All right. Good afternoon, everybody. We have RingCentral here this afternoon. Vlad Shmunis, our CEO, and Vaibhav Agarwal, CFO. Welcome back to the conference, guys.
Thank you.
Thank you.
Awesome. Vlad, let's start with you. How do you define the next chapter for RingCentral, and what are the two to three priorities that will matter most to making it successful?
Next chapter is to embed AI across the whole portfolio, externally, internally, to become just like an AI-native company. Obviously, we predate AI as an analysis, but we were a stat-native company, and we've moved the industry with that. I think we have a very good chance of doing this again, leveraging the best of AI. Again, that will be both customer-facing, as well as a major revamp of all internal operations to become substantially more efficient and to better serve the needs of the market as well.
Okay. All right. Good level set. ARR from customers using at least one of your native paid AI products now represents 13% of ARR, and I think it's doubled over the past year. I guess, what's driving the adoption today, and I guess, what needs to happen for it to broaden meaningfully across your customer base? Maybe we start with you, Vlad, and then we'll move to Vaibhav.
Yeah, look, we're happy with this progress, but there is a lot more to do. It's only 13%. It's not 30%, it's not 90%. Over time, we see it touching the entire portfolio. Doubling year over year is good. Hopefully, that will continue and maybe even accelerate. Look, there is a lot of demand. AI is absolutely becoming the new reality for everyone, certainly for vast majority of our customer base. How do you get to them? We have 600,000 companies, logos on the platform, so that's a lot. To get through to all of them at the right time, so many times we just get to talk with them around renewal times.
Obviously, we're working on PLG type motions as well. We just think that there is a lot of goodness to come, and I think we've shared before, but about half of our new business, if you're looking at new logos, are taking AI. So our AI. Okay? I would say that would be the goalpost is to get from 13% to at least 50%, and then we'll go from there.
Is there anything to read across in terms of the types of customers that you're landing in these earlier stages, within the installed base or even from a new logo standpoint?
Yeah. Look, so firstly, so far, our AI has been skewing more down market. There are some very notable exceptions, but majority, I think we're now 16,000 logos last quarter. Can tell you it's meaningfully more than that now. Most of them are smaller businesses, are SMBs. Where it plays the best, but frankly, this applies to almost our entire base anyway, is they are like B2C businesses, which makes us a B2B2C. Okay? So where Ring shines is when we're serving businesses whose customers are consumers who are calling or texting. We have massive amounts of traffic on our network. Some of the numbers we've shared, 40 billion minutes. We're sitting on billions of transcripts and recordings. We're processing billions of text SMS messages. That's across the globe.
O kay.
As long as consumers are calling their service providers, business providers, we are in a very good position to get in the middle of those transactions and to apply AI before a human picks up, assist a human if there is a human transfer, and then do post-processing and generate insights after the call is done. Okay, so basically, back to your question is, so what makes these customers special is they're the ones dealing with consumers, and they're the ones who need help putting an AI agent at every stage of this consumer-provider interaction, and we're just very strongly positioned there.
Okay, great. I want to talk about AIR. I was at Enterprise Connect, I think, last year when you guys introduced that product. I got to see the demo. It was really compelling. Fast-forward to today, more than 16,000 paying customers. Maybe talk to us about, once a customer adopts AIR, what does expansion typically look like? Is it primarily greater usage, deployment across more parts of the organization, adoption of additional RingCentral products? And Vaibhav, feel free to opine as well.
No. Look, like I say, for now. It's been mostly SMB, so expansion is usage. And there's absolute expansion, just a lot more usage, so it's in the wake of an outcomes-based approach. But it also resulted in stickier customers. Stickier customers over time will take more product. With larger enterprises, sky's the limit because they can be deploying multiple agents. We have various business models that we offer, but in the end, it is about outcomes, and there is just more outcomes to be had with larger enterprises.
Vaibhav, maybe just expound here. You've highlighted stronger ARPU and net retention among customers that are adopting at least one AI product. How should we interpret those benefits today, and what additional proof points would demonstrate that AI is becoming material to RingCentral's overall growth rate?
Yeah. No, that's a good question, and thanks for hosting us, always. Good to see you. That's a key metric for us, the RCAI adopting customers. The reason for that is as we are selling more bundled products, that metric shows paid AI adoption. It's showing product penetration, and it's frankly showing quality of growth that Vlad talked about. That metric is showing importantly that it's showing the monetization potential of AI, and it's showing that customers are more sticky because that cohort of customers has better ARPUs and higher net retention rates. It's also demonstrating that customers are not just experimenting with AI. They are paying for these AI products. They are expanding the use on the platform. They are getting benefits of it, and they are staying longer.
As the adoption of AI products increases over time, it'll start becoming an increasing driver of growth.
