Any questions, please reach out to your Morgan Stanley sales representative. For everybody, I'm Meta Marshall. I cover the communication software space here at Morgan Stanley. We're delighted to have Bandwidth here today. David Morken, CEO and Founder, and Daryl Raiford, CFO. All right. Perfect. David and Daryl, great to have you guys back at TMT. Since your IPO, you guys have been fairly focused on voice as the great communication channel, an idea that seems to be kind of regaining more traction as we look to AI and kind of voice assistance. Can you speak to how Bandwidth can best meet the needs of a voice-centric world?
I can, and I will, but I just want to start by saying thank you for having us back. We're delighted to be here and love working with our Morgan Stanley team. It is fast becoming a more voice-centric world, as you're pointing out, with the advent of voice agents. It's exciting to have a universal platform and our Maestro product ready to embrace the emerging trend. There are really five reasons why, if you're an enterprise building a voice agent, you want to come to Bandwidth. It's because of the high fidelity of our service, the low latency, the intelligent routing, the seamless app integration, and the great customer service. Those five elements are fundamental to why we're a good platform. We're also global, which is helpful so 65-plus countries, if you bring a voice agent to our Maestro platform, you are available instantly all over. So t hat's a very good thing. Enterprises that are doing cloud migrations between CCaaS providers and UCaaS providers, it's really helpful to them to already be on our platform. Those are some of the reasons why, in a voice-centric world for an enterprise right now, we are a great fit.
Do you see it kind of becoming this voice-centric world?
So, we have majored in voice for a long time. In fact, I remember something long ago, Alexa, in the earliest days, we supported the launch of it. Gosh, was that ahead of its time. Extraordinary to think back on that time. Yes, things are moving in our direction regarding voice. It feels almost like when work from home was happening, it just benefited us suddenly and fundamentally and pulled a lot of demand forward. What we're seeing among enterprises now is that same pull to migrate to the cloud if they haven't already or to launch a voice agent is pulling demand in a similar fashion. Certainly, that was a phenomenon that came and went quickly, thank God, in my opinion. But I think voice agents are really an exciting dynamic that's similar.
Got it. The voice focus of the platform has always been there in kind of those five things you laid out. The platform has evolved a lot over the past couple of years. Can you just kind of speak to kind of how that platform that may have started as voice-centric has kind of expanded?
When we began, it was domestic, U.S. That's all. Now we're global. That would be the first thing that's important to understand if you're a team out there. The platform has also now supported lots of pre-integrations. Five9, Genesys, all the UCaaS, CCaaS that are relevant. And what we're seeing are migrations between them. If you're already on the platform, you can very elegantly move, and that's vital. If you've got a hybrid engagement or you're still premise-based, the platform now supports very quickly onboarding, deep analytics once you're on board, and super high-touch professional customer support. The platform has gotten much better at all those things over the years.
Okay, got it. And then, how has it, maybe at the time of the IPO, which is now a while back, there was a lot of focus on kind of the UCaaS, CCaaS kind of customer base or kind of that cloud customer base. You have expanded a lot to kind of help enterprise customers. Just can you say one, kind of the success that you have seen there, but kind of what tailoring you did to better go after that opportunity?
So, every enterprise customer has an international component to the conversation. So, the global reach of the universal platform is part and parcel of the value prop for all of the major enterprise solutions that we've offered. Having messaging side by side so that there's both messaging and voice would be the second component that's made us more attractive to enterprise and will grow from 5% of revenue to 10% by 2026 enterprise. It's growing very quickly. But the expansion of the footprint and the capabilities to include messaging are the two that really jump out that have attracted large enterprises.
Got it. And then, maybe kind of going back to that core group that we had talked a lot about over the years, which are these kind of communication software vendors or the cloud vendors and kind of providing underlying connectivity services. Just where are you? You know, we've gone through some digestion, as you noted, but where are you kind of seeing continued traction with these customers?
They were flat two years ago. Then last year, we're now at single-digit growth. We'll grow more this year than we did in 2024 with those customers. They will contribute to our growth as we've guided and will continue to grow with them. They are also engaging in lots of great innovation around voice agents in their contact centers, in their UCaaS solutions. It'll be interesting to see how we support those from a usage-based model perspective. We do love those customers and innovating with them and supporting them.
