All right, we're going to go ahead and get started. Thanks, everyone, for joining as we begin to wrap up the conference. I appreciate everybody staying. Got a nice, pretty full room here for the end of the day on the last day. I do appreciate that. I'm Adam Tindle, and this is part of my connected devices coverage here at Raymond James. Very happy to have the team from Arlo here. CEO Matt, CFO Kurt, are both here in attendance.
In terms of the format, Matt's going to give us a high-level overview. I know we've got some that are newer to the story. If you're looking at the stock chart, obviously it's been, over a long period of time, a very strong evolution of the company, and that's starting to get recognized in the market. A very timely presentation for us to have. We will have some time for questions at the end, but I'm sure Matt would love if you would raise your hand during the presentation and ask any questions that you have. Of course, we'll have a breakout afterwards in Cordova. So with that, Matt, thanks for joining us. Take it away.
Appreciate it. Thank you, Adam. Like Adam said, this is a subset of the earnings deck we gave on Thursday of last week. I will run you through some of the highlights of the company. I would encourage anybody who wants the kind of the deeper, fuller story, number one, feel free to ask me a question during this presentation. We did a webcast of our earnings deck last week, and it has a full voiceover and everything. Within about 30-40 minutes, you would get the full update on Arlo as a company.
First of all, who are we? Arlo, if you are familiar with the brand, started off as really a hardware company. It was developed inside of NETGEAR when they were looking for a way to get people to upgrade their wireless routers. They came up with a video product, which uses up a lot of bandwidth to try and get people to upgrade. I think at the time it was 802.11n version of Wi-Fi.
They hit upon, instead of streaming video for entertainment purposes or others, the idea of using streaming video for security. Arlo was born just over 10 years ago as a division inside of NETGEAR. It took off like a rocket, created a whole new market segment in the security business for DIY security, and was growing so fast that ultimately NETGEAR decided to spin it into its own company, which we did in 2018. Like I said, it started a whole market segment of DIY home security.
The real reason for that is it was the first product in the market that provided full video security access, battery operated, so you did not have to run power cables to where you wanted to place the cameras, fully Wi-Fi, so you did not have to run Ethernet cables to the cameras either. You had full flexibility to set up an entire security system in less than 10 minutes and actually get visual cues of not only did my front door open, but who opened it, and full access from your mobile phone. Since the spin, we have become really a services business.
Over the three years or so after spin, it became very clear that the security industry is one of services first, and the hardware is really just the physical instantiation of that service, a way to create a relationship with that end user that lasts over time. Our average user stays a subscriber for about seven and a half years. It is very sticky. Our churn rate is one of the lowest in the industry, which I'll touch on. Really today, we are a subscription software company in the home security space and a little bit of small business.
Most of our culture at Arlo is around innovation. Like I said, we invented the category about 10 years ago. We are a leader in AI. We've actually been selling consumer subscription AI services since 2018. Before any of the buzz of AI has come out, we're actually monetizing subscription services that are based on AI features for nearly seven years now.
Quick highlights in the company to kind of show you where we are today if we take a quick snapshot. We have about 11 million registered accounts. From that, we have about four and a half. In fact, at the end of last quarter, we had 4.6 million paid subscribers, and that's growing quite quickly year over year. Annual recurring revenue from subscription services is about $250 million. We have over a quarter of a billion dollars in recurring revenue, and that's growing at 20% or more year over year. In fact, and I'll touch on this later, our guide for 2025 shows us growing at 25% or more year over year.
We're seeing an acceleration of our subscription businesses and ARR. Service gross margin exited at about 80% gross margin. If you take just our subscription, direct and retail subscription revenue, because we do other service revenue like NRE, non-recurring engineering, and a couple of other things. If you look at just the subscription piece, which is the vast majority of our service revenue, that's around 92% margin from an annual perspective.
Our LTV to CAC ratio, for those of you that are subscription service companies or SaaS company investors, our LTV to CAC ratio is four. Obviously, anything over three is considered world class, and we're actually at a four, sometimes even higher than that going forward. What started as really a hardware business inside of a hardware company has been transformed into a services business that's actually growing quite quickly.
