Calix, Inc. (CALX) Earnings Call Transcript & Summary

Good morning, everyone, and afternoon for those who are joining us from the U.S. East Coast. Welcome to the Calix 2020 Investor Day. My name is Tom Dinges. I'm the Head of Investor Relations here for Calix. I want to thank everybody who is joining us here in person in San Jose as well as those of you that are joining us on the webcast. We really appreciate your interest in Calix. We've got a lot of great content for you today. Before we get going, I do have to go through a few housekeeping items with you. In the presentation today, we will be making forward-looking statements. In addition, we will also be disclosing or discussing non-GAAP financial information. I'm not going to go through the full disclaimer for you. They're available on our website. They're also available in all of our filings with the SEC, both quarterly and annual. And a full reconciliation of all the numbers that we're discussing today is also available on our Investor Relations website. Before I turn it over to Carl, one other housekeeping item just for all of you, the full slide deck will be available shortly after we finish the formal presentation today. So we've got a lot of great content for you today. We are going to take a couple of breaks, going to have a lunch break here as well. I'll come up and announce those as they come up. We're going to finish the day today at 2:00 sharp. And for those of you that are here, we'll have a product showcase open out [ here. If you guys could take a look at the lobby, you'll see ] what we've got there for us. So with that, it's my pleasure to now introduce Calix' CEO and President, Carl Russo. Carl?

Thanks, Tom. Good morning and good afternoon to those of you who are on the webinar and East Coast. First of all, let me say thank you to those of you who are here and braving the coronavirus headwinds. I suspect that we will have a sold-out webinar because there were a lot of folks that were looking to attend and in the last moment said, "Oops, coronavirus is bad. We're dialing in." So for those of you who are here, appreciate it. Hopefully, we won't get in that hotspot that shows up on the news. But again, thank you. The goal for today -- for all of you is there are 4 key points that I'm going to make sure you all understand. The first is that we are positioned in front of not 1 but 2 secular disruptions that are going through the industry. The second is that we have the right technology to advantage us and our customers. In our case, we call them platforms. And obviously, our belief is that this will enable our customers to succeed in front of these 2 waves of disruption and build different and more valuable business models. Obviously, our belief is that they will share a percentage of that increased value with us in some form or another. The third point I want to make sure that we're making is that we have the right people, the right team in these companies to take advantage of this. Many of you are aware of the diversity that exists in Calix, and we've been building over the years. You're going to get a chance to see the key leaders in the company today, and I would encourage you to dig deep because I believe we have the right team to take advantage of us. And the last point I'll make is when you take #1, #2 and #3 and put them together, if we execute against this vision then this should ultimate roll out into a financial model that continues to build value that all of our stakeholders can benefit from as well. So those are the 4 points you're going to continue to hear me come back up and speak to. And I would ask that you mark those down and see if in fact we achieved that mission. So without further ado, disruptions. They don't come along often, but they do come along. And I thought it would be useful to give you a sense for disruptions in the past in this industry. I suspect a few of you are as old as I am, but some of you may recognize the things that are up there. When I grew up, I grew up with this. Let me just see a show of hands. How many here grew up with that? Michael you grew up with that because you're in Canada. Paul's was smaller. There it is. That's a TV with antennas. By the way, that's really cool for me and for those who are old enough. We have this new thing called the Internet. Have you heard of it? It's free. You get the Internet for free because it's supported by advertising. What a business model. Remember that one? It was free, and it was supported by advertising. And now for a word from our sponsors. When I grew up, I had 3 VHF channels with pretty good reception. And we had 3 UHF channels, which weren't so good. And we would occasionally orient the rabbit ears and go accordingly. And cable came along with a non-free model that ended up disrupting a free model. Why? Well, there were 2 many reasons that cable disrupted the original free-TV model. One was better quality. So they promise to, in fact, get rid of the antennas twirling and deliver a much higher quality signal. And the second was, they promised to do better than those 6 channels, and they would win a municipal franchise. There were families like the Cox family or the Roberts family, maybe these ring a bell. They went off, and they built little cable franchises. By the way, they would negotiate with the municipality. And the municipality would say, okay, we're going to give you the franchise, you have to bill to everybody, and you also have to give us a free municipal channel, which if you ever watch the free municipal channel, there was some fun stuff on those channels. But that's what they did. And this disruption started, by the way, in the 60s. And so not 1, not 2, 3 decades to traverse through the industry. Along the way, $1 trillion of capital was dislocated, growth torn apart over the course of this industry. That's a disruption, and it was a disruption at the physical layer. It was driven by a different physical layer, cable, coax. So that's one example of a disruption. Here's a second. Yes, I have one of these too. The phone industry has ultimately been disrupted by the cellular phone industry. How many of you had that Motorola phone in your career? Come on put the hands up. Okay. What did you hear way back when the cellular phone industry is always going to be a specialty industry, which -- by the way, most disruptions start with vertical markets. It's going to be a specialty industry. Why? Because people are not going to accept the crappy voice quality that exists on these cellphones, plus the fact if you go too far away, you can't get connectivity. It's just not going to happen. And that was an early prevailing wisdom of the existing industry. What happened? Well, in fact, utility, the ability to carry it with you, won out over quality. And then as utility continued to win out and more subscribers came on, guess what? The service providers figured out how to do what, raise the quality. And so today, through -- remember, we used to call this cord cutting. Today, cord cutting means something different. But back then, cord cutting was getting rid of phones, but many of us still have, in fact, a plain old telephone service at our house. This really started in earnest in the late '70s from a cellular standpoint and guess what? Not 1, not 2, 3 decades and trillions of dollars torn up, built, et cetera. Anybody here remember McCaw Cellular? Craig McCaw. Craig's ex-wife is my next-door neighbor in Santa Barbara. She did okay in the divorce. This has now led to smartphones, which is an intersection of some of the things we just saw, trillions of dollars. How about this one? Now we're coming more modern, IT versus connection-oriented protocols. Anybody remember this one? How many of you are -- have a technical background? Michel remembers it. Michel is putting his hand up because Michel by the way is Michel Langlois, who prior to being here used to run Junos software at Juniper and prior to being there used to run IOS software at Cisco. So he loves this sort of thing over here. He used one of these packet disguise. So what was the issue? The issue was this is a nondeterministic protocol, you're telling me that we're going to take our valuable data from like financial firms and just put it out there in a packet, and it will find its way over to the other side. That's for scientists sharing files, that's not for real applications. But we know what happened. Over time, the facility that was borne of a level-playing field and ultimately immediate independent transport layer, because IP doesn't care whether it runs on twisted pair of copper, fiberoptics, coax, wireless, satellite, doesn't matter, that utility allowed the performance to continue to come up. And every formerly "It will never work on IP" collapsed into it like the event horizon on a black hole, just slowly but surely disappeared. So where -- this doesn't exist anymore except in the battles of some of the service providers that still have SONET rings and ATM, but it has receded to the point where it's irrelevant technically and long since irrelevant from any value standpoint. And not 1, not 2, 3 decades and trillions of dollars torn up and created in that disruption. So these disruptions occur, we have a tendency to not remember them. But if you think back, most of you have gone through these and live through them and had a sense for the words you hear at the beginning of a disruption, what happens at the end, where venture capitalists invest, where public investors invest, the cycles, et cetera. So hopefully, that gives you some sense for what's happening now. And the first disruption that we're focused on is this. And you say, wait a minute, that's not a disruption. That happened in the '80s. I used to have like a Wang desktop connected to a DEC PDP-11, and they were connected over LAN. Sound familiar? Well, in fact, what's happening today is exactly client server. Every single device that you're carrying around has 4 things in common. It has an LCD screen, it has an IP address, it has a wireless PI and it has a control mechanism. And then if you are typing right now -- I'm sure not recording my words, of course, send note to your friend, you also have phones that have a touch screen. You also have Alexa or other things. You have cars. By the way, how many of you have an electric car? I don't have electric car. It's more like a device than it is a car, full screen. What about an HDTV? How many of you have an HDTV? Why is an HDTV any different than this? Just expand it. It's the same, wireless physical interface. Makes sense? That's what's going on over here. And by the way, if you assume a LAN connection, a broadband connection, the data that you're displaying on your device isn't really resident in the device. Where is it? It's over here in applications or content that exists in the cloud. That is a client-server model, that is not a peer-to-peer model, nor is it a stand-alone compute model. And guess what? Just like client server, there is an infrastructure in between. And this is the second disruption. So why is this first one a disruption? This first one is a disruption because guess what? Remember, that video, coax. Remember the cellular phone. And when you had cellular phone, a new feature would show up. There'll be a new feature. You could do call waiting or hook flashes or whatever. Did you go buy a new cellular phone that had that feature? No. It was rolled out to you from the service provider. If you've got a new channel on coax was it because you decided I'm going to go to YouTube and look at something different? No, they added a new channel, they added new content that was rolled out to you. By the way, occasionally with your cable supplier, they got into a contract with some over-the-top content player. And you wouldn't be able to get that content for a period of time until they resolve the payment terms. Remember these things. What's the point? All of the power was in the providers' hand, not in your hands. Now it's driven by the subscriber, you choose. So this client-server model with an assumed broadband connection has put the power in the subscriber's hands. What's happened on the infrastructure side? Here's the second disruption. The second disruption is not coax, not twisted pair, it's not cellular. It's a combination of 2 things, a wireless physical interface out here that's talking to all of these wireless devices that you now have because who wants to have a cable right behind you when you drive down the road. But by the way, who wants to put a new HDTV up and knock a hole in the wall to go string cable to it? So literally, if you think about your home, virtually, every one of those devices is connected wirelessly. So there is an antenna over here that's really important, that's providing that connectivity. And the most efficient way to move those bits back to the data center is over a piece of fiber. So you have a virtually unlimited physical carrying capacity combined with the untethered nature of wireless. This infrastructure is going to eliminate every other infrastructure. Copper, coax, I don't care what it is, it's all collapsing into this infrastructure because it is the most efficient physical layer. So there's 2 disruptions. The client-server model, driven by the devices that we're all carrying around connecting to the cloud, enabling the subscriber. And at the physical layer, you can build an entirely different network. What's the moral of the story? Here's the moral of the story. By converging this infrastructure, you will see in the course of this morning, different examples of this converged infrastructure. How you build this network enables you to build an entirely new network with entirely new attributes, costs, performance. Stated simply, all services are with the lowest cost per bit per mile infrastructure. And if you can start to think about how do I address these empowered subscribers, I can now start to figure out how to build not a new network, but a new business model. And for most of you, you are aware that service providers don't necessarily approach this from a marketing perspective because for 30, 40, 50, 60 years, they've been a legislated Five 9s company. They roll out services. This is an entirely new opportunity, which you're going to hear more about today as well for us to build those new business models with our customers, and then, in fact, share in some of that value. So with that, this is the theory. So you've seen previous examples of disruptions. You can see the 2 disruptions that are converging in the access infrastructure. And so this is what we are pursuing. So this is the theory of what we're pursuing. Now you're going to see it applied, and I'm going to ask Michael Weening to kick us off in the application section. So Michael, all yours.