Why are you, I mean, those customers in general are not accelerating. So why are you kind of accelerating if they're not?
So share is good. The particular use cases we support at them may be growing more than others that are supported by others. We are excited about having grown with them and what we do with them. International also contributes. We have a unique footprint where others may not.
Okay, Okay, perfect. And then how, you know, how Maestro has been another great product for you, kind of providing that vendor agnostic layer and kind of integration layer. Maybe for the audience, if you could just kind of explain the Maestro platform to start.
Maestro is an orchestration layer. It sits on top of our universal global platform. What it represents to an enterprise is the ability to get from a premise-based solution to the cloud and then, once you're there, have the flexibility to plug in third-party apps. If you're using Verizon and you're plugging into a premise-based piece of equipment, you're not going to be able to, if you build a voice agent, you literally can't plug it in. Maestro has a solution called AI Bridge, which lets you take advantage in three different ways of pre-integrations with Google Dialogflow, with Cognigy, with things that are already available in the Maestro platform. If you want to engage with VoIP protocols directly, whether that's SIP or else or otherwise, you can. If you want to consume an API at Maestro, you can. If you want to take advantage of the pre-integrations, you can. The flexibility of consuming and providing voice and messaging through Maestro means you can intelligently route calls without hard-coded or premise-based solutions. So, most of the time, the enterprises that we win are leaving Verizon for the first time. In the case of Southwest Airlines, that was a 35-year relationship. Maestro is what they moved to.
Okay. I mean, how are you choosing? You mentioned Google. You mentioned Cognigy. Just how are you choosing who to partner with for Maestro or to do these integrations with? And how much of that is your decision versus kind of customers asking you for it?
Two different parts of your question. The first part, whether it's Genesys or Five9 or Zoom, those integrations are just best of breed, Gartner Magic Quadrant, folks we already work with. They need to already be integrated with. An enterprise doesn't have to go through all that work when they come to Maestro. The second part of your question, in terms of go-to-market, we're targeting the large enterprise with our value prop. We're a direct outbound sales motion. We will take anybody that wants what we offer for the right reasons. We're excited about having a channel that's just been recently successful in adding more larger opportunities than we've ever seen before. We love leading our customers. We love being collaborative with our customers. That includes integrating with them if they have something to offer Maestro. But that, it also means welcoming aboard any Global 2000 customer that wants to be a part of what we're doing.
Got it. Yeah, I mean, I guess the question is more from a perspective of there's a lot of AI startups that would love for you to kind of give them this integration platform so that then they can go to customers. Just how are you deciding, there's hundreds of them, there's one of you, you guys have so many resources? Kind of how to make that decision?
That's a fair question. Thanks for clarifying that. The bias is to triage toward the biggest first, most known. That said, enterprises are building voice agents. They come to Maestro to give that agent a real voice globally. Yes, there's no question we don't have the smallest emerging AI startups as part of the ecosystem out of the gate.
Okay, all right, perfect. Particularly with so many CCaaS customers, you mentioned Genesys, Five9, Zoom, anybody, you know, can you continue to kind of play that agnostic partner who is making no decisions? you know or do customers sometimes try to call your hand, see kind of where your bias is? Just you've done it in the past of being able to be this agnostic layer. But how do you kind of continue to do that?
We thrive being agnostic. Five9 or Genesys love when a customer wants to bring Bandwidth to the solution as opposed to Verizon or as opposed to AT&T or Lumen for obvious reasons. They love the fact that we have an agnostic approach. We do not go upstack and have our own CCaaS offering. We value our role as a platform. But make no mistake about it, bring your own carrier, bring Bandwidth to the integration is something that the largest CCaaS and UCaaS partners prefer.
Yeah, Okay. You know, you noted kind of having evolved this kind of the go-to-market over time to better target enterprise customers. Just you know, how has that taken place? Are there different initiatives for landing customers versus expanding customers?
