Axon Enterprise, Inc. (AXON) Earnings Call Transcript & Summary

Hello, everyone, and thank you for joining Axon's executive team today for our second quarter 2026 earnings conference call. Before we get started, I'll note that our remarks today are intended to build upon our most recent shareholder letter and investor materials, which you can find on our investor website at investor.axon.com. During this call, we will discuss our business outlook and make forward-looking statements. These comments are based on our expectations as of today and are not guarantees of future performance. All forward-looking statements are subject to risks and uncertainties that could cause actual results to differ materially as discussed in our SEC filings. We will also discuss certain non-GAAP financial measures. Descriptions and reconciliations of GAAP -- to GAAP are included in our shareholder letter and available on our investor website. Now as always, before we kick it over to Rick, we have a quick video to get us started. Let's pull it up. [Presentation]

All right. Thank you, Erik, and great job to the team. I love those videos. They just get me pumped up. So lucky to get to work here. I'd like to thank all of our analysts and investors who are joining us today. As you can see, we had another record quarter. Again, I was just looking at the numbers, like 39%, wow. Okay. So we had a record quarter, and I could not be more proud of this team. And none of this, none of it happens without our customers. You guys are the foundation, and we can't do any of this without you. This morning, my cyber truck drove me to work. I sat behind the wheel and never touched it. A few years ago, that was a keynote fantasy. Today, it's my commute. And I've stopped noticing. That's how this goes, right? The impossible becomes remarkable, and then it becomes Tuesday. Two days ago, OpenAI announced an unreleased model had solved 10 open problems in mathematics, problems that the best mathematicians alive have been stuck on for decades, one of them unsolved for at least 27 years. And the compute cost to do all this to do 10 of these unsolved like PhD career-level problems was $2,000 or $200 per problem. So for many years, the critique was that AI, sure, it could remix human work, but it can't create anything new. Well, that critique is finished. A Fields medalist, which is math's highest prize, it's been called the Nobel of Math, said he would have published a comparable result without hesitation. And there are many that said any of these problems might on their own have qualified a human had they solved it for the fields metal. And yet within a day, a model from another lab had reproduced half of the list, meaning it could solve half those problems on its own. So this isn't about one company getting lucky. The whole field just crossed the line at once. The cost of an original breakthrough has just collapsed. So why does this belong on an Axon earnings call? Well, because an idea on a page is not an outcome on the street. Every one of these models perceives the world largely through strokes on a keyboard. When Axon deploys AI, we perceive the world through the largest connected network of sensors and public safety, millions of eyes and ears already in the field, already running and built for this moment. The intelligence doesn't replace the person, the human making a decision or making the call, but we hand them a clearer picture of what's happening in the seconds that matter most. And then it does the one thing software alone can never do. It acts. A TASER device can stop a violent incident without taking a life. Drones and robotics let us send a machine where a person would otherwise have to go. They both do the same thing. They lower the risk for everyone, the officer, the bystander and the person that the police were sent to stop. Every person is safer. See it, understand it, act on it. This is the chain of events. It's the same chain that drove me to work this morning, right? Software alone can't close it. Hardware alone can't. But in public safety, Axon holds every link in this chain. Now when we changed our name from TASER to Axon, we were very specific about that. And Axon is the nerve fiber that carries the signal from the sensors to the brain and the decision from the brain back out to the body so it can act. We didn't pick a name that just sounded good. We picked a description of what we were building, and we've been busy building it. At the edge, sensors can see what's happening, whether it's body cameras, drones, fixed and in-car cameras, license plate readers, 911 AI assistance. Behind them is the largest customer-controlled data repository in public safety. Then AI and real-time operations can turn these signals into understanding and then the means to act. TASER devices, drones, communication, training. At the heart of it, of course, is our mission to protect life. This doesn't just add up, it compounds. Every sensor adds a signal. Every workflow adds context. Every outcome improves the next one. Brookhaven, Georgia shows you exactly what this looks like when you connect it end to end. Drones on the scene in 53 seconds, coverage across 96% of their city. Burglaries down 45%. Look, no single product on the list produces these numbers. The system does. Data shouldn't sit trapped in a silo waiting for someone to find it. It should move securely through the workflows where it does something, a city's cameras to its 911 center. It's -- to its officers to real-time intelligence. It's evidence to courtroom, a frontline worker in danger to help that arrives in seconds. And all of this depends on trust. We've earned that over years of stewarding our customers' data and never treating it as our own. We'll keep deploying this technology responsibly and thoughtfully alongside our customers and their communities with people at the center of every critical decision. In public safety, power without trust isn't an opportunity, it's a liability. I see this coming together everywhere I look. A few weeks ago, a customer bought body cameras for municipal employees, not for the police, for building inspectors. A few weeks before that, some of our largest enterprise pilots began putting Axon Body Mini on frontline workers. And during the World Cup, our ecosystem supported agencies across North America through one of the most complex public safety missions in the world. So where we sit today, the world is being flooded with intelligence. What's scarce is intelligence connected to the real world, trusted by institutions that depend on it and aimed at problems that matter. That's what we've spent 30 years building. There is no company better positioned to benefit from the intelligence explosion than Axon. And through us, our customers and their communities for whom we operationalize this intelligence with physical hardware and thoughtful design and controls to help maximize benefit with appropriate oversight. What an exciting time to be alive and an even better time to be at Axon. And with that, I'd like to turn it over to Josh Isner.

