Axon is a fascinating company. It was founded 33 years ago by its current CEO – Rick Smith – after two of his friends tragically passed away in a road rage incident. Smith’s sorrow sparked an intense energy within him that morphed into motivation for inspiring change. He then dedicated his career to the initial Axon goal of providing better public safety tools to keep the world safe. He wanted to equip everyone with tools to safely de-escalate, rather than lethally injure. His answer? Tweaking and popularizing the Taser. 33 years later, Axon is a $50B company where he’s still CEO.

Axon, which was called The Taser Company until 2017, did not invent this hardware, but it did change the design and go-to-market strategies that led to its broad proliferation. The inventor was actually Jack Cover, a NASA scientist who used gunpowder to shoot the canisters. That decision meant Tasers were regulated like any other firearm and prevented meaningful traction. Smith sought to find a loophole that wasn’t legally treated this strictly and enabled more relaxed restrictions. He knew this technology could be an invaluable asset without needing deadly force. Together, Smith & Cover redesigned the Taser to use compressed nitrogen instead of gunpowder, which unlocked its use for civilians without a gun permit.

Interestingly, Smith and Axon initially focused on the consumer market and found very little success. Because of this, the company was on the brink of bankruptcy throughout its early years, even relying on the family to plug liquidity gaps and avoid shutting down in 1994. Like any generational founder, Smith dug in. He bobbed, weaved, contemplated and clawed until Axon finally found the product and go-to-market strategy that drove its success.  The change? Pivoting to a focus on state and local law enforcement, which became Axon’s primary growth driver for decades and is still its bread and butter today – with plenty of runway remaining. Early on, he even made himself a public test dummy in a famous marketing stunt to convince customers of this product’s value (& non-lethal incapacitation). He was and is all in.

As we’ll learn throughout this piece, Axon is now the global Taser leader by a mile. It has leveraged this great success to expand into several product categories that round out its positioning as a holistic, central operating system for its customers. Over the last few decades, they’ve carefully built a large and connected product platform well beyond the Taser to cater to a wide range of public safety needs.

Their increasingly platform-wide pursuit is why they’re now called Axon, as this refers to nerve fibers that connect the human brain and foster movement. Axon wants to be the company to connect public safety workflows, departments and systems into one collaborative force for enhanced security. This goes for both hardware and software, with Evidence.com (Axon Evidence) marking its software debut 18 years ago. And as we’ll see throughout the piece, the magic of Axon is offering this better safety while also generating meaningful cost and labor savings for law enforcement customers.

The founder-led culture can be described as responsibly ambitious. They’re a bunch of friendly yet maniacally focused sci-fi nerds, as evidenced by Smith rocking his Star Trek-themed attire at the most recent investor day. They host grand events that resemble Berkshire Hathaway or Dreamforce, yet do so only if they can generate enough sales pipeline to make them worthwhile for shareholders. They take innovation and growth very seriously, but have strict guardrails in place to ensure stakeholders are still cared for.

Management is also exceedingly competitive. They are driven to outspend, out-hire and out-innovate the competitive landscape. They will openly criticize themselves when they think they can do better, and show a consistent ability to quickly fix any operational issues as they pop up. For example, the company was quite blunt about a disappointing start to bookings in 2024. They said they weren’t focused or proactive enough in sales motions (despite what was a pretty good quarter), and fixed that for Q1 2025, delivering bookings momentum that surpassed expectations and set them up for a great year. Some companies will search for any excuse they can to explain bad results. Others, like Axon, overcome very real headwinds and execute. They can’t stand losing and are obsessively determined to beat everyone.

Axon’s leadership is wonderfully paranoid, constantly looking over their collective shoulders and searching for new threats to address. They are allergic to complacency despite holding dominant positions across several product categories. That starts and ends with CEO Rick Smith and extends to every corner of Axon’s operations. Simply put, this company is a lean, mean fighting machine focused on the right things and led by the right people. And they don’t want us to rely on their internal opinions. They want stakeholders to hold them accountable to their abstract goal of improving public safety by setting a concrete target of cutting police-related deaths in half by 2032. They won’t declare a win unless it’s objective and they’re energized by all the tools they have to secure that sweet, sweet victory. We’ll dig into all of that throughout the piece.

Rick Smith’s resume is simply graduating from Harvard and founding Axon. One of his key priorities over the last handful of years has been building out the rest of his executive team. He wanted to deepen the bench of talent to confidently hand off day-to-day work so that he could focus on big growth opportunities and customer relationships. Axon’s President is Josh Isner. I love his background because it screams loyalty and alignment with the company mission and culture. Like Smith, he finished school at Harvard and immediately began climbing the ladder at Axon. 17 years later, he is running earnings calls and core Axon operations. He’s energetic on these calls and certainly a tad feisty, as he’s happy to remind a bearish analyst when their forecasts were way off during a Q&A.

The COO/CFO is Brittany Bagley. Ideally, I think a company of this size should have a separate person for each role, but this isn’t unprecedented by any means. Bagley was the CFO at Sonos for 5 years before this and was a Managing Director at KKR before that. The CTO and Chief Product Officer (CPO) is Jeff Kunins. He was just CPO for 4 years and added the CTO title in 2023. Before Axon, he was a VP of Alexa and Kindle at Amazon and a GM of Messenger and Social at Microsoft, among other things. Finally, Cameron Brooks was named the Chief Revenue Officer almost 2 years ago. He spent 6 years at AWS as the GM and Managing Director of their EMEA region and was a Director of key divisions at IBM like Watson. Internal employee surveys indicate strong employee support for leadership, and the company routinely earns “great place to work” accolades from various 3rd-party research organizations.

Axon sells hardware like Tasers and body-worn cameras (BWCs) and software to form an end-to-end product suite. All products are bundled into various subscriptions, where hardware revenue is largely recognized upfront and software components over the course of the deal. 96% of its revenue is tied to active subscriptions, which, when paired with sky-high renewal rates, leads to great demand visibility. Recurring monthly subscription fees range from $99/month to $600/month. The firm loves to offer free trials to show off what it views as a superior offering, as these trials frequently close the sale. Like all good subscription-dominant companies, they’re fixated on elite price-to-value. They will not opportunistically take price at every turn and instead focus on improving product and customer value wherever possible. When (and only when) that leads to an overly aggressive price-to-value gap, they will take price. And? As a nod to the mission-critical nature of their platform, they’ve never dealt with material churn when doing so.

The Total Addressable Market (TAM) sits at an impressive $160B. They have so many growth levers at their disposal, while their most mature U.S. state and local law enforcement niche ($19B TAM) is just 15% penetrated. The 15% figure is before factoring in all of the product innovation and up-selling available to capture larger deals and more budget. And they’re already basically in every single one of these departments across the nation (about 17,000 of them), with cross-selling the biggest remaining driver of their growth. They already dominate. They’re already ubiquitous. For a sector like policing where vendor reputation and trust are especially important, that gives them the ability to win in other product categories. As we’ll explore later, they are doing just that. 

Across its emerging enterprise ($48B), international ($78B), federal ($12B) and civilian ($5B), the company is in excellent shape in terms of finding more opportunities to drive sustainable, multi-year growth. They don’t need to reinvent the wheel to find more demand, as the highly relevant product adjacencies are bountiful.

In terms of broad categories, Axon splits its connected device revenue into Tasers, personal sensors (like BWCs) and platform solutions (like drones).

Before we get into Axon’s various products, we need to talk about the sensors that augment their value. Without these tools, Axon’s value proposition would be far more limited. Its Taser and other offerings can provide a mountain of valuable data to improve processes and outcomes. To capitalize on this, the assets need to actually know when to start collecting that knowledge. And that’s where sensors come in.

The sensors are deeply embedded across its vehicles, officers, city infrastructure and community volunteer networks and tie very closely to these pieces of hardware. The sensor-to-hardware interoperability ensures that data collection reliably happens exactly as needed and greatly helps customers grasp more complete information to do their jobs better. These automatically jumpstart recordings, log weapon activity, activate first responders and more. Officers don’t need to remember to do all of this on their own while they work through dangerous and rapid situations. They can focus on doing their jobs in the field, knowing the sensor base will understand situations to a point of reliably and automatically activating cameras in real-time.

Considering all of this, Axon boasting the world’s largest connected sensor network means the high-fidelity context it delivers to its customers is unmatched. How do these sensors know when to start streaming footage? Through a series of triggers that can prompt these cameras to turn on. Types of triggers include:

The firm provides this wireless communication technology through a product called Axon Signal. Leadership will frequently call this the “eyes and ears” for everything else it does, and I think it’s clear why that is. The boatload of quality data its sensors provide turbocharged insight gleaning across its full ecosystem of products, helping sharpen things like drone response quality, vehicle identification and so much more. As we talk through every single product in this connected devices section, keep in mind that sensors accompany all of them and bring their maximum potential to life.

