Good morning, everybody. I'm Tim Horan. I am the Digital Infrastructure and Communications Analyst here at Oppenheimer. Massive changes in the market thanks to AI and new technologies. Wi-Fi 7, the cloud is really transforming how we do networking both on a public and a private basis. Stan Kovler used to do my job, so he knows it really, really well. He is VP Finance and Corporate Development at Extreme Networks. Extreme is one of the leaders in private networking, both wireless and wired, switching basis. They have relatively low market share, and they're going to gain lots and lots of market share, right, Stan?
That's right, Tim. You got it.
We're going to go through all the different products and services and how they're going to gain share. We think they're really, really well positioned. I guess with Stan, how do you describe the company and the company's strategy and what your core competencies are and competitive advantages? I guess we can touch on it at the same time and maybe touch on it first. How do you kind of measure the TAM and what you think the TAM can grow at and, yeah, share you can gain within that TAM? I know that's like an hour-long question, but I can just tell it.
I'll package it. I'll try to package it up for you so we can have a good conversation. What I would say is, Tim, we are trying to solve complex networking problems inclusive of various security threats and just make it super simple for customers to do networking, right? If you were to distill what we do and how we do it, we are trying to bring an element of simplicity, reliability, value, and a lot of innovation. A lot of that innovation was historically packaged in capabilities that we had in networking related to our fabric technology. What that means is that we have a unique way to automate network connectivity, create different groups of customers where you can have different divisions of a company or a city infrastructure that work differently.
For example, my stomping ground in New Jersey, sometimes you have this building of a municipality that houses both the city government and they have some finance people that collect property taxes and whatnot, or the water bill. The other side of the building is your police department. With our type of network, you can create one network for those different environments. You have logical separations where people feel like they have their own network, but it's actually running on the same physical infrastructure. That's how you really bring value to the industry and to our customers instead of having discrete networks for all these different functions. We could do that at the corporate level. We could do that at the local government level. We do it for schools. That's some of who our major customers are.
We do really well in public sector environments because of this posture that we have for value-oriented, highly automated solutions. A lot of customers don't have massive IT departments in these fields of work, and they really appreciate the solutions that we bring to the table. We also have some high-end capabilities as well, where if you needed to fill a stadium or venue like MetLife Stadium, which we have now, we won the Wi-Fi of MetLife Stadium. We used to have just the wired part of the network. We do this for many of the NFL teams. We can do it for NASCAR, NHL. We do it for Major League Baseball, football or soccer teams in the U.S. and also in Europe as well, like Liverpool, Manchester.
We can give you super high performance Wi-Fi for a 60,000, 90,000 person stadium that is really high quality and also gives you a lot of analytics capabilities. You can see exactly what you're doing while you're at the game. Someone could be ordering a hot dog or a beer at the game, and someone else could be sending uploads of what they're doing at the game to their friends and family or social media. This is really what we excel at, giving, you talked about private networks. We give the owners or the operators of the private network, it's really the local area network, an incredible amount of power and visibility, and our solutions are both cloud-driven, meaning you can log into a cloud instance, like a website, and see everything that's happening on your network.
That was very powerful in the form of a productivity gain that you would be able to achieve. For customers, for example, another vertical that we do well in is retail, multiple stores. Kroger has thousands of stores that we service in the U.S., and we do the Wi-Fi there. They can manage a lot of them from a centralized location. You don't have to send your IT folks to individual stores when things are not working well. Now what we're doing is we're taking all of these things that we have, our fabric, our cloud-driven capabilities, and we've added the power of AI, specifically AI agents. We now have a solution called Extreme Platform One. It takes all of this goodness that we've had before and created an AI-enabled core, Agentic AI, where you can have a conversational interface with the AI.
You could just tell the AI what you'd like it to do on your network. I know that's something that you want to get into, this new innovation that we've come out with that is just on the cusp, it just GA'd last month, and we're very excited about it. We had really good results coming out of FY2025 and Q4 on the businesses not even related to Platform One, and we're excited to bring the next generation of networking to market with this capability.
Sorry, Stan, I was confused there a second. The Platform One was generally available last month, but there was something else that was driving results that you mentioned?