Thanks a lot, Rick, and good afternoon, everybody. As Rick said, we can't do any of this without our customers. Every day, our focus is on delighting them because when we do that well, everything else follows. We also can't do any of this without our team. They have embraced our mission and are working hard every day to deliver outsized societal outcomes. And there have been many of those in the last few months, including on the biggest stage in sports. I am so proud of the Dedrone team for their work to support the World Cup. This was one of the most complex and comprehensive installments we've ever worked on. The expectations were clear, simultaneous delivery and activation across 11 U.S. host cities, more than 50 additional sites and surrounding areas, international deployments, multiple agencies with very little margin for error. The team executed with incredible precision and I believe this deployment further established Axon as the technology leader in public safety. Dedrone is delivering tremendous growth and the opportunity continues to expand across state and local public safety, international, federal, enterprise and corrections. It is also one of the areas where we're investing in most aggressively, and the team is earning that investment. This is a business that was still in its infancy just a year ago and is quickly becoming one of our largest product lines surpassing $100 million in quarterly revenue. Next, I want to give a shout out to our international team. International bookings came in at roughly 3x the prior year in Q2, driven by several large deals. What's especially exciting is some of our international customers are adopting our newest offerings right alongside our core products. In fact, they may be even moving faster than some of our U.S. customers, 3 of our top 5 AI Era Plan deals in the quarter came from international customers. We're also seeing international markets fuel expansion in TASER. In Q2, we won a TASER 10 deal in the Middle East comparable in scale to our largest current U.S. state and local deployment and another in Europe comparable to a top 5 U.S. agency deployment. These are massive opportunities showing up more and more, and we're just starting to scratch the surface. State and local public safety and corrections continues to show strength as well. We signed not 1, but 2 9-figure agreements with major cities in the quarter. One of those was our largest ever individual TASER order. We also signed 2 8-figure agreements with major state corrections customers and another groundbreaker, our first full scope Axon 911 customer. This customer will deploy a combined Prepared and Carbyne solution to power the call to closure vision we laid out when we announced these acquisitions, committing to that future just months before both deals closed. And lastly, in the U.S., we continue to see more and more customers who bet on other vendors coming to Axon for body camera pain relief. We expect body camera shipments to be up 20% to 30% sequentially in Q3 and 15% to 20% quarter-over-quarter. The success across markets led to our TASER volume being the highest in a quarter to date. We've now booked more TASER 10 units than we booked over the lifetime of TASER 7 with a ton of pipeline still in front of us. That may sound crazy for a product category that has been around for 3 decades. Well, it's real, and it speaks to how much opportunity remains. Next, enterprise. We've been talking about this moment for a long time. The point when those early conversations and pilot programs begin converting into meaningful enterprise deployments. That moment is here. Enterprise bookings were also up about 3x from a year ago, and the volume of opportunities continues to grow. And last but not least, I'm excited to see our federal business showing a lot of promise. We are engaged in opportunities across the federal civilian space, and the team is executing with intensity as we approach the end of the fiscal year. Federal was part of the story with Dedrone in the World Cup, and we also closed a major deployment of Axon Assistant with a federal agency on our body cameras because they are using them as body cameras. And just yesterday, we were named one of the participants in the $1.5 billion DHS program for counter-UAS. I am looking forward to talking more and more about our progress in the federal space. As far as bookings go, every indicator I'm watching looks as good as it ever has. Q2 gross bookings were up 20%. That's on top of nearly 50% bookings growth in Q2 last year. On a 5-year normalized basis, which adjusts for contract duration, bookings were up even more, over 30%. That matters because some of our newer markets like international and enterprise signed deals in the 1- to 5-year range compared to some of the 10-year agreements we sometimes sign with state and local customers. Normalizing for duration gives us a clearer view of the underlying demand across markets. And as I review the pipeline, I see a clear line of sight to finishing the year with 30 -- in the 30% range of normalized bookings. And before I hand it over, I want to spend a minute on how we're operating internally because this kind of execution doesn't happen by accident. A little over a year ago, we began flattening the organization to empower people closest to the work, reducing layers of friction and moving faster. We paired that organizational change with disciplined investment in AI tools for our teams. We didn't rush to deploy every tool or ask people to use AI for the sake of using AI. We rolled it out deliberately, and we focused on where it can make our people more effective. That has evolved in the regular hackathons and recognition for teams finding new ways to move faster and spend more time on the work that matters. We're seeing that operation model -- our operating model pay off in execution. I believe that discipline is a big part of why we've continued to raise our own expectations while navigating supply chain disruptions, higher component costs and tariffs, all while scaling to meet demand and never missing a beat for our customers. It is coming together nicely. The team is earning. The mission is fueling them. Our customers can see it, and we are investing in Axon, and they are investing in Axon as their trusted technology partner. We are a next play organization, and we are stacking next plays like never before. With that, I'll turn it over to you, Brittany.