It wouldn’t feel right to start the product portion of this deep dive with anything other than the Taser. This company owns 95% of the entire market and has a presence in virtually every single state and local police agency in the USA. And while Axon’s Taser has a decades-long history, it makes sense to focus this section on its newest Taser 10 model. It views this as a key unlock for less lethal policing and reaching its 2032 goal of cutting fatalities in half. In my mind, that opinion is the right one and is based on a few exciting technological upgrades that make the Taser a better option in more scenarios. Their goal is for this to be the de facto weapon of choice for officers instead of firearms, and this launch got them a lot closer to that objective. There are three main reasons why.

First, the Taser 10 model now holds 10 shots and 1 cartridge per shot. Previously, this hardware allowed for 2 shots and fired 2 cartridges at a time in order to work. As these move towards the target, they naturally spread and make accuracy rates much lower than a gun. That obviously makes officers more hesitant to select this option in high-stakes, rapidly evolving situations. With Taser 10, officers have more shots on goal and accuracy rates that can eclipse 90% with proper training vs. 65%-70% for previous models and even lower (national average around 35%) for guns. 

The second major upgrade comes in the form of a near doubling in range. The previous model and its sub-scale competitors had 25 feet of range, while this has 45 feet. While that may sound subtle, it isn’t. In the policing world, it’s a game-changer.  It means an officer deciding to use a Taser in high-stakes situations can now select this weapon before it’s nearly too late to actually use it. Officers need confidence that Tasers give them enough time to think through situations, as that can be the difference between life and death. They need to feel better about using this weapon in up-close confrontations or combat, and this range boost helps mightily in that regard. Not just that, but a Taser doesn’t jam if it’s pressed up against an officer's vest and it doesn’t take a life if fired at the wrong part of a suspect’s body. Both of those factors are clear benefits that a traditional firearm doesn’t match, which is important.

The third and last upgrade is still in beta testing but comes with so much promise. Axon will soon fully launch its upgraded “Apollo Cartridge.” Sticking with the wonderfully nerdy sci-fi culture theme, Apollo is moon-landing inspired. Why? Because Axon views its policing death reduction goal as a “moon shot” and believes this will be a big piece of it getting there. Why? Apollo solves a key Taser issue. Up until now, these weapons have been pretty useless in cold-weather climates with people wearing thick clothing. Tasers must be powerful enough to pierce the skin and not powerful enough to keep moving deeper thereafter. Otherwise, people will get seriously injured and the Taser utility drops. The Apollo cartridges rely on clothing and skin piercing differently. When the projectile cartridge feels penetration resistance change, it is designed to deploy flaps on the Taser spear to slow it down. This way, the weapon works if a suspect is wearing nothing or a winter coat. They will likely ramp Apollo production in the coming months, which will represent a key up-selling opportunity across their gigantic install base. Looking ahead, Apollo, more Tasers per department, upgrade momentum and key growth areas like international and corrections centers (more later) can fuel many more years of brisk compounding for its oldest product category. It’s still growing at a nearly 20% Y/Y clip decades after initial launch.

Despite all of these advantages, some officers still struggle to use the Taser effectively in the field. Knowing this, Axon has worked hard to standardize virtual reality (VR) training modules that simulate a plethora of real-life situations and boost concept retention by 45%. These modules can be perpetually accessed, making training more frequent than a typical annual session. They also look and feel like video games, effectively giving officers a more engaging way to learn. When customers deploy both VR and physical training, employ a Master Instructor and follow behavioral recommendations like firing until an adversary is controlled (very different than a gun), programs are routinely successful and strong traction ensues. Axon has worked very hard to communicate this playbook, and it has already had a material impact on Taser outcomes.

With all of this in mind, it’s no wonder as to why Taser 10 has been a home-run. Specifically, on an apples-to-apples timeline basis, the Taser 10 adoption rate is 2x higher than prior models. It is driving frequent upgrades to higher subscription tiers and exceeding management expectations quarter after quarter. The upgrades are shrinking the normal 5-year upgrade cycle and the model is still supply constrained 3 years after launching, as they simply can’t keep up with insatiable demand levels. They’re rapidly automating and expanding manufacturing to try to service all interested buyers.

When we hear personal sensors, think of everything physically on an officer’s or employee’s uniform. This includes Axon’s leading body-worn camera (BWC) called Axon Body 4 (AB4) and its weapon holster sensors.

Whether it’s a gunshot audibly activating a BWC to begin recording or drawing a weapon doing the same thing, Axon’s BWC product ensures officers are recording more of their work, driving enhanced accountability and transparency. While one may read this and think “do officers really want this?” The answer is a resounding yes. The vast, vast majority of officers are just trying to safely do their jobs. Despite this, they deal with false accusations constantly, with the AB4 serving as a vital evidence collector and conduit into the rest of the Axon Ecosystem (more later). Now they simply record what happened instead of recounting it in the future. This way, review boards and judges don’t need to subjectively assess a recounting, they can actually see what happened. This was the second problem Axon solved after Tasers, and it now boasts a commanding 85% market share in major U.S. cities. As we’ll see throughout the software section, Axon has worked hard to layer in several add-ons that extend the utility lead this product enjoys over its peers.

The efficacy of this hard work and its integrated sensor, hardware and software network means $1 invested in AB4 from a police department saves them $4 on average via lower non-compliance and legal costs as well as faster, more efficient response and resource deployment.

The rest is for paid subscribers. It includes detailed commentary on the rest of the large product suite, growth opportunities, risks, financials, modeling and my conclusion.

If you'd like to read those sections, enjoy full earnings reviews on 40+ companies this season, track my portfolio/performance and join the Discord community, you can access 25% annual discounts below.

Upgrade & Save

Non-drone Platform Solutions include its in-car sensors and cameras called Axon Fleet. This boasts low-light automated license plate reading (ALPR) technology that can go beyond traditional IDing variables like car make, with other attributes like damage, roof racks, and bumper stickers. Axon Fleet 3 (latest model) can see across 3 lanes vs. 1 lane previously and comes with a real-time streaming camera as well. While ALPR is great at automating static plate reading, matching it against hot lists and alerting officers when there’s a match… that’s about it. If, for example, a department’s command center wants to activate a live stream to get more eyes on a pressing situation, they need this type of camera and they have it with Axon. The live stream can capture wider feeds than ALPR and essentially turns police car fleets into eyes for central command centers.

Axon Fleet greatly helps with total police coverage in a jurisdiction and offers departments a more complete understanding of the area they’re responsible for protecting. But Axon now goes even further than this. It saw that competitors like Motorola and Flock Safety had an established base of fixed cameras attached to existing infrastructure around urban areas. And players like Motorola have used this to scale their ALPR capabilities beyond what Axon currently has today. That gap is now quickly closing, with two important product launches called Axon Outpost and Axon Lightpost. Outpost is a fixed camera system that boasts dual-ALPR and a live streaming camera system that it borrowed directly from Axon Fleet. It plugs into an existing building or a pole using either wiring or solar power and without needing to retrofit that fixed infrastructure pre-installation. Axon Lightpost is the byproduct of a partnership with Ubicquia that can access existing power from streetlights to turn them into “public safety nodes” (again without any retrofitting). Doing this requires maneuvering slow bureaucratic processes and cutting through tape, which Ubicquia has already done for the shared customer base. That means their customers can immediately plug into millions of cameras around U.S. urban centers, allowing police departments to enjoy more seamlessly scaled visibility.

It’s clear that these two products will turbocharge ALPR and live-streaming coverage to catch up to Motorola’s more scaled offering, which feels inevitable to me. Axon simply has not been focused here in previous years. They dominate wherever they focus. This is now a priority.

And there are clear reasons to believe Axon can quickly win.  Installation of Outpost and Lightpost can be done in minutes and without any formal mounting. They just plug it into an existing outlet to extend its ALPR and live-streaming coverage. That makes these products easier to move across locations if a department needs more eyes somewhere for a concert or fair, for example. Next, Axon’s offerings are default ALPR and live-streaming and plug perfectly into the powerful and end-to-end software suite we will assess in section 3 of this piece. Motorola and others require separate subscriptions to access live-streaming and pieces of their software to emulate the full value Axon provides. Next, while vendor lock is a risk we’ll get into on the software side, it’s a non-issue and even a strength here. Axon’s “Works with Axon” initiative ensures customers can access the software and tools within Lightpost and Outpost with 3rd-party cameras from trusted partners like Bosch and Axis. Motorola generally makes their customers use their own cameras and infrastructure, meaning they require more rip-and-replace with more operational disruption if a potential customer doesn’t want to give up their familiar camera systems.