Thanks, Carl. So Carl talked about the 2 disruptions that are happening, that converged infrastructure, everything crossing into fiber and then the shift in power that's happening on that empowered subscriber. I'm going to pull apart this empowered subscriber a little bit more because that's the life that I live, which is actually spending a lot of time in the data and out with customers and how they address that subscriber. And it starts with what he said, was if you think about where we came from, it was a broadcast-only world, right? What the subscriber got in the past was you took whatever the service providers provided you. If you've got yourself a new high-definition television as I did back in the early '90s, right, the only place where you get content for it was from my cable provider and that would drive my behavior and what I would buy, right? I bought their broadband. I bought all those things. I really had no choice. I think it's safe to say that as we go into this next stage of the world, choice is something that the subscriber's overwhelmed with. There is no shortage of choice. I have all of these different choices that I can make at any point in time, and it's a very complex market. And so as we start to think as the balance of power goes to subscriber, then we're really determining who I pick and that will determine who is the winner and the loser in the future. I think the best way to talk about this empowered subscriber is actually also talk about the changes in their behaviors, and I'm going to talk about my sons. I have 2 boys who are up in university. They share a condominium. And if I think about them from a persona point of view, so if I was to actually take the data and look inside that condo at what goes on, there's really 2 personas. The first one is a gaming persona. And that gaming persona is all around "I want to have it fastest, lowest latency network possible." It's a reason why every month, despite my consciously arguing about the fact they don't need 1 gig, that they have a 1 gig network, right? They don't care about IPTV. They don't care about all those kinds of things. They want to have the lowest latency. It's not about downloading a 4 gig file or a big file, it's actually about I want to have milliseconds in between, and I want 1 gig or I want 2 gigs or I want 10 gigs because that's going to be the determination between whether or not I win my game, right? So that's the first persona that's driving the behavior of who we select from a network point of view. And believe me when we selected the service provider it was all about them actually figuring out who had the lowest latency network. That was the decision that they make. Then the second one is the consumption of content. In the past, content would drive a lot of our behaviors. If you actually were a great content player, you could get a lot of subscribers. Those days are gone. Or if they're not gone, they're well on the way to gone. I look at my children. So again, we go and buy them 1 gig service. What did they throw in for free? IPTV. They have 2 big screen LCDs in that condo. They lived there for 3 years, and they've been turned on twice. The only times they've been turned on is when I was there and stayed there for the weekend. That's it. The IPTV boxes are disconnected. They -- and I keep trying to return them, saying can I get a lower -- can I get a 1 gig fee and a lower fee if I return them? And they keep saying, no, you can't. You have -- it's a bundle and that if you want to lower cost on your 1 gig network you actually have to take IPTV. They'll come to our house, and there's a big 80-inch television there, and they sit on the couch with their laptop in front of them watching YouTube and Netflix and all that kind of stuff. And I keep saying, you know you can watch that on the 80-inch television that's literally right there. And they say, why would I do that? Because that's the second shift. They want their content where they want it and they don't care about [ quality ]. They don't care about that it's -- I cared about, hey, I got an 80-inch television, I want to see this great HD content. They don't care. It's actually about I want to be able to move around the house, I want to go back to my bedroom, I want to go sit in the backyard, I want to have an amazing WiFi experience. So that always-on experience anywhere that I go is really important. And those behavioral shifts are now putting the power with the subscriber and completely changing what makes a successful service provider. So as we've been thinking about how we build our company and the products we've been developing over the last 10 years that have consumed hundreds of millions of dollars investment, it's actually changed the way that we're operating as a company. And we think that in the future, the CSP that's going to be successful -- there's going to be 2 CSPs. And this market is bifurcating, you have to make a decision. You're either going to be one or the other, you can't be both. And so the first one is the wireless wholesaler -- or sorry, the wholesaler. I have to make a decision that if I'm not going to own the antenna in the home, if I actually don't want to be in there and own the subscriber, then what I'm going to do is I am a wholesaler. Now a lot of our CSPs have actually made the determination to be a wholesaler without actually consciously making that decision. They've actually decided they're not going to go into all these kind of things, and Carl has been preaching to them about that for about 10 years, right? But that you actually -- if you want to be a successful wholesaler, you have to make a conscious choice. And that means you better build the most operationally efficient fiber network that's possible, that's to run autonomously, all those different component parts. And we see that market happening a lot in Europe and other places, it's called open access. I become a wireless wholesaler -- sorry, a wireless wholesaler -- becoming a wholesaler, I have all these ISPs on top of me, but I am building a wildly efficient network. And to do that, I'm going to gun for lowest cost per bit per mile. That is my goal. And so we think that to win, there's really 3 elements to it. First, it better be always on because that's your value proposition. You have a highly reliable network. Just take down -- let's go back to those boys, go take down a gamer and have an up and down experience. You're going to lose that customer. So it better be very reliable. Second thing is simple to operate, hopefully autonomous, run by neural network with no people at all. You have to make it as simple as possible so you can get your OpEx way down. And then third, I have to be able to update it as the market changes and enhance it's -- at a devops pace. So that's the first one. I make a determination to be a wholesaler. Now the second choice is that going to be a -- someone who's actually focused on the experience that they're having, and I'm going to go compete head-to-head. Whether it's against the over-the-top players or others, I'm going to compete as a retailer. So first, I'm going to have a great network. I need to do that. It's really important that I still have lowest cost per bit per mile. Even as a retailer, because I want an access to that network, I want to be reliable, and I can make some great margin there. But then on top of that, I want to build experiences. When we think about experiences, we think success really comes down to 3 things: real-time subscriber insights. It is really important. We're going to talk a lot about data. The company with the best data wins. You have to have insights into what's going on at home. Let's go back to those boys. So how successful are you going to be if you build out a parental controls application and you fire it into those -- my 2 sons in the condo? Not very. They're never going to pick that up, and you're probably going to annoy them, right? So instead, you have to understand that it's a gaming, you have to build your offerings, you have to attract them. Offering them IPTV clearly doesn't work either, right? So figure out what does that subscriber need. The second thing is you better have cool technology. The days of the service provider being 18 to 24 months to get technology, that's a failing business proposition, and that's the reason why people are going to Best Buy. So you better start getting cool. And then the other one is build out those great experience so you can attract and keep your customer in the long term. And from a product point of view, that's what we've done. So as you look at building out that lowest cost per bit per mile network, that's the Intelligent Access Edge. And as you look at taking that and then layering on top experience, that's Revenue Edge. For the next couple of hours that's what we're going to talk about. We're going to walk you through our business strategy and where we're going as a company and how we meet those realities. And so the first is the Intelligent Access Edge. So let's pull apart a little bit this lowest cost per bit per mile, right? If I think about that, what does it really mean? If I'm a lower service provider, what I'm very focused on is all CapEx. When I came -- I was at Salesforce, I was at Microsoft, I was at Bell Canada. And when I was Bell Canada, it was always about CapEx, CapEx, CapEx, right? And that's what most service providers are taught, CapEx, CapEx, CapEx. The problem is if that you actually just focused on CapEx and don't think about the long term, especially in this next-generation network you're missing out for the cost of ownership, right? You're missing out how do I actually get the lowest OpEx possible so I can run this as an autonomous network and be wildly competitive in the future to -- regardless of what happens. So for us, as we think about lowest cost per bit per mile, how do we create this OpEx model that, yes, give a CapEx component, but you're really focused on what's the long-term value? And building long-term value really comes down to those 3 attributes: always on, super simple to operate and then it can be altered at a really quick pace. And to talk about what we built, I'm going to invite Michel Langlois who is going to walk you through the intelligent access Platform.

Thank you, Michael. Okay. Let's talk about the imperative for the business. So what we're going to try to do is to explain you what we built, why we built it this way and where are we with the result. And we're really excited about this. So 3 imperative for the business. So let's translate that in product. Always on, and I remember Shane when we started that, that's probably the attribute which is the hardest to do in the industry. You try to build a network which is -- has no failure. We tried to be able to service that are always distributed, never interrupt. And you have also to build those services in such a way that if you try to upgrade the system, what we call the maintenance window, there's no more such thing moving forward. Because if the services are going to be always up, there is no more maintenance window that you can turn off services and upgrade the network. So we have to look at this in the context of service provider today. Simple to operate. Three dimensions you'll see across the whole presentation: simplify the network, which is an element and reduce the complexity in the architecture of the network; simplify the operation, essentially automate, reduce the human factor for the result; and simplify the business. Is there better ways to look into delivering those services than established thinking that date years ago? Last one is for the service provider to succeed, they have to catch up the pace of innovation. If you really look into a service provider, they define services, they validate that services, they integrate the services and then they roll out the truck and turn on the services. It's a 2.5 years cycle on average. We have to reduce, essentially compress everything, time to test, time to integrate, time to deploy, and that was a mission we have. So in order to do this, we decided to build a new platform. We didn't find it in the industry. We look everywhere from the past, from the cloud, from compute space and decided that we have to invent there. And that was a journey that started 6 to 7 years ago. Key principle of that platform. For Calix to innovate, we wanted to be completely transparent, agnostic to the type of silicon or electronic that exists. We don't build these -- we don't build those components. So what we wanted to do is as the technology evolve and evolve fast and faster, we wanted to be able to pick up anything from any vendor, from any type of technology, from any type of supplier and reduce the time to build those products. Because we're guilty also. It takes years for us in average in the industry to build anything. So we wanted to compress that. We built this platform in such a way that we could instantiate capability across any type of system, will it be centralized, distributed, disaggregated, stand alone, you name it. We didn't want to build anything that constrained the capabilities and access because of our setup. That's the key element of AXOS. Third aspect which is really interesting in the simplification of operation, the tricky thing was to say you have a service layer. What is a service? It comes with a definition, what it's supposed to do. We took an approach which is standard based on an object module. So we try to standardize everything in a way that you could mix and match vendor, if you want. This is the same ways to define the services. Second thing is you have to take that definition and basically define it as a workflow, how you are going to turn on the services. Every time we have built services we have looked into instead of 20-step to do something and you simplify, reduce to a really relatively few steps to turn on the services because it's service activation you're trying to do. Third is once you have that definition, and it's automated, don't touch it. If the system evolved, get replaced, you never have to go back there and retrain or relearn a new syntax or a new type of interaction, if you want. It's done. Top of it after that is how do we innovate and accelerate the pace of innovation. We have a few choices. We could decide to build everything ourselves, which limits us from an investment, or partner, obviously leverage the industry. And the way we build the software was we have access to all those sources. We can drill new capabilities inside the container. The container essentially are well defined. The advantage of container, you can virtualize for scale after that. Or you could do it in such a way that inside that container the key element is to limit the failure. If there's a fault and there's spots because we're still writing software, in reality, it's contained. It doesn't impact the rest of the system, doesn't impact the service. So how do you do this is by containment and failure are contained. And what's happened -- what the advantage of that model is the module that evolved that you need to change, you need to replace, you need to substitute, you can do all those operations, take out the block and replace without shutting down the services of the system. That became essentially the AXOS software and the foundation as a platform. So let's go to the next step now. We have a platform, we created a foundation. It's called the Intelligent Access Edge, and it has a set of attribute or categories, if you want. And it's similarly to what Shane has done also on the Revenue Edge. You'll see some kind of common thinking, a common framework. For us to bring capabilities in the network, it starts with system. And I'll talk a little bit about the design philosophy and the offering in a few. With those systems come capabilities. Now what are we trying to do? Reduce the time to test, the time to integrate, the time to turn around the services, the time to diagnose, the time to restore. All of this is up in tools that we have built for service provider, not only thinking they're engineer or trying to leave the reality that they are on the ground. So basically build capabilities for them, not just for engineer. On top of it, we look at the network in itself. And we say, service provider are kind of interesting for us. They are the only person, only entity that exist that could see everything about all type of traffic, all type of usage, all type of subscriber, what you do with your session, map it to flows. Essentially -- they're essentially the aggregator of all experience because everything transit to this. So for us, we look at visibility, and we look at ways into how we can optimize the experience if we leverage all this data which is out there. This is the ultimate data-led collector, if you want. Last is how do we make them successful, and Matt will talk about it in a minute in terms of giving them the tools, the knowledge and the learning and the best practice to basically get the best result with their investment. Last but not the least is the promises of transforming the network. We look at the network and to -- it used to be a demark that we call the edge, the access edge. Next to it, there was a lot of network element adjacent to it, network router, lawful intercept, BNG, you name it. Is there better ways to bring intelligence inside so we can get closer to the unified network? And what's -- and by doing this you change the economy. So I'll talk a little bit about those capabilities in the few, but I can tell you what's interesting is when you go that [ fast ], you can transform the network, you can transform the operation, you can transform the business. It's the next frontier, essentially, build the last network you're going to lead. So that's the framing and that's the foundation. Let's talk about the system. So we have essentially a different type of system. They have different density. They have different capabilities. They have different performance. They have different features. They are optimized for the best deployment and the best placement in the network. They can reside at the central office or scale. They can reside outside in the outside plant to be closer to the service delivery, in the cabinet. They can also live in really hard environment and we called it [indiscernible], where they can be installed in a post or buried in the ground. We have people that do that and survive every aspect of physical climate, if you want. What's interesting with those system, and this is one of the key to understand how we do product here, is we build those system for scaling up and scaling down. They are the same. If you're Tier 1 or Tier 2 or a Tier 3 or a smaller operator, you have access to all that technology, meaning we didn't build different system for the different size of customer. Second thing is these system can be deployed for wireline user, MSO, mobile, next-generation service provider. We didn't build differences then for the class of customer that we have. By doing this, we have now an operating system and set of capabilities that can go across everywhere in the network and for different type of scale as well. That's unique in the industry. Also interesting, they really well optimize for fiber technology, technology of the past, technology of today, technology of the future because we've done this level of abstraction, if you want. So that's unique in the industry. Let's talk about those modules. As we're looking again into the lowest cost per price per bit, we look at this into why don't we integrate, reduce the network complexity and increase essentially the aspect of service delivery. So we have 3 modules that we built over the past few years. The first one is essentially big capability that used to live into the network rout adjacent to access and bring them in. So by doing this, you reduce network -- the number of network elements, but you also open the door to new more advanced services for our customers. You're thinking about Tripleplay, for example. You're thinking about IPTV. By doing this, we're opening the door for this. And we're pretty proud into what we have achieved over there. If you like connection oriented, something Carl was mentioning, you will be familiar to something called MPLS or tunneling services. So if you wanted to connect your access network across other networks, these are technology that you bring easily into what we call the PI router. So we took a subset of it, only the relevant pieces for access and brought it into AXOS and the intelligent edge. Last, within the segment subscriber management, and why we did this? I think we can solve the problem more locally to find better policy, enforce SLE and make sure that if they're outside the range we can correct the network to offer that level of services. So that's what we did with our software, brought to you by AXOS. Let's talk about the other aspect of the offering, which is the tools. Remember, the [ problem ], can we reduce the time to deploy, the time to test, the time to integrate and essentially isolate flow and restore services faster than ever? We have 3 [ tools ] that we did. The first one is around integration with existing operational system. So you all know service provider has system to manage equipment and their services, and they come in different flavor. If you're from wireline space versus MSO, DOCSIS, for example, or others, you want to ride essentially in the future things like SDN. There's different ways you can provision those system. We support all of that. And we do it in such a way that we're not forcing the service provider to learn a different language. Sandbox. It's interesting, Sandbox. We took a page of the compute industry. Because if you're in virtualization, everybody do Sandbox. And we said, why don't we use that? And we did that first internally in R&D because we couldn't spend enough money to build every network out there to simulate the customer. It becomes really costly, those giant lab, et cetera. So we decided to virtualize our own equipment to the point that we can go to market and test our product at the different level of scale using this environment. Then smart people say, why don't you offer that to our customer to do the same? They don't have to spend the same CapEx, and they are allowed to do it. And by the way, we just realized that the Sandbox can simulate every aspect of workflow that the customer is going to have to exercise over them. Why don't we provide that to them? And you'll see in the result of this when Matt talks about the way our customers have used this environment, how this thing changed the industry completely. It's quite -- squeezing every aspect of time to integrate and deploy. Our commitment has always been the same, provide the best tool for advanced services, diagnostic, et cetera. We're going to continue to do it as well. We have the ability now to bring to market these tool as well, which lead into probably the most exciting thing. I talk about visibility in the network. We see all traffic, subscriber, session, flows, et cetera. Traditional way that's being export all this information up there to a network management platform. That's traditional ways that you do this and you literally rely on the knowledge of those operators in their head to figure out alarm and what to do, et cetera. People on that side went into "There has to be a better way. We need to move to the new edge." And that's what created this offering that I would like to invite Matt to say -- let's talk about how you change industry with insight.