Yeah, I would say in Q4, we had a lot of success and momentum from just all the business that we're generating based on what I discussed before, which was the power of our cloud-driven and fabric solutions. We saw a lot of momentum not just in the U.S., but in our international business. We're running business in Asia, and Europe is coming back for us. From a product standpoint, we're seeing for a couple of quarters in a row now the wireless business coming back. We have two components from a revenue standpoint of our business. We have a product sale that we do. Attached to that product sale is various services that we offer, support services or maintenance services if you want to call into tech support, or next business day replacements type of services where a product, let's say, breaks down. We support the customer.
We send them a replacement within a day or, in some cases, hours. Customers pay for that element in the service and subscription part of our revenue. From a management perspective, we do these massive networks for customers. You have to be able to manage them. We offer software for management. Some of that can run on-premise on a server. That's a site engine. Some of it works in a cloud instance, like hosted on a local data center instance of Azure or AWS or even the private cloud of a customer. For that software, we charge as a service. Customers subscribe to a three-year subscription for that software. What that gives us is, in addition to the product sale, we have about 36% of our revenue that's recurring based on the software and subscriptions that we're selling. That's the part of our business that really re-accelerated in Q4.
We measure our SaaS business in the form of SaaS ARR. We have $208 million in revenue in our SaaS ARR that grew 24% year- over- year. It was a nice 20% plus growth rate. We talked about for FY2026 that if you look at the midpoint of our guidance, it's about 8% total growth. That SaaS ARR, we talked about on the call, has legs to grow somewhere in that 20% range for this fiscal year.
Got it. Now, I think you've changed some of your SaaS licensing agreements. I think now most of your new equipment sales have to come with a SaaS license. Is that correct?
Starting in FY2026, we are requiring customers to buy our new, for new customers, when you buy products from Extreme, we are promoting this new product that we have, which is a combination of the AI-based subscription services and also the support that's bundled into one license that we call Extreme Platform One. This is where you can purchase that Agentic AI capability to help you run your networks and drive productivity.
Are you charging? Now everyone's going to have to take this. I think previously only about 70% of people took the support.
That's right. Support gets a boost in that sense as well. Part of the reason for it, I think, we've talked a little bit about this before, but if you think about support, you can do support from the classic tech support model, which is you call in physically and speak to a human, and they help you troubleshoot your network. With Platform One and with AI, those capabilities are fundamentally built into the platform. They're built into the conversational interface. If you have an issue with your network, the first thing that you will do with this technology is you will go into your bar, the chat bar, which is similar to any AI like ChatGPT, and customers can type in to diagnose what the network problem is. Tim might be having an issue with his phone.
It's not working well, or your IT administrator might call in and say, "We can't process a payment. There's something wrong with payment processing." You can write that into the conversational interface, give it a little bit of detail, and understand what is happening with your network. It gives you an answer in a matter of seconds. The remediation of your network issues goes from hours to minutes, to some cases, seconds, because the power of the AI and all the data that you're feeding into this model, it's producing these results and giving you that answer. What will happen is that over time, your level of service and satisfaction will grow, but the calls into our tech support will drop. We have a 9.5 rating in our customer CSAT. We have the best tech support in the industry. People love it.
You still have to spend the time calling in. One of these ancillary benefits of using AI as a productivity gain is that the mean time to resolution gets cut to a minuscule amount of time.
Right. That's a great, great point. The customers, they probably oftentimes could use it. They have a few network people working in-house, so I'm assuming your average customer would probably have four or five. I don't know. Maybe a larger, much more amount of people. I'm assuming if this works well, they can cut that number down at some amount, 30%, 50%, 70% down if it's more automated. Is that a way to save them money also?
It is. I think the other way is that let's say that you have a person that leaves or retires or just a load on your organization grows. When you fill in your next staff to replace or add a new person, you can add a person who is arguably less experienced and then use the power of AI to fill in that knowledge gap. As your network engineers leave or age out or just run out of human capacity because your networking needs are growing, that next person that you're adding will be potentially a lower cost resource because you can do more with less with the power of the software.
How much more will Extreme Platform One cost than fabric cost historically? Is there a step up in price for customers also?