Thanks, Josh. We continue to see the power of the Axon Ecosystem. We're not just one product, but a system that all works together, getting better all the time to solve real problems for our customers. From before a call comes in to during an incident to closing a case, the sensor, software and data work together. This is what you're seeing in our results. Revenue was $904 million, up 35% year-over-year, marking our tenth consecutive quarter of growth above 30%, and the demand remains broad-based. Software & Services revenue increased 36% year-over-year to $398 million. Axon Evidence remains the backbone of our software business and the secure data foundation powering the vision Rick described. Increasingly, our newer software offerings are also meaningful contributors. More than 1/3 of our software revenue now comes from offerings beyond our core evidence platform, consisting of real-time operations, which includes Fusus, productivity applications, including records, the AI Era Plan, counter-drone software and Axon 911. Together, that part of our software business grew roughly 70% year-over-year with the AI Era Plan as a true standout, growing almost 700%. And these businesses are still in early innings. That breadth is supporting strong expansion within our customer base. Net revenue retention reached 126% and annual recurring revenue increased 39% to $1.6 billion. Our software fed by sensors and enhanced by data is a core pillar for our customers' workflows. We're enhancing these with AI and the numbers continue to show that this is sticky and growing. On the devices side, Connected Devices revenue grew 35% to $507 million, primarily driven by Dedrone, counter-drone, TASER 10 and Axon Body 4. Platform Solutions revenue grew 123% to $150 million with Dedrone driving much of that performance. Counter-drone was a new category for Axon added at the end of 2024, and it has quickly become a consistent contributor to our results, validating our acquisition strategy alongside our organic product development. We continue to see strong demand across customer markets and expect the business to continue to scale. TASER 10 remained a standout in the quarter, as Josh highlighted. Turning to margins. Adjusted gross margin was 62.9%, up 130 basis points sequentially. This increase in gross margin was primarily driven by tariff refunds, partially offset by increased mix from newer product offerings, which are still scaling. Adjusted EBITDA was $242 million or a 26.8% margin. Operating cash flow improved to $20 million from an outflow of $92 million in the prior year. This resulted in a free cash flow outflow of $1 million and reflected continued efforts around free cash flow conversion while still making substantial inventory investments to support customer demand and reduce supply chain risk as we scale. We continue to have line of sight to strong free cash flow generation consistent with our Q1 guidance of $450 million for the full year with strong seasonality expected in the fourth quarter. As we've noted before, we also expect long-term free cash flow conversion to improve from these levels, but need to get through this period of inventory investment, which we expect to moderate after this year. Future contracted bookings grew more than 40% year-over-year to $15.1 billion, reflecting the broad-based momentum we're seeing across products and end markets. As Josh mentioned, we continue to feel very good about the momentum we're seeing in the second half of the year. We're raising our full year revenue guidance to a range of 32% to 34%, 200 basis points above our prior range of 30% to 32%. Based on the strong demand pipeline and results in the first half, we expect seasonality similar to last year, where Q4 is our strongest quarter, both in terms of revenue and year-over-year growth. We continue to expect full year adjusted EBITDA margin of approximately 25.5%. We received a refund in Q2 for tariffs, but we are also facing increasing component costs for the year, especially on memory, which leaves our full year guidance unchanged. We expect Q3 adjusted EBITDA margin to reflect the impact of those memory costs with no benefit from tariff refunds before we scale margins back in Q4 to these levels to hit our full year target. We are also thrilled with the performance of our counter-drone segment. But as we scale this and other new businesses rapidly, we are absorbing the mix impact. Over time, these are well worth the investment on the bottom line as well as the top. While there are a lot of moving pieces in our business, we continue to deliver for our customers, our shareholders and our employees and are very proud of our results this quarter. As you think about a Rule of 45 (sic) [ Rule of 40 ], we delivered above 60 and are looking forward to continuing to drive a strong second half of the year. With that, I'll turn it over for questions.