Thanks, Michel. So you're going to hear a lot about persona-aligned analytics and intelligence.

Turn your mic on.

Here we go. Hopefully, everyone can hear me. And so I want to start by talking a little bit about how we actually built this remote monitoring solution. So it's a SaaS platform. By the way, it can also be delivered as a managed service. We've worked with thousands of service providers for the last 20 years to help them deploy and optimize their networks. We've developed a lot of insights over that time period around what impact service levels, what's irrelevant, what do they need to focus on to deliver an amazing service. We took those insights and build them into a machine learning algorithm that literally scans all the data that Michel was talking about, millions of alarms and messages a day, isolates the few hundred that need intervention and then generates reports for the team. In this case, it's the network operations persona that tell them where the problem is, what's causing the problem and what do they need to do to resolve it. As we've rolled this technology out to our customers, they are seeing 90% reductions in the number of interventions that our operations team needs to implement. And because of the insights that we deliver, along with this information, we're actually reducing their network issue resolution time by 50% or more. That is incredible, incredible operating leverage. And that's the kind of promise that these kinds of analytics are bringing to our customers today. Now Michel talked through a lot of great capabilities, incredible new technology, and we recognize that for many CSPs, getting from where they are today to where they need to be in the future is a daunting task. And our goal is to help them accelerate that transformation by leveraging our expertise, the same expertise I talked about that went into the machine-learning algorithms and the remote monitoring solution. We deliver those through 3 primary enablement services. The first one is education services. They're online, on demand or in-person, instructor-led. We educate them about the technology, we certify their teams, and we make sure they have the knowledge they need to go implement and operate this new technology. Professional services. Some of our customers, they want us to actually deploy a team on the ground to work with them as they go through the transformation. We can help them plan, we can help them deploy, we can help them operate. We leverage the methodologies, the tools and the expertise to make sure that they're operating in a way that is best practices and really drives simplicity in their operations and generates great returns. And then finally, customer success services. This is where we take deep network operations expertise, people on our teams, and we literally align them to the customer as a coach, as a guide for the duration of the entire time that they run their network. They interact on an ongoing basis to share best practices. They look at how they're operating. They're looking at the data from the network, and they're constantly advising them on ways that they can optimize their operations. Success services is not a new concept. Salesforce is known from being one of the leaders in deploying this. It's a SaaS-oriented, persona-oriented services. We are now applying this into the domains in our industry to help our customers generate incredible returns. So what's the goal of all of this? Simplification. Michel mentioned it, Carl mentioned it, Michael mentioned it. Carl likes to talk about the 2 brothers CSP: one brother -- he has been saying this for years, one brother to operate the network, one brother to go out and sell and market the services. Well, this is actually a picture of a network operations center for one of our customers who has deployed an AXOS network and is rolling it out across their entire subscriber base at a scale of 75,000 subscribers. They are going to run this network operation center with one person. One person, why? Because of all the automation, the intelligence, the analytics, the standard workflows that Michel talked about, one person. A 10,000-subscriber service provider typically will have 3, 4, 5, 6 people running their network operations center at all times. When you simplify the architecture of the network, when you simplify the operations, this is how you can actually run your business. And that's how our customers are starting to run their business today. In fact, customers who are rolling out the Intelligent Access Edge are actually realizing reductions in new service integration and system turnups of 75% or more. They're seeing their total network operations costs improve by 50% or more. And as Michael said, whether you're a wholesaler or a retailer, this is the foundation of your business. And so I'm going to share a few examples of our customers who are adopting and deploying Intelligent Access Edge. I'm going to start with a wholesaler, CityFibre, based in the U.K. 2 years ago, they announced an initiative to roll out fiber broadband network services powered by AXOS and Intelligent Access Edge to 8 million subscribers in the U.K. What's really fascinating about this is that they've received $4 billion in investment over the last several years. It is a true wholesale play. They build the network. They delivered the broadband service. The ISPs run their subscriber service on top of it. And they are committed to driving and building and deploying the lowest cost per bit per mile network in the United Kingdom. Plan is to get to 56 cities. They're deploying in 18 today. They're live in 11. They're serving 1,500 customers. They range in size from very large ISPs like Vodafone down to very small ones. They're very targeted like NGC Networks who serves only very targeted business subscribers. What's really fascinating is that they look at their early deployments, and they compare them to running a traditional fiber network, and they are seeing 70% operation savings. That's what they're expecting to deliver every time they roll out to a new market. What's really particularly encouraging is the statistic of around 96% of customers are either extremely or very satisfied. Many of you who are familiar with our industry, you know what our typical satisfaction rates are for broadband services. They're not 96%. And they're not going to stop here. At our conference, ConneXions in October, they announced that they're going to be the first CSP in not only the U.K. but all of Europe to roll out an NG-PON2-powered network, the ability to use all of this great technology, coupled with NG-PON2 and deploy wireless, mobile, enterprise, SMB services all over 1 physical network. And in fact, recently, they announced that Three, a mobile carrier in the U.K., is going to use this network to backhaul all of their 5G network traffic. A truly incredible story, shows the power of the Intelligent Access Edge. So we'll switch gears. A full-service retail provider, Forked Deer. That's right, it's pronounced Forked Deer. They're in Tennessee. It's an organization that's been around for 80 years. They started by bringing electricity to rural parts of Tennessee. That was their mission. 2 years ago, they announced that they're going to get into the broadband business. They want to help close the digital divide. They announced that 2 years ago. Starting from a standstill, not in the business, to a full network up and running delivering services in 8 months. I'll repeat that, in 8 months. Now how do they achieve that? Well, as Michel discussed, they saw an 80% reduction in integration time. So they had existing billing systems for their customers. They were able to very quickly integrate those systems to work with AXOS so they could actually go provision, bill, et cetera, for the service. They saw a 75% reduction in system turn-up time versus what we see traditionally with most service providers. And the service is so high quality, they have a 94% take rate for managed WiFi service. By the way, they compete with AT&T, one of the largest national carriers. 94% take rate. By the way, that service is powered by the Calix Cloud and our data center premises systems. 94% take rate in a competitive market. Most CLECs would tell you if they get to 50%, they are through. Now what's really fascinating about this is CityFibre, 8 million subscribers; Forked Deer, 1,000. It's the same system, same platform, same software, the same tools. That's the power of a platform. You can scale up and you can scale down. Michel made the point, we're not building systems for different segments in the market, and we're not also building different systems for different classes of service providers. That's the power of a software platform and the solution. Final example, another full-service retailer. Verizon is, but we tend to think of them as much more. We think of them as a pathfinder. This chart is actually from their Investor Day presentation from February. And it depicts their strategy for their One Fiber initiative. This is the initiative that's driving their network strategy to build One Fiber network. They've already started to roll it out, as you can see, lots of progress. That One Fiber network will start by supporting their 5G and 4G rollouts, converge a lot of their old copper infrastructure, deliver enterprise and wholesale services, SMB and eventually residential. Now we're thrilled because they've deployed over 2,500 E9 system shelves to support this rollout, and we have 100% share in this initiative. It's the heart of their strategy, and we're thrilled to be partnering with them on it. We're also very lucky because they've actually come, their VPs of network planning and strategy, and keynoted our ConneXions conference for the last 4 years. And every year, they played out their strategy, they've talked about their progress and they've shared their results. And this chart talks a lot about what Michel talked about, but this is actually their content from our conference. So what do you see here on the left side of the chart? You see functions like access, aggregation, lawful intercept, edge router. Traditionally, for Verizon and other providers, these are all discrete systems. So what have we done? We virtualized these functions. We've collapsed them onto an E9-2. We eliminate all of these elements, as Michel called them. Clearly, there's some CapEx savings there. What's really significant is you're now operating one system with virtualized functions instead of all of these systems in your network. And because you're leveraging NG-PON2, Verizon is able to deliver a 40-gig symmetrical service that can support residential, small and medium business, enterprise and wireless on one fiber network. So think about that. Instead of operating 4 physical networks, you're operating 1. A lot of CapEx savings, but it's really all about operational savings. Verizon has always been a pathfinder. They announced FiOs in 2004, started rolling it out in 2005. The primary reason for that initiative, the business case justification was not new services. It was operational efficiency. And they built one of the most efficient networks in the world, and they operate one of the most efficient networks in the world today. What is their expectation? 80% reduction in operating expense on top of one of the most efficient networks in the world today. It's not my chart, it's not our chart, this is their chart. This is what they're seeing. It is the power of the Intelligent Access Edge solution. So with that, I'd like to invite Tom back up to talk about AXOS.

Thanks, Matt. We are going to take a quick break here for 15 minutes. I think you can grab a cup of coffee that's here, and then we're going to assemble back in this room. At 11:10, is that when you're coming back at? We're going to come back at 11:10 Pacific time. Okay. [Break]

At this point, we're going to get back to the presentations for the day. It is my pleasure to reintroduce Michael Weening, who is going to take you through Revenue Edge. Michael?

Great. Thanks. So in the first section, we talked about the Intelligent Access Edge and how do you build that lowest cost per bit per mile network to be wildly efficient as a wholesaler or if you choose to be a full-service service provider, as a full-service service provider with a great network. But on top of that, we're now going to move into that subscriber, we're going to move into experiences. Because in the end, if you want to differentiate in a very crowded market that's getting more and more crowded with that subscriber having lots of choices, you need to actually differentiate on experiences. And to understand where we're going from a product point of view, I wanted to take a step back and I want to explain the market from our perspective. If you think about competition in the CSP market, you would normally start out and say, "It's MSOs competing with telcos and telcos with MSOs, and they're beating each other up trying to win subscribers." We have a very different view of what competition is in this marketplace, and we call them the home invaders. We believe that the ultimate competitor in this market is now the Goggles and the Amazons. The simple reality is that Google and Amazon look at the home, and they come up to that doorway and then when they get to that doorway, they actually don't have a lot of insight into the home. And so they're looking at it as an awesome opportunity to take the next step using Wi-Fi to get inside that home, own it and put a data probe in the home. They're not doing it because they want to provide WiFi or they want to provide WiFi services or they want to extend the fiber or any of the things that a CSP cares about. They could not care less about those things. They care about one thing, data, how do I get more insights so I can help you buy more amazon.com services or I can advertise to you in a very different way. It is a data probe. So they're playing this very different game to the traditional CSP, who is looking at them and still fighting this war of how do I win the connection. And unfortunately, those legacy CSPs are focused on connectivity, reliability, and you give me a phone call and I'll support you. That's the value proposition. And by the way, to give you really ugly boxes that's in basement and there's nothing cool about it whatsoever, and they're getting crushed by these guys. These guys are completely playing a different game because they don't care about it. So what's valuable on the left-hand side is not valuable on the right-hand side. And so there is this war going, and that, in our view, is the real competition. And so as these competitors, Amazon and Google who have a huge amount of money to come in to market, what are they doing? They're taking those R&D dollars, and they're investing them in amazing design, fast innovation. They're constantly innovating. They're throwing some services on top of it. So Amazon bought Eero and then they have parental controls, they've virus protection, things like that. Are they doing it for the reason why CSP wants to do it? Do they care about that revenue? Absolutely not. It's actually just another way to attract the subscribers so they can get the data probe in the home. It is all about winning the data. So that, in our view, is how there's competition, and that's the war that's being won. And over the last 3 years, we've been talking a lot about that's the competitor, the home invader, and that's what they need to pay attention to. And so as we go into this next stage, we have been focused on what is it that a winning CSP needs to do to. It's actually beat the home invaders, and we believe it's very possible. We believe that where they currently are, they can actually win. And the reason why is if they actually shift their mindset and approach the market in a very different way, move away from pipes and connections and just supported the value proposition to being an agile competitor that they can win, and it starts with kind of these things. First, they have to understand that the position that they have in the home is invaluable and must be defended at all costs. They're at the end of that fiber. They're already in the home. They're not Amazon and Google who is banging on the door, trying to get in the home. They are already there. They already have a trusted relationship, and all they have to do is understand how valuable that is and focus on it, make it core to their strategy. And sadly, not all of them are. So they still have -- they have to make that shift. That's the first step to actually determine whether or not they're going to be an innovative successful CSP. The second, you have to understand the table stakes is cool WiFi. You cannot actually put WiFi that takes 18 to 24 months. You can't take 18 to 24 months to actually bring technology in because what's going to happen is my sons are not going to wait. What are they going to do? They're going to go to Best Buy and they're going to buy Amazon and Google. They're going to tire of that ugly black box. They need to actually bring cool technology to market fast, and they need to keep up and they need to be first. Third, they need to understand data. I come from data background. My background is Salesforce. I was responsible for over 200,000 customers, and everything that we did was data. Data -- the company with the best data and insights wins, flat out. And these guys have all that data. That's why Google and Amazon want to be there. The problem is they don't know what to do with it. The world is filled with people who go and buy lots of data scientists at $400,000 a year each. And then they go and ask them all these questions. And in the end, what happens is the report lands up on the CEO's desk and nothing gets done. When you actually use Google technology, for example, AdWords, you actually don't use AdWords to put in an ad. So I'm a marketer. I put in a purchase for an AdWords, and then what I do is I wait for it to go off through a BI person who spends 2 weeks building a report and then sending it back to me 3 weeks later for me to take an action. That's not how Google works. Everything is real-time behavioral analytics on the fly, fast, right? This is not how they think. They are thinking data warehouses, BI guide, slow, slow, slow. So they need to completely change their minds to real-time behavioral analytics to understand the subscriber. They need to understand my kids in the home. And let's go back to them. They need to understand gamer, one great performance, all these different things and here's the experiences that they want. Don't try to sell them controls. At least I hoped on some parental control because I don't know what they're doing, but [ it's natural ]. So leverage analytics to understand the subscriber, build great experiences around what it is that they want, sell them what they want, and then the last piece, manage the experience. So worst case, they call you, right? Best case, you have an omnichannel experience. And that omnichannel experience is on the web. It's on a branded application that you can interact with so you can actually push notifications to them, you could talk to them, they could talk to you. All those different component parts, right? And if they have a problem -- so for example, if they subscribe to your parental applications issue or application and you don't know what you do as many of us parents don't know how we kind of screw these things up and you're having a problem getting it right and your daughter is screaming at you because of the fact she can't get through YouTube, well, when you call in, they're going to say, no problem, I can fix it for you. They reach into that application. They fix it remotely, and you're up and running. So it's a managed experience. So I tried through this interaction, but in a worst-case scenario, I know I always disrupt the CSP. This is what builds a winning CSP. This is what's going to change. This is the reason why when you -- especially when you lay over the data, the real-time behavioral analytics, they can beat Google and Amazon. They won't lose that place. But they need to be smart enough to actually understand that they have to be proactive and completely change how they go to market. And so that's what we've done. Over the last few years, we built out the Revenue Edge. That's really focused on 4 things: real-time subscriber insights. Again, I come back to I'm a Salesforce guy. I am all about data. I've built big data warehouses, I understand analytics. We have built one of the best analytics engines that exist in this industry, flat out, absolutely. And it's incredibly powerful because it's democratization of data. The second thing is cool technology fast. Be first. That's a big shift for them because they're always last. Third, amazing experiences; and fourth, killer brand. You have to have a great brand. You're up against Google and Amazon. You can't actually be that laggard brand. So how do you actually go that way? And so now I'm going to invite Shane up, who's going to walk you through how the Revenue Edge comes together and how we make all those things happen.