Yeah, we're expecting that over time, we're trying to achieve a 10% - 20% uplift in the price for customers. Part of it is also to finance a lot of the additional AI that costs us money to run. When you do queries on the AI, that runs additional cycles on the environment that we're running back in Azure or AWS. We essentially have to pay for those cycles. That's why the cost is going up for the customers. We're giving them that capability. As we all know, these data centers are very expensive to build. They charge us and other service providers money for that. We have to recoup some of our costs using this AI technology. What we did was from a pricing standpoint to ease customer adoption is, number one, there's backwards compatibility to the older system, to ExtremeCloud IQ.
When you upgrade, you can test things out, and it's a gradual upgrade cycle. Some of our customers are reporting that the upgrade was a 45-minute cycle to upgrade from ExtremeCloud IQ to Extreme Platform One. It's relatively quick for a large organization. The other thing that we're doing is from a pricing standpoint, we've split up our customer base into various cohorts. You might have been a wired customer, but you may not have been a subscriber to cloud-driven management. You were just a support customer. That's one type of customer that will now migrate as their support contract runs out. We can offer them, "Hey, you know, why don't you go into Platform One? This is a new capability that we have." It could be when you buy new equipment.
You're renewing your hardware, and that's an opportunity for us to uplift the customer into Platform One. We're going to be very strategic because the idea is to bring everyone into this new paradigm. We don't necessarily have to get everyone, you know, it's not going to be a 20% uplift on day one. We're trying to get people to that level over a certain amount of time, you know, maybe a three-year cycle, and gradually step up so that we can pay for all these additional capabilities.
Is there an ability with Platform One to add more AI, more security, more services over time? Yeah.
Absolutely. One of the things that we've already added into it is we have capabilities that we launched a year ago, which are Universal ZTNA. That's not something that we had before. We had network access control. Universal ZTNA is something that we added as a security feature. Over time, we can add additional capabilities, such as security features, more automation, capabilities that really improve management, maybe add additional devices like non-Extreme devices to manage in this platform. Once we have this capability and once we're ingesting a lot of that data from a network, we can tie into a lot of different other systems. We can also look into other systems that are maybe like the supervisory systems for our organization, like a ServiceNow environment. We plug into that type of environment to feed data into those other systems.
Overall demand is pretty amazing in these arenas. I think pretty soon, hopefully, they're going to have facial recognition. I know some places have, but it'll be ubiquitous. They'll have grab and go shopping so you don't have to wait an hour to get a hot dog and a beer. I'm sure there's dozens of other applications. Where are we, you see all the stadiums. Where are they in adopting facial recognition to get in and grab and go and other things?
We're starting to roll out. I mean, I think that, you know, the grab and go is what we see at the Gillette Stadium, at the Patriots. They've got grab and go. Met Stadium has some facial recognition. I think, I'm not sure if Yankee has it, but we're seeing that adoption as well. You need a very strong network. You need a lot of security. The other thing is just the proliferation of usage where you have video cameras all over. You are going not just to basic video surveillance, like 1080p, but there could be 2K, 4K cameras that get really specific. This is also true at airports. We're seeing a lot of upgrades, not just in venues, but we're seeing airports make upgrades to their network infrastructure as well for those physical security needs. It's a big driver, I would say.
Video in general is one of the bigger drivers from a bandwidth perspective and from a use case perspective of why people need a modern network infrastructure.
Yeah, I would greatly think so. I got dozens and dozens of more questions for you, and I got to focus them down. You know, I've been a big proponent of Wi-Fi 7 upgrade cycle. You guys, I think, are very early in that product. I think right now I have much more flow share than your peers for Wi-Fi 7. Can you just step back? What are you seeing in Wi-Fi 7 in terms of improved quality and capacity and latency?
Yeah, great question. Wi-Fi 7, and I would say Wi-Fi 6 before that, there's a significant shift in the capabilities that you can offer on these technologies because traditionally, we've operated Wi-Fi on two bands. We've had 2.5 GW and 5 GW bands of wireless. Now, with 6E and 7, we've added the 6 GW band. You're expanding the information highway to another lane, and you can do a lot of things on that lane where it can be sensors, it can be internal systems, security systems, or it could be just load balancing where you send certain users on that 6 band. That's one of the benefits that you get on upgrading to this technology. It's a future-proof investment.