The last Axon edge is the most concrete. It delivers major value through undercutting potential substitutes. Outpost is $2,749/year and Lightpost is $3,249/year vs. $3,000 minimum for Motorola’s base ALPR offering. If customers want Motorola to provide live-streaming and its own integrated software suite, those subscriptions quickly and meaningfully rise over $4,000/year. Flock Safety (another entrenched player) is even more expensive than Motorola when considering all the 3rd-party software licenses, integrations and functionality. Sources say this type of package would be at least $5K/year. So… Axon is easier to install… more interoperable… more portable… and cheaper on a head-to-head feature comparison. That’s interesting. In Axon’s more mature categories, it doesn’t compete on price. It’s typically more expensive than sub-scale alternatives, as it can lean on its highly sticky ecosystem and interoperable feature breadth to stand out from the pack. Conversely, in this area, it’s the hunter, not the hunted. And it’s hunting with a major pricing advantage. Whether it’s this edge, the 3P integration advantage, the installation ease or the connection to a better platform, I think Axon is going to lead here just like it does in everything else we’ve reviewed so far. In my mind, it’s simply a matter of time.

Axon bundles the Fleet Cameras, Lightpost, Outpost and a few pieces of software like Fusus (so much more later) into a suite called Vehicle Intelligence. It’s like ALPR on steroids (also with live streaming). This launched last fall and effectively upgrades Axon’s existing ALPR capabilities with the fixed ecosystem and easy 3P integrations to vastly bolster vehicle identification processes and expedite resolutions… without relying so heavily on dangerous high-speed chases. Early interest in the bundle has been uniformly strong and revenue should ramp meaningfully this year.

“We can offer this suite at a far more reasonable price than competitors. This capability is complementary to the other products we build… Agencies that took us 10, 15 years to win on body cameras and digital evidence management are asking us for very large deployments out of the gate of this product.” – Axon President Josh Isner Spring 2025

As I’ll work through in the “Axon Fusus" section, Axon is rapidly expanding the deployment of cameras (or eyes) to make it a superior partner for expediting and enhancing resource deployment decisions. Whether it’s through Outpost, Lightpost, Works with Axon or other initiatives, they are making sure a customer’s single view of public safety is most powerfully enjoyed with Axon.

Skydio:

Axon technically includes some of this revenue in its software category, but the passage makes more here. Just know that this piece of platform solutions contributes to both major revenue buckets.

Axon Air is the name of the operating system purpose-built for drone programs. It automates the collection of flight logs, tracks drone pilot certifications and records deployment history and uses all of this to drive fleet work optimization and compliance. It also leverages all of this lucrative telemetry to handle drone maintenance schedules and ensure these assets deliver maximum up-time. Furthermore, with Axon Air, a drone pilot can use partner drones to simply pick a spot for a drone to go. The big partner here is Skydio, which is an outdoor, wide-area search and rescue drone company. While Axon wants to build everything internally, that is not realistic. It saw how far Skydio was ahead of the competition, contemplated how expensive and time-consuming it would have been to get there, and decided to partner. It also participated in the company’s series E round and owns probably around 2%-3% of it.

So what did they see in Skydio? Per leadership, this is the “only viable solution in the market if you’re flying below 200 feet” thanks to best-in-class autonomy, object locking/tracking and elite collision avoidance. That collision avoidance is made possible by something called “PathFinder,” which picks the optimal route and flies. This can work without any light thanks to its “NightSense” tool and can be easily used by beginners without expert knowledge on how to fly these things. All they have to do is tell the drone where to go. Skydio can also handle high-altitude flights and extreme weather conditions and autonomously dock and charge these models upon human command too. These docks can be installed on fixed surfaces or on patrol cars, effectively giving officers their own “private air force” between this feature and its connected software.

Because Skydio can take aforementioned destination instructions, map its physical surroundings and safely move from point A to point B more frequently than others, leadership is confident in their models being best-in-class. While that’s subjective, their leading market share is not.

Skydio drones are the market leader in the USA. Interestingly, this utility doesn’t rely only on point-to-point (P2P) radio connection. It also can use cellular, which means near-infinite range and connectivity without a line of sight with the drone. Pilots can easily hand off between P2P when they need lower latency feeds and cellular connections when they need more range. All of this, which is powered by Nvidia neural networks, gives the Skydio team confidence in their clear product leadership in the USA.

Skydio builds its drones and robotic docks to autonomously station and charge upon human command. Axon is effectively the exclusive Skydio channel selling partner and gets a referral fee.  You may be wondering why Axon is willing to concede this portion of demand to a partner? Two compelling reasons. First, they deeply respect Skydio and I would not be surprised if they end up owning it. Second, bundling best-in-class U.S. drones makes Axon’s suite more compelling; it provides more opportunity to lace its high-quality software throughout this high-quality hardware to create a higher-quality ecosystem. They routinely up-sell important software products that we’ll dive into soon.

China is where competition was the fiercest. But? Those have been deemed a national security risk in the USA, which was obviously good for Skydio’s competitive positioning. Axon even leaned into this with a SkySwap program, paying customers the replacement value for their drones in exchange for Axon Air and other drone product credits. That went a long way in fortifying Skydio’s piece of the rapidly growing market amid sharp regulatory change.

Skydio thinks the last decade was spent driving perfect product-market fit and collecting needed licenses to unleash this product. They’ve spent the last few quarters (with Axon at their side) beefing up manufacturing capacity and preparing for this hardware to keep selling like crazy.

Sky-Hero:

Axon also bought another drone company called Sky-Hero in 2023. While the name doesn’t sound like it, these are tactical rugged drones for indoor environments. You can beat these up like crazy and they will still fly perfectly, making them well-suited for indoor flying in tight quarters where collisions are inevitable. The model is built with polycarbonate and protected rotors and utilizes analog video signals, enabling connectivity through thick walls. It even boasts a “turtle mode” to course correct if the drone flips upside down. They can feed location data without GPS, using (you guessed it) Axon ecosystem sensors. This is a lot more niche and mostly for special operations teams and other military use cases. The skillset for this product is delightfully complementary to Skydio.

Counter Drone:

Drones are a very exciting technology that can augment talent-constrained police departments and improve public safety outcomes. But? Like with any emerging category, there will be bad actors as people hijack the well-intentioned innovation. Drones will be no different.

Knowing this, Axon bought Dedrone late in 2024 to power its set of counter drone offerings. Now, Axon is able to map all drones in a geofenced area and alert responders as soon as any unauthorized flight takes place. This is unsurprisingly sensor-powered and serves as yet another example of them delivering value across the entire product suite. These assets enable data collection through the hardware, and Axon’s software suite knows exactly what to do with that insight from there. This connected hardware and software can be installed in fixed locations (FixedSite) or portably in stadiums, concert venues and more (via Dedrone CityWide). Beyond the core observability capabilities, Dedrone’s technology can actively jam a drone to take it out of the sky or send it to a desired location via a product called Dedrone Defender. And notably, it can be installed very easily by simply telling Dedrone what area needs protection and offering locations to place their hardware.

The founder (still at Axon) showcased the hardware portion of this tool at Axon Week 2025 to give an idea of what this actually looks like.

Notably, Defender is only legal for federal-level agencies. Not state and local. They think that permission, with the World Cup and other major events coming, will be a key growth unlock and view it as inevitable. There’s a ton of pent-up demand whenever this approval comes. Even before this, however, traction is very strong. The world saw the weird New Jersey drones issue and the drone-assisted assassination attempt on our president, which are both serving as tailwinds for proliferation. There are 3,000 illegal drone flights per day in the USA, and this is Axon’s response.

DFR – Tying it all Together:

Axon neatly ties all of these products and software tools like Fusus and Axon 911 (I promise, much more soon) into a tightly integrated bundle called “Drone as First Responder” (DFR). Just like Outpost and Lightpost give Axon a uniquely diverse and broad ability to grow its base of ALPR and other sensors to offer better area coverage, so does this. It ties perfectly into command center workflows and data feeds from all of these other sensors so operators know exactly when and where they need more sight. And? They can place that added vision wherever they want to via data-driven decisions and easy drone piloting. This just gives a taste of how truly unified Axon’s end-to-end platform is, with the software section making this equation glaringly obvious.

Drones also give Axon another secret weapon in the form of helping departments do far more with less. This leads us to the pressing issue of policing labor shortages. 14% of police departments are fully-staffed and applications per available role are plummeting. Most aren’t even at 80% capacity and labor takes up more than 95% of a typical budget. While Axon has built a great company, its share of the budget opportunity is generally in the low single-digit percentages of total and only gets up to 5% max. That’s an enticing opportunity in my mind. If Axon can provide great tools to make officers more efficient and log all of this information in an organized fashion to query and use, then I believe the firm can command a much higher percentage of the overall budget than previously thought. Drones are a massive unlock in this regard and can even lower officer deployment rates by 25% due to their presence at the scene of a crime naturally deterring bad actors. Policing is starved for sources of incremental efficiency and DFR can provide exactly that.

All of the aforementioned geopolitical momentum for pieces of the DFR suite combine to foster fantastic momentum for this product since launching. They have 9-figure deals in the pipeline and so many interested parties to pursue.