Thanks. So we talked about some key attributes. We said the first one is really real-time subscriber. So when we sat back and looked at the framework that you need, the building blocks, the first obvious one for us is edge data science. We want to give the service provider the insights into the subscriber real time what's going on. The second one we said is about rapid adoption of new technology, the latest technology, cool technology. That really takes 2 forms. That's both the systems that we create, the hardware, whether that's a WiFi 6, but that's also the software. This is about speed. It's about innovation at the speed of software. So that speaks to what we're doing with the AXOS services that we're offering. And as Michael said, you got to have good enablement. We got to make sure that our customers are successful, they can adopt this, they can deploy it, they can offer those differentiated experiences. So a key part of this is being able to mirror the insights of the systems with the software with enablement. That's our foundation. We think everybody needs that, every CSP needs that, every customer needs that. On top of that, though, we can start to offer differentiated experience, software for consumption, value-add software that increases their revenue opportunity. So that's really the last piece of work. So let me talk a little bit more about how we do each of those. So the first one is we want to give -- we think the service providers are incredibly well positioned, given where they are in the network to understand their customers. They've got great relationships. They're in a great place. They just don't have the platform they need to get to insights. So that's what we built. It was how do we build and how do we collect that. For us, data has to be real-time at a subscriber granularity. It has to be about your actual behaviors, what are you actually doing, and it has to be actionable. Those 3 things give us something that the service provider can actually work with. So for us, it was about building a web scale cloud platform to collect data. We want to associate the subscriber with these devices, with these applications, what you're actually trying to do. And now what we want to do is then put behavioral analytics, predictive analytics against that and drive workflows. So on top of that, we build persona-based SaaS services. One of the first persona is marketing. I'm a marketing person. I want to do churn prevention. I wanted to do upsell. I want to do acquisition. First of all, I've got to understand my customer, the segments. Are you in Silicon? So look at my family. I telecommute a couple of times a week. I'm out of the office 3 days a week. I'm home on weekends. I stream video in the evening. I stream video in the weekends. I don't have any children. That's a set of personas and profiles, the way I behave, the way my wife behaves when we are at home, but it also changes during the day. When I'm at home 8 to 5, I'm a telecommuter. I care a lot about Zoom. In the evenings, I care a lot about YouTube TV. On the weekends, I care a lot about Netflix. Those are different behaviors. If I have children that are gamers, all right -- I happen to be a gamer. Do I care a lot about home security? Do I care a lot about home automation? I'm a Luddite. I have no home -- IoT devices in my house. I've got friends who got 35, 50 IoT devices, very different behavior. I want to learn about this. I want to learn about those devices. I want to learn about those behaviors. I want to start to create profiles. And then I want to decide, are you a streamer, are you a gamer, are you a telecommuter, and figure out, are you a person that's highly likely to churn. You're having a bad experience. You're starting to speed test. You're bumping up against your limit. You've got poor WiFi. You're probably pretty unhappy. You start to roll up. If you're a high ARPU customer, I probably care a lot about you, and I want to do churn prevention. I'm a horrible person to try and sell parental controls to. I have no children. I don't care about parental controls. Many of my friends do. I also happen to be a Luddite when it comes to IoT. I don't care about home automation, don't sell me home automation. Home security, those are interesting to me. I care a lot about my house. I live in San Jose. Home security would be good. How do you get those profiles? How do I understand those behaviors? How do I understand what you're doing and then drive the workflow? That is churn prevention. That is upsell/cross-sell. That is acquisition. Same thing on support. I'm looking for customers that are having bad experiences. I'm either going to fix it, drive a troubleshooting-type of workflow or make it better. I may want to upsell. The answer may be, "Hey, if you had a higher bandwidth service, you'd be a lot happier." I want to drive those profiles. I want to drive those workflows, and you want to do that from a cloud-based platform. So that's kind of the first part of this. So where does the data come from? Well, I'm in a great location. I'm in your home. I'm already there as a search, right? I'm giving you services. I'm putting great cool technology into your house. I really want to do a couple of different things. These are my storefronts. This is how I deal with you, other than on the phone, on the app. These are my storefronts, right? So this is how I get to offer you great new services. It's also how I get to understand your behaviors. This is a great place to understand what devices are connecting to WiFi, what devices are connecting to IoT, what you're doing, what your applications are, what are your behaviors. I start to understand that profile. But I want to start to have that latest cool technology to have better experience. So much like Michel talked about, I want to abstract myself from the underlying hardware. So with EXOS, much like with AXOS, I want to be abstracted. I want to take the latest hardware. I don't care if it's Broadcom or Qualcomm or Intel. In this case, WiFi 6 is a huge disruption. If you look at what's going on in every consumer device, the latest iPhone, the latest Galaxy, the latest TV, everybody is moving to WiFi 6. Why? It's a great WiFi experience. There's a lot of underlying capabilities that are fantastic. So that means I want to be able to adopt. One of our key strategies was to make sure we were first to market with incredible 12x12 11ax capabilities. I want to continue to adopt technology, but also I want to launch software and software innovation. So we built an OS that not only leverage that same type of abstraction, always on, but also is container based and really designed on how do I quickly launch new services and give the service provider the ability to launch new services and how do I continue to adopt the latest technology and put it into [indiscernible]. So as I've got these storefronts, I've got the greatest cool technology from a hardware perspective. I want to make sure I'm continuing to give them the right experiences, the right software. The first one is everybody wants incredible [ WiFi experience ]. If I can't give you the best WiFi experience, you're going to think I'm horrible as a service provider. It doesn't matter whether it's my problem, your problem. At the end of the day, it's my problem as a CSP. You're not happy. So I want to be able to see what's going on. I want to understand the WiFi. I want to be able to instrument it. I want to have the right intelligence and the algorithms and the capabilities to give you that best WiFi experience. If I'm giving you the best WiFi experience, I want you to be able to see it's the best WiFi experience. I want you to be able to easily test it and go, "Hey, this is really cool. I'm getting 1 gig. I'm really happy about this. I'm getting a 1 gig on Wi-Fi to my iPhone." That's pretty cool. Next part of it, though, is how do I interact? Traditionally, service providers, I'm waiting for you to phone me to go you're unhappy. A big part of this was easy -- and other things and this whole architecture is to get, "Why don't you be proactive instead of reactive?" So I want to give you an app, I want to give you an app to start to work with this and interact with this. So we've developed CommandIQ, Android and Apple, to give our service providers the way to have their subscribers interact with it from a workflow, from an interaction perspective. It's also a great branding opportunity for them to put their brand on it and extend their brand into the service -- into the consumer. I want Alexa Voice Service. I want you to be able to talk to me. I want custom skills. I want to be able to be your CSP. I want you to also start thinking this is core innovative technology. You don't think of CSPs having an app, first with WiFi 6, Alexa Voice Service. You've got to change that mentality. We've got to give them those capabilities. The last piece is, if I've got this great app, I can start to use notification messaging. I can start to change that 2-way interaction using that app to start to communicate with you and start -- what I send you to make things -- make you aware of what I'm doing, the services I'm offering, tips and tricks, new services and really go to that 2-way communication. So we think everybody needs the underlying insights, systems and the services, and those really form the foundation. On top of that then, I can start to look at new revenue opportunities, differentiated experiences. Two of the obvious ones for us was security in experience. Everybody wants to feel secure in what they're doing. It's fundamental in what we're trying to do. It's fundamental to the way we use these services. So I want to be protected from viruses, from malware, malicious websites, intrusion, all of those capabilities. I want to be able to offer those building blocks on top of that foundation as new interesting software. I want to have the insights into that. And I want to be able to easily deliver that new software to you. The second part of it is, I want to give you the control to be able to define the experience you want. I want you to be able to decide what content do I want to filter, how do -- where I want to place limits in place -- time limits in place, anything about managing that experience. These are really our first 2 examples. Much like we're going to deliver additional edge systems and WiFi 6 and we're going to keep driving that, we're going to deliver additional software suites on top of what we're doing. And you're going to see more and more software suites from us throughout the year. Okay. And with that, I'm going to turn it over to Matt Collins again, who's going to talk to you about incredible enablement that we have around this for our customers.