From how the technology actually works, from how these access points and how the data gets streamed from a user or between devices, it's become a lot more deterministic and more reliable. The increased reliability of Wi-Fi puts it into an interesting position because before Wi-Fi 7, there was this notion that private 5G could gain share in wireless LAN environments. I think with Wi-Fi 6E and Wi-Fi 7, that story gets kicked down the road a little bit, kicked the can down the road because the increased reliability is one of the use cases of 5G. The other additional benefit was just how far the technology could reach. If you have a lot of distance, then 5G is something that you might want to use. For shorter distances, Wi-Fi is the preferred technology because you have more bandwidth and you can cover more users.
The third benefit is that in addition to the normal speeds and feeds, like the 4x faster that you can achieve on 7, the deterministic nature of the technology really improves density. That means that you can support more devices in a particular location. You can support more density. We're all carrying more and more devices with us. You talk about more and more devices at stadiums and other venues. It happens everywhere you go, office buildings, schools, airports, manufacturing facilities, hospitality. Anywhere you go, there are more and more devices. There's just that natural need of density and being able to support density. Where we are in Wi-Fi 7 is that we talked about 30% of our wireless business coming from Wi-Fi 7 now. That's a lot faster than you would expect from the broader industry.
When you look at the industry analyst forecasts, the industry analysts expected about half the market to be on Wi-Fi 7 next year. That's next fiscal 2026 for us, and we're at 30% now. I think you're seeing the adoption. The other advantage that we have is at the end of FY2024, a year ago at this point, we wrote down our Wi-Fi 6E and older technology hardware for Wi-Fi. One of the reasons was we didn't want to offer our customers older technology as they move forward and make these forward investments. There was a lot of over-ordering during the supply chain crunch when the entire industry just got way too much supply of components because first we went into a supply chain shortage three years ago. A year or two ago, that shortage alleviated in a very rapid fashion. Many of our competitors still have Wi-Fi 6 inventory.
What we're seeing is that in some cases, we will go up against a competitor who will offer older technology like Wi-Fi 6. We can come in with 6E and 7, and the customer would like to future-proof their network. We have a natural advantage right now. I would call it a tactical advantage. Being able to offer more freely, having that flexibility to offer 6E and 7, whereas a lot of our competitors are still trying to work down their Wi-Fi 6 inventory, that's also giving us a tactical advantage in the marketplace.
A lot to digest there. Do you think we're entering another, I mean, it's obviously incredibly cyclical. Is there a chance we enter a period of shortage of Wi-Fi 7 again?
No, I don't think there's a shortage of Wi-Fi 7. As we look at the availability of supply, there's chips that's being built for this technology now. I don't think there's a shortage coming up of this technology.
All this stuff is the chicken and the egg. It sounds like this is doubling, tripling the capacity. Latency is a little bit lower with a lot more devices. It enables all those new use cases that we touched on, some of them. A lot of those use cases are really AI-driven. I mean, if you're looking at, you know, a lot of it's facial recognition and image recognition, right, which is all AI. Are there, you know, other AI-driven new applications or demands on the networks?
Yeah, so I think from a productivity standpoint, that's the biggest demand on AI. AI also helps you for network planning. If you have an event coming up or if you have growth or an expansion plan coming up, a lot of these network planning tools you can add into AI. What we're seeing in various industries is as companies build out for AI, for example, the physical infrastructure required to support AI, the data center infrastructure, you obviously need a lot of planning for that. We're seeing a lot of our customers really develop new capabilities. In the manufacturing sector, you've got an expansion in the semi-cap sector. That's driving growth. You've got the power sector also trying to support the power requirements of AI. That's also adding growth.
