Good afternoon everyone. A little loud. Good afternoon everyone. For the next section, we have Calix CFO, Cory Sindelar. Cory, thank you for being here. Appreciate the time you're spending with us. Maybe just starting off, the migration to Calix 3.0 was clearly a massive undertaking, and congratulations on getting it done. Now that we're on the other side of this and customers are seeing what the AI native platform can do, how are those early conversations going about Calix One adoption?
About a year and a half ago, we made a determination that we needed to re-architect the platform. That culminated in the first quarter where we migrated off of AWS onto a Google platform. The benefit of us moving to Google platform was not only to improve the scalability of the platform in terms of cost, but it gave us the ability to enter into some new markets. Part of which was to help us to do a private cloud to go into large customers, to go international. It also enabled us to bring AI functionality into the platform and bring automated workflows to the service provider marketplace. That's what just happened. In March, we finished the migration from AWS to GCP.
As you know, we release our software every 91 days on the second Tuesday of the second month, which happened to be last week. We've done our first major release with the new cloud platform, and we are now at that point engaging with the early customers, those that had already bought the three clouds from Calix. I kinda joke around saying, "These are the customers that are all in on Calix. If we built a margarita machine, they would buy it." These customers are now taking advantage of the three clouds that they had in the Calix One contract, and we are working with our customer success teams to go demonstrate the value and the productivity enhancements associated with the agentic workflows.
It's going in a very controlled manner, meaning we're going in and measuring kinda where their productivity is today, and then as they deploy the new functionality, what improvements are they seeing. That's kinda where we're at. Still very early days in terms of bringing it along. As you see, as we build to our show ConneXions in October, more and more of those outcomes will become more visible as we do press releases and share that with the public.
Got it. Maybe the flip side of this is one of the debates we hear from investors is around the defensibility in a world where AI tools are becoming more accessible, AI disruption, right? You've talked about security trust and half a billion workflows or proprietary data as is your moat. You know, maybe just extrapolate on that a little bit and expand, like, what is it about the way you've architectured the platform that makes it really hard for someone else to replicate, even as AI tools become more commoditized?
If you're gonna allow agents to actually work autonomously in a network, you're gonna have to construct them in a way that they don't hallucinate and that they are in a controlled manner. Building that trust and security layer puts the provisioning around them, and then you are using the collective data collected from our 1,200 customers to train the LLMs in order to allow them to take action inside of the service provider network. The platform is obviously have to bring context from other data sources. Through A2A and MCP, we're actually pulling all that data together, combined with the knowledge of our customer activities. That allows us then to provide these agentic workflows in a way that don't hallucinate and get to the outcomes that we're looking for.
No, got it. Maybe another aspect of this is you made the decision that customers can't access any meaningful AI features unless they sign up for Calix One.
Right.
Collapse all their contracts into one. You know, maybe can you just give us an update in terms of how customer discussions have been trending since that decision? How has feedback been on that front?
Yeah. Customers do not have to go to Calix One out of the gate, right? They can still purchase our clouds on an individual basis, and maybe I kind of explained that just a little bit. Prior to Calix One, we sold our cloud offerings on a persona basis. Think about it as a job description. If you're in the call center, you're using Service Cloud. If you're in operations or a network role, you're using Operations Cloud. If you're a marketeer and you're doing the commercial arrangements and you're looking to bring on new subscribers, you're using Marketing Cloud. All of those solutions were oriented to a role, to a person. If you're using agentic workflows, it's really a horizontal workflow. It's touching all aspects of the enterprise.
To get to that, you really have to have all three clouds. It makes no sense 'cause, you know, if you take an example of any of these workflows, it's gonna touch every bit of those different three clouds. What we've said is as we migrate forward with these agentic workflows, we're gonna sell it as Calix One, which is taking all three clouds, putting it together into a commercial model that then looks at it as all subscribers are equal, and we can move forward. It doesn't preclude you. You can still buy it the old way. Those customers that don't see the value yet in terms of agentic workflows, can still buy the old way. For the ones that are early adopters, it was easy. They were very happy to sign the Calix One contract.