Yeah, so net revenue retention was 124%, 122%, 112%, I think, net of political. We do expand really well with the customers we have. Our logo retention rate is insane. I mean, it's still like 99% plus year- over- year. The direct motion of our outbound sales team has been consistent for years. But the channel's really only grown over the last two years. We haven't had strong ISV or SI support until very recently. On the last earnings callbacks, I said we've got more larger opportunities than we've ever had before. And that's really from the channel effectiveness. Southwest, which I mentioned, was an SI that brought us to the table to displace an incumbent. I think what you should expect of us is a channel that continues to grow and a direct enterprise-focused outbound sales motion at much larger opportunities than some of the other CPaaS folks.
OK, got it. You've noted the net retention. You also have just kind of one of the highest average customer sizes. And you've noted your Maestro customers are even larger. Just what is it about that platform that's just better suited for kind of larger customers?
I think it's flexibility. Again, these large enterprises are bringing into a call flow 10 to 15 different AI use cases from third-party vendors. If you're trying to do that with Verizon or AT&T or Lumen, it's just insane. You're trying to do sentiment, you're trying to do transcription, you're trying to do recording, you're doing forwarding in an intelligent way. You just can't do that with an incumbent. I'll give you a quick example from our earnings call. We've got a large hotel chain that has gotten a voice agent that's an artificial voice agent. We used to make a penny and a half for a five-minute call to the front desk. That's now making us $0.045 because instead of just connecting you to the front desk, we're simultaneously doing sentiment, recording, transcription, and forwarding all at the same time in a way that the AI agent, the voice agent, resolves your call as a guest for far less OPEX than the hotel used to have to pay, but are paying us $0.045 for that five-minute call instead of $0.015.
Okay, got it. Maybe turning back to today, kind of given the expected cyclical reduction in political campaign messaging revenue, what specific growth drivers will offset kind of the headwinds to achieve the 8%-11% normalized revenue growth rate for 2025?
Daryl, you want to take that?
Yeah, certainly. Thank you, Meta. And it's a pleasure. Thank you again for having us. It's a pleasure.
Pleasure to have you.
We achieved 25% revenue growth last year. We were very pleased with that. We've guided essentially flat revenue growth at the reported line because of the cyclical nature of the political campaign revenue of $62 million that will be absent, we expect to be absent in 2025. That leads to 9.5%-10% organic growth at the midpoint, 8%-11%, as you said. We think that that is, we know, we can tell by published reports from others that it's outpacing the market of our peers, of our competitors. We feel really good about that number in the sense that our enterprise business grew 29% last year. We're expecting likewise growth, if not slightly more, embedded in our guide. Our global voice communications plans customers, which are our largest customers, is 70% of our revenue, grew at 3% last year off of flat, off of essentially being flat the prior year from that, from 2023 into 2024. We're expecting nearly doubling. And that doubling, there was an earlier question on that, that doubling is coming from usage patterns as well as the software fees that are involved and you know, along with the Maestro platform and things along those lines. So really, we feel like the growing mix, the growing of the, say, customer category and the growing richness of the underlying revenue itself with the software fees and the platform fees allows us to grow it above market or at least above our peers in terms of that 8% or 11%.
Got it. I mean, you mentioned kind of additional software content. Just can you give a sense of how you're monetizing Maestro?
There is a platform fee. That software platform fee then has usage revenue on top of it. But i t is a software platform fee with gross margins that are consistent with what you would expect in SaaS.
What we like about it is there's not a strong thrusted strategy to say, let's monetize to the maximum extent a software fee. Let's use it as a, let's use our Maestro platform and our capabilities in the orchestration layer to attract and onboard new one-plus, multiple one-plus million dollar accounts, which our enterprise customers turn into, and drive the usage across the platform. You see that in our gross margin. You see that's why our gross margin keeps expanding with that as well.
Got it. I mean, you guys laid out kind of the 8%-11%, as we just kind of talked about, for 2025. Just how do we think about this kind of versus the 15%-20% annualized targets from the analyst day in 2023, understanding kind of 2024 being a growth year beyond that?