Thanks very much, Shane. Hopefully, everyone can hear me this time. So WiFi 6, cutting-edge technology, software-based services, mobile apps. As most of you probably know, that's a little different than the way most CSPs are running their business today. Similar to the access edge, we know this is a transformation. So we've developed a full set of services to help enable our service providers to learn the technology, deploy the technology, and of course, customer success. With the Intelligent Access Edge, I talked about the operations persona. In this case, we have experts in marketing who understand how to use data to segment, to target, to create campaigns, partner with our customers over several years, work with them on a continuous basis to help them use this capability to transform their marketing. Customer support. We have world-class experts who understand how to run a world-class customer support function. Many of our customers don't know that. So we assigned them a coach, a success professional who shares their knowledge, who can actually look into the accounts of cloud, see what the service provider professional sees, coach them, guide them, help them optimize how they're leveraging the capabilities. But we also recognize that for most of our customers, this isn't just launching a new business. What you see here, this is actually a new business. And so we've made the commitment to invest in a market-activation program, which is designed to give them everything they need to launch this new business. So what do we mean? Whether it's digital marketing, direct marketing, whether it's social, whether it's quick turn-up guides, even things like subscriber self-help videos, we produce all of that, and we give it to our customers so they can turn around and in minutes, have marketing campaigns up and running that will compete with the biggest players they're fighting. Many of them are very small. They don't have big marketing teams. They don't have big marketing budgets. Why would we go have 1,000 service providers create a self-help video for their subscribers on how to use the CommandIQ app? You shouldn't. So that's what we've done. We go out and we actually create those videos. These are campaign concepts that we've created that talk about these new services. This is a specific campaign that we built with one of our best customers, basically the providing possible campaign. Every day, every week, we're working with those teams. Our success teams take these assets, they go, they meet, they work with the service provider, and they help them get in market and differentiate themselves. And we're even doing new and innovative things like this, for instance, is a packaging sleeve. Some of our service provider customers actually have a retail presence at their headquarter office and whatever town they're in, whatever city. People come in. This enables them to have the same quality of packaging, positioning that they can put their brand on the most -- our brand is not on it, that their subscribers would see when they go to Best Buy. This is really about helping them raise their game. That's our objective. The message we put in market is elevate. Well, what do we mean? The folks you see here on this side, if you've ever been to our ConneXions conference, or you've watched any of the content, we created a fictional family called The Millers. The goal there was to essentially help use them as a lens on the subscriber's life to help educate our customers, the CSP, as to what's happening in the life of a subscriber, what do they want and how do they need to rethink the way they operate their business. And what Shane presented is a set of tools and technologies and capabilities that's meant to elevate every aspect of their business. Michael said, look, they're already at sort of center of the home. We want them to become central to their life, elevate their service. Well, what do we mean? I'll give you an example. From the not-too-distant past, let's say, news hits, there's a security hole in Roku, right? We all remember that. And go into marketing cloud, look across their entire subscriber base and see who has a Roku device. We have a prebuilt really nice campaign built around ProtectIQ. They go, they send an e-mail out, whatever methodology they want to do, saying, "Hey, you might be interested in ProtectIQ that will secure your entire home network. So if there is a risk on one particular device, we've got your back." That's elevating their service. When I was a consultant at McKinsey, I used to tell my customers, you want your marketing to be so good, it feels like a service. And that's what that is. They're not marketing. They're reaching out to you and saying, "We understand what your experience is like. We understand what your risk is." They're not using the data to go sell it and monetize it. They're using it to drive a better experience and elevate the service they're offering, elevate their brand. Shane showed you the protect -- the CommandIQ app. Everything that we provide to the service provider is about them reemphasizing their brand. The BLAST system, our brand is not on it, but we let them put their brand on it. The CommandIQ app, we have a service that helps them personalize it, put their brand on it. So the subscriber isn't seeing Plume, they're not seeing Amazon, they're not seeing Google. The brands they see on the cutting-edge technology is the service provider brand. That's a way for them to completely reposition themselves. They're not the "Oh, yes, you guys, you hook up to my house, you deliver the broadband service". No, I'm actually putting WiFi 6 systems in your home, I'm giving you cutting-edge apps and I'm given you cutting-edge services, completely reposition themselves, elevate the experience. As Shane said, the Calix Cloud, we put all of the data in the home network at their hands in persona-aligned dashboards. What does that mean? If there's a problem with one of these things, they can actually deliver a managed service. I believe that's their biggest differentiator. To give you an example, I'm running parental controls. My child -- I set it up so they can only access Facebook at certain times of the day. Oops, I didn't set it up right. She can't get on Facebook. I don't know what to do. Well, if you're working through a service provider, they can look and say, "Hey, here's how you set up your parental controls. Would you like me to help you fix that? Oh, by the way, I have a great video I can send you that walks you through everything." Now you go into a consumer store, a big-box store and you buy a parental control solution, who's going to help you? You're going to call Circle? No. So that is the heart of what we're looking to do for the service provider. They are going to be able to deliver a managed service to their biggest differentiator against any of the big technology providers. And of course, the end game is, as they do this, we help them elevate their revenue, their ARPU. So I'd like to share a few customer examples. This is a solution set that applies to whether you're an MSO, you're an electric co-op, you're a CLEC, whatever type of service provider. So I'm going to start with ALLO. ALLO is a CLEC founded in 2003, primarily located in Nebraska. They're also going into Colorado and other markets. So they started by rolling out an incredible fiber network. But they didn't stop there. They're one of the earliest adopters of the Revenue Edge. They bought the full suite. They're rolling out GigaSpire systems, right, left and center. So they used the Revenue Edge Solution, specifically the EDGE Insights to really optimize the subscriber experience. There's actually a capability built in called self-heal. It's machine learning that analyzes the home network. It identifies things like, "Hey, you got a microwave that's on the same channel as this device that you like to use. By the way, we can fix that." It automatically fix it -- fixes it. They saw a 92% reduction in WiFi issues. This is not just managing, it's being proactive. They're literally fixing the problem before the subscriber knows that is a problem. By the way, they're switching the channels at night when they know, by the way, no one in the home is using any devices. It's incredible intelligence, it's the kind of tools that traditionally only the web-scale giants had. We're putting it in their hands. By the way, when they roll out this new service, every market they go into, their expectation is at least a 50% market share. Every market they go into, they're competing with big national players like Charter, like CenturyLink, and they're winning. And their customer satisfaction is one of the highest in the industry, one of the highest in the industry. And what's fascinating is they shared with us, when they rolled out the EDGE Insights and started to use them, within 3 months, their NPS score went up 15 points, 15 points. So if you're familiar with NPS, that's not bad. So Allo, progressive CLEC, but this isn't just for the aggressor. Another example. Range is an ILEC. They've been in business for a long time, as you can see. They've also fully adopted the Revenue Edge. Again, starting with the EDGE Insights, they deployed support cloud, marketing cloud. They saw a 30% reduction in their operating costs. Well, what does that mean? Well, when you're dealing with issues more intelligently, you reduce truck rolls. So you don't have to pay someone to get in truck and drive to someone's home. By the way, your subscribers don't want to be there either waiting for you to show up. Fewer phone calls, shorter call durations, 30% reduction. By the way, when they looked at that investment and they compared it to what they saw in terms of cost reductions, ARPU increases and subscriber acquisition, 173% ROI in just 6 months. Now you're all in the investment community, that's a pretty good return. 173% ROI. And what's most fascinating is the quality of their service is so high that their monthly churn rate for their subscribers who they're serving through the Revenue Edge, 0.2%. Again, in this industry, if your churn rate is 10x that, you're generally pretty darn happy. Go look at monthly churn statistics for our industry, 0.2%. And what I particularly love about Range is they were one of the first ones to go out and say, "We're going to brand our mobile app. We're going to put our logo on it we're going to put our colors on it so every time one of our subscribers interacts with the service, that's making them really happy, 0.2% churn, they're going to see Range." They're not going to see Plume, they're not going to see Amazon, they're not going to see Google, they're going to see Range. So third example. Both of those organizations that I talked about, their primary mission is delivering a great service and driving profits and revenue. Triangle, they're co-op up in Montana, also a very competitive market. Their primary competitor is Spectrum, 1,700 subscribers, one of the earliest adopters of the Revenue Edge. Now whether you're trying to make subscribers happy or shareholders happy or both, you want to have an efficient operation and you want to grow your business. In this case, if you're a member on the organization, you can take those profits and reinvest it in services that you provide. So Triangle, again, one of the earliest adopters. One of the things that they did that was very innovated -- innovative is they took this solution and its capability and they completely revamped their go-to-market. They rolled out targeted marketing campaigns, focused on upsell. They saw a 40% increase in their marketing efficiency and a 28% reduction in the cost to drive upgrades. But what's most amazing is, over the course of the time that they've been working with us on the Revenue Edge, they saw a 60% increase in ARPU, growing double digits every year. And this solution was so transformational for them that they were able to take their Tier 1 tech support professionals and turn them into CSRs. There was such a dramatic drop in the number of issues that they had to deal with and a number of tickets that they literally could take relatively well-paid tech support personnel and drive them into roles where they could be selling and marketing their service, and that's the power of Revenue Edge. So with that, I'd like to invite Carl back up here to talk about some of the observations from his side.

Thanks, Matt. By the way, that 62% increase in ARPU came on what incremental cost in the network, do you think? The network's there. This is frequent flyer miles stuff. I've already got the airplane, I have an empty seat. It doesn't cost you any more to put a person in it. It's literally incremental revenue with virtually no cost. And if you're also lowering your costs, that incremental revenue is actually coming at greater than 100 points of margin, take that back to your finance class. So quite interesting stuff. We started off today by talking about not 1 but 2 disruptions. I trust you can all understand the disruptions that we spoke of, but to go back over them, it's the device enabled-subscriber connecting to the content and applications in the cloud. By the way, let me say that again, the device-enabled subscriber, like Alex over here, connected to content and applications in the cloud. Guess what's in between those 2 things, a service provider. The second disruption, literally an end-state network, wireless endpoint sitting on top of an unlimited transport mechanism, fiber, by the way, shared over a PONs. Guess who builds that infrastructure, the service provider. Those 2 disruptions are changing the entire industry and will change everything above it as well. So I trust we've done a good job of helping you see those disruptions and the fact that they will continue. The bad news is you're all 10 years late to the party. But the good news, we got another 20 years to go. So this is really just starting. That's the first piece. The second piece is, hopefully, you understood that this requires technology and, we believe, platforms. So you heard that over and over again, platforms. Why did you hear platforms? Because in order to address this correctly, in order for it to be deployed, in order for us to, frankly, close the digital divide, these systems not only have to scale up to large service providers, but scale down to the very smallest of service providers. You heard Matt earlier talk about 8 million to 1,000. There is not a company in this industry that has built anything that's capable of spanning that scale. We do, it's in our nature. Secondly, it doesn't matter the type of service provider. Whether it's a wireless provider or cable MSO, an ILEC, a greenfield, a brand-new service provider, an electric co-op that's getting into the business, a wireless Internet service provider, it does not matter. Same technology, same platforms. So we're not building a box for the telcos and a box for the cable MSOs with its own operating system over here and its own embedded operating system over there. It is literally one set of platforms with hardware deployment that's the same hardware deployment scenarios at the [ CEO ] data center, out in the outside plant and ultimately, in an embedded [ CL ] data and those 3 units are outside. You got a chance to lay hands on them if you really want to get back to your hardware roots. But we have the right technologies and the right platforms. And you're seeing all of these here and the results, the effects that they're having on the service provider. So I trust that's the second thing that we spoke of and that that's compelling. The third piece is huge disruption, great platforms. They don't sell themselves, they don't build themselves, they don't help our customers transform their very business models. I'm sure it's not lost upon you but this is a very different service provider than the ones we grew up with. It's a wildly different go to market. It's a wildly different engagement for those subscribers. And you're seeing results of churn rates. In rehearsal, not to be somewhat dark, Matt mentioned a 0.2% churn rate. And I said, yes, that's roughly the mortality rate. But people die, that's a churn. What's my point? That churn rate is so low that no one is leaving that service. 0.2 is effectively 0. That does not exist in our industry. So the platforms are there, the disruption is there, what about the team? Well, let's go back and sort of build it back up. We talked about the Intelligent Access Edge. Well, that's led -- that group is led by Michel Langlois. What's his background? His background is nearly a Silicon Valley legacy at this point in time. And I say that in the best of terms. Prior to being here, for 5 years, he led Junos, thereabouts, at Juniper. Prior to that, grew up at Cisco and ultimately ended up leading IOS at Cisco. That's a deep networking engineering leadership talent. Let's look at the Revenue Edge, a rapidly evolving space that's connected to broadband. Shane leads that. What's Shane's background? Well, in addition to coming from some start-ups and getting used to the rough and tumble notion of start-ups, before joining Calix, led the broadband products group at CommScope. So over here, deep networking experience or with a cable experience, broadband, facing the subscriber experience. The next thing you heard over and over again, data. Well, if you can understand this transformation, we can sort of -- the blind squirrel hunting around trying to find a nut, but not only are we trying to put data in the hands of our customers. We have to use data to understand where to go, who to sell to, how to find these prospects. And so our Chief Marketing Officer, Matt, who was just up previous to me, comes from a deep data background. Now he mentioned he was a consultant at McKinsey, let's not hold that against him. But in fact, he came out of IBM, Dun & Bradstreet, Merck. Doing what? Using data and analytics to achieve focused marketing, helping buyers through their journey. That level of data expertise almost a sine qua non for us to be successful. And then, Michael. Where does Michael come from? Well, previous to this, headed up global customer success at Salesforce. Prior to that at Microsoft, Bell Mobility. So he's been on both the service provider side as well as data side, not a hardware guy. We have fun with him. Not a hardware guy, pure software guy and comes from the service side. What's most important about Michael's role? This transformation has to happen at our customers. Who's going to help our customers? Well, it's the front end of our business, our sales and service organizations. Without that transformation -- and we've been through a big transformation, so have they, that's what Michael has led us through and continues to lead us through. So I submit that we not only have a diverse team, but it's a highly talented team, a deeply experienced team coming together on this mission to help us down this journey and more importantly, help our customers down that journey. Last person that you haven't met, and you'll meet him right after 1:00 when the market closes, is our CFO, Cory Sindelar. Cory's background, actually also diverse and necessarily so. If you think about this from an accounting standpoint, let's talk about rev rec. Revenue recognition on hardware, you book it, you ship it, you recognize it. What about professional services where you have percent complete? You have an entirely different approach. What about licensed assets? An entirely different approach. What about cloud platforms and software-as-a-service? Another whole approach. What about getting through this transition with a balance sheet that had about [ $1, $2.98 ] on it for a while? How does that feel? Let's look at Cory's background? Legato, EMC, Ikanos, doing silicon, Violin Memory. Tough times, by the way, put to use here. This has been a very interesting transformation for a number of years with not a lot of net underneath it. Bringing that all together, bring this down together, bring the opportunity together in front of this disruption. And we'll come back after lunch and talk about how all this resolves itself into the financial model going forward. So 2 new disruptions, the right technologies, the right team. Do I have your agreement? If not, Allen Ludden will not be able to say the password, which is fajitas. So I will keep you here until I have your agreement. So hopefully, you understand what I just said. When we come back, we'll talk about the financial model. Tom, it's all yours.

Thanks, Carl. We're running a little bit ahead of schedule, but we're going to get back on schedule. So we're going to take a lunch break now until 1:00 p.m. Pacific time. We will start the webcast again at 1:00 p.m. Pacific time sharp. For those of you that are in the room, lunch is served right outside the door. So we'll be back at 1. [Break]

Good afternoon, everyone, and welcome back. Those of you joining us back on the webcast, welcome back. It is now my pleasure to introduce our Chief Financial Officer, Cory Sindelar, who will take us through the financial update. With that, Cory?