Time to stop saying “more later.” Let’s talk about software. Axon Cloud Services is its suite of apps, operating systems and infrastructure to power everything Axon does on this side of the business.

Axon Evidence is the gold standard for DEMS. This is what can seamlessly scale and neatly ingest all of the vast piles of unstructured data that Axon sensors collect from its connected devices.

It can take this mess of insight and make it orderly, understandable and leverageable across every software workflow it provides. This is the conduit between the content its connected devices provide and its software being able to reliably make sense of it all. Whether it’s data from drones, Outpost, Axon Fleet, holster sensors, interview rooms, or any other part of its sensor base, this turns chaos into actionable signal and gives police operators a far more complete understanding of the information flowing into command centers. It’s also what integrates with the justice department and court systems via Axon Justice, cutting time required for legal case prep by automating and operationalizing open yet secure data sharing between Axon’s base of customers and these stakeholders. For a department in Scotland, this alone reduced their court backlog by 80%. Want your DEMS to be interoperable with the people constantly bugging you for evidence and information regarding various cases? Then you want Axon. 

Motorola sort of copies this with a very basic link-sharing feature, but it pales in comparison to what Axon provides. Simply put, Axon Evidence is where knowledge comes to life. Knowledge is power, better resource deployment decisions, faster time-to-resolution, lower costs and safer outcomes. Knowledge is everything and Axon Evidence delivers it in abundance.

Axon technically has a productivity division in its software suite. This topic blends into Axon Evidence and belongs here because it helps Axon Evidence do everything referenced above. Axon Records is the product to know for this bucket. While Axon Evidence provides the scaled repository for data, Axon Records is its sidekick Record Management System (RMS). It can easily search through the DEMS database and pull information to populate reports for easier compliance.

Axon Fusus:

In 2024, Axon bought a company called Fusus to power its Real-time Crime Center (RTCC). While Axon Evidence and Records are fantastic for ingesting information and making sense of it, Fusus takes this a step further by arming command center workers with real-time information on where issues are popping up, what resources are available and how to best deploy them. With this in mind, I think it becomes more clear and exciting why Lightpost, Outpost and DFR are such big focus areas. Not only are they solid independent revenue drivers, but they give Fusus incredible amounts of actionable information. There are several places where it’s clear that Axon successfully groups products in ways that make them better together and drives platform adoption. I think Fusus is one of the best examples of where this comes together.

It effectively takes the mountain of sensor data feeds and gives a “single pane view,” or a map of all available assets (officers, cameras, partners & drones) and their locations. Want to know where the closest officer is to send to a scene? Fusus has you covered. Want to understand drone availability to optimally place them where needed and to help perfect DFR workflows? Fusus has you covered. Want to alert all available officers as soon as an ALPR cam picks up a match? Fusus has you covered. Want to track criminal movement and activate Outpost and Lightpost infrastructure whenever a person of interest is in the area? Fusus has you covered. Simply put, this is why knowledge equals power for Axon Evidence specifically. It’s how Axon arms law enforcement to help them make faster and better decisions. It’s how Axon lets its customers act methodically and effectively as dangerous situations unfold.

And I think this is where Axon will distance itself from Motorola, Flock and all other companies trying to offer a holistic, complete view of a given jurisdiction beyond simply ALPR-based solutions. Why? A great interface, great product value and superior 3rd-party interoperability.

Let’s stick with the interoperability note for a second. First, the “Works with Axon” product that has been so popular for products like Lightpost works here too. Customers do not need to access Fusus only through Axon’s hardware. They can keep their trusted 3rd-party systems in a way that makes Fusus easier and more convenient to adopt than competing solutions from others like Motorola. And they significantly amplify their video feed reach with a couple of key partners. The first one is with the Amazon-owned smart doorbell and camera company called Ring. They have millions of users and can turn neighborhoods into connected networks to help keep people safe. Next, Axon also has a newer partnership with the Citizen public safety app. This is a police department’s connection to the public and their interface for requesting information. It’s sort of like a next-generation process for anonymous tips with more data monitors and filters to ensure a quality data pipeline. It lets departments send alerts to ask for help from people close to an area of interest. They allow customers to anonymously “Go Live” or start filming and feeding that information into the Citizen app for officers to use. 

Both of these partners and their user networks are now neatly tied into Axon Fusus. This way, through a Fusus tool called Axon Community Request, the sought-after information is shared right within Fusus. Officers simply circle an area and everyone within it is notified and asked (never forced) to contribute. And? Axon can now embed the feeds that any user opts in to provide. Both apps have millions of users, giving Axon a dramatic boost to its overall coverage by forming a 1P/3P partnership network and serving as the gatekeeper between those two sides. That creates a ton of stickiness and unique value, as Axon’s total coverage network now spans a boatload of 3rd-party sensors from popular consumer offerings that competition can’t match.

Fusus makes pretty much every other software product it offers better. This is how the Vehicle Intelligence Suite (which this is a big piece of) is able to stay organized as plate reading and live-streaming video feeds exponentially grow. This is how Vehicle Intelligence can do things like ID a stolen car 32 minutes faster than before. Fusus helps everywhere.

With this in mind, it’s probably not surprising to hear how rapidly this is growing. Integrated feeds rose 70% Y/Y the last time we got an update last spring, with international markets such as Brazil and Chile boasting 100%+ Y/Y growth. That strong initial international adoption has been a pleasant surprise for them, as they now have another powerful selling tool beyond Taser 10 and BWCs.

Axon 911:

Axon Fusus’s real-time intelligence capabilities make it an instrumental piece of Axon 911, which is why the Fusus founder runs this division for Axon. I love that he’s still around. Up until now, Axon’s end-to-end service relied on clunky, antiquated Computer-Aided Dispatch (CAD) software that shared data with their DEMS and other software products slowly, unreliably and incompletely.  

Right now, when someone dials 911, the operator listens, frantically types needed information into the command system and passes this off to a dispatcher. From there, the dispatcher reads everything and relays the info to police officers with deployment instructions. This is slow and leads to natural human error. Axon wanted to improve the pace and quality of information flowing into Axon Evidence and every other software tool they provide. That wasn't their reality and they’ve worked hard to make it so.

Axon tried to fix this in the past with a product called Axon Dispatch, but ran into immense friction. Their approach was to try and convince police departments and other customers to rip and replace their existing CAD infrastructure in favor of this. The value creation didn’t justify the disruption and Axon struggled to find traction… so they shut it down and contemplated other ways to tackle this opportunity. The ultimate decision settled on purchasing companies called Prepared and Carbyne.

Prepared provides seamless and AI-powered software add-ons embedded deeply into legacy CAD systems. There is no disruption associated with onboarding this product like there was for Axon Dispatch. Customers simply plug the device into their audio feed, download the software and enjoy an immediate boost to productivity with no operational risk. This automates the collection of 911 data and report writing so that the operator on the phone can focus on taking care of the victim. It can auto-translate and autonomously handle 50% of non-critical calls without any manual direction or orchestration, which is proving to be very popular. Employees simply need to verify the work. The data is immediately shared with Axon Evidence and the RTCC, allowing resource deployment to be even sharper and even more expeditious than before. This is Axon shedding reliance on dated partners that shouldn’t be relied on.

Carbyne was the other recent acquisition. This is the infrastructure provider they hope to eventually find traction where Axon Dispatch couldn’t. And I think they’ve got a great chance. With Carbyne, Axon provides an expansion and cloud-native call-tracking ecosystem that can eventually displace legacy CADs. The modern plumbing means all of the data from these calls is usable and closely tied into the rest of Axon’s ecosystem more quickly. This product can also create a secure, iPhone-based link to communicate via text and even activate available iPhone sensors without any app or hardware installation required and with an easy process for sharing this with deployed officers. Pretty cool. There’s a reason why the word “eventually” is in bold above. Carbyne integrates with legacy CAD systems and allows departments to use them in parallel for as long as they’d like to. They can spend a lot of time getting comfortable with Carbyne and gradually shift to modern cloud systems, something that highly regulated police departments should find compelling. 

Carbyne and Prepared together can cut high-priority response times from 7+ minutes to 2 minutes by transforming the slow 911 call process laid out above into the following:

Together, Prepared + Carbyne give Axon a more unified platform and $5B in incremental TAM to pursue. The thought process is that Prepared will be an easy sell. From there, moving some workloads from legacy CAD to Carbyne should be a lot easier thanks to the less disruptive transition. They believe that Carbyne will quickly prove its value and win more workloads as operators see how much better it is than what they have. They think that will ramp to a point where budget renewals routinely include cutting legacy CAD use in favor of this. They’re confident they’ll be able to show customers how much they’ve saved with Carbyne and how much they’re pouring into on-premise CAD solutions while they use them less and less. That should make Carbyne up-sells more frictionless. So, the ultimate goal is the same as Axon Dispatch, but the path to getting there is much more realistic. And vitally, both executive teams have stayed with Axon post-close to bring this vision to life.