Real quick, just to give you a little bit of history of where we've been and to show you how fast this company has been transformed. You'll see a couple of our kind of key metrics on our subscription side. This is how things like registered accounts, paid accounts, obviously recurring revenue, and our gross margin has been growing. You can see the trajectory it's been on, but also how fast this happened.
This is a look back just over the last five years. Again, 10.8 was at the end of the year. We're already over 11. 4.6 was our last reported paid accounts. We're quickly marching towards 5 million, and then we'll start to our push to 10 million paid accounts. Over $250 million in ARR, like I mentioned, you can see how fast that's scaling.
Service margin, as we've been growing and increasing our recurring revenue, at the same time, we're actually increasing gross margin on the service and, of course, company gross margin. We exited at a velocity over 80%, and that's continuing to go up. Especially, like I mentioned, our service subscription gross margin is 92%, and that continues to mix into the overall service revenue w e expect this to continue.
Real quick, we decided to kind of compare ourselves to a lot of SaaS companies. One of the reasons we did this is Arlo, often with investors or people who knew the company in the early days, either as part of NETGEAR or right after spin, still think of us a little bit as a hardware company, even though we've transitioned out of that over the past couple of years.
If you look at all of the SaaS companies that are public and look at how long it took them to reach $250 million in ARR, we're in a very small cohort of those that did it in six years or less. There are some pretty big companies here. The acceleration of our subscription business is absolutely outstanding. The other thing we've looked at, as we looked at our guide going forward, we looked at how many companies are actually going to be in the Rule of 40. For those of you who are SaaS investors, you know there's a Rule of 40 where you take kind of the growth rate and operating income and see if it meets 40 or not. Our guide for 2025 is actually Rule of 40.
That puts us in a class, if you add the 20% growth, it puts us in a class of only four other public companies, public SaaS companies in the world right now. Very fast growth, really separating from the pack, and clearly a SaaS company that's growing quite quickly. If you look at the TAM, what market do we play in? Obviously, security and DIY security. That's a relatively large market.
If you take the hardware part of that TAM and the services part of that TAM, you're looking at roughly around $25 billion or so from a market perspective. We are now also moving into the broader smart home category, which adds another $13 billion. There are several adjacencies that we'll start to be addressing over the next 12 to 24 months. This is a fast-growing market. It's a very large market.
These numbers are US only in the dark blue. You can put a multiple on those when you start looking internationally or you start looking into things like small business, age-in- place market. Other things our platform could actually address. Like I said, our history is really around innovation. We are a product technology and consumer experience-led company.
We invented the space about 10 years ago, and every year we bring multiple innovations to the market and stay ahead of many of the competition that's out there. Some of that innovation is on the hardware side. That could be RF, power design, new optics, other things that we do to make sure that our cameras and our other devices are top of class. Increasingly, it's been around AI, platform development, subscription services, and other capabilities that are bringing that user experience to the end user.
If you take a look at our platform, and this is really the asset we've built over the last 10 years and a big part of the company value, it is a fully operational scaled cloud platform specifically built for low-latency applications like home security or small business security, but can be leveraged into many different markets. We just happen to be using it right now specifically for scaling the DIY security business through multiple channels.
When I talk about scale, we send out about 170 billion AI alerts per year. That is obviously a lot. Those are AI-driven. So that's off our subscription services where we're actually able to characterize what's happening in a scene and actually convey that to the user so they can filter out the different notifications that they don't want and get the ones that they really care about.
We upload about 1,700 hours of video footage from our cameras per minute. That is about two to three times the size of YouTube, just to give you an idea of the scale of the platform we are driving. Part of what we do is we obviously have a lot of APIs that we use internally, but one of the fastest parts of our business from a growth perspective is what we call strategic accounts.
Yes, we use our platform to go with the Arlo brand through retail and Arlo.com and through some partners, but also we are powering the security platforms for other companies. One big example of that is Verisure in Europe, where their entire cloud backend for video storage, video AI, and other security components from those cameras we are actually running on their behalf on a background as a partner.
If you add up all that as well, we're doing about 26 billion API calls per year across the platform. We've shipped over nearly now about 35 million devices. This is a business that's operating at scale. Part of the expansion you're seeing in gross margins is us being able to leverage our current business as the top line continues to grow. If you look at the services that we sell to users, so our subscription services that we sell to users after they've purchased some product from us, we have basically three plans. It's really two plans. The first plan has a single or unlimited cam plan that provides all of our AI capabilities.