Thanks, team. I think we're all up in gallery view. We have our -- one note today. We're going to try to take one question. And then if we have time, we'll come back to everyone around the circle. So we're going to start with Meta Marshall at Morgan Stanley.

Congrats on the quarter. I guess, Rick, I wanted to ask, just as customers gain comfort with AI, where are your customers asking you to innovate around? And are there any areas where you're still finding any resistance from customers?

Sure. Change is always hard. And we -- this is a very risk-averse customer base when it comes to tech. So I would say, actually, some of the new areas we're being asked a lot about are around privacy, data security. I'm sure you all have seen the de-Flock movement, which is -- has some broader anti-surveillance sort of momentum. Now for me, this feels a bit familiar, like when TASERs first came out, there was a massive backlash. I view this as sort of the birthing process almost of a new -- any new real technology in public safety is going to go through an adjustment period. And I would say the critics aren't all wrong. There have been some mistakes, things that could have been handled better in terms of data privacy and security and a couple of high-profile uses where police, employees were abusing their access to license plate reader data to do things like track an ex. But those are exactly the sorts of things where I think our approach to data privacy, where we have a critic-based system where we bring in critical thinkers that are going to challenge us early. I think we're about to post a new episode of the Boldly Go podcast with a couple of folks off of our EEAC, our Equity and Ethics Advisory Council (sic) [ Ethics and Equity Advisory Council ]. And I think that's become a real superpower. Like our customers, when we talk about trust, it's not just hand-wavy. It's -- we've hit a point where our trust with our customers is an incredibly valuable and fragile asset. If we break that trust, it would be very damaging to the overall business. And if we protect and cherish and continue to earn that trust, it makes our business more durable. So I would say we're seeing a significant number of agencies want to come over to Axon and convert from existing license plate readers to us, specifically because of the controls and privacy work that we've done around the space. And for example, one of the things we're doing is we'll be building AI tools to watch for anomalous search behavior so that we're watching out for potentially abusive patterns and people have access to the data. So I would say, in general, there are still -- there are places where there's concerns about AI, and we just have to be thoughtful and understand those concerns and say, look, this is a tool to amplify human ability, not to replace human judgment.