“For the first time, intelligence from the 911 call, from cameras, from sensors, from cell phones, all converge in real time, powered by AI at every turn, instantly understood and shared across the Axon ecosystem.” – Axon CEO Rick Smith

Various Products:

For another encouraging sign that Axon’s leadership team is appropriately paranoid and perpetually driven, they’ve been an early mover in AI for public safety. Their objective within this exciting area of innovation has centered on delivering value-rich software add-ons that make their existing products and workflows better. It uses AI to devour insights generated by Axon’s leading sensor and data scale to make customers better & more efficient at their jobs. And thanks to that same leading data scale, it can train (with partners) and season the algorithms and agents embedded in these tools to make sure they’re improving faster than alternatives. So… an early mover and a faster mover in terms of pushing the roadmap forward and making Axon’s tools better. In the competition section, we’ll get into where Axon has formidable foes and even companies ahead of them in some areas. AI is not one of those areas.

Draft One was Axon’s first big product entry into AI. This capitalizes on the rich, complete flow of information Axon receives to actually help officers draft incident reports. It logs and summarizes BWC audio feeds from AB4 and includes any other piece of information relevant to a given case. This can handle data from multiple transcripts and languages too. The product deploys OpenAI’s latest models and sharply augments the value that existing products like Axon Records provides. Not only do Axon customers have a central database and RMS to organize that data, but they also have an AI companion that can create the tedious reports officers need to do for easier compliance. With this, cops can cut report writing time by more than 50% and move from spending 50% of their time on this mundane administrative work to 15% of it. They can now simply review, edit and submit with direct Axon Evidence and Axon Justice integrations. That’s important to note. Because police departments are tightly regulated, Axon’s AI tools are built with human guardrails in place, creating a bit of user friction but ensuring compliance with strict laws. With all of this in mind, it’s no wonder that this has been its fastest-growing software product ever, ramping 2x quicker than any other previous offering and easily eclipsing $100M in bookings in year 1.

Axon wasn’t done in this area. Far from it. Since the Draft One and Brief One launches, it has been briskly shipping more tools to extend its customer value proposition. The company provides Redaction Studio to automatically blur faces or other info to preserve privacy, which helps greatly in automating Axon Evidence and general investigation compliance. This cuts redaction time by 75%. From there, it boasts something called Smart Search or Detection. This is a far more targeted search function within Axon Evidence where an officer can search hours and hours of evidence data for a face or an object or any other meaningful part of a crime scene. That profoundly impacts productivity for Axon Evidence and Justice, Axon 911’s ability to locate plates more quickly and Axon Fusus’s decision-making efficacy.

But wait, there’s more. Policy Chat is sort of the ChatGPT of public safety (and it uses OpenAI models). It’s trained on a police department’s own rules and regulations, allowing officers to conversationally ask questions in the field to stay compliant. This reduces the rate of thrown-out cases from policy violations and mistakes by making sure officers are equipped with needed knowledge more frequently. The amount of information packed into their policy documents is daunting and overwhelming to memorize. Now, they don’t need to. Instead, in real-time, users can provide information on a given situation and ask for advice on how to handle it or if their proposed response is a good decision. The company actually didn’t think this tool would be useful to officers, but they were wrong. Like any good management team, they listened instead of assuming and now have another AI tool enjoying rapid traction because of it.

Last but not least is the Axon Assistant. This is an AI companion built right into their BWCs. This enables general Q&A to run background checks on a person, for example. It also includes a Policy Chat integration and a real-time translator for two-way conversations with persons of interest. That’s huge. Right now, officers have to figure out what language a person is speaking, call for help and wait for a translator. That can take several minutes during rapidly evolving, uber-intense life-or-death situations. And now? They can simply press a button on their BWC to get the conversation rolling in real-time… without needing to use a separate piece of hardware like an iPhone and with this information pushed right to the rest of Axon’s software ecosystem.

“Now, the Axon Body Cam is not just a recording device. It’s a completely connected communication support too.” – Axon CEO Rick Smith

Needed Value:

Remember when we cited drones as an efficiency force multiplier for helping talent-starved police departments do more with less? This is too – and is arguably even more impactful in that regard. The setup is perfect. 95% of a department’s budget is allocated to a labor bucket, while most agencies struggle to maintain 80% capacity and 75% of them think AI is a solution to this problem. This is how officers stop spending half of their time on tedious administrative tasks and can pour their energy into actual fieldwork. This is how that fieldwork comes with fewer issues and legal battles. This is how safety gets better without costs exploding higher.

Companies that can help existing officers deliver a step-change in productivity, I believe, will be rewarded with larger and larger portions of a department’s available budget. And Axon is the company doing this, which makes me think their current $600 subscription ceiling is only the beginning. Both AI and drones will push that maximum higher in the coming years and will support a longer cross-selling growth runway for its core end market, in my opinion. For evidence of tangible present value, Draft One saves an average user 11 hours per week, while Redaction Assistant saves 10 hours per week and Smart Search saves 6 hours. This is creating real value today, which is why real AI revenue contributions are coming today. It's worth noting that this is Axon's internal data and there are skeptics about how accurate it is. I have every reason to trust the company, and the strong AI bookings trends discussed later certainly provide supporting evidence.

“At some of our largest deployments, we're still far, far less than 1% of the police department budget.” – Axon President Josh Isner

AI Go-To-Market Approach:

Axon’s approach to selling AI tools minimizes friction and inspires seamless adoption wherever possible. Public safety sales cycles are very long, which means cross-selling new products can take years. Axon knew they would be rapidly delivering new AI products over the coming years, and didn’t want to be prisoner to these frustratingly slow cycles.

That’s why it packages all AI tooling in its “AI Era Plan.” This promises customers all current and future AI offerings that Axon introduces over the course of their 5-year or 10-year contracts. They pay a fixed fee for all of it, with Axon ensuring all future products are easily onboarded with little disruption. This may sound like Axon giving too good of a deal, which is fair. At the same time, I think that providing the good deal is a strategically sound policy for now. Leading its sector in AI innovation and bundled value is extremely important for Axon’s long-term growth engine and will simply create even stickier and more loyal customers. This frictionless adoption approach does sacrifice some near-term monetization but it also should rapidly entrench Axon’s AI product suite as a standard in its industry, fortifying its already elite retention levels and bolstering its value proposition. Down the road, if customers are using several of its AI products and enjoying the profoundly positive impact on time and money savings they’re already having, future pricing power is inevitable. I think this GTM structure makes a ton of sense, and clearly so do Axon’s customers. During Q2 2025, less than a year into the package’s introduction, this delivered $150M in net new bookings. That’s already about 19% of total for that period and should be around 10% of total for 2025. That’s excellent progress and places Axon firmly near the top of the leaderboard for scaled AI app layer monetization. During the most recent earnings call and leadership interviews, the end-of-2025 pipeline was called “loaded”, while they expect a meaningful ramp throughout 2026.

Why is Axon having so much success? They (and I) think it goes back to their product development philosophy. They’re humble, open-minded and eager to seek customer feedback. They do not assume they know what their clients want. Leadership is constantly listening to who they serve to tell them what problem to solve for them next. They’re not guessing… They're gladly following instructions. Furthermore, as Smith puts it, they’re not giving customers raw AI tools and letting them freely use them however they decide to. Axon is giving them concrete AI use cases that are proven to derive better outcomes for everyone involved. The reality is that customers generally need to be shown how to use these tools. They don’t need freedom in this realm; they need guidance. Considering this is a company that thrives on openly fielding customer feedback, that’s how Axon structures its AI applications.

I think that these last two sections make the true completeness and interoperability of Axon’s full product suite loud and clear. They masterfully tie workflows together to make sure their high-quality standalone products combine to form something that truly nobody in their space can fully match. From their sensor army… to Axon Evidence intake… to unleashing data in Fusus… to triaging and contemplating that data to make better decisions across Axon 911, DFR and more… to deploying AI systems to amplify the value of all these workflows… to logging all of that information for court systems to close cases faster. They are powerfully end-to-end point solution displacers who save customers time, money and life. Looking ahead, consistently excellent traction across all new product categories should merely become more and more obvious. 10% of their 2025 bookings will likely come from these emerging categories, with the bucket rising 3x Y/Y (Q4 earnings this month). Everything they buy is flawlessly integrated with a level of ease that emulates organic growth. Virtually everything they build either works, or they learn and quickly iterate until it does.

If companies like CrowdStrike and Zscaler are proven digital endpoint and network-centric cybersecurity platforms… Axon is the clear and proven platform in the physical security space. This is why their net revenue retention (NRR) remains well over the world-class 120% benchmark at their large scale and again why they’ve never lost a major customer.