You can step up to the premium plan, which also includes professional monitoring, cellular backup, battery backup, and really a full user experience that you would get from any other big security company. While they charge $60 to $70 a month, our top plan is $25 a month if you buy it on an annual basis. Even today, we are about half the price of what a traditional security provider would charge. We think there is a lot of potential pricing opportunity options for additional tiers of plans as we go forward. If you look, I captured a couple of stats here just for review as well. Our monthly average revenue per user for subscribers has now risen to $7.50 roughly.
If you go back about a year, and these are new subscribers coming in after we made a couple of plan changes at the beginning of this year. Last year was about $12. The year before that was about $10 to $11, and the year before that was about $7 to $8. One of the things you're seeing is ARPU expansion.
That's being driven off our AI services and AI capabilities that the users are really responding to, and they're mixing up their plans, and we're changing the plan tiers to actually include more AI capabilities because we're seeing users really demand that in the field. We have an annual plan and then monthly. Most of our users are still on monthly plans, no contract. I'll show you, we still have world-class churn, even though 75% roughly or more of our customers are actually on monthly t hey can churn any single month.
When you look at churn from other companies where they may have locked their customers into a three-year contract, it is very different. Ours is a real transparent organic churn rate where others may be locked into certain contracts.
As we launched Arlo Secure 5, which is our new service plan that we launched in October, November, and saw that mix up in ARPU, part of that was we saw a 20% lift in users actually selecting the premier plan when we made that change, indicating an additional ARPU increase. If you look at the devices that we ship that are active in the field, roughly 65% of those are actually tied to a paid plan already. We have a pretty good attach rate, conversion rate.
When somebody buys hardware, how likely are they to actually pay for service, even if they just buy it at a Walmart or a Best Buy. Real quick, I wanted to dive into one of the specific areas of our platform, and that's our AI capabilities. This is something we've been investing in, like I said, since 2018. In fact, the original models that we wrote were built in 2017. It seems so long ago. We've been doing object detection so we can tell the difference between a vehicle or a person or a pet. We can detect packages at your front door. We can detect if somebody took the package from your front door. We're also doing now fire detection.
What's interesting is with fire detection, often the flames can trigger the image analysis and a detection and alert for fire three to six, seven minutes ahead of smoke actually hitting a smoke alarm and going off. This is important for several areas of the home, but we've had some small business requests for fire detection as well. We also built a very robust recognition engine. We can do full person recognition.
Think of facial recognition, determining who came in the front door, who came home, being able to trigger either alerts or automations based on that. We've been doing that for quite a while, but we've also added vehicle recognition. If a car pulls in your driveway, we can detect whose vehicle that is. Is it your wife's vehicle? Is it your daughter driving home from school?
You can ask for alerts on specific vehicles or, for instance, say, I only want alerts if it's an unknown vehicle. A vehicle pulls in and I don't know, I don't recognize it, then give me an alert. We also do a lot of audio recognition, we have smoke and carbon monoxide alarms, we can also tell the difference between the two using AI. You will see us roll out gunshot detection, scream detection, dog barking. There is a ton of AI components coming on the audio path as well.
We can even listen for car alarms and other things and give you specific alerts. The one that we've innovated the most, and this is what drove a lot of interest in Arlo Secure 5, is we're the only company in the world to offer what's called a personal AI engine.
What happens is you can come into the system and just type in what you want to detect. You can type in, "Did I leave my garbage cans on the street?" or "Did I leave the back gate open?" or "Did I leave the lights on?" What we do is we take that text and our AI engine actually starts to develop a model around what does a garbage can look like, what is a street, what is a garbage can on the street, and starts to build a model to detect what you've asked our engine to do.
You can then add images, you can add an image of your garbage cans on the street and one without and actually help train it. That'll get it from about 88% accuracy to over 95% accuracy. It actually learns over time and gets smarter over time.
This opens up unlimited AI capabilities for security or smart home use cases. What's interesting is what we're doing with our platform is every single person user on our platform, these are specific to. This is not going into our general larger model that we use for general detections.