And then maybe just to just add, I think you asked about where the demand is. I think there's sort of 3 broad themes that are always present. Number one, and this is anything that helps them get response faster, effective correct response faster and many of the things you see us talking about anchor on that. Number two, is anything that helps save officer time to get them more time back to do their jobs in community. And number three, is anything that drives officer and community safety directly. And so again, across all of our lines of business, across all of our categories and all the investments you see us do in AI, almost all of them line up directly on those 3 broad categories.

Up next, we have Mike Ng at Goldman Sachs.

I just have 2 as well. First, just on the momentum that you're seeing in Dedrone, counter-drone, maybe you could just talk a little bit about other use cases that you're seeing beyond World Cup. I think it's natural for some people to think that you guys saw some outsized demand because of the event. What are some of the other private infrastructure, other use case demands that you're seeing on a go-forward basis to give you confidence around that greater than $100 million quarterly revenue run rate? And then second, just as a housekeeping item. On professional services within Software & Services, historically, we've thought of that as something that is tied very closely to fleet. I would imagine there are much broader use cases for professional services now. So just wondering if you could talk about what drives that.

Let me take the first one, and then I'll let the team take the second one. So look, counter-drone, there's this war going on in the Middle East and the one in Russia, where like you look at what the Ukrainians have been able to do to these Wildberries warehouses now, they're just decimating that. The Iranians are threatening to go after American infrastructure and data centers. But these aren't confined to just tools of war, right? You can have -- our customers get it. I think the World Cup is certainly an accelerant, but every data center understands, especially with now some of the rising controversy around data centers that any disgruntled person with a 3D printer, access to his drone in the Internet could build something that could cause real problems. I think people -- anyone who's prone to gun violence could be prone to drone violence. And so I think -- and it's not just us saying that, right? I think our customers broadly, even in enterprise, like we have deployments at CEO's houses as part of their security protocol. We're seeing it at data centers and corporate campuses. It's at the World Cup now, but I think it's going to be every Friday night football game in 5 or 10 years. So I think the World Cup was certainly a surge, but it's part of just a normalizing acceptance that there's a 3-dimensional aspect to safety that we need to be able to police.

So I'll add a little bit to that. Michael, thanks for the question. So ultimately, one of the more exciting things about our Dedrone business is, yes, federal is a piece of it for sure, but we're seeing far more opportunity internationally. And so I would not necessarily tie any -- much of our Dedrone revenue to one specific event. Those trailers, for example, that we sent to the World Cup are already being repurposed by the customer for other use cases. And so we certainly expect the Dedrone demand to endure. Now it's important to recognize that these are large revenue shipments of hardware. So quarter-to-quarter, they can swing a little. But when you add it all up, I don't think it will look like the World Cup was an outlier. And on top of that, as Rick said, right now, it's really enterprise, federal and international as drone mitigation becomes legalized for state and local, again, we view that as another potential surge in demand here. So we're feeling really, really good about Dedrone beyond the World Cup. Now second, on the professional services, we view this as an enabler to customer happiness, but also customer -- the usage of our products. So a couple of years ago, professional services meant show up and deploy body cams or TASERs or integrate to a record system and so forth. Now professional services are showing up and starting to build things on site for customers using our AI tools. And so very much like some other companies in some other markets, we really view the opportunity here to start to delight our customers with more and more on-site that could be rapidly built and custom deployed, just as Rick talked about at Axon Week around Axon Gravity. So professional services will certainly be a strategic enabler to more usage and more purchases of products.

Up next, we have Jonathan Ho at William Blair.

I really appreciate the additional disclosure on your 5-year annualized deals. One thing I would like to understand a little bit better is when we commented about the 9-figure deals that you signed with the municipal police, some of these are likely longer-term deals. Can you help us understand, first, like how much you've really expanded with these customers versus prior deals? And what this potentially looks like as you broaden your engagement with the entire customer base?