Homegrown:

While Axon dominates across many pockets of its field and boasts world-class churn rates, there is still competition everywhere. And as the company fixated on never growing complacent, they closely follow all of them and take every threat seriously.

First, a lot of customers still run on antiquated homegrown software. Axon has a very easy time coming in and displacing apps that haven’t been updated since the iPhone came out. This gives them an ability to create sharp value from seemingly very basic changes and dramatic value with the full power of their suite. 

In terms of displacing existing 3rd-party vendors, they battle against different companies for market share in different areas. Motorola comes the closest, but there’s no company with the richness or totality that Axon’s umbrella encompasses. And that’s very convenient for this company. As we’ll explore, there are firms with products in some places that are about as good as what Axon provides and in some cases cheaper. But? If a company wants a true operating system to handle all needed workloads cohesively rather than in disparate siloes…  if they want the value that stems from that setup… they want Axon. This is again why they love free trials. They’re a great way to concretely show customers how effective their solutions are. And that’s why they’re able to fetch premium pricing for their entrenched offerings and why they’ve shown such a consistent ability to successfully ramp new tools as they come out.

Motorola:

Motorola is the king of radio-based officer communication. Their APX radios are the de facto solution for the industry and usually sit right next to Axon’s body cams on an officer’s uniform. Motorola does have body cams, but Axon is dominant there just like they are with Tasers and the associated sensor-based data. They’re also bigger than Axon is in ALPR scale, which is important for a lot of customers. These titans are fierce rivals and only integrate with each other because customers make them. And just like Axon, Motorola has worked very hard to leverage its subsections of hardware dominance to expand deeply into software and do so much more for its customers. 

Motorola has a competing DEMS called CommandCentral that is its response to Axon Evidence’s scaled ability to ingest ridiculous amounts of disparate and unstructured data to harmoniously use. It has two Dispatch-related offerings for large and small departments, which combine CAD systems that uniquely boast cellular-free connections, an Axon Justice competitor, jail management and administrative record help too. Their Axon 911 competitor is more mature and their hybrid offering (not solely cloud) is comforting for departments hesitant to open their data up to public spaces. That should help Motorola defend market share, although Carbyne letting customers move to the cloud as slowly as they want could quickly change that. And similar to Axon Fusus, Motorola’s RTCC is called CommandCentral Aware. On the AI side, Motorola has translation tools, something that resembles Axon Policy Chat and “Assisted Narrative” that aims to compete with Draft One.

It’s one thing to have all of this, but how do these products compare to Axon? Again, Motorola is well ahead in APX Radios and Axon isn’t trying to change that. Motorola is also ahead on ALPR scale, but, in my opinion, Axon’s Outpost and Lightpost launches close the entire competitive gap and should let Axon quickly gain ground as it undercuts Motorola with a modestly better product and pricing.

Axon Evidence is the gold standard for DEMS with 80% U.S. market share. It comes with far better data sharing workflows with the justice system and a better overall user experience. Motorola’s data sharing processes are based on antiquated and often vulnerable link sharing. Axon Evidence is far more scaled, mature and popular today. That is a commonly held opinion across the industry. Their dispatch offerings are ahead of the newer Axon 911 bundle, but I see every reason to believe Axon will move past them eventually. Their sensor network is simply far broader than what Motorola has already and their willingness to include popular vendors through “Works with Axon” tears down friction that Motorola has created for years. Motorola may have a larger ALPR scale, but they don’t have the compelling 3rd-party partnerships with Ring or other companies that give Axon a massive injection of valuable data. And they don’t have fixed camera systems that are as compelling as what Axon now provides. Oh… and they don’t have a rapidly growing base of drones perfectly integrated into their platform to deploy for more eyes when needed. This is also why I think Axon Fusus will stay ahead of anything Motorola does with CommandCentral Aware. 

The value of an RTCC is based on the completeness of information it enjoys, the simplicity in which that information is presented and the actionable recommendations that stem from it. Axon is the clear leader in all of those things in my opinion. Its ability to map and make sense of everything is better because it includes more assets in the network and more coverage. The same is true for AI. Motorola’s Draft One competitor isn’t really an autonomous report writer. It’s best described as an “editor and fact checker,” which naturally limits its time-saving capabilities.

All in all, Motorola leads in a radio niche Axon doesn’t want and in an ALPR space where Axon is quickly closing the gap with tools that should allow it to create a new lead over time. Their end-to-end platform is more unified, provides more use cases, delivers a more modern interface and creates more value than the pieces of Axon’s platform Motorola can stitch together.

Others:

There are point solutions sporadically splayed throughout Axon’s ecosystem, but none of them come nearly as close as Motorola to matching its full set of capabilities. And I think that is worth highlighting again. It’s very hard for companies selling 1 or 2 of Axon’s products to win market share unless those products are dramatically better. Otherwise, Axon can bundle and drive more operational productivity than these other vendors. In the rare cases where these point solutions are better than anything Axon can provide, it buys them or partners. In all other cases, it just wins the lion’s share of the opportunity. It’s sort of like SentinelOne vs. Microsoft. Does SentinelOne have arguably a better endpoint product? Yes. But Microsoft does 50 things pretty well and SentinelOne does 2-3 things very well. Companies routinely opt into the Microsoft of this equation. 

NICE Investigate is a DEMS competitor and has some solid dispatch offerings. They’re excellent at 911 call trend analysis, which prompted Axon to buy Prepared (which it thinks is the leader). They even have live translation tools that can be embedded in BWCs. They lack the rest of Axon’s software suite utility, its massive 1P/3P sensor network, AI add-ons like Draft One and its budding DFR segment.

Genetec also has a solid DEMS offering and Genetec also has an RTCC. I just go back to how massive of an edge data scale and organization both are for a product like this and Genetec doesn’t touch Axon in that arena.

Veritone is probably the most capable AI competitor with an established DEMS offering. They’re also conveniently vendor agnostic, as their DEMS can integrate data from Axon, Motorola and others. They boast redaction assistants, facial recognition, great trackers and less vendor lock. If a company (for some reason) wants to use 8 different vendors to do everything instead of Axon, this is a better option. They’re more open-source and lace the “Works with Axon” throughout their entire suite rather than in bits and pieces. The world is moving towards full platforms, and that’s advantage Axon. 

Carbyne will need to replace several established legacy CADs like Tyler Technologies and Hexagon Safety, which have both been around for a long time. These CADs are also all trying to rapidly infuse software add-ons throughout their offerings but they all have a long way to go. Candidly, I don’t think they’re capable of modernizing and catching up as AI and more modern software in general take over this industry. Axon just moves faster and does more.

There is a plethora of drone competitors in the market, but most other big players focus on other end markets and use cases. Skydio is the big boy for U.S. wide-area search and rescue, and doesn’t need to compete with Chinese leaders in this market following the U.S. ban on their new drones. Parrot is a capable competitor here and is better than Skydio for optics and imaging quality. On the other hand, Skydio is leaps and bounds ahead in terms of autonomy, and that gap is larger than the one Parrot enjoys over Skydio’s cameras. Brinc is another competitor to focus on. They’re much larger than Sky-Hero for indoor drones thanks to an early partnership with Motorola. Their drones also serve as indoor/outdoor hybrids, but Skydio is considered superior for its core use cases. Teal Drones and Anduril are out there too but, for the most part, they’re going after different use cases and end markets.

Axon is the Taser king, but there are a few tiny alternatives with different product designs. Wrap makes a net gun that can capture a suspect without a lot of pain. They also debuted a BWC recently, but don't really have any scale or any kind of platform breadth to match Axon. There have been others in the past, but they've failed due to poor market fit or from being found guilty of stealing Axon's IP.

Finally, while Axon partners with OpenAI, they’re often asked about how they can compete with pure-play AI companies that can build agents. Good question. As I’ve argued in recent enterprise software landscape articles, those who overcome AI risks and use this technology to drive a net tailwind will be those with the most use-case breadth, experience and data. The more weird fringe use cases, the more cross-product collaboration and the more proprietary data a company has, the less likely it is for Anthropic to competitively match it. Axon has a fortress of sensitive, highly regulated data at its disposal and delivers product diversity in spades. I think they’re as well positioned to use AI as an accelerant as arguably any other software company I look at. The sharp AI Era Plan ramp is enticing evidence with (I think) much more to come.

While Motorola is the closest, Axon’s DNA is outworking and beating every single competitor. They make it a core mission, a perpetually driving source of motivation… and they take it personally. The decades of trust they’ve painstakingly built with loyal customers gives them the ability to do more for their customers. And? That’s why everything they introduce does so well and why their competitive landscape is so compelling. They don’t have to win against Amazon, Apple, Meta, Microsoft or Google. They’re the highly capable, rapidly moving industry giant.