These models are actually personal, fully encrypted, and none of the information or the model itself is shared with any other user on our platform. Our platform's moving from a couple of very large models that we leverage for all the users to what will be hundreds of thousands and eventually millions of micro models that users can actually train on their own. It's a huge innovation in the platform and something that drove a lot of demand in Q4. What's coming?
We did announce last week a couple of things that are coming in Arlo Secure 6. One is smart home integration. We're going to start to add a lot more cross triggers. Think of using your sound bar to bark like a dog if somebody comes to your front door really loud or doing other things with other capabilities. Turn all the lights if somebody is in your backyard. Other things to actually mitigate or deter t hat's coming.
Full video and audio descriptions. You'll see some examples on the left. Instead of car detected, it'll say a blue SUV pulled into the driveway. Two people were pulling packages out of the back and somebody walked into the home wearing a red shirt. You'll be able to go back and actually search on that.
You could search on red shirt and suddenly all the videos in your platform would actually show that had a red shirt in the video. Very specific. We will be rolling a beta threat assessment. This will be one of the biggest innovations, I think, in the next year.
We think we're years ahead of other people where we have a model that we're developing will actually take a look at everything that's happening on the video, the audio path, and other sensors in the home and actually score from a threat assessment perspective how significant that is across multiple scores. Maybe personal safety, property damage, other things. That score will then change how the system alerts you on what's happening. Is it an escalated alert? Is it a normal alert? Is it muted so it doesn't bother you?
If the call center, professional monitors have to respond, they will get a lot more metadata and be able to respond a lot faster because they'll have a threat assessment score and they know how to prioritize the calls that are coming in. All of this will come out later this year. In fact, some of this we might launch a little bit early because we're actually ahead on our development cycle.
Again, a lot of the value we provide is that innovation cycle that's driving consumer demand to the subscription services and staying ahead of any of the other user experiences that are out on the market. I mentioned this already. A lot of the security companies don't provide churn metrics. They don't like to share it because it's a lot higher than ours.
If you look at just consumer services in general, the ones that actually do report churn metrics, you can see that we have a world-class retention from a subscription services company. This helps drive our LTV of over $700 per user. This is something we focus on very much. Some of this, I think, is driven by the subscription services and the great user experience and everything else.
Some of this is driven because security is an extraordinarily sticky subscription service and one of the last things that consumers tend to cancel. Real quick, just on Q4 metrics, just to give an update. Again, there is a lot more information in the deck that we provided on Thursday or if you want to watch the webcast. Just a couple of ideas. We have already talked about hitting 4.6 million paid subscribers. That is up significantly.
$257 million in ARR. That's up 22% year over year. Service revenues to total revenue mix is now over 50%. Service gross margin actually hit 82% in the quarter of Q4. We see that continuing as we go forward. Real quick on the guidance, as I mentioned earlier. We've given guidance for both Q1, which was up from consensus in the quarter, and also full year, which was also significantly up. We decided to give guidance not only in the typical metrics like revenue and EPS, but to help people peg how fast the growth is happening on the services business, which is really driving the financial metrics of the company.
We decided to show that we're going to hit $300 million or more in service revenue and that our service gross margin will stay above 80% as we go through that year and drive that growth. This is contemplating roughly a 25% growth on ARR, which is an acceleration of ARR growth year over year. We did do a long-range plan a couple of years ago, which was 5 million, $300 million in ARR and over 10% operating margin. Our goal was to hit that in 2027. We're going to hit that probably midpoint this year. It was pretty clear to a lot of our investors that these targets, which seemed huge at the time when we put out the original long-range plan, we call it LRP1, it seemed like these were going to take five years to hit.
We hit them in basically two and a half or will hit them in about two and a half. We actually set out, as of last year, we set out a new long-range plan and actually doubled or more than doubled every key metric. I mentioned on the call on Thursday, it's very likely given our current trajectory, we'll probably hit these early as well. Yes?
Can you talk a little bit about who you're taking market share from with this growth?
Yeah. The question is, who are we taking market share from as part of this growth? I think there's two major categories. One is there's just a lot of new market. If you look, the penetration for security, we call it smart security, security experiences that are connected to the internet, is around 18%. Some subcategories may be about 20%.