They're looking forward to seeing the value being created through agentic workflows, and we're in that DV proving stage at the moment.
No, got it. Maybe if we think about 80% of the base that does not leverage all three clouds today, how are you thinking about them transitioning then? Is it more going to be, at least in terms of your viewpoint to date, do you expect them to move to Calix One as they think about procuring another cloud? Or do you see them doing more of the piecemeal till they get comfortable and moving to Calix One? I have a follow-up there.
The way we wanna monetize the AI is not through selling more clouds. That ultimately is not the goal. The goal of Calix AI is to increase their productivity so that they can do more and/or do things faster. One of the issues we have with service providers is their ability to do more things. Calix has been able to deliver a lot of different functionality and capabilities, and the way at which they consume it is gated by know-how and/or capacity. The promise of AI is to unlock that, allow them to do more, and add more subscribers at a faster rate or introduce new services that they aren't today. For us, that's the way we monetize it. 'Cause all of our operating system, platform, managed services are on a per subscriber basis.
The quickest way for us to drive more software revenue is to increase the rate at which they're adding subscribers, not so much selling that third cloud.
Yeah. No, makes sense. Maybe as the follow-up to this is, like what's the milestone that investors should be looking at? Is RPO the right number to be looking at? As a second part to that question, how should we think about the cadence in terms of a potential inflection.
Yeah.
In that milestone?
Yeah. I, out of the gate, RPO is the tell-tale sign. It's probably a leading indicator for the progress on our software contracts. You're also gonna then see it in the Service Cloud margin line as we make more progress there. You know, from an early day perspective, and we're still now building and proving out to folks the benefits of the AI through the DV creation, I expect, and like I said on the last earnings call, we expect that RPO to re-accelerate in the second half, particularly as more of those proof points come into the existence and being shared with others. Ultimately, it leads up to our ConneXions show in October down in Vegas, where we typically do very well with our customers.
Fourth quarter, I would expect to be a big quarter in terms of signing contracts, culminating kind of in that experience.
Oh, got it. Maybe just on the pricing side, like for the average customer who's going from one or two clouds to Calix One, like or help us frame how we should see the uplift look like.
Yeah, I mean, each one of our offerings is fit into a price volume curve, so the more subscribers you add, the better price you're gonna get. Generically, I'll speak kind of at a high level to give you a sense for what you're talking about. Let's say a cloud's $0.50. If you had two clouds, you're paying $1. You wanna get the third cloud, you're at $1.50. A new sub's coming in at $5.
Yeah.
Like I said, chasing the $0.50 is really not what you're after. You're looking for bringing forward the $5 per sub forward as opposed to trying to sell that third cloud.
No, got it. As you bring in new logos, does the bundle approach actually make the value proposition clearer for them? Or does it feel like a bigger initial commitment?
Haven't got to that point yet. Like I said before, the commercial arrangement is such that they do not have to buy it all at once. If they aren't convinced that agentic workflows is for them at the moment, they can start wherever they want to. They can start with just deploying our premises systems with EXOS. They can start on the access side with AXOS. They can take one of our clouds and start there. Inevitably, we aren't gated by them having to buy the whole package initially. They can start where they feel appropriate. You've seen that in some of our prior letters where we try to outline for folks those platform journey.
You typically see that they will start with Support Cloud. Ultimately, as they move forward from Support Clouds, they are then going to add managed services onto them. We would expect that a customer can come in and start their platform journey at any point.
Yeah. Got it. Maybe changing gears a little bit here. You know, when we talk to investors, the Tier 1 opportunity is one of the more exciting parts that we're hearing now, from the investors that we speak to. You know, you mentioned expecting to land an anchor customer this year. Can you help us understand what that engagement looks like? You know, is the entry point more likely through an MDU or small business where the barriers are lower, and then it expands from there?