“We're a very, very competitive group of people. We like to compete, we like to win.” – Axon President Josh Isner

Axon is in the early innings of tapping into its core state and local niche’s runway. The cross-selling opportunity remains massive, with most customers on subscription plans ranging from $50-$100/month and rapid progress towards moving them to more expensive tiers. The police budget CAGR continues to compound at a solid 5% clip (for 40 years), and Axon continues to build on that growth through budget unlocks, platform-wide adoption and consistent market share gains.

And while all of this is true, it has its sights set on many more markets. They are fixated on building a juggernaut and that requires new growth outlets. They know their products can drive real value well outside of U.S. state and local law enforcement use cases, and they’ve been hard at work on laying the foundation for these opportunities to eventually become their largest revenue generators.

Global expansion is the most obvious source of runway elongation for Axon. While the work and regulatory landscape is different in the USA than everywhere else, these tools can be easily retrofitted to comply and drive value everywhere. While this used to be a risky opinion, the progress enjoyed over the last couple of years has made it much more certain. That has a lot to do with hiring CRO Cameron Brooks. Brooks came to Axon with a deep understanding of the cloud environment and regulatory backdrop in Europe, which was a needed skillset. He built out Axon’s EMEA team, leaned into channel partnerships and borrowed the best of Axon’s USA go-to-market. The result has been an encouraging cloud “unlock” of a few EU member nations. It’s no secret. Europe’s cloud migration journey has taken longer than it has in the USA. Axon thought it could get more proactively involved in driving adoption of Axon Cloud Services (ACS) and get customers more comfortable with change. It’s working. Axon just got their very first major EU department to go “all in” on ACS and its unified software bundle with a 9-figure deal. They see this as a profoundly positive first domino to fall, with many now expected to follow suit as they see Axon work its efficiency magic. Generally speaking, bookings for the segment are growing at a triple-digit clip and generated 2 of their 10 largest deals last quarter (including the big EU cloud deal).

Axon thinks this segment of its business will be even larger than the U.S. state and local business in the coming years. The deals available to them are simply much larger and more consolidated than in the USA. For context, Italy’s largest police department is 5x larger than the NYPD and 10x larger than the 2nd-largest U.S. department. 2 wins in Italy represented 33% of the total officers in the USA. They’re massive, which means deals can be massive. The 9-figure contract cited above was not even with one of the larger nations there. They can turn this into a giant business much more quickly than the piecemeal process that played out in the USA.

Momentum is expected to stay rapid, as immigration issues and antiquated technology inspire customers to seek better tools.

While international is the most obvious growth segment, enterprise is arguably the most exciting. They’ve been working on product market fit for a few years and inked their first large contract with UPS in Q4 2024. The deal included integrating 300,000 disparate video streams from their infrastructure, with “Works with Axon” ensuring seamless integrations and no rip-and-replace. Axon is seamlessly feeding all of this data into UPS’s Security Operations Center (SOC), providing them with the same single-pane view that its police customers receive. Fusus was the secret weapon for that. It was a truly massive deal, with many more potentially in the works. For context, there are 900,000 cops in the USA and Walmart alone employs more than 2M people. The deals here could be their largest overall, and, for evidence, this UPS deal was their largest ever despite being the very first major enterprise contract they’ve signed. Another large win came during Q2 2025, with the pipeline swelling and adoption finally building.

This is a good start, but there are reasons to think the next inning will be even better. Its very first enterprise-focused product was called the Axon Body Workforce (ABW) Cam. It was clunky, heavy and seemingly better suited for policing use cases. For those reasons, adoption was limited, just like for all of its other hardware at the beginning. It was in the learning phase, knowing an open mind and a capable team would get them where they needed to go.

This led to the ABW Mini – built in tandem with key retail partners. It will soon be broadly available, with a much lighter and smaller design, a panic button, AI translation tools and situational awareness support similar to Policy Chat. It boasts tighter integrations with Axon Evidence, Axon Fusus and Axon 911 and plugs into Axon’s Retail Crime Hub (Its RTCC for retailers). This Hub is offered in tandem with a New Zealand-based partner called Auror that uplevels Axon’s suspect recognition technologies. With Auror’s capabilities now part of Fusus, Axon can further automate private sector data sharing. It can also ID a suspect in a ski mask based on their shoe color or watch if they’ve ever been in another location in Auror’s network. This capability has landed them CVS, Walmart, Kroger, Ulta and several other impressive customers, and I would not be surprised to see Axon fully purchase this impressive company down the road (if they can).

In early testing with Auror, this is already materially eliminating violence towards employees, which was a key objective. It’s also saving 3-5 hours per incident or the equivalent of 5% of New Zealand’s policing budget in labor. Cops no longer need to go to a store, talk to witnesses, collect video footage, sift through all of it and file paperwork. Auror, the Retail Crime Hub and key add-ons like Smart Search and Draft One sharply expedite all of that work.

With $100B/year (and rising) in theft-related losses for the sector, there's a massive opportunity for Axon to fix these issues and fetch considerable revenue while still helping make their customers a lot more money. I love win-wins.

A lot of the theft comes from inadequate access to evidence, which is exactly where Axon will help. Simply put, this now should be an easy sell… and not just for retail. Lee Health used Axon’s products in their workplace to cut violence by 99% and save $37,000/year in forgone medical claims. In Scotland, tighter retail and justice system communication helped their Dundee court system cut their court backlog by 80%.

In terms of competition, Axis Communications (a Works with Axon partner) offers a capable alternative. Motorola’s hardware here is 60% heavier than Axon’s and offers limited functionality. It’s similar to Axon’s first enterprise cam iteration that was too clunky and not purpose-built for enterprise customers. I think the most interesting competitor is Utility Associates, which actually builds the cameras into the fabric of an employee uniform. That’s a capability Axon should look to copy or acquire. For all of these alternatives, software and AI add-ons are limited, the connection to justice systems is less tight and the ecosystem is less complete.

Federal is not as exciting as International or Enterprise and doesn’t command nearly as much of their focus. Government agencies, especially on the Department of Defense (DoD) side, command more custom and complex builds than their other end markets, and it’s just not worth it to spend those resources to build what they want a lot of the time. That’s a big reason why most of their traction to date has been with civilian agencies.

While Federal isn’t the most exciting topic for Axon, it has given us a great datapoint for gauging the mission-critical nature of Axon’s platform since last April. The Department of Government Efficiency (DOGE) cuts had zero material impact on them with one contract in total not being renewed. They were not the source of cuts because their products are not a source of bloat and not discretionary. Despite this resilience, they baked a lot of prudence into 2025 guidance for this segment because of DOGE and the government shutdown. They didn’t expect a lot and have been pleasantly surprised with growth that outperformed pessimistic expectations through 2025. 

There’s still a lot of work to do and more reason to spend that needed time on other segments, but they think a FedRAMP authorization for Fusus (they have it for their other software) could be a key accelerant and also think the stimulus and added budget from “One Big Beautiful Bill” will help in 2026 too.

Finally, their Corrections segment is small but increasingly mighty. Their Tasers can help reduce excessive force rates by 70% and their evidence logging processes are highly coveted in these environments. That’s why bookings are growing at a 100%+ Y/Y clip and why this was the source of 2 of its 10 largest deals during Q2 2025.

The biggest risk here is valuation. It's not crazy expensive but it is still a tad pricy. It now trades for 47x EBITDA with an EBITDA growth multiple (using next 2 years of growth estimates) of a little over 1.6x. It also trades for 66x forward FCF, although growth for that metric is expected to be 85% over the next two years, so that growth multiple is under 1x. As we'll see in the financials section, FCF is very noisy for Axon and EBITDA is a better metric.

Regardless… any way you slice it… it’s still modestly expensive. I think the last 25 pages explain why I think it deserves the premium. As I’ll talk through at the end of this piece, I have decided to add this to my portfolio, but I’ve kept the initial bite intentionally small. I’m selfishly rooting for more near-term multiple contraction so I can keep adding to my existing stake. I think it’s prudent to accept the risk of not owning quite as much as I want to if it just zooms higher from here. I think it’s equally prudent to choose to guard against the risk of adding too quickly and aggressively. 

Dilution is going to continue. Share count is likely going to grow by a 3% compounded annual clip over the coming years. They are not focused on buybacks and remain firmly in growth mode to capture the opportunities that lie before them. I support this approach. They should not be using finite capital to buy back shares today. Their scaled growth is fantastic and they should be leaning into it in ways that align employee incentives with more Axon revenue and profit. This is not changing and it is something bulls must get comfortable with, as it will mean profit growth modestly trails profit per share growth going forward. In the chart below, 2022 and 2023 show what years without large executive stock awards look like. At the same time, SBC being higher means Smith is delivering on revenue and market cap goals that get shareholders paid. That means the higher the stock profits, the more SBC there will be.