Thank you, Tom. So we've been sharing this information for about 6 quarters. What we've been saying is that as our platforms become greater percentage of our business, you're going to see improved financial performance on these 4 metrics. We're going to spend a little time to drill down in each one of those. Deliberate revenue growth. What do we mean by deliberate? We believe that we -- our goal is to go out identify and win every prospect and customer that is aligned -- strategically aligned with our vision. How are we doing? We're doing pretty well. We've got 340 customers -- new customers in the last 3 years, bringing our total count up to about 1,600 customers. Those are some really good numbers. If you put it in the context that there's only 10,000 service providers globally, it's pretty remarkable. The nature of these new customers are across the board in terms of size and type. But they are traditionally going to be more of our small customers, right? We're going to follow the mix of our customer base. Matt has talked to you about a few of those today. We've got new customers such as Forked Deer who is an electric co-op, who had no broadband subscribers when they first joined us; and with CityFibre who comes in -- who is now being a wholesaler, and they're one of our customers. Now largest of those is Verizon. You've seen this slide before, last quarter. Let me just reorient you about the slide. This is our revenue recast based on customer size. This is determined by a number of subscribers our customers have. Our small customers are customers that have broadband subscribers of 250,000 or less; medium are 250,000 to 2.5 million; and our large customers have 2.5 million subscribers or more. I bring this slide up and repeat it one more time to bring one point. Over the last 3 years, if you take those 340 new customers plus a large percentage of our existing subscribers, and they come to our platforms, you're seeing growth in our small customer base. For the same duration, you're seeing our dependence on a few large customers diminish. The combination of these 2 trends means we have a more predictable revenue stream today than we had in the past. Let's talk about gross margin expansion. All day long, you can talk about -- you've heard how our platforms are helping our customers. Every one of those products, services have a gross margin that is higher than our traditional business. We're talking about software, cloud, the aligned services that attach to those cloud products. Even the systems that the platforms run on have a higher gross margin than our traditional business. So I'm going to make kind of an easy assertion for me, probably the easiest assertion I'm going to make today. Our platforms increase as a percentage of our business, gross margins go up. It's that simple. I'm going to call it, Carl, that's the good news. You want the better news. The better news is our platforms are not at scale today. Take our clouds, for example. While our cloud margin is better than our corporate gross average, it isn't what a mature cloud company is in the mid -- in the, call it, 80% or so. What does that mean? That means as we continue to scale our platforms, even within those margins, those margins are going to improve over time. We've talked a lot about platforms, platform adoption. Well, let's put a little visual to exactly what kind of success we're seeing. I think we've seen pretty good success. Over the last 4 years, we've seen customer adoption of our platforms at 145% CAGR. I'll also point out that these growth curves mirror product life cycles. Take AXOS for instance. First platform into the marketplace, access infrastructure, maybe a little bit hard to say where you're going to implement, but it's growing steadily. Our second platform comes in the marketplace, cloud, easier to make that decision, easier to deploy, more rapid, in market less time, has already exceeded our AXOS count. And last, our last platform, EXOS/Revenue EDGE, latest in the marketplace, and it is growing like a weed. There's no counts over here. But you can certainly know there's at least 340 because all of those customers came over to use our platforms. These are hundreds. So it gives you kind of an idea of kind of where our growth rate is. But let's not get too exuberant. While these are our customer signings, our seeds for success, this is a leading indicator, there are number of reasons why revenue is delayed behind this. So the first one being, if you signed our contract, you adopt our platform. You need to deploy it, right? This is a pay-as-you-succeed model. All of our platforms monetize on subscriber count. So let's take an example of AXOS. I decided I'm going to go to buy AXOS. I don't go out and rip out my entire infrastructure overnight to deploy AXOS. A year comes up, I'm going to go do these new deployments. You know what? This portion of the network's kind of old, I'm going to time it up. I'm going to expand it to that. The next year comes along, I expand it a little bit more. Time goes along, I add a little bit more. Every time they're bringing more subscribers onto the network using AXOS, we get a piece of that, so it takes a little bit of time. Second thing is by signing up with Calix, you're signing a 3-year agreement, a 3-year agreement. There are certain components of that revenue that's recognized over time, specifically, the cloud solutions and certain of our services. Specifically managed services, maintenance and those lines are ratable over time. So again, they're signing a contract, revenue is being delayed. Third time -- third point, this is a land and expand strategy. These are the seeds, these are the landing points. It provides us an opportunity to upsell other platforms that they have. So they may start with Calix Cloud, marketing cloud, support cloud, then go to EXOS. We have the ability to cross-sell. Furthermore, as Shane brings out new service suites, those can be dropped on the platform. That's the beautiful part about a platform, the ability to add additional products on top on an existing base. So to put this all together, this is our seeds of success. Now it's a matter of time before we're going to harvest the fruits of our labor. This is an interesting chart. Maybe more interesting to me because I understand kind of all that went in and out of this gross margin line. We have traversed quite a bit in terms of transforming our company over the last 3 years. First off, we traversed our deployment services mess. At the same time, we were eliminating SKUs, and we're simplifying our traditional business. We disposed of a cabinet business product line. We've moved our supply chain out of China. We started to use ODM partners to do some of our designs for some of our systems. We invested in customer success earlier last year -- first of last year, which is essential to us to continue to grow our platforms. More importantly, we've got about 2 years now of running our platforms where we have data, right? We have an understanding of the way customers are adopting platforms, how long it takes for them to onboard those customers. Once they're onboarded, how fast are they adopting or through 1 year, [ set the year of ] renewals? So we're getting it at the rate at which they're deploying and what is the uptake. So we're starting to collect that data. So we need to have all that data before, so it really kind of made it hard for us to come and have a day like today to provide you some target margin metrics. But we didn't have run time on our model and there was a lot of stuff going on under the covers, right? So the good news about all of that, all of that noise, all that transition, all of that stuff that was going on, all those moving parts, we still increased gross margin year-over-year for the last 2 years. Third point is disciplined OpEx investment. And on this one, I would have to say that the team has done a fantastic job. Let's start with G&A. G&A has been running above $36 million a year, just the last year. Last year, we had a change in accounting pronouncement that had us capitalize certain internal and external cost around our ERP implementation. But if you ex that out, we've been roughly at the same level. As we look into 2020, we provided some information that we'll be back around that level, slightly higher as we continue to -- as Oracle gets depreciated and some of those people costs are coming back onto the balance sheet -- coming back through the P&L, rather. But in short term, we'll get that back down to about a $9 million quarterly run rate. Sales and marketing has been pretty constant through these last 3 years, constant in absolute dollars. But now what's going on within the department, they've been rebalancing the workforce, taking advantage of what we see in terms of the upcoming market opportunity. And we've done all of that in the confines of that investment decisions. But we're now at a point where sales and marketing needs to grow. So we've said we're going to let it grow. We're going to let it grow at the rate of revenue. The last is R&D. In 2017, we completed the large major development of the platforms. We decided to reduce our investment in R&D, taking it down by about 40%. But when we were talking about how we were going to resize R&D, we set a target. We said, "Hey, we want to be at $20 million a quarter run rate." After the engineering team actually executed the actions in the early part of 2018, they have done that. So all in all, I think the team has done a phenomenal job in terms of being able to maintain discipline over our OpEx investment decisions. Okay. We're going to get to the why you're all here, why you want to talk about our target financial model. Before we get into it, I want to make 1 caveat. We are not trying to change any of the numbers that The Street had published for 2020, okay? So the numbers and the targets that I'm about to share relate to years 2021 to 2023. That's kind of our planning horizon for providing these targets and it's with all the information that we have today, [indiscernible] our platforms in the business into the marketplace and where we're at. So ready? Here we go. Number one, revenue. Revenue is going to grow. Our belief is that it will grow between 5% and 10% per annum. That's based on our understanding about how we think the traditional business will roll off and the growth that we will see from our platform businesses behind it. Similarly, non-GAAP gross margin is expected to grow between 100 and 200 basis points a year. It also follows the same revenue targets. So these numbers are based on our beliefs around where our revenue are. I would also further go and say, I would not create a straight line. Each quarter, there may be variability. We may have a down quarter. But each year, we would expect over a year, in the time horizon, margins up into the right. Let's talk about G&A. Once we get to 2021, we believe we can hold G&A at $9 million a quarter until it becomes 7% of revenue. Likewise, R&D, we're going to hold them at $20 million a quarter until such time their number becomes 32% of product gross profit. So we're basically going to run those organizations flat, let revenue grow underneath of it and let them grow into the model. Sales and marketing is different. Sales and marketing, we know we need to make investments in sales and marketing to capitalize on the opportunities ahead of us. We're going to let it grow between -- grow with revenue, somewhere between 16% and 18%. To the extent we find ways that we can actually accelerate our business model, we'll do so. It might end up being at the higher end of this range. If you combine these 3 things together, what you end up having is a substantial part of our P&L that is being held flat, a part of it growing with revenue, providing a little bit of leverage into the model. But the most important point, most important point is our platforms are in the market today. Customers are adopting them. We have a huge opportunity in front of us. We need to invest in the front end -- front part of the business to go capture that opportunity through riding the disruption . One last slide. These are our balance sheet metrics. They're pretty good. I'll put them up against anybody in the industry. We're providing a little bit of insight to our balance sheet, more so, to give you some idea around our capital efficiency. And the other point I'll add on is, is our platforms allow these kind of results. Our platform business is going to have a more linear model. It's going to be less of a box ship where everything comes at the end of the quarter. It's going to give you the ability to have a better DSO, better forecasting of our supply chain. All of these things improve because of platforms. And more importantly for the folks here is, as we grow, we're not going to need a lot of growth capital. We're not going to have to go raise additional money to fund the growth of the business. We think between the cash on hand, cash flow generated from operations and our line of credit, we have what we need to be able to accomplish our objectives. With that, I'm going to turn it back over to Carl.

Thanks, Cory. Why don't you stay up here with me? So I just want to emphasize couple of items. The OpEx efficiency and the commitments that we're driving OpEx to as we obviously expand revenue and expand gross margins and hold OpEx, OpEx is going to undergrow that. You all figured that out. So it does not -- you can go model to your heart's content. Feel free to buttonhole Cory in the hallway and say, "I did some numbers." But I think you'll start to understand what we believe we can drive from a value standpoint. I want to also emphasize that what you're seeing here is sort of the next 3 to 4 years. And so there's an implicit statement in there that we're watching the traditional business roll off, the new business growing. We have enough data on these 2 to now say this is what we think we can run at. At some point in time, this traditional business stops shrinking, there is a 0 that it eventually gets to. So we're trying to put brackets around this to allow you to start to understand what we think we can deliver in the next 3 to 4 years. And so I want to go back to where I started, not 1 but 2 disruptions. You've heard this before. The right platforms, the right team, and now you can see how this monetizes in our model. So with that, I'm going to invite the leadership team back up. And we'll have a Q&A session. Hopefully, everybody's got their microphones turned on. And when we're done with the Q&A session, we're going to close down the webinar. Those of you who are here, the showcase will be open. And off we go. Fair enough?

Hang on a second. You have to be on the microphone.

My apologies but I forget, have you -- the numbers you just shared with us you said were for '21 on out. Can you remind me, did you give guidance on gross margin for 2020 previously?

We didn't provide annual guidance. Our model's out there. And I think we're -- we understand what they are and...

Okay. So you're okay with the consensus number?

Yes.

To your right.

So Carl, 2 years ago, you had the successful news that you began working with Verizon and you're rolling that out, and we've heard CityFibre. To your sense -- all the success that we've heard of is from smaller customers. Now that we've got enough real-life data of what you're able to do for everybody, I'm surprised we haven't seen more endorsements, more adoptions, more customers of the higher -- larger size, even just as a test rollout or a market-by-market approach. When we look at this outlook here, is that dependent on that needs to happen or it's going to just continue to be many of these smaller customers continuing to eat away from the bottom up?

So good question. And you're right. To answer the question, I'm actually going to frame it back to disruptions because what it is actually a sign of is the immaturity of the market. So when you see disruptions, you rarely see larger inertial incumbents moving there first. Verizon is actually a highly unusual service player. For those of you who have been following the industry, you have to go back to fiber to the prem. Fiber to the prem, we've been doing fiber to the prem since before we were actually founded. Wait a minute, how did that work? Well, when we were founded, we acquired a company that was founded prior to us, which was OSI. We've been doing literally fiber to the prem since 1996. Verizon came in, in 2004 and built FiOS, by the way, justified as legend would have it, based upon simply a lower cost network to deliver voice. Fiber reduced their cost by so much it made sense to do. They sold off the properties they didn't want to build. You know how well that worked out for the people that bought them. But that's what they did. Verizon was doing fiber in 2004. When was the next large service provider you heard doing fiber to prem? So you really have to get a sense for scope and scale without these disruptions and how long they take. So we're so -- I said earlier, the bad news is you're all 10 years late. The good news is you're 20 years early from when it will be done. You're literally on that scale. Now let's take the numbers. We're literally opening new accounts over 100 a year. It's data-driven. We're understanding where to go target them, execute against that accordingly. If you're opening 100 new customers, by definition, the majority of them are going to be smaller because there aren't a 100 large customers. There's these many large customers. There's a lot more medium-sized companies and there is tons of small ones. As you heard Cory allude to and Matt, we're opening customers in all of those segments, but obviously, many, many more of them here. Now directly to your question, I'll approach it almost from sort of a Tier 1-ish type of question. We're going to continue to call on all those folks that we think are aligned and continue to help them come on to the platforms. There's clearly going to be a predisposition to smaller service providers than larger at the rates that we're trying to add service providers to it. So long story short, we do not count on any one service provider of any type moving these numbers. And what you're hearing today is we have enough run time now. By the way, we did the 145% slide -- Cory did that slide just so we could have an acronym called CAGR, which I quipped the other day is also I think you might have coronavirus if you say CAGR too many times. That 145% growth rate coming off of here, combined with this, is allowing us to say, okay, for the next few years, we can see what's happening. It's going to be, we believe, driven very predictably. So it's not like we're trying to avoid large service providers. But even inside of large service providers, you have to understand the compression that's going on in our industry. So if you go outside and you look at an E9-2, you take an E9-2 and you build upon -- you build split ratios, you do all those things, there is 5,000 subscribers on the end of that 1 system. You have to get your head wrapped around what these new networks look like. So the good news there is the opposite of the bad news. We're leaving the environment where any one customer -- why put my head on the pillow at night and wonder who's blowing up tomorrow. Just -- I don't even think about it any more literally. I sleep very well. So I'm just trying to cover this every way I can. And by the way, if any of you are looking at us from a trading standpoint to make a trade, I would hope this entire presentation dissuaded you from that. This is about investing over a long period of time and understanding this disruption and deciding whether or not we're well placed. I believe us to be. This is not about what the latest big deal that's going on. And increasingly, in a disruption, big deals get smaller at a rate because you're getting compression. But the other thing that's happening in those big deals, typically, is they're box deals and the margin's rolling down very fast. So if you look at what Cory just put up there from a margin standpoint, what we're doing, I challenge you to go look at other communications box companies in the industry and actually go look at their margins over the last 5 years. Chart them, tell me what you see. So it's all about whole model there. So hopefully, that answers your question without hammering it too hard.

Can you give us an idea of what percentage of revenues comes from platforms in '18 and '19?

From -- came from platforms in '18 and '19?

Yes. Percentage of total revenues from platforms in '18 and '19.

Unfortunately, I'm going to be the one that gives this answer because I'm the one that's going to take this arrow and I'm going to be snarky. Yes, I could. No, I won't. The best that we can do, in my view, is to do exactly what Cory did on the growth and give you that sense. And unfortunately, because of the nature of the platform business and the reason we did not put dimensions on it is I can't emphasize enough the land and expand nature of this that tie to our customers' success and then the ratable revenue recognition. We are laser-focused on that curve. We were out adding new customers as fast as we can. And we're looking at how fast we're expanding inside of it, knowing that the rest will take care of itself. So unfortunately, the answer is no, I would not.