There is a huge greenfield ahead of us in the United States alone, let alone internationally. That is one. Two, the market is really starting to move mass market. The next 50 million households of capture are going to start happening soon this year. In fact, we are seeing Walmart and other kind of mainstream customer outlets start to really grow in their share of the market.
That is a signal to us that this is going mass market and we are going to see a lot of growth. First, I would say some of the growth, a good part of the growth, is coming from just new customers coming into smart security and actually adopting this for the first time, whether it is a home or an apartment or some other thing that they are trying to do like a small business.
From a competitive share capture, the biggest share capture is from traditional security companies. So think of the ADTs, the Vivints, the Brink's, the Resideo that are charging $50, $70 a month. And our top tier on an annual basis is only $25 a month. And the service is substantially better from a technology perspective, but also a user experience perspective. If you double-click just into then the DIY space, we're seeing customers like Ring start to lose some market share. Google's losing share.
Our plan is to capture additional share throughout this year. Last week, I mentioned on the call, we're going to have our largest product launch in company history in Q3 for the holiday season where we're refreshing every single line and actually expanding our product portfolio. We're already confirmed to double our shelf space at one of our key retailers.
I think there's more to come. I think we're capturing some share in our subsegment in the DIY kind of smart area. We're definitely share shifting share from traditional security just because the value is so much better and the experience is so much better. I would tell you that there's just an under-penetrated broadband household market that is just starting to go mass market. I think that'll be the next leg of kind of major growth. Yes?
This Q3 launch, is that hardware going to work with the new software? Yes. Or do you wait to buy the new hardware?
Yeah. Great question. The question is, the Q3 launch that I'm talking about of the new hardware that's coming, does that work with the new software or do the old cameras work with the new software? One of the strategic decisions we've made as a company is that we're really a services company and the hardware is the enabler. That's what fills our funnel at the top of the funnel.
Actually, all of the services I've shown you to date actually continue to work on cameras that we shipped eight years ago. Because it's a cloud platform, even our oldest cameras can actually take advantage of all these newest features. That's, I think, been part of our success of bringing even older users into the subscription realm and going forward. Really, when people make a choice around buying new hardware, it's the hardware capabilities or tend what we see is they're expanding their system from two cameras to four cameras. They're just kind of building out because they have a new floodlight or there's a new form factor they want.
Almost all of our features that we roll out can roll out against all 35 million devices that we've shipped. Here is the new long-range plan, like I said. We are targeting this by 2030. Again, on the current trajectory, we are going to hit it pretty early. We decided not to update it again because we've updated a couple of times. We are going to clear LRP1 way early. We think we are on track to clear LRP2 early as well. In summary, just to kind of wrap up, one is we are a company that has a SaaS platform, right, and an innovation engine that's running very quickly.
I'll tell you, being focused in the space, meaning not being part of a large company and sometimes I'm an ad company and then I've got to build a general purpose AI and then I've got all this other stuff I have to do, we wake up every morning and just work on consumer small business security and maximizing the potential of our platform to monetize and create great experiences for our end users. That has led us outpace both on a growth perspective, but also on an innovation perspective. The platform we've built is a huge part of that. We have a big device launch, like I said, coming in Q4 that we think is going to add additional growth as we exit this year.
I mentioned already, we've already got commitments from our retail side to expand our shelf share and potentially grow unit velocity through that channel. About half of our business is through strategic partners. And those are, like I said, like Verisure or other service providers that actually sell or go-to-market with our platform in the back. We announced three new strategic partners on the earnings call.
You should expect, our hope is within the next two, three quarters, we're going to announce some additional strategic partners to grow that piece of our business even faster. Kurt and I, when we look out over the next five years, we think about 60% of the incremental growth will actually come from additional strategic partners on a global basis. $300 million in service revenue, which was a surprise. A lot of people thought we were going to target maybe $280 million, $290 million.
We think it's going to be 300 million or more. That growth and acceleration of our service business continues. Like I mentioned earlier, these metrics, including our growth, actually put us in a very rare space where even in a somewhat muted consumer environment and what people are not sure a very volatile consumer market for 2025, our guidance puts us into a Rule of 40 company this year. I don't think that's been reflected, obviously, in the stock or anything else.
Right on time, Matt. Good for you. We're out of time, but Cordova 4 is going to be.