You know, we were moving forward with our Calix One platform and agentic workflow. Michael went off to Mobile World Congress in March and thought that conversation would gravitate around the use of AI and workflows with those large customers. His takeaway from coming back from Mobile Congress was that these conversations with large telcos inevitably move to MDU and small business. That's the pain point that they are experiencing today. We are extremely excited about the opportunity to develop our MDU offering into the marketplace. We see it as a significant TAM expansion to the work we're doing today. It's another $10 billion market TAM. We see that the profile of that business very similar to the existing business inside of the company with our regional service providers.
It's a highly fragmented market, where it's being served today by enterprise solutions coming down, and our approach will be bringing a consumer-led managed service offering from the bottom that is focused on tenant subscriber experience. It's approached by the same way in terms of a sales force, in terms of a direct model. We think we have the ability to double the size of our company by approaching MDU, and the added kicker is, by approaching that market, is that we can then also service Tier 1s. If that's the way we go into a Tier 1, as Michael would say, we're going in in the side door, solving a pain point, that's what we'll do. Our platform, once installed, makes it very easy then to extend anywhere else inside the enterprise.
The fact that we solve an MDU pain point, our technology stack is fully stood up. They can then, by definition, through our platform, move into small business. They can add, you know, outdoor Wi-Fi. They can add all kinds of things in matters of days and weeks, as opposed to months and years, because there's no incremental IT work involved once our platform is put in place. That's our approach. We've demonstrated that in a couple locations where we've been that easy button for large service providers, and as long as we continue to show that we are the easiest way forward, I think we'll have the ability to penetrate these larger telcos.
Maybe can you just touch on the differentiation versus the competitive market? It sounds like there's less of a rip-and-replace motion if they go with you and implement it, but I guess why are they displacing the existing, the existing incumbent?
In MDU specifically?
Yeah. Yeah, exactly.
Yeah, exactly. 'Cause it's, again, the way that market's been put together, right? They are serving it with enterprise-class technology being pieced together. Our approach has been taking the residential package that has been consumer-oriented and bringing it up with our anchor customers like Zentro, who's that's where they started, right? They were serving the MDU space. They were looking for us to come in and provide them a little bit more functionality to round out the solution. Ultimately, we can provide a solution that is managed, at a lower price point than what is being offered today, and through the management, give that tenant experience a better option. You might ask yourself, "Well, why today?
Why are you approaching MDU today, and why didn't you do that two years ago or three years ago? The answer is, the MDU is a microcosm of a small town, and it takes all of the technology that we have developed over the last, you know, 10 years to bring it to bear into the marketplace. A community, you know, in an apartment tower, for instance, has our apartments. Well, that's the SmartHome application. When they're in the property, they wanna be able to roam freely. They wanna go into the parking garages, the workout facilities, the tennis courts. Maybe this apartment has retail on the first floor. They wanna be able to go down to the retail shops, and their Wi-Fi just works everywhere they go. That's SmartTown.
The property's gonna have paywalls and different networks, and that technology comes right out of our small business application. The landlord property management company's gonna have its own network. You're gonna have a tenant network. You're gonna have a guest network. If it has retail, they'll have a couple different networks themselves. Managing all those networks in kinda the one cloud is what comes out of the small business application. You're putting all of that technology now together into an MDU package. It is here more so because it is there. It's now that we've put it all together to have these pieces. We have to augment it with a few more pieces in terms, like, of a landlord app for managing tenants as they come and go.
We're at the point where we can actually sell this application into the marketplace, and so we're excited about being able to do that. Today we've got 10s of 1,000 of apartments under management, and we see our ability to get to millions in not too long a timeframe. Pretty exciting. The economics are similar to our regional profile. We are pivoting 100% and trying to make a lot of inroads into that marketplace.
No, very interesting. As you think about the potential, 'cause you described it as kind of like a land-and-expand motion once you're in, now it opens up the door to other avenues. Is there low-hanging fruit there that you think is going to be, you know, first up at bat in terms of being an addressable opportunity at the Tier 1s, or is it still a little bit opaque at this time?
The two adjacencies from MDU would be then small business. After that, then, you know, we'll have to see kinda where that evolves to. Ultimately I think that that's the easiest way in, is through MDU at the moment.
Got it. As we think about kind of the broader opportunity there, like, what are the key milestones investors should be looking out over the next 12 months?