Okay. How about this one? So after you've landed and now you're going to expand, when you expand, can you give us an idea of a typical aggregate revenue opportunity by customer, by broadband subscribers, which is the way you like to talk about it? So let's say we have a platform, AXOS, Calix Cloud, whatever, sub-250,000 subscribers, a typical revenue opportunity once you've expanded would be X to Y, et cetera.

I'll come at it 2 different ways. If you look at traditional access networkings environments and you look at the CapEx that goes into it, you can sort of put yourself to sort of $100-ish per subscriber on both ends of the CapEx side. But I think your question is a little more refined and you're also trying to understand the platforms and what they look like since they're ratably recognized. And so I won't answer the question directly, but I'll direct you to what I said at the earlier side. If we can help our customers succeed -- and you saw some of the success metrics that you saw Matt presenting, for example, with ARPU raising, et cetera. Obviously, we're not charging more than whatever their ARPU is. So there's something less than that. And we think there's some percentage of that, that customers would be willing to share. Narrowing it down beyond that, I would not do. So I appreciate that's 2 attempts, 3 and we're right out. But I apologize, Chris.

Is there any questions you're prepared to answer?

Yes. I'll think of them. I'm 63 and devilishly handsome young man.

Okay. Well, we'll give it another try. So if I can summarize your answer to the first question, your 5% to 10% growth range does not contemplate the addition of any major new Tier 1 customers, of the sort of opportunities you have with Verizon and CityFibre currently. Is that correct?

It does not -- it neither contemplates them nor eliminates them. It just looks at it statistically in what we've seen. So the important part here is we have enough run time on this model now where we think we understand how this grows.

Yes. The current model...

Well, it is not dependent upon...

With your current -- right. Understanding these things could be step functions in both directions, which is why I understand your reticence there. However, let me paint a different scenario, which is, is it possible that now is exactly the time you should be pursuing these major opportunities? Because you've got your largest competitor in the market in the midst of strategic upheaval and your second largest in the midst of political upheaval. And your third kind of on the mat as well. So there would seem to be -- and a brand-new platform that customers seem to like. So there would seem to be a lot of opportunity here. And while I understand that part of the problem amongst your historic competitors has been, to a greater degree, of customer concentration. It seems like you should be able to curate a Tier 1 and added to the stable every year of the sort that you like. And I wonder why you are not giving us more message along those lines.

You actually have enough to answer your question in and of the question you asked. Because that which put them in that position is the major reason why we will tread carefully. Because the challenge that you have in a disruption is the first thing that happens is larger inertial customers drive to the lowest common denominator, which is price, and they don't see the technology coming. So I can take you back to every one of those disruptions I showed you. I can take you back to IP and ATM. What happened in the ATM market when the IP market started driving? ATM started to get overcrowded, commoditized, competition, revenues per blank went down, margins dropped precipitously. And you see this over and over again in a disruption. So actually, the worst thing we could do is to go chase those deals. It's counterintuitive because there is revenue there, but there's little value there. Now let me give you a corollary that is valuable. We could very well add a large customer, a name that you would all go, "Wow, look at that, they've got so and so". And they may -- they would never become a 10% customer. We can have a large customer go and deploy Revenue Edge. Remember, all these software platforms we have are abstractable. So in our future, we may not necessarily be deploying hardware. And likely, if you're looking at customers that are larger, they might be the first ones that we might do a software-only model with. We could do a very large customer software deal that would come close to being 10%, you'd never see it. It would be very important, but you'd never see it. But we have to be very careful about saying no to things that are counter-strategic. And when you see that disruption and you actually -- I think you know this from your history. If you go chase the void of a competitor, you -- don't be surprised if you don't meet the same fate, right? The customer is not changing their behaviors. So let me give you a corollary. So let's take 2 of the folks -- or 3 of the folks or whatever you were talking about, I don't remember, which. They were nearby different things.

It was Nokia, Huawei and...

Political, financial, something like that. I didn't mention them. So -- but let's take an example where there's a large service provider that has historically had a 2-vendor model. And they might be, let's just say, on the continent of Europe. And they've had a vendor H and a vendor N.

They're sort of done with copper and they're moving to fiber.

But they've had those 2 vendors. Okay. And they decided, maybe we're a little concerned with the political H thing, so we're going to go do it with another vendor, a third vendor. But because we're concerned with H, they might become a second vendor. So everybody is doing okay with that? Does that sound like a reasonable statement? It could be a big deal. Well, here's my first question to you. If they deployed H and N, do you think they were deploying H and M based upon value or based upon price competition. And we know the answer, that's a price competition. So when they go look for vendor number three, what do you think they're likely to be picking it on. Price, okay. So let's just say -- if we think about it, it's a revenue opportunity. And maybe we look at it and say, let's go get that revenue. Michael bounds in, the ever-enthusiastic leader of the front end of the business, and said, "I've got a deal for you," right, which, by the way, I will tell you, it's not in our culture. Michael has led the team to the point where the team does a very good job of selecting out things that are not aligned. But let's just say he had coffee and Krispy Kreme donuts one morning, he's full of energy, which, by the way, he is known to do.

There's a story there.

There is a story there. That's true. So -- and we look at that. Well, we've all been in networking before. How long does it take -- so you're the service provider. You're running the deal. How long does it take you to issue the RFP, narrow it down and select the first vendor to go on the ladder?

A year.

A year. Let's say 9 months, right? Let's say it's Christian because he does stuff faster. So it's 9 months. Now you select us. He said, "Okay, we're going to bring you to the lab," right? In and the lab, they have somebody like Alex. He was educated technologically. And you hired Alex in the lab to do what?

Break it.

Break it. You didn't hire him to approve it. He doesn't get marks for approving it. He gets marks for breaking it. And if he breaks it in a way that you would never use it in, it doesn't matter. He is going to break the thing. And he's going to stand there like Mr. Clean, which if your hairline recedes fast enough, maybe you'll look like. So Mr. Clean there in front of the lab going, "Not letting you -- sorry. They give it back to you. We have a complex system. We hand it over to Michel. Michel -- by the way, not that Michel hasn't been through this several hundred times in his career. We hand back and say how long does it take you to make the changes, integrate it, test it and hand it back to Mr. Clean.

2 to 3 months.

2 to 3 months. He was very fast. Let's give them 3 months, okay? So this cycle happens a couple of times. So let's just add another 6 months. So maybe it's 15 months at this point in time. Meanwhile, you're sitting there, we're negotiating the contract. So at least we're going to do something in parallel. So now we're negotiating the contract, he's doing this. It's like a time waste because we're not making progress. So 15 months end, we're now at the point of saying let's go do what? Well, you know the acronym, FOA, first office application. Well, how do you do a first office application? Oh, we can't do the first office application until it's integrated into the OSS and BSS. How many of you have been around OSS, BSSs inside of telecom service provider? How long does it take to integrate?

I think I get the point.

6 months? No, no, no. You may have, but I'm not done. I appreciate the fact. No, you're not getting out of this room. It's 6 to 9 months to get it integrated. So you're 24 months out. Now here's -- now we're the third vendor, maybe the second vendor. No, not that the news moves every day, but can you tell me what the geopolitical environment is going to look like in 24 months? Well, just say it's the same as it is now. Okay. That H company, you didn't rip out all the H stuff. You just said, "You know what, folks, in this portion of the network that's H, we're going to keep buying H cards, but you're not going to put H anywhere else." They're stuck here. They're capped, which means the remainder of the network is now available to something. But there's an N, which has created sort of squeeze out of there. By the way, N is very clever. N knows that you brought a third guy in. N is going to sit there with you and not move your pricing right up until you just start to deploy them. And then they're going to come back and say, "Hey, we have a discount." And all of a sudden, burp, down goes the pricing. Now we're back saying, "Oh, the RFP, it said that." "Yes, I know, but you know our existing vendor came in and cut the price." Does this story sound familiar to you? By the way, he has consumed dozens of Krispy Kremes across this process. The good news is he's exerted so much energy, he's actually lost 40 pounds. And we've gotten nowhere. So we're exerting all this energy to go chase what appears to be the chimera of revenue. If you're right in front of a disruption, the worst thing you can do is do anything other than focus on that disruption. We do not have time for long yeses or long nos. You want a short yes or a short no. It's either aligned or it isn't. Boom, boom, sign here, sixth copy is yours. Fun experience? The third, in case you don't remember this name, [indiscernible] deals, where 1 call closes and the transport switches place. We didn't do any trial of things. You want to do a trial, buy it. When you're in front of a disruption, you have to be that acute or you're actually squandering your future strategic value because we're chasing revenue over here. And by the way, if we were ever going to do that, we would have done it already when the traditional business did a big step down on revenue. And to our credit and to the team here, this team has done a phenomenal job of building people at the front end of this business that recognize a misaligned customer and stop there. I can't emphasize how important that is. If what you do is instead have your salespeople out calling on everyone, you're not data-driven, you're not sorting the company, you're not doing all these pieces, they will go out and create a selling motion, an inertia, an action. And then it comes back into somebody like Steve, and Steve is the captain of the order prevention desk. He's the one that says if it goes on or if that price is too low or whatever. Yes, except you just exerted how much selling? And eventually, he's like a goalie, somebody's going to get one past him. It's just a bad way of going about it. So comprehensively, it's really important that you understand the selections that we're trying to make. This is very much about us trying to find people that are aligned and wanting to go the same way and getting them onboard. And by the way, all you have to ask yourself is, are there enough of them? We've already answered that question for ourselves. That was a question we had 5 years ago. It is not a question we have anymore. We're beating that to death. By the way, just in case you're worried about it being beat to death, Michael is going to beat it to death some more.

Probably a little bit different tack because you've done a great job beating that to death. Even after he said I submit multiple times. So the other thing is that we don't push upgrades. So why push an upgrade from the old technology to the new technology? Actually, no. We educate and we educate the customer and we educate the customer and educate the customer. And then at some point, whether it's because they see one of their peers or they see a success story or they go to a webinar and they understand what's going on, then finally it's going to click, and they're going to say, "Ah, my leadership team now really understands why we want to do that or there's a business need, whatever it is, we're now ready to be aligned and actually go to the next step," which is why it's so important with the marketing organization and everything we're doing in customer success, everything we're doing with our press releases and the case studies and all those things to educate. We have an event -- in this disruption, our role is evangelists. We are evangelizing new business models, new approaches to the market. And not everybody is going to get it, so you have to align to the ones who get it, amplify their success so that the next wave gets it. It's a classic adoption curve, right? And then what generally happens is, at some point, the laggards -- so over that period of time, the laggards come along, whether dragged, kicking and screaming, or whatever it is, or they go out of business.

Well, the service providers out there, let's take another example, that are losing subscribers under ARPU pressures and are fighting churn will eventually realize that these other service providers, they must be doing something different. So let's just focus on helping them win and we'll get on with the rest. So I'm...

Okay. Wait -- so I'm going to add one more thing. So let's just talk about what you said with regards to losing subscribers and adding subscribers, right? What's fascinating inside the telecom industry is that everybody focuses on adds, right? Adds, you've got to have adds. I hear this nonstop. I'm all about acquisition. And I turned to them -- and this is the same thing when I was at a large carrier, first thing I did once I took over the organization, I went through all the data. I sat down with the president and I say, do you know that I am a sales rep? Through last year, it was 200% of plan, and he sold 1,000 units and lost 2,000 to his 200% of plan because the competition is all focused on acquisition. You're not focused on the right things. Acquisitions shouldn't be your first focus. I hear all these companies Say "I got to acquire, I got to acquire. Look at how many I added." I'll tell you what, you know when Matt showed up 0.2% churn, that's the only churn that should be acceptable. Instead, what does this industry do? They actually take annual churn and they turn it into monthly churn so they don't feel bad about themselves. Let's be really clear, 2% monthly churn is awful, absolutely awful. It's 25% a year you're flipping your base. At what cost, right? So back to the revenue, and I think we talked about how we're going to educate them. Start with churn, focus on actually keeping your existing customers. Then what do you do next? Find out all your customers who are on this level of the service when they could actually be on this level of service. And the difference between the 2 is profit. It's all profit just sitting there latent, waiting for you to make money. So fix your churn problem, then raise your revenue by actually maximizing your existing investments and then go to acquisition. Instead, they all do it the same way. I start with acquisition because that's the only thing they understand, because they're like -- you know what, it's a leaky bucket approach. It's wrong. And so when we were at Bell Canada, in the 3 years I was there, we went from 5 years of decline to 14 quarters of double-digit growth and a 25% increase in ARPU. How did we do that? Churn, then maximizing revenue and then acquisition. So these types of philosophies are the things that as an evangelist in the business, we've talked a lot about the AXOS side, about the Revenue Edge side, that's our mission, to help these guys be wildly successful and go to profitability. And if they're a community-based organization, we always say the same thing to them. I get it. You actually don't care about profits. But boy, what's the chance of you being reelected when you actually put a couple of hundred dollars in the jeans of every single community member, right, of the co-op members and you actually increased your investment in fiber and all these different elements you made a community better? Really high.

So by the way, we have roughly 30 minutes left. So I think that's time for one more of Tim's questions. Just kidding.

Super fast. Follow-up is can you describe kind of what the circumstances that define the upper and lower end of your growth range? Is that macro? If we're looking at 5% to 10% growth range, what are the -- what drives you to the high end versus the low end?

From my perspective, the larger component of that is roll-off of traditional customers. So to the extent that, that roll-off happens faster, we'll probably be at the lower end. If the roll-off happens slower, we'll be at the higher end. There are some things that we can do from an amplification standpoint that are yet in front of us. So you saw the early suites that were being talked about, those have potentials. So there's just potentials for other things that we're not talking about today that relates to future products. We're clear about our innovation cycles now that the platforms are built. We do something that you do not see in this industry. Our innovations are released to the market every 91 days. There are 4 cadence releases every year. They are on the second Tuesday of the second week of this -- second Tuesday of the second month of each quarter. Those dates are known. And so there's lots of innovations coming, none of which we're talking today about. And so clearly, there's opportunities that are ahead of us that we think will continue to expand that. So to me, the upside is more about how fast new things have taken off, in addition to how well we're executing on the plan. The thing that pulls it back down is potentially faster roll-offs on traditional. You need a microphone?