Yes, absolutely. So, at the end of 2022, for the very first of 2023 analyst day, we set out medium-term targets over the next four years from that point of 15%-20% revenue CAGR growth. And that would take us through the end of 2026. Last year, we experienced 25% revenue growth. This year, we'll experience less. We certainly understand in 2026 the recurring nature of the cyclical political campaign revenue. And so taking all that into account, I think that we're on track for that. I recently saw a published report. I didn't do the calculation myself. I saw a published report that through 2024, our CAGR was 16% or something on the top line. I think we're kind of in that zip code. We expect to achieve our medium-term targets.
All right, perfect. You know, one of the things you highlighted last year kind of at your customer event or product event was that you're now kind of a directly connected aggregator. Just how is this enhancing your kind of positioning and messaging and kind of changing the structure or P&L structure of the business?
So, deliverability goes up, reliability goes up, costs go down, scale goes up. We had a season in political, for example, Meta, that made Black Friday and Cyber Monday look quaint. The presidential election was bonkers. Direct connects help you scale infinitely, really, really high.
You saw a little bit of that occurring in 2024. We became a direct connect aggregator at the front end of the fourth quarter. We achieved a record margin of 57%, non-GAAP margin of 57% last year. It's implied in our guide. Our guide this year, Meta, you had said the 8%-11% organic revenue growth. We're guiding up our EBITDA and EBITDA margin higher than that. It's implied that gross margin is growing. One of the gross margin drivers is the direct connect relationship we have with the Tier one in the United States, which essentially removes wholesaler costs, completely eliminates it on that side of the business. That just, that just accrues to our benefit.
Is there a path to kind of add additional relationships there?
There is a path.
Okay. Another big opportunity that you alluded to earlier, David, was kind of the international opportunity for Bandwidth. Just what specific regions or markets are you prioritizing for expansion in 2025?
We're in 65 plus for full PSTN replacement. What we're focused on right now is really cross-selling and upselling into those markets more than greenfield new countries. The opportunity to have our existing customers really involved in each one of those is the focus for 2025. We've come through seasons of adding additional countries. We're really focused right now on maximizing each of them. We don't see anybody else following us around the world in a way that makes us feel like we need to continue outer regions as much as maximizing the impact with our existing customer base in countries where we already are.
Is there an effective sales motion? I mean, it seems like a natural kind of area where you would be able to kind of cross-sell with customers. But is there, is it just kind of the amount of time and making sure that you're making contact? Or w hat are those initiatives?
It is a natural upsell, cross-sell motion. You can align incentives with your sales team to go after it. Yes, it is what we think is the right thing to do in this season with the footprint that we have.
All right, perfect. Maybe, Daryl, coming back to you on the profitability outlook, just where do you expect to kind of see the most leverage in the cost structure this year? What are you doing to make increment, where are you kind of making those incremental investments?
The leverage is going to continue from 2024 into 2025 with gross margin providing the largest benefit. We grew gross margin 2 percentage points last year to that 57%. It is clearly implied in our guide it will grow again. The dollar volume of that causes us to be able to afford OPEX growth. We grew OPEX 6% last year. We would expect that we are going to grow OPEX again this year. Half of that last year and half of it again this year is on R&D and innovation. AI is one of those investment areas. That still allows us to disproportionately grow our profit to our revenue. Last year, while our revenue grew 25%, our EBITDA grew 70% and 5 points of margin, of EBITDA margin. We expect that to continue. You know, you mentioned our medium-term targets, our revenue targets. Our EBITDA target has us above 20% EBITDA margin. And we believe we're on track for that through the end of 2026.
Got it. I mean, David, when it comes to AI and just kind of making investments, how do you measure kind of where to make those investments, what type of ROI you're expecting to see?
We think voice agents are the area. Supporting onboarding of voice agents, especially from enterprise, is the right focus. Maestro, AI Bridge, and what will support the emergence of highly vertical voice agents is where we're putting our money and our time. The developments are happening at light speed and are really exciting. The intuitive, effective voice agent in the enterprise is giving knowledge workers in verticals like an Iron Man suit to do their jobs more effectively. It's extraordinary. It's happening fast. So that's our focus right now. We're a voice-focused company. It aligns really well with what we already know. But that's where the preponderance of our time and attention and money is going in the AI realm.
Okay, got it. Daryl, maybe back to you. Just kind of what underpins the confidence towards driving gross margins towards or above 60% and EBITDA margins above 20%?
