect of it. The views of the experts expressed in the transcript are those of the experts and they are not endorsed by, nor do they represent the opinion of Sacra. Sacra reserves all copyright, intellectual property rights in the transcript. Any modification, copying, displaying, distributing, transmitting, publishing, licensing, creating derivative works from, or selling any transcript is strictly prohibited.