You need a microphone?

Yes. I had a really simple question.

Oh, Alex, I'm sorry.

It'll take just a sec. I should have asked earlier Cory, I hate to focus on this here, but just as a departure point to the guidance you gave on the '21 plus model. I assume you're comfortable with consensus -- the question, are you comfortable with consensus revenue numbers?

On 2020?

2020, just as a departure point relative to the guidance you're giving.

Let me be semantically correct before Cory answers. You mean on his thoughts and targets, not guidance? Because we didn't give any guidance.

We haven't given guidance.

I recognize that.

Okay.

I recognize that. But to the extent you're giving a model for '21, right, 5% to 10% growth. As a departure point, so the 2020, right...

That's what would I use in my models.

What you used in your models, what consensus has? I assume you're comfortable. I figured I'd give you some relief after the last one.

You didn't give me any relief. You gave Tim some relief.

You mentioned that you weren't sure if there would be enough customers who want to go along for the ride that can set up and you convince yourself that there is -- that there are. What can we do to understand how big that market is, how fast those people are growing and how healthy they are? I continue to struggle with it because it's so many small people that it's hard for me to add them together and believe -- you talked about 1 or 2 that's been great, but then it's hard to extrapolate that across.

Well -- so there's 2 things I think that we can do. And one that we are doing, which is we're giving you a sense for the number of new customers we're adding. Because here's what you do know, we are not acquiring any new customers today with our legacy products. Every single new customer is being acquired with our go-forward platforms. So that's an indicator. The second thing you could do is stop doing what you're doing and go start a service provider and we'll sell to you. I mean that would at least add one. But I don't know if there's any other -- I mean, Matt, are there some data sources that you would point folks to that might help?

Well, I mean, I think you can certainly follow all of the reports that are out there around the different government programs, right? So certainly, the existing CAF programs. Certainly, the RDOF process that some of you are aware of is going to be kicking off later this year. There are hundreds of potential bidders who are coming together in consortiums themselves, and we believe a very high percentage of those would be new service providers in the market. And I think that's going to be a really good indicator because, as Carl said, our new customers are adopting our new platforms. And the way that the rules are set up for those programs, they actually will be better -- they will be written in such a way -- or they are written in such a way that folks who are going to be providing gigabit fiber-based type services are going to be very highly favored. So I would start to look at those kinds of programs, who's applying, who's winning. They will all give you an indicator about who are the newest entrants into the market and they will be a lot like many of the companies that we talked about today.

Just to use the U.S. as an example, do you know -- is there a place where we're going to see -- or do you know like how many broadband people there are that aren't part of legacy...

Yes, certainly. I mean we can certainly talk offline. But I mean there's -- we look at sources, for instance of every electric cooperative in the country, as an example. And we know who from the FCC is currently providing a broadband service. We know who's applied.

We know, but if you start there -- the FCC has a list of all broadband service providers. Yes. So if that's what you're looking for, quite good if you start there.

So yes, I'm giving you digital insight into how you can determine also who might be entrants coming into the marketplace.

And get together with your neighbors and start up a co-op, we'll be ready to talk to you. Other questions?

I want to try to ask this question a third time around based on what Christian and Tim said. At what point would you be comfortable giving out more platform information? Is it when reaches a certain size? Or -- I understand why you're not doing it now, but it's kind of useful information, obviously, for us to have. But at what point would you be comfortable, when it's 10% of revenue, when it's 20%? When -- just trying to understand that philosophically.

Yes. Philosophically, it's been asked. And the challenges, there are things that we think are useful and things that we think are competitively useful and things that come under the heading of events that are competitively useful or as Monty Python used to say, right out. So not now, not soon, potentially not ever. Now there's another answer to that, which is if what we're saying turns into reality. Here's what I think. I think long before we start to do that sort of information sharing, most of our investors that are understanding the company will have already figured it out. Does that make sense? You will have already started to flip the model and understand what's going on underneath the covers. You don't have to be a Wharton econometric grad to do that math, and I think you'll be able to reverse footing. And if I'm right, it will almost be like a fait accompli. So it's not going to happen in the near term because we're so utterly focused on this disruption and what we're doing. And we have no interest in giving folks a clear view into the what or the how of this. We're sharing enough that we think helps you understand the why and what we're doing. So I appreciate -- I appreciate you asking. I understand it. Clearly, we look at indicators. And look, what you just saw, this was a -- not a big debate, but a reasonably big debate to share that CAGR. Even without "white taxes," that sort of starts to understand what's happening and what the leading indicators are. So I apologize for yet again giving you an indirect answer. The improvement I've demonstrated is I've given you an indirect answer in 1/10 the amount of time that I gave to Tim.

On the customer growth CAGR, can you give us an idea of -- you talked about -- at length about customer acquisition time at large customers. Can you give us a time frame for the new customers that you are aligned with, the customer acquisition time frame, the 1-call closes, if you will.

Well, our customer acquisitions can vary wildly by type segment. So I'll just give you some clue. Then I'll turn over to Matt and Michael. I'll let them address it. So municipalities is an example of our customer segments. They can have a gestation period of years. But Matt, Michael, do you guys want to speak to that and sort of give some views?

I think the answer is exactly what you said, is that there's -- you can have a 90-day close, you can actually have a 4-day -- a 4-year close, right? It really comes down to where is the customer in their business cycle, right? Are they an existing customer who is under competitive pressure and looking for an alternative and they understand it, right? Is there a change of leadership who understand the power of data? Are they actually a co-op that decided, "I'd like to get into broadband. I'm not sure what that is. Can we have a conversation," right? And so the key thing is -- if you follow the philosophies of sales, is actually have them wide and broad and then have a whole bunch of them closing in each quarter at very different cycles.

Well, a follow-up to the question that was already asked earlier. We spoke at length on we didn't want to target big Tier 1s that -- for a 24-month time frame, and now we're going to go get a 2,000 subscriber municipality and take 4 years of time to do it.

By the way, that wasn't the answer. The answer wasn't actually that we don't want to go after a 24-month sales cycle Tier 1. The answer was actually going after that Tier 1's bad business, right? And it's the fact that you spend 24 months and then, at the end of it, you have really crappy margins because you're the third man in and you had no value.

So really -- I just want to make sure I understand that. So the real point of that conversation was we have no interest in being price aggressive with what we believe are leading-edge products that nobody else has, but if somebody wants to pay full price, we'll take as long as the process takes.

Or if someone is strategically aligned and actually sees value in it, right? So they actually have to understand the platform. Here's -- again, we're evangelizing a different process. So if you just want to buy a dumb box and buy a cheap PON, that's not us, right? If you want to actually go through and address your operating model and build out a new business that you can get to what we talked about, which is the lowest cost per bit per mile, highly operationally efficient with a lot of flexibility business model in the future, then that's the conversations that we're going to have. And as with regards to Tier 1s, I would just say that what's going to happen is that I do see that as some of the successes continue to be expanded out and the TCOs prove out, other people are going to start looking and saying, "Wow, look what Calix is doing. Okay. Now I finally get it," right? So there's also the settlement is that -- and those who aren't strategically aligned, at a future day, they're going to start realizing there is a potential.

But Michael, I would just say though, on average, for a small service provider, the time from decision to go do something, close a deal to deploy is typically much smaller than what Carl described. They don't have the teams like the ones that he talked about, right, procurement to technical evaluation. I gave you an example of Forked Deer. We provided it as an example because we thought it was indicative, right? So 2 years ago, they announced. There's a period of time through which they did the evaluation decision. They kicked off. They deployed in 8 months, they're up and running. So all of that happened within a shorter time frame than what Carl described in that large service provider.

Let me add one other piece that might help frame a little bit of what we're dealing with. You heard that, in that chart, the units were hundreds in the 145% CAGR. And by the way, we've made announcements in the past about roughly where we were with different deployments and then we've stopped. So you have a sense for what those look like. But there's another piece in here that is a set. In addition to adding new customers, we're also approaching something -- close to 2,000 customers now. I, for one, believe the vast majority of those customers will come along with us on this journey. Some may come today. Some may, as you heard Matt and Michael talk about, 3 years from now go, "Oh, I could have [ VA ]. Let's start." I don't know where in that process we'll do it. And the other piece you heard Michael say is they'll do it this way. They'll insert it in a [ prime revenue ] build. We don't need to go unplug things that are doing -- they may have E series that are running [ DXA ], all the greenlights are flashing. It's doing a great job. Don't take it out. Go take those dollars and put it into the Revenue Edge, put it into places where you're going to materially affect the business. When I was listening to what you're asking me, a thought occurred to me. Another way I was thinking about the answer to the question is what are the odds that what we're talking about is going to shift their business model. And the ones you were talking about, Tim, Mr. Blutarsky's grade point average, 0.0.

[indiscernible]

No, no, no.

I mean new ones.

The ones you were talking about, right? All they're going to do is look for the same thing they were looking for. They're not going to shift their business model. Verizon, shifting the business model. So that's also what drives this because, remember, we're focused on bringing platforms to the service providers that are in front of this disruption and helping them build an entirely different business model. That is part of the selection criteria.

Yes, part of the selection. If I may, to the CAGR slide portion. You've got to look at what products we're talking about, the adoption decision to say I'm going to change my access network and SAAS. The adoption cycle to adopt the cloud, significantly short, we're talking months. The adoption cycle to pick a new piece of [ CPE ] and start deploying your first unit, much shorter. The adoption cycle of new services, whether you add modules or software, you can decide in weeks you want to do that. So you've got to look at what is the product that you're talking about, what is the platform that you're talking about. Intelligent access, the cloud component, once you're integrating, you've integrated the cloud, turning on new capabilities, your second cloud -- or I've got marketing cloud and I want support cloud, that's instantaneous. The same thing, I want to decide on -- I love the GigaSpire. I'm ready to go, let's go. Again, very short decision cycle. Software and new software suites, instantaneous, right?

Okay. So we're about to run out of time. We can take one more question. Tim -- Tom, I'll let you pick it. And then, by the way, we're going to adjourn. And look, we're going to -- the showcase will be open. We can continue from there.

Two, if I can. Just when we look at the company right now and 3 years out, we're going to expect it to be 20% 30%-plus percent larger. And you talked about the customer counts that you're growing. To get from here to there, is it that the customer count -- so you should be adding 20%, 30% more customers or that those customers that are already in will be growing on top of the same kind of annual 100, 120 new customers?

Yes, yes. I mean it's just -- well, because it's -- look, the difference between what we were and what we are today, you have to think about this in much more terms the way a software company runs. It's a land and expand model. That's what I can give you. So we're going to be very focused on adding new customers. But also very focused on our customer success inside of our existing customers and expanding entirely.

But the point is -- so from what we see, which ability to disclose customer count increase, is that there doesn't need to be an increase, so it's not a, "Oh, my God, they only added 100. The last....

By the way, oh my God, anyone adding 100 new service providers a year, good luck finding anything like that in our industry.

Cool, that's an unbelievable rate.

But I think it's also the expand point that Cory was making, is as our solutions are deployed more broadly across our subscriber base and they go from being an Intelligent Access Edge customer to being a Revenue Edge customer to find the actual suites on top of that, a lot of the growth will come from that expansion within existing customers.

We added 0 customers.

Right, which we will have 0 visibility on. That's my point. Just trying to make sure we establish the right criteria.

So if we added 0 customers in any given quarter, we'd still grow, right? Obviously, that's not our goal. Okay. Tom.

Carl, can I just -- I apologize for [indiscernible] but this is important relative to this new customer number. If I recall, once upon a time, I recognize the numbers have changed over the years, it's something on the order of 1,500 IOCs, colloquially speaking, in the United States. The CLECs community was a fraction of that number. And there's not -- last time I checked, there's not much of what we used to consider CLECs left. So when you talk about adding new customers, there's new customers as in the existing customers have been around for decades that you never sold to. And then there's new customers that weren't in the business of comp services, whether it be ALLOs or all the electric cooperatives that are relatively new. So I mean, frankly, 100 new customers a year at the context of where the industry is at strikes me as not a lot, strikes me as it will be increasingly hard as we go forward. But the question would be, what is your penetration rate to the existing universe of customers before all of these relatively new folks entered into the arena?

Okay. So I'm going to give you an answer to your question, then we're going to call a halt to the proceedings before Tom comes over and gives me a hook. So you heard a comment earlier that, look, there's on the order of 8,000 to 10,000 service providers worldwide of different types than ones that might be coming in. So there's your universe. You'll see additional service providers coming in that are greenfield that don't exist. So it's going to continue to grow. But no matter how you measure this, it's a high-growth rate for new customer acquisition, but we're going to stay very much focused upon it. Because it's more central to a land and expand strategy, along with bringing our set of customers that we already have today along with us. So I'm sorry for being [ impossible ], but you can ask questions afterwards. I want to thank everybody on the webinar for attending, especially want to thank the folks that are here physically in attendance and braving these pieces and for, look, your patience being investors and understanding where we're going. We're very happy to be in a position now where we have enough run time to say, look, here's what we think we see in these next few years, and we wanted to share with you. But we also wanted to make sure you understood the opportunity that we're pursuing, the size of it, the technology advantage we believe we have. But for me, most personally and what is flattering most of all is the team of people I get to work with every day. It's an extraordinary team of experienced, diverse professionals leading, by the way, an extraordinary team of diverse professionals in the organization -- in these various organizations, structurally underneath them. Please get outside and get a chance to engage with some of them. So thank you all for attending. Thank you very much. Tom?

All yours.