Well, let me circle back to what David just said. I will get to that, which is it is really exciting when we talk about our AI investments that we will get to the end result. When we talk about the AI investments, all the building blocks are falling into place. We have invested in released Maestro. We have invested in released AI Bridge. We have invested in released Universal Platform. All these building blocks are necessary to enable what we are doing when David talks about AI investment. Those elements then get to your question of what we have been doing in terms of our cloud architecture to allow for improved gross margin, lower cost, improved gross margin, along with growing revenue and scale, along with international growth, and along with product mix, where, of course, we have been very clear over the last 18-24 months. We enjoy, we had 46% growth in messaging revenue last year. We enjoy that to the extent that it's a higher margin business than what we're doing versus the overall aggregate business. You take all those kind of important voice investments and then the product mix, the international mix, and the like, and you put that together. It gives us good confidence in gross margin growth.
David, I know we talked a little bit about go-to-market earlier. But with a richer sale, you know, this is not necessarily saying, you know, hey, I bet you you did not like who you used before. We could be kind of this great service agent. As it becomes a more platform sale and there are more products, does that change who that Bandwidth sales rep is? Does that change the go-to-market? Has it really been the land is the same, and then we can hand it off to a different team, you know, and kind of scale it over time?
We have focused on large enterprise. And once they're on board, we have handed them to a team that focuses on supporting them well, upselling, cross-selling, increasing the net revenue retention accordingly. We have watched that grow from $140,000 a year on average when Daryl joined to now $218,000. That has been really effective. The team that goes out and identifies appropriate customers that can benefit from Maestro is not the team that then grows necessarily with that customer over time, over two years on average, to get to the share of wallet that we would expect to persist going forward. I think that the channel and the direct team do have an opportunity, as we expand the platform, to go mid-market, to certainly target a broader swath. But w e really do well with enterprise teams and with getting them to the scale and reliability and quality, whether that's large health care or large finance. That's been our effective path to market. And we're good at that. I think the AI moment that we're in with voice agents scales massively. When I talk about the permanent charge we're paid in that example, the impact of enterprises deploying AI, I think, is going to be meaningful. We should continue to focus there.
Got it. Daryl, you guys have gotten the balance sheet to a much improved position. Just how are you thinking about whatever last overhang there are, just kind of explaining that to investors and then just how you're thinking about kind of capital allocation going forward?
Yeah, thank you. The state of the balance sheet is very strong. We just recently essentially completed the final repurchase of nearly $450 million of our April 1, 2026 convertible notes. We have outstanding $250 million due right at the first of April in 2028. If you take our net debt calculation of around 250-ish million and take out about 100-ish million of cash, say, just doing a quick calculation, 150 over last 12 months EBITDA of $82 million leaves us at two times, which is a very, very good net debt ratio for a company of our size, a mature company of our size. Our cash flow last year grew, our free cash flow grew 206% to $59 million. It's implied again in our guidance this year that it will continue to grow. We're looking forward to improving the balance sheet. You know, we have sufficient cash on hand. We have access to $150 million undrawn line of credit with Bank of America and Wells Fargo. We are looking forward to cash flow generation. We do not think there is any issue in several more years with the $250 million. Carrying two times leverage is very reasonable. We have no other capital allocation strategy other than to keep the company healthy. We are not proposing any dividends or anything like that.
Okay, perfect. And then just the last question we've been asking everybody this week is just, are there any ways in which you're using AI internally and just any productivity that you've seen from those?
Yeah. So f raud prevention, campaign registration. When a customer is coming on board and having to qualify with a carrier for a messaging campaign, AI has been fantastic in pre-qualifying campaigns rather than submitting them all the way through and getting rejected. So t hose are two areas that are already underway that are exciting. On the go-to-market side, the SDR activity, increasing the productivity there has been fun. The targets that we have are areas where we have repetitive, uncreative knowledge work that you can improve the quality of life of the bandmate, make them more effective. We're excited about it. It's early, but it's exciting.
Okay, perfect. All right. David, Daryl, thank you so much for being here today and sharing the Bandwidth story.
Thank you, Meta.
Thank you, Meta.