Understood. Vlad, let's talk about Customer Engagement Bundle. I think that's now 10,000 customers in a relatively short period of time. What unmet need was that addressing, and how do you see its role alongside RingEX, RingCX, and more of the traditional contact center offerings?
Yeah. No, it was 10,000. It's, like I say a bit more than that now. We're late in the quarter. Yeah. No, it's our fastest growing product ever. It actually has a lot of AI in it as well. It sits between EX and CX. It's officially part of the EX. We call it a bundle. But what it is, it is a lightweight contact center for generally speaking, with some exceptions, but generally speaking, for smaller businesses who don't have dedicated human contact center agents. So that's a rapidly emerging category, and it's actually fairly well-protected, so to call it, from the AI threat because everyone, ourselves included, with a formal dedicated contact center, you know what? You just want to have less agents because they're expensive, and AI agents can do as good a job, maybe a better job for a lot cheaper.
Okay. This is why we're investing so much into our agentic voice AI. But with the bundle, people who pick up the phone, they have other jobs. They can be dental technicians. They can be realtors, maybe junior realtors, right? Any walk of life, insurance agents. They have things to do, and yet people are calling them, and they're calling their business number or texting their business. Somebody needs to pick up, not necessarily that particular person. Many times, the interaction can be handled by an AI agent and progressively more so. That's kind of where it sits in the portfolio. It's basically for our higher-end PBX customers who are not yet full enterprise, and that's just a sweet spot there.
Okay. Vaibhav, let's switch back to you. The newer AI-led products are growing rapidly. We've talked about it extensively, but subscription revenue growth is still holding in sort of the mid-single-digit range. What needs to happen for that product-level momentum to begin moving the company-wide growth rate more meaningfully?
Yeah. Look, first of all, we've gotten growth. The growth rates that have stabilized at a sustainable level. If you look at the last five or six quarters, we've been in a similar range in terms of subscription revenue growth. I think that's point number one. Number two is we have a durable core base that is a recurring revenue model with strong retention rates and high ARPU, so that's the foundation. Our AI portfolio, like Vlad mentioned, is growing strong. It's growing rapidly, and it's showing customer traction. Having said that, it's growing off a small base relative to the overall company, which is at $2.8 billion in ARR. Look, as AI adoption grows across the base and as we acquire new logos, it'll become a more meaningful portion of the overall ARR and be an incremental layer for growth in the future.
Okay, great. Vlad, bigger picture one for you. As communications, contact center, and AI come together, customers have more choices across established platforms and AI-native point solutions. What does owning the underlying voice network create a meaningful advantage, or where does it create a meaningful advantage? Where does RingCentral still need to prove its differentiation in the market?
Look, so that's where the magic is. There is a network. There is also traffic on that network, and there is data that's flowing through this network. Some of it is transient, but especially with AI, a lot of it gets transcripted. So what was transient now becomes fixed, and then all the recordings like that. It's really a combination of those. If you were to ask me in the end, what is your one differentiator? It's the fact that we can seamlessly combine AI agents with human agents, with human representatives, and we can make them work together as a cohesive team. Some of this team, again, is just us humans, and some of this team is new AI agents, which are obviously very rapidly evolving. If you look at that, you have AI natives, purely agentic companies. That's all great.
They only have so much traffic, most of them. Tech-wise, they simply don't have, they cannot light up human-facing endpoints. That's a very heavy lift. It took us a couple of decades, for example, okay? Anybody who has tried, you're talking years, if not decades. So that's it. That's the differentiator. Then you have people coming in from the human side, your contact center providers. That's fine. They're kind of trying to do the same thing. Thing is, they don't have the breadth of the base. So what we're seeing is pretty competitive up there. In the enterprise, it's very competitive, and we have our head in the Ring, too. But we just find it easier to grow and to deliver value at scale in lower segments of the market with SMB, which is like 40% of U.S. economy is that.
You talk about growth being, say, still stuck in single digits. That is true. But about 2/3 of our business is double digits and Rule of 40+ , which is exactly the non-enterprise portion. Look, we're hoping that we can turn this enterprise thing around, and do that sooner than later. But the backbone of the business, we started out as an SMB company. That backbone is alive and well and healthy, and double-digit growth, Rule of 40+ . I also want to make sure that people don't forget that.
Yeah, absolutely. You recently expanded the NICE partnership and restructured the Avaya relationship as well. How do those changes strengthen RingCentral's position in the enterprise market, to your point, and what would make these partnerships meaningful contributors to growth?