Sure thing. I think it's following a trend that we expect to see in a lot of places, right? It's what we've talked about with this idea that we sell products, we delight the customer and then we earn the right to sell more. And ultimately, I think that's what's happened in those couple of large deals. And they are meaningful deployments or expanded deployments of our products. It's not simply a renewal. And so oftentimes, there's more hardware and a lot more software in those deals. And I think it's something we can start to see internationally here and there as well, potentially this year. And ultimately, like our bookings, I feel equally good about them as I did at this time and on this call last year. We're very excited to see how we got out of the gate in the first half of the year. I was just at our Sales Summit all week in Scottsdale here and talking to our state and local team. And the attitude is we're expecting a really massive second half of the year, especially just like last year, a little more weighted towards Q4, but certainly plenty of room for growth. And when you add it up, we'll be in the 30% range again for bookings. And so we've got it all in front of us, and now it's time to go do the work.

Up next, we have Trevor Walsh at Citizens.

Josh, great to hear about the full scope Prepared-Carbyne deal that you signed. That seems a little bit counter kind of how you guys set up the go-to-market there in terms of, I think, landing more with the prepared use case of getting -- bringing some AI capability and then shifting to Carbyne later, which is more of a larger lift and shift of larger systems and databases. Are you seeing maybe an early trend that that's now changing and people are going to go whole kit and caboodle from the outset? And then kind of follow-on to that, what -- if and when they do that full scope deal, what's the uplift look like, generally speaking, if a customer already has some flavor of maybe an OSP type of bundle?

Sure thing. So the one thing that's becoming apparent to us is customers are actively looking for alternatives in 911. And we're excited that we're in exactly the right place here to help them by virtue of the Prepared and Carbyne acquisitions. One of the things we did early on to your question, Trevor, is when we acquired Carbyne, we actually integrated them into Prepared, not into Axon. So now this apparatus is Axon 911. And so now we are seeing a little more combination in certain places of the desire to do both of these at once. But that's kind of the beauty of it, right? Like some customers are going to be ready for that. But a lot of customers are going to buy one or the other upfront with a clear intention of moving away from their current vendors. And so I am wildly bullish on the 911 space and our ability to win there. The APCO Conference was this week and our Prepared team was telling us how much momentum they're picking up and how universally they're hearing that more alternatives and competition in that market is welcome. And so we're excited to compete.

Up next, we have Joe Cardoso at JPMorgan.

Maybe I wanted to touch on the body camera shipments commentary that you made in terms of sequential growth going into 3Q. I guess first question is more so any way you can kind of contextualize the growth that you're seeing from there and maybe bifurcating it between Axon Body Mini and maybe Axon Body 4. And then as a second part to that question, just curious how you're thinking about the Axon Body Mini rollout, now that it's at GA. I think your expectations around adoption, at least in terms of some of the early conversations was maybe more measured and like thinking about it as more of a tranche deployment, but at least some of the early data points that you guys had in your prepared comments as well as in the presentation material, it does seem like things could be moving a bit faster and more widespread. So just curious if that's a fair read and how we should be thinking about that and if there's any limitations we should keep in mind?

Yes. Thanks, Joe, and thanks for hosting us at your conference a couple of months ago here. Ultimately, on the body cameras, I wouldn't consider AI -- or I'm sorry, AB Mini as a big driver of that guidance that I mentioned for next quarter. I think there's some in there, but the real story there is we are winning in public safety commonly across all of our markets. I think there's a few different flavors of those deals. There's renewals in there. There's expansions in there. Like I said in the prepared remarks, there's also winbacks in there where more and more customers that made a different decision upfront are migrating to Axon. And so I think it's every bit as much of a commentary around U.S. state and local as it is enterprise. And then, of course, international is also showing some real signs of life. And then as it's been reported publicly, we've got some exciting federal opportunities in the body camera space as well. So I think hitting on all cylinders and certainly very pleased that this late in the product life cycle that we continue to see so much demand from so many different places, but I don't want that to be mistaken for a lack of demand in U.S. state and local. It's certainly there. And then last thing I'll say is on the quarter-to-quarter shipments. Just remember that the body camera shipments are a function of our bookings, right? So in Q1, we tend to have seasonally the lightest quarter in bookings of the year. And thus, when we ship those, since a lot of the deals come in late in the quarter, we ship them in Q2, that's a reflection of Q1 bookings. And so as we go through the year, you should certainly expect those body camera shipments to pick up. And I think if I were betting, I'd say it'd be closer to 30% sequentially next quarter versus this quarter. But we are giving the range of 20% to 30%.

