RingCentral, Inc.RNGNYSE
Loading
Operator: Good afternoon, and welcome to the RingCentral Fourth Quarter 2025 Earnings Conference Call. [Operator Instructions] Please note, this event is being recorded. I would now like to turn the conference over to Steven Horwitz, Vice President of Investor Relations. Please go ahead.
Steven Horwitz: Thank you. Good afternoon, and welcome to RingCentral's Fourth Quarter and Full Year 2025 Earnings Conference Call. Joining me today are Vlad Shmunis, Founder, Chairman and CEO; Kira Makagon, President and Chief Operating Officer; and Vaibhav Agarwal, Chief Financial Officer. Our remarks today include forward-looking statements regarding the company's business operations, financial performance and outlook. These statements are subject to risks and uncertainties, some of which are beyond our control and are not guarantees of future performance. Actual results may differ materially from our forward-looking statements, and we undertake no obligation to update these statements after this call. If the call is replayed after today, the information presented may not contain current or accurate information. For a complete discussion of the risks and uncertainties related to our business, please refer to the information contained in our filings with the Securities and Exchange Commission as well as today's earnings release. Unless otherwise indicated, all measures that follow are non-GAAP with year-over-year comparisons. A reconciliation of all GAAP to non-GAAP results is provided with our earnings release and in the slide presentation, which you can find under the financial results section at ir.ringcentral.com. With that, I'll turn the call over to Vlad.
Vladimir Shmunis: Good afternoon, and thank you for joining us. Before I begin, let me work with welcome Mahmoud ElAssir to our Board of Directors. As Senior Vice President and Chief Technology Officer at UnitedHealth Group, Mahmoud leads technology infrastructure, platforms and services, including corporate systems. Previously, Mahmoud held senior leadership roles at Google and Verizon, where he led major AI and cloud network and platform transformation initiatives, powering global enterprise and consumer services. Mahmoud brings deep expertise in AI native platforms, cloud infrastructure, real-time data systems, security and large-scale product engineering. This perspective will be invaluable as we scale RingCentral through the next phase of our AI-led evolution. Moving on to the results. We had a strong Q4 capping a solid 2025, in which we met or exceeded all our key operating metrics. Total revenue for the year grew nearly 5% and subscription revenue grew just over 5.5%. Of particular note, we generated record free cash flow of more than $0.5 billion, up 32% versus 2024. This translates to over $5.80 of free cash flow per share in 2025. I am also pleased with the progress we have made in meaningfully reducing the value of new shares granted by over 35% year-over-year. Managing SBC is a key priority. Our steady-state SBC target is 3% to 4% of annual revenue as we expect to achieve it in the next 3 to 4 years. Improving profitability, combined with the reduction in SBC translated to our full year of positive GAAP operating margin. We achieved nearly 5% GAAP operating margin in 2025, which we expect to approximately double in 2026. We are targeting to achieve approximately 20% GAAP operating margin in the next 3 to 4 years. With this in mind, we are now in a position to expand and diversify our overall capital allocation strategy. I'm happy to share that, today, we announced our first-ever quarterly dividend of $0.075 per share. We believe that these strong results are not an aberration but an early sign of good things yet to come, as RingCentral transforms itself into an agentic voice AI company. Here is why. RingCentral is an acknowledged leader in cloud-based business communications. We have built a $2.5 billion business from scratch by making human connections simpler, cheaper and more reliable. Our global platform is carrier grade, secure and regulatory compliant. It is trusted by 0.5 million businesses and over 8 million end users worldwide and supports tens of billions of minutes and billions of calls and SMS messages annually. Critical technical requirements are low latency and bullet-proof reliability, with the system designed to avoid any downtime even during maintenance windows. Scratchy voice or worse, no dial tone are simply not acceptable. With a multibillion-dollar cumulative investment and thousands of highly specialized real-time communication specialists, expanding and improving our cloud-native platform over the last 2 decades, this asset is a strong differentiator as the world gets transformed by AI. Simply put, RingCentral's investment and know-how serve a mission-critical need, and they're very hard and likely not cost effective to replicate. Far from being outdated by forthcoming AI agents, RingCentral's platform is a natural bedrock for emerging agentic voice AI. Looking forward, consumers communicate with their providers predominantly via voice and text, and these interactions are growing. When a consumer calls or texts their business provider, it can be answered by a human or AI agent. In either case, it is RingCentral's platform that makes this interaction possible. And as workflows gradually incorporate more AI agents, RingCentral is in a strong position to provide additional value by incorporating agentic voice AI at the very top of the B2C communications funnel. At the recent Investor Product Day, I laid out our vision for RingCentral 3.0, whereby RingCentral is well on its way to transforming itself into a leading agentic voice AI platform. With agentic voice AI, we are now in a position to not only make connections but also to add significant value to those interactions themselves, before, during and after every call or text messaging, thus, enabling businesses to answer more calls more efficiently, garner more leads and process more inquiries at a higher quality. This makes our service substantially stickier and more valuable to our customers as we are able to answer questions, provide insights and analyze conversations for better customer experience and outcomes. While it is still early, recent results are encouraging. Firstly, our pure AI ARR revenues have almost tripled year-over-year and has contributed significantly toward us meeting our stated