Going into Tier 1 are longer sales cycles, so the next 12 months might be an ambitious statement, given the fact that they are usually, you know, 18-month sales cycles. Ultimately it'll culminate in greater RPO growth. Tier 1s typically do not like to have their name in press releases, but somehow analysts are smart enough to figure that out, and somehow it gets out there. That's how I think you would see it. I don't know that we would do a press release in particular unless the Tier 1 was willing to share it, which they typically don't.
No, got it. Maybe shifting gears to the international market. You know, you mentioned Google Cloud partnership opens up the ability to do private instances and host in market. How are you thinking about the sequencing there? Is Europe the first focus?
Are there markets in Asia-Pacific where you'd further or you would look to? How should we think about the revenue contribution timeline?
That's great. We love our partnership with Google. Calix will be an orderable line item in the Google Marketplace, because we had talked about how GCP allows us to create a private instance within Google for our cloud. If you look at us going into a large Tier 1, they are going to have their own contract with Google. Their volumes are going to be greater than what we have. Their pricing is likely to be better. If you are talking to a CIO of a large Tier 1, they are going to want to be able to burn down that Google commitment by utilizing these workflows on their own private instance. We facilitated that. We will then just be a line item within that Google Marketplace.
The, the great news about that is the margin profile turns out to be 100% because large Tier 1 is actually buying the compute and the store and everything else that creates our paths in terms of the cloud operation from it. As we enter in the Tier 1 market, we are engaging Tier 1s around the world. It probably starts a little bit closer to home here in, in the North America market. Conversations are being held around the world with large telcos. I don't think one is more aligned with Calix than another. You know, whether you're looking at when you say Asia, I'm using more of the Five Eyes.
Yeah.
As opposed to some of the other countries that may not be geopolitically aligned.
Yeah. No, makes sense. When you think about that motion now going into these international markets and trying to kinda spread out there, how are you thinking about the associated investments to go after it?
Yeah. We're making those investments in the selling and marketing resources. It's a small footprint to start in terms of people. You're gonna have a less than, you know, 12 people that would be focused on adding those Tier 1s who have the relationships that are accustomed to.
Yeah.
In that environment.
Makes sense. Maybe just sticking on that, investment. You know, I think you spent 80% of R&D on the platform migration in 2025, which obviously constrained feature output, and now that capacity has freed up. You know, how quickly does the innovation engine ramp back up after kind of that allocation there?
That's a great question. You know, we had said about last fall that we would be ramping up forward investing in R&D to make these investments on agentic workflows. We did get the cloud migrated over, and the funnel for new features is off the charts. So much so that we're actually looking at doing monthly releases, something that we've not done before. I told you that last week was our first major release post the go live on GCP. Features are coming fast and furious. We'll do a June release, a July release, another major release in August, all the way probably through November. Sit back and watch. You're gonna see a ramp in terms of the innovation engine of Calix as now all our engineers are focused on delivering on these product sets.
Sticking to Tuesdays?
Sticking on Tuesdays. Second Tuesday of the month.
No, no, makes sense. Maybe just on the migration, you've been transparent that there were a few bumps, particularly around installer workflows. As we sit here today, are those fully resolved, and should investors think of the platform as fully stable and ready to scale from here?
Yes, that's true. I still think we're working inside the new model, and so I've explained to folks that it's a brand-new platform. We're optimizing it. Expected, like for example, the software and services margins to snap back a little bit more. I think the optimization of the current platform will take a little bit more time than that. Our current expectation for that is that we would be somewhere between last year's Q3, Q4 margin profile. As we look to Q3 of this year, we should be setting new records for software and services. Obviously, we didn't go through this pain without expecting to see margin improvement on the overall platform and scalability. We should be able to grow from there as we move through Q3 into the future periods.
No, got it. Let me just pause there and see if there's any questions in the room. If you have a question, please raise your hand.
Hey, good afternoon. Appreciate the time. Can you walk through the factors that are impacting gross margins from the standpoint of the price uplift to account for the memory offset by the not having to run dual clouds?
Where all that, kind of levels out?