Yeah, look, with Avaya, basically, we are transitioning this ACO base to our native RingEX. It is just a customer-friendly gesture. Look, I do not want to speak for Avaya, but it is a very different company now than it was when we started out back then, 2018, I think. So kind of going on nine years. Look, we are still their exclusive provider for UCaaS, so in as much as there is a need, it would be met through us. But they seem to be retrenching into the very high-end enterprise on-prem, and more power, we wish them luck on that. With NICE inContact, it is different. We have been reselling NICE, not the NICE, we have been reselling inContact part of NICE.
Yeah.
Yes. It has been very successful. We built a multi-hundred million dollar portfolio on that. The latest announcement was that they are now saying, "Hey, why can't we also offer this complete solution as well?" It is a very good solution. It is two clear industry leaders coming together with best-in-class UCaaS, which is us, and best-in-class CCaaS, which is them. Having this combined solution out there, which is otherwise not available. It has been early, and again, they are playing also high-end enterprise, so very long sales cycles.
Right.
They are now enabled, and we continue to be enabled with them, so we will see what happens.
Okay, good. Vaibhav, let us talk about margins and investment philosophy. RingCentral has been investing more than $250 million annually in R&D, but you are also expanding margins. How do you balance investment in the core platform with the opportunity across newer AI products, and where do you see the greatest potential returns today?
Yeah, absolutely. Look, I mean, the strength of the cash flows and our expanding operating margins allow us to both invest in product and innovation and show a better profile. We are doing it at scale. In terms of how we are able to do it's a very deliberate and a conscious effort. Every investment decision in the company goes through very strict ROI criteria, whether it's time to pay back or LTV, CAC-type metrics or gross margins. Everything goes through a hurdle before we make the decision. In terms of balancing the investments between core and AI, look, we have a large core base of customers, over 0.5 million customers that we have.
Look, core is the foundation of the telephony infrastructure, security, and reliability upon which AI is built. That's a very unique asset for us. We continue to assign dollars to that to protect and strengthen that portfolio, if you will. Increasingly, we are making more investments in the AI product portfolio, AIR, AVA, ACE, and AIR Pro that Vlad touched on earlier. I think essentially for us, it's not a choice between core and AI. Core is what the AI is built on, and AI is strengthening the value of the core. Look, at the end, we have strong free cash flows, a strong margin profile, and that allows us the flexibility to make these investments.
Okay. As usage-based AI products start to scale, compute and model costs become a larger part of that equation. Right? How are you thinking about pricing and model optimization to preserve attractive unit economics while continuing to deliver a compelling ROI to the customer?
Yeah, we're doing it now.
Yeah.
We'll be doing more of it. Look, we are arbitraging amongst the models as of late. More and more open source comes in. My general belief is that token pricing will be coming down, at least effective token pricing. There always will be bleeding edge misses or whatever it is they come up with next that will be worth a lot of money, and as it should. But at least in our use cases, you don't necessarily need that much of that stuff. Maybe you do to get Elon to Mars or something, but that's not our thing. We are able to do quite a bit with frontier models, but cost effectively, and more and more we'll be looking at open source and differences in pricing that do well.
What does that look like today? Just in terms of the complexion of open sources versus frontier and even on the frontier side, it almost sounds like maybe you're not using Opus 5, but you might be using Opus 4.6 or something to that effect.
Yeah. So what does it look like?
Yeah, in terms of like the[crosstalk].
We experiment and we see. Look, again, we want to do world-class job and present world-class experience at the least possible cost. There is constant experimentation and research and it helps that every couple of months something new comes out and something that was bleeding edge is not bleeding edge anymore, so prices go down. We think it's a good thing.
Yeah. Okay.
Yeah, and I'll make maybe two other points. When you talk about unit economics, there are really three drivers. There is pricing, there is the costs associated with LLM and infrastructure that Vlad talked about, and then there is the GTM scale and the cost. On pricing, look, we are being very deliberate. The idea is value-based monetization. Customers are paying for what they are consuming and the benefits they are getting. Frankly, a lot of customers are paying less prices for these products because the ROI is very tangible, both from a revenue standpoint and from a cost efficiency standpoint. That's visible in the metric, the RCAI metric, wherein both ARPU and net retention rates are higher. That's the pricing part.
On the cost part, like Vlad said, we are doing model orchestration. There is inference efficiency that's getting built in and the infrastructure cost. Over time, with scale, I think these costs will come down. As the models are training, inference is getting better and the price per unit is coming lower and expected to come down even lower with open source models. Then on the GTM side, our current GTM infrastructure is selling multi-products now. We don't need to add incremental sellers. There is further efficiency that we are seeing from a go-to-market standpoint.
Yeah, maybe expound on that a little bit. Talk to us about a little bit of how you're deploying AI internally, how mature that motion is relative to maybe six, 12 months ago.
It's a lot more mature. We had a press release on this. We actually ensured that entire product and technology workforce, so that's all of the PMs plus all of the engineers, QAs included, they had to complete an AI native project. That's actually an effort I'm pretty proud of.