As you plan for these new applications, new plans, greenfield expansion, AI is a great tool for understanding what the networking needs are. Typically, when you deploy a new network, you do a survey and you go around and you understand what the parameters are, requirements of what you're operating on. From a Wi-Fi standpoint, where are you going to position the access points physically? Mapping becomes very important. These tools of mapping, understanding where to position the tech, and running various simulations, AI is great for that. You understand how your network will run in that simulated way before you actually deploy a live network. You can do that scenario planning with AI. Reporting as well, you can come in and you can prepare a custom report that runs every day and it repeats itself every day.
Come to the office and the AI reports to you exactly how your network is running, if you had any outages overnight, any upgrades that happened overnight that may have affected you. It just automates your work a lot and gives you that additional productivity that you need so that your IT staff becomes more strategic and they can focus on planning or project-based work instead of maintenance and just plugging holes.
Switching gears a little bit now to maybe your competitive advantage, there's been some consolidation in the industry. I guess what do you think the overall industry is growing at and how do you gain share in the industry at this point? We've been trying to gain share for a long time, but you're still relatively small. I know it's hard to do in this industry given a lot of issues, but yeah, how do you gain share now and what's the overall market growing at and what do you do to gain share?
The overall market, I would say if you take all of enterprise networking into account, it's probably still growing mid-single digits. Our ability to gain share really is on land and expand. Part of that is with our product differentiation that you mentioned. A great example of that is, let's say, with our recent example was the government of Japan. It's been a long time coming. We've developed a sovereign opportunity over several years. We started with one ministry with the Japanese government two years ago, and we've had to switch channel partners over time. We've built a relationship with channel partners in Japan that allow us to tap into the sale into the government ministries of Japan and working alongside the CIO over there. We've really expanded into the judiciary, theology, like various sectors of the Japanese government with those capabilities.
What they like is the security and simplicity of what fabric brings. This is what they really like. We're being installed in lieu of the largest competitor that we have in the industry that's being displaced with this customer. That's been a real nice growth opportunity for us. Since developing that over the past two years, we see additional growth vectors. There are many other ministries or parts of the IT infrastructure in Japan that we can go get. When we really partner and become strategic to these customers, this is when we have an opportunity to gain meaningful share in sizable chunks. I would say that we have developed this upmarket strategy, and moving upmarket allows us to take those meaningful chunks of share because the market we talked about TAM earlier, we're playing in a market that is probably $35 billion in size.
We as a $1.1 billion company last year, you know, on a run rate, I would say Q4, Q1, if you look at the guidance, the run rate is about $1.2 billion. At that run rate, we really need to take meaningful share out of that $35, $36 billion market to show up in more than 10, 20 basis points of share gain at a time. You know, 10% share, we would more than double the size of the company. In fact, we'd essentially triple the size of the company to be a 10% market share company. That speaks to how much growth potential there is, but also that we really need to become more strategic to our customers in order to take these larger chunks of market share out there. We are seeing a lot of greenfield, a lot of interesting greenfield investments in the hospitality sector.
On the call, we talked about opportunities in the Middle East where there is expansion in the hospitality sector. We had some wins in the energy sector. Manufacturing, we mentioned as being a strong cycle, and a second derivative effect of AI is providing growth opportunities there. Our strength in education is very prevalent. Specifically now, we've had two years of share gain in the higher ed space. Our success in K through 12 is well documented. Where we're adding share in education is that university sector. That's the second leg of growth there. This is what we're trying to do in moving up market, to gain these larger customers, gain the trust of larger customers.
You've done real well historically in education, sports, retail. Now you're entering manufacturing. It sounds like you think you're on the cusp of gaining share in most of these industries.
We are. What's interesting is that the biggest opportunity that we have to gain share is when there's market disruption or our customers are in a position where they're upgrading their infrastructure. One of our largest competitors talked about $43 billion coming up for refresh over the next five years in their installed base. We view that as our opportunity as well. Typically, customers just run their networks. If there's no issue or no major issue in the network, you'll be hard-pressed to upgrade your network until your equipment is fully depreciated or you're up for a renewal on your service and subscription. With this refresh cycle of our large customer, many of those customers will start to look elsewhere. Some always are perennial customers of the largest player in the industry.