I think I just talked through the service and support gross margins of they bottomed, obviously, in the first quarter, and I think we get to new highs starting in the third quarter moving forward. As it relates to memory, that's our favorite topic of the week, month, and year. We are being encumbered by increased memory pricing, no different than anybody else in the industry. We have chosen to pass that along to our customers in the form of a surcharge. We did that as it changes, we can then concomitantly pass that on to our customer. If it continues to go up, surcharges will go up. If they go down, they will get the benefit of any relief that we see in terms of memory pricing without having to change price.
Our estimate for margin impact for the year as a result of memory was 200 basis points to gross margin. Since we aren't margin stacking it, you're effectively pushing through a surcharge at zero margin. That's having that headwind to it. At the same time, we took our revenue growth profile up as a consequence for that increased pricing through a form of a surcharge. Where does it go? I, well, I would say ultimately, your crystal ball is as good as mine. I do not think we're at peak pricing yet in terms of memory. I do, based on from what we understand from our supply chain, is that it will continue up into the third quarter and then flatten out a little bit in the fourth.
We do think in sometime in 2027 that you'll start seeing additional capacity come online that actually helps up then with the supply. At the same time, companies such as Calix and others are optimizing their memory footprint, shrinking the software to help reduce that price. You know, when memory was a $1 a gig, you didn't care actually having an extra gig in your appliance. At $25, it's a whole different story, so you're gonna make that change, and you're gonna reduce that footprint. I think those supply-demand dynamics will continue, and we'll start seeing something in 2027. Ultimately, I think this plays out that we'll get to a new normal for memory. In the beginning of 2028, we'll probably start seeing the first bits of Wi-Fi 8. I see this playing out in one of two ways.
We'll simply take that new normal for memory pricing, embed that in the Wi-Fi 8 pricing, and as customer migrate from Wi-Fi 7 to Wi-Fi 8, it just goes away. If Wi-Fi 8 pushes out later in time or the new normal for memory accelerates earlier in 2027, not what I'm thinking is gonna happen, but if it did, well, then we would just simply increase the price on Wi-Fi 7 and put the surcharges behind us. This headwind that we're facing here in the short term actually unwinds itself, right? Ultimately, what'll happen is the gross margin and the revenue at zero margin will dissipate. That 200 basis points will reverse itself. The incremental revenue growth will then contract a little bit.
It'll be our goal to just continue to grow through all the other avenues we talked about in terms of BEAD and through the MDU, through the footprint expansion that we have on the business. That'll offset that slowdown in revenue growth. But that's how I see this ultimately playing out.
[inaudible] Constraints and memory pricing and other macro factors might be impacting your customers' cadence of deployment of fiber passing and fiber connecting? Is your customer set getting a little bit more reluctant or unable to deploy as much as they would because of component constraints?
No. I mean, well, let me frame that a little differently. The component shortage hasn't led to a supply chain shortage. We can still produce and supply. The increased price has an impact on what they may buy. For example, if they had a plan to go deploy $2 million worth of CapEx during the year, fiber costs are higher, labor costs are higher. We just increased our equipment costs on them. They may not be able to pass 1,000 homes. Might be 800. The competitive landscape may be such that somebody's coming right behind them, and they need to get it done, and maybe they go back to their board and increase the capital to $2.2 million. Kind of hard to say, but, you know, these price increases have an impact of some kind.
At this point, it's not because of the lack of supply. I mean, our number one priority is for us to be in a position where when our customers have a new subscriber to turn on, they have the equipment to do so and start generating revenue. Of course, we benefit then from the recurring revenue from having that new subscriber on the network. Ultimately, we would like to be in a position such that if there's an opportunity to take share, we can. We are aggressively in the marketplace, making sure that we have adequate supply, not only to serve our customers, but take advantage of the situation if we can.
Any other questions? No, I don't see any. Okay. Maybe shifting gears from there, and maybe your second favorite topic, BEAD. I think, you know, is it fair to say that the peak numbers that you guys were estimating for BEAD has come down a little bit? I'll caveat that with, I think some of the analysis that you guys were doing has changed a little bit. I guess where we're getting a question from investors is around has peak numbers come down or
Has the timing changed in terms of when deployments are occurring now?