There’s rising focus on the amount of M&A Axon has conducted over the last couple of years to expand into Axon 911 and drones. Some think this is a distraction from the core or even a sign that organic growth is slowing more than desired. This skepticism from bears actually surprised me. How many examples of Axon flawlessly integrating M&A into a suite in a way that greatly accelerates that product's adoption do they need? How many times do they need to see Axon make 1+1=3 and generate fantastic returns on these purchases? I’ve personally seen enough, and I sleep extremely well assuming this team can effectively integrate new pieces. The only blunders they’ve ever had entailed their Vievu BWC purchase taking longer than expected and their My911 purchase not scaling as expected. But? That team is still with Axon and delivering considerable value within the newly formed Axon 911 product, so I wouldn't even call that a mistake.

Substantiation of their strong M&A capabilities comes with the uniformly positive scaling of virtually every single new product. It’s not like their track record in this department is anything but elite. On the organic growth front, I also reject that concern. Axon is not buying revenue. They’re not buying large businesses with loads of tech debt and bloat that slow them down. They’re buying great products very early on in their growth curves and making sure those growth curves become far more lucrative. This is a common source of concern voiced by skeptics and is a source I don’t put much weight into at all.

“Across the board, our acquisitions over the past year have outperformed initial bookings expectations, which are showing up in adoption of these new products.” – Axon President Joshua Isner

This one is a bit sensitive. Axon does have products that can be packaged in ways that create a surveillance state. American culture will surely not tolerate that and would rapidly push back. They need to thread the needle of driving better safety without being seen as an enabler of a police state. And I think they do a great job here. They deploy an opt-in model for data sharing and will never make an exception. From there, they obsessively and proactively redact identities and other information to stay compliant. The company also has an Ethics and Equity Advisory Council that serves as a needed check and balance to ensure the focus isn’t only on monetization, but responsible monetization. They do as well as they can here, but there is still headline risk that stems from being a policing vendor.

Other risks:

If the first 5 sections are the cause, section 6 is the effect.

Axon has been delivering stable-to-accelerating growth for a decade despite rapid scaling. As you can see below, this is a 25%+ Y/Y revenue growth machine and has been for a long time.

Revenue is a solid metric to track for Axon, but it’s not the best datapoint. It’s influenced by the timing of lumpy hardware sales and fluctuating procurement cycles. Considering this, Axon loves to tell people to focus on Cloud annual recurring revenue (ARR) and future contracted bookings. Cloud ARR offers a view of expected software revenue over the next 12 months; future contracted bookings is similar to a remaining performance obligations (RPO), except it also includes contracts with legally mandated termination clauses that prevent Axon from counting it as RPO. Bookings growth tends to lead ARR growth by a quarter, offering a great hint of what to expect for that metric looking ahead 90 days. The one noisy thing about bookings is that it doesn't account for longer contract duration. Growing trust and confidence in Axon’s suite is leading to 5-year contract demand shifting to 10-year. If a contract is signed for a 10-year period vs. an identical deal signed for 5 years, that’s doubling the bookings. Thankfully, this isn’t the source of the resiliently strong bookings expansion for Axon. They also disclosed bookings normalized for duration last quarter, with growth there meaningfully accelerating Y/Y.

Next, net revenue retention (NRR) offers a view of existing customer demand growth from a 12-month period. It excludes all hardware revenue, offering a great view of demand growth from the existing user base. So many software firms struggled to maintain 120% or even 110% NRR following the pandemic bubble bursting. Axon’s ability to comfortably remain at 120%+ is a testament to the need for its platform, its cross-selling success and its customer loyalty.

“We see the healthiest and best indicators across our business we have ever had, and I'm confident we are on the right track as we scale up the investment in our products and sales team to go after the entirety of what's in front of us.” – Axon President Josh Isner Q3 2025

Finally, Axon made a series of accounting changes during Q1 2025 that moved various products into different groupings. It moved from Axon Cloud, Sensors and Taser buckets to Software & Services and Connected Devices. Pre-change buckets had revenue from both newly formed segments that were moved without precisely walking us through what went where. We can guess on reconciliation and get pretty close, but we can’t be exact. Data for those specific revenue buckets before 2024 should be taken as close approximations.

What Metrics to Focus On?

There are a lot of weird things to unpack for Axon margins as they like to do things their own way. This might be one of the rare companies where adjusted EBITDA is actually the best metric to use. That’s not because it’s perfect. It excludes very real stock-based compensation, which is a large expense for Axon. At the same time, stock comp needs more context too in this case. When their stock price rises and equity packages vest or vesting becomes more probable, Axon is on the hook for paying the taxes associated with that compensation. That counts against net income and FCF, meaning its profitability actually suffers when its stock price succeeds. While Axon adds non-cash stock comp back to its non-GAAP income, the cash comp expense is left in there. That makes non-GAAP net income partially tied to share price movement and a bit noisy. Relatedly, timing of lumpy up-front hardware cash collections and changes in inventory make FCF less useful than it is for a pure software firm.

Ok so FCF, GAAP net income and net income are all a bit wonky in this case, so what about EBIT? That would work well here, but they don’t make non-GAAP adjustments like pretty much all of their peers do, which makes an apples-to-apples comparison more difficult. While we could manually back that expense out of EBIT each quarter to get a better number, it's very hard to do forward modeling for SBC (because it is tied to share price).

So... we're left with EBITDA. Sorry, Charlie Munger. There’s absolutely nothing shady or aggressive about the way Axon calculates any of its profit numbers. There are just several items totally unrelated to its core business that make common metrics spike and tank on a regular basis for irrelevant reasons.

Margin Headwinds & Tailwinds:

On gross margin, the most recent quarter was the first full tariff period for the company. That is why connected devices GPM is currently falling Y/Y, which is expected to continue for a little while. Software and services GPM is expanding and should keep doing so, but that’s not currently enough to prevent the Y/Y GPM contraction seen below. Once the one-time tariff impact is lapped, that headwind will go away and GPM expansion, driven by structural mix-shift, should resume.

Moving down the income statement, Axon is not focused on margin expansion right now. They’re dedicated to aggressively making the investments needed to support their right to win the majority of a truly massive opportunity. These investments mainly show up in R&D and sales, with G&A still a source of operating leverage through this all. They are firmly in rapid growth mode, and are able to allocate funds in a way that enables this expansion while still delivering stable and strong 25% EBITDA margins. This is probably where margins will be for the next little while, with significant room for expansion whenever the rapid influx of new growth opportunities finally starts to slow down. Hopefully that takes a while; the uniformly great traction they’re enjoying across the board probably means it will.

Axon has beaten revenue and EBITDA consensus estimates in each of the last 25 quarters. That EBITDA streak remained in place despite tariff headwinds that weren’t baked into its original 2025 outlook. They’re a machine and they know what analysts want: Outperforming results and targets that drive upward profit revisions and justify a more premium multiple.

“Yeah, I mean, we have a lot of confidence. I think, as Josh said, we are just seeing more momentum than ever in our bookings. We talked about it a little bit. We're expecting a really big bookings quarter in Q4” – Axon CFO Brittany Bagley a few weeks ago 

For long-term guidance, Isner gave us a hint during the Q3 call that I personally found exciting. He said “over the next 3-5 years, I think you’ll see the same type of growth you’re seeing now.” That's a quote. Why is that so encouraging? Axon will likely do around 32% Y/Y revenue growth for 2025. Analysts expect its CAGR to slow to 25% for the next three years. If they maintain their 2025 growth rate, Axon’s 2028 revenue would be 16% higher than current consensus (and they LOVE to sandbag). This is part of the reason why Axon is expensive and fetches a premium multiple. There's a high degree of confidence in this company surpassing consensus assumption.

Axon is no longer crazy expensive but it is still a tad pricy. It now trades for 47x forward EBITDA with the EBITDA growth multiple (using next 2 years of growth estimates) a little over 1.6x. The average for its high-growth peers is close to 2x. It also trades for 66x forward FCF. Growth for that very noisy Axon metric is expected to be 85% over the next two years, so that growth multiple is under 1x.

The table below represents the range of scenarios I see for Axon over the next few years. It’s intentionally wide because who knows what 2028 will look like. It's intentionally simple for the same reason. I'm firmly in the Buffett camp of thinking forward modeling should not take longer than 15 minutes. I also think it’s appropriate to bake at least a little more multiple contraction into even the most optimistic scenarios. In terms of what I view as most likely outcome, it’s somewhere between “Average” and “Bull.”

I am deeply impressed with this company. I think they can compound revenue at a high-20% clip for at least the next few years while generating at least a little operating leverage. I think their customer stickiness is elite and see the competitive landscape as highly favorable. Every single new product they launch works to incrementally extend their already massive runway and distance their interoperable platform from the pack. And? If the last few decades are any indication, they're more than capable of taking advantage.

The valuation is still a little stretched, but I like this enough to own a small piece (2.2% of holdings). As of right now, I plan to keep accumulating more shares if the valuation continues to decline. Long Axon... hopefully for a long time.