Anything cool come out of that?
Yeah, lots of cool stuff came out of that. I tell you, not everything that they were doing was Ring related. It was just first step. Moving forward, everything we do is AI native because now we've established the baseline and now there is no excuse. "Well, I don't know how it works. I don't know how to extract from Jira and then check back in if it's AI generated." There's just a lot of friction, just education. This is friction because people are fighting just because how they do this. We're doing a lot of education. We're working closely with OpenAI to this day. We're working with [audio distortion] X as X and X as Cursor, just to name a few. We're always pretty close to Google.
There is absolutely stuff going on and more and more, not more and more, but P&T is all AI today. Analytics is all AI, both generated by AI and consumed by AI. Other parts of the org, work in progress, but next time we chat I'm sure I'll have more to report there. We are seeing major efficiencies. Major efficiencies.
Okay, great. All right. Vaibhav, let's talk about margins. You expect to reach 20% GAAP operating margin, I think, in two-three years. How much of the remaining improvement comes from underlying operating leverage versus lower SBC and other factors? How much flexibility do you retain to reinvest if the growth opportunity develops faster?
Yeah, no, thanks for the call out there. Look, our GAAP operating margins are growing faster than non-GAAP, and the reason that's happening is there are three drivers for it. There is operating leverage in the business, then there are structural efficiencies that we are gaining with the use of AI that Vlad touched on, then we are being very disciplined on SBC and dilution. It's a combination of those three things that are allowing us to improve operating margins. If you look at our history, we've improved margins, call it about 200 basis points every year. Again, we haven't laid out targets for 2027 and beyond, but our expectation certainly is that we'll continue on that journey.
In terms of SBC, we are very focused, disciplined on grants, and focused on getting to the 3%- 4% target that we've laid out for the medium term. It'll be a combination of improvements in operating margins over time, SBC reduction as we go on and continue the discipline on shared grants. The two things put together give us the confidence to be able to achieve the 20% GAAP operating margin.
Okay, and then I know free cash flow per share has become central to how you manage the business. As you look over the next several years, what gives you the confidence it can continue to compound from here?
Yeah. I'll maybe answer that. I'll give you two perspectives. One of which is, look, we are absolutely confident in the compounding free cash flow and free cash flow per share model. Again, the drivers there are, if you look at our base, we have a large base of customers. It's a durable revenue model, recurring revenue model with strong retention rates and gross margins. That's number one. Then there's a lot of operating leverage in the business. We have 80% gross margins. Our fixed cost base doesn't need to grow in the same proportion as revenue. As operating margin increases, that will drive more free cash flows. On the other side, we are managing SBC and dilution and being very disciplined on capital allocation in terms of buybacks, paying down debt, and dividends, which are incremental ways of compounding free cash flow.
When you put the top line, expanding margins, discipline on capital allocation, that's resulting in over $7 of free cash flow per share, which is where we've guided for this year, and I believe that's the best amongst our peer group. Frankly, that gave us the confidence to introduce the dividend and then expand the dividends this past quarter. I think that's one perspective. The other perspective is, look, illustratively, and again, we haven't given out long-term guidance, but if I were to project the current run rate of this year's guidance of over $600 million, over the next three years, we can generate close to $2 billion of free cash flow. That 30% of the enterprise value of the company gives us a lot of flexibility.
We can become debt-free, we can certainly pay dividends, expand dividends over time, and keep investing in the growth of the business. Net-net, there's a lot of flexibility. We are very confident in continuing to compound free cash flow and free cash flow per share over time.
Okay, very clear. Vlad, maybe to close out here, we have talked a lot about AI, but stepping back, what do you think changes most about how businesses communicate with their customers over the next five years, and where do you want RingCentral to sit in that ecosystem?
Yeah. Well, sit in the middle of it. Look, again, lots and lots of B2B, and in particular, B2B2C communications is going through Ring today, and we are still a 20% shareholder. That has not changed, by the way. We intend to build on this. How it will change qualitatively, look, there will be a lot more AI, and AI is not only deflecting calls or resolving calls, but it is also making humans smarter about how they deal with things. What we are seeing is there are just lots and lots of learnings that are coming out now once you expose it to AI. Just really, really good at pattern matching, AI is, and better than us humans, even today.
Who knows what is going to be coming out, but I think that in the end, businesses who embrace it, not just Ring, but our customers and customers of our customers, but populations that embraces it will be at a major competitive advantage. Certainly we intend the company to be on that side of the equation. Its future is very interesting. I think it is pretty bright. It is going to be different.
Okay. Well, that is a great place to leave it. Really appreciate both of your time. Thanks for coming out.
Thank you.
Very efficient. Thank you.
Thanks, Vlad. I appreciate it.