Yeah. At the end of the day, we've been trying to caution investors not to really get into stock because of BEAD. We don't need BEAD to grow. BEAD's a nice tail breeze, tailwind, if and whenever it happens. We don't feel any less bullish or more bullish about it. Our estimates haven't changed. Unfortunately, our stock trades quite a bit with BEAD announcements and sentiment, right? Somebody says something negative about BEAD, we're being traded down, I don't understand that dynamic at all. BEAD is a favorable line item that'll happen in time. Whenever it happens, we'll benefit just fine. Specifically, nothing has changed from our estimates. Our estimates for low 10s of millions of dollars of revenue this year hasn't changed. Backlog's got coverage. I think it ships this year. That's great.
If other people have their BEAD money slipping out into 2027, so be it. I don't think that affects us at all. As we're looking out in 2027, excuse me, nothing has changed from the tops down number. What we're working to is looking at what the backlog is, how it's developing, and I suspect that it'll continue to mature. I mean, 2027 is a long ways away. They're still working through a lot of different things within the BEAD program. I said I'm not any more or less bullish on the program. I think we'll do fantastic as it goes along, and I don't really care when it starts.
No, got it. Just on that same topic, though, you've mentioned that if BEAD does surprise to the upside, you'd want to reinvest some of that into sales and marketing for MDU and international opportunities that we just talked about. You know, should investors think of BEAD upside as effectively self-funding the expansion into these new TAMs with operating leverage still improving on the core business underneath?
Yeah, I mean, that was a statement of really our OpEx model. Our OpEx model, as we outlined in our Investor Day, was one of, as we move into 2027, we will slow the growth of OpEx such that it is a portion of the revenue growth. As we get to 2028, you're gonna start seeing more of the full effects of AI being adopted internally at Calix in terms of productivity. The areas that you're gonna see us grow is we're gonna have to build a little bit more into the sales capacity to attack the two new markets that we're entering into, but it's in there in the context of inside the revenue growth. Ultimately, in 2028, we think the operating margin, the operating margins get into the 20s. That OpEx itself rolls down, drifts down to a 40 or less.
The margins stabilize and get back into the 60s overall. I think that's ultimately where we get to in 2028. BEAD is just another source of revenue, and it just fits into the model. Like I said, I mean, BEAD's a tailwind, a tail breeze maybe, in which case we don't really need it to grow. It's just, it's there. When it happens, we'll do well.
Got it. Then maybe just on satellite technology as a competitive threat, obviously is pretty discussed topic. Maybe in the last 30 seconds here, help us think out through how you guys think about that narrative.
Starlink is great. I think it actually helps. For some of our customers, they need a competitive push to help them move along. Why would you try Starlink? You have a crappy internet experience. If you enjoy your internet and satisfied with your internet, you aren't calling Starlink, right? You take somebody like Tombigbee Fiber, who's down in Mississippi that has 30,000 subscribers with a 92 NPS. They aren't calling Starlink. However, if you're a service provider in somewhere in rural that you're the only game in town, and you haven't been servicing your service providers, and you suck, well, yes, you should have your margins compressed. You should have an ARPU battle. You should have let Starlink come in in there. What should that do?
That should invite them to come to Calix and talk about how you can actually improve your game, focus on NPS, focus on the subscriber, and keep Starlink at bay. It's really incumbent on the service providers to provide a better service, and if they don't, gladly that Starlink is there to provide that competitive nudge to make them go do something. The only other place we see it is really around a backup play. Kind of an example of where my boss lives in an affluent community, and they had an eight-hour outage. What do you think that affluent community is doing? Out there signing up Starlink as a backup alternative because they can't live eight hours without internet. It doesn't matter when you're coming from it that, from that perspective. Starlink's a good thing.
We embrace it and know all the internet traffic is not going to space. The economics just don't make sense.
No. Thank you, Cory. I think we're at time. I appreciate it, and thank you everyone for joining us today.
Thank you, everyone. Thank you, Jeff.