There's a good enough opportunity for us to go after the other customers who are open to doing business with us or other networking providers. We're in a very strong competitive position to win those new opportunities. There's also disruption from the HP and Juniper merger. As those companies come together and figure out their product portfolio, there's a lot of disruption coming from their personnel consolidation and cost-cutting efforts as they try to create synergies. From both this opportunity and also from Cisco changing its partner program, we think there are opportunities that are opening up with new channel partners that would like to do business with Extreme. We have a nice story for them in terms of a product sale. We have very flexible commercial models that we've introduced over the past several years where if a customer wants to buy in a traditional manner, they can.
If the customers want to buy as a service, they can do that. We're also growing our business with managed service providers if customers outsource their IT. I think there are a lot of opportunities for both customers and partners to get access to this technology. The technical advantages matter, but how this is consumed is also important.
Yeah, we didn't really discuss your go-to-market now. You have a few new products there. Anything you want to, you know, the channel, anything you want to elaborate on the channel? I know you have the automated managed services program also. You know, how's that going? Yeah.
The managed services program is going really well. We doubled the size of the program in terms of partners that are signed up. About half of them are transacting. Decent contribution to our SaaS ARR if you look at a year-over-year contribution. That's really benefited us in that line item of our financials. The other thing that's helping us, our SaaS ARR, is that talking about the manufacturing sector, our marquee win there about a year ago was in a disaggregated model where customers that meet a threshold of potential purchases can go directly to our ODMs and buy Extreme compatible equipment. We then license the operating system. We provide subscription management and support capabilities on top of that hardware. As those deployments by this customer have grown, so have our ability to recognize that ratable service and support that we're providing them.
That also factors into the SaaS ARR growth. When we talked about Wi-Fi 7 and wireless coming back, we've had two quarters in a row growth in SaaS and wireless. Those are some of the factors that are helping our SaaS ARR. In general, we also have some customers that purchase from us on an as-a-service model. There's some hospitality or casino customers that just like to match their cash flows of what they're buying from Extreme and how they're generating revenue from their clients, from their patrons. There's an opportunity there. All of these factors are driving the recovery and resurgent growth in SaaS ARR.
Good stuff. Anything else, Stan, that we should really, we only have three minutes left, but I didn't ask or that you really want to convey yet?
Yeah, I mean, I think the other thing is we're obviously with these models in place, and we saw it in Q4, we're generating a lot of cash flow. On a run-rate basis, our EBITDA exit in Q4 was about $200 million in EBITDA. That gives us a lot of flexibility. We continue to buy back shares. We've done that for a number of quarters now, just adding a lot of value to shareholders from that standpoint, trying to offset dilution. We have a lot of flexibility in terms of our balance sheet of how we add value and how we deploy our cash.
Dan, I mean, do you think you can grow revenues double-digit with everything we've kind of talked about? How should we think about where EBITDA margins could be five years from now or longer term?
Yeah, nice question. When I think about double-digit growth, it's certainly possible for us to achieve total company double-digit growth. It's certainly within the realm of our opportunity. We haven't forecasted that for the market. We haven't guided that in FY2026. It's still early days. It's still Q1. We have a lot of opportunity, but obviously, there's still the macro environment to take into account. We didn't really talk a lot about tariffs. Right now, we're exempt from tariffs. That could always change. That would require us to react to that market dynamic. Our remedy for that would be to raise price. That also factors into our outlook. The part of our business that we feel really good about double-digit growth is subscription and support. We feel good about double-digit growth over time there, and certainly, strong growth prospects within subscription.
If you look at the business longer term, there's no reason why we shouldn't have a 20% operating or 20%+ EBITDA margin. That's something that we talked about several years ago when we had our Investor Day. We've got another Investor Day coming up this November. We'll be able to talk to investors about how the business is positioned for growth and what kind of margin profile we should expect going forward off of that.
Stan, have we locked down the date on that? Is it in-person or virtual?
It is, yeah. It's up on our website. We're welcoming people to NASDAQ marketplace November 10th about noon time. November 10th, which is a Monday, we'll see you and investors that want to attend in person at the NASDAQ MarketSite.
Great. With that, I think we'll end it. Thanks everybody for joining. Stan, thanks so much for the education. I really appreciate it. We'll talk soon.
Thank you, Tim. Thanks for having us.