Got it. And then maybe just the second part on Axon Mini there.

Sure. On Axon Mini, it's driving a lot of interest. We've kicked off -- I was just talking to our Head of Enterprise yesterday. We've kicked off a number of trials with the AB Mini that are starting to gain momentum and we're kind of brands you've heard of in the market and brands you might see if you visit their stores. So there'll be that growing trend. But it does take time, right? And ultimately, we got to do one thing, put one foot in front of the other for a while, succeed very well in the trials then capture our first orders. And then just like in public safety early on, that second cycle of expansion is where we see the highest volume. So right now, it's about seeding the market in a number of different large enterprises with maybe even hundreds of body cameras per deployment, even though some will certainly be in the thousands and then building from there just as more interest is coming online with Fusus and Outpost and Evidence.com in general with a number of these enterprise customers.

Up next, we have Keith Housum at Northcoast.

In terms of the Outpost and Lightpost now being out there for several months now, perhaps you guys can touch base on a little bit what you're seeing in terms of adoption. Obviously, we saw some takeaways that you had in the quarter. Perhaps just talk about success there and the lineup that you guys have.

Yes. So we -- so Keith, in the first half of 2025, I believe our Outpost, Lightpost bookings were somewhere around single-digit millions. From 2H last year through 1H of this year, so a full rolling 12 months, we're right around $100 million of bookings already in Outpost and Lightpost. So that business is certainly growing. We're very proud of how we've built these products, as Rick said, keeping privacy and information security and auditing and all the accountability features around how this data is used. We're very proud of what we've built there, and we think that is a major tailwind to adoption of these products as customers are looking for an alternative to what they've been using historically, and they're seeing a very responsibly built product that protects everybody with Outpost and Lightpost.

Up next, we have Andrew Sherman at TD Cowen.

Josh, great to hear the international commentary, especially the 3 big AI deals. Were you surprised by the early action there? What is really turning the light bulb on for them with some of the newer products, including AI? And are there more of these in the second half pipeline?

Yes. The second half pipeline is full with a lot of exciting international opportunities, no question about that. In terms of what they're seeing internationally in the AI Era Plan, I'd probably point to a couple of things. Number one, we've enabled Draft One in more markets now. And so that is picking up a lot of steam. And that does take some work because reporting is different in every country. So there's some work to do to bring countries online, but that work has happened and continues to happen. The second one is the translator. Whenever people are speaking 2 different languages, the real-time translator via the Axon Assistant is the thing that's driving a lot of the interest there as well. So was I surprised? I mean, our international team, it's just incredible to see the amount of progress we're making there, the quality of the team, the work ethic, the delivery, just everything that it takes to build an exciting business internationally. And so not a lot surprises me about our sales teams these days. They're pretty effective.

Up next, we have Brenden Rogers at Wolfe.

In the past, you've talked about longer sales cycles for AI and the OSP bundles with the bigger price tags on those. Are those cycles shortening now that the products have been in the market for longer periods of time, customers might be more familiar, more used to AI? Or are you kind of noticing these conversations are still taking a long time and maybe that's driving some of the more budget flush conversations?

In terms of timing, I don't know that there's a material difference because ultimately, even like when we're selling TASERs just stand-alone 15 years ago, you still have to go through a government process to buy something. And so I would say, if anything, yes, certainly, there's more acceptance of the products upfront, but it's maybe a little offset by just the sheer size of some of these deals. And when you're up at these levels of contract value, you do have to go through several approvals and a lot of contract negotiation and so forth. So I think one way to look at it is, clearly, the team had a fantastic Q4 last year, and they got out of the gate fast. And while delivering the results you've seen in the first half, they've rebuilt the pipeline for the second half. And so a lot of the work we've done in the first half, you'll see come through clearly in Q4 this year with another massive bookings quarter and even some buildup to that in Q3. So I'd say everything -- generally within a year, but still got to go through the process and do all the right things, and our team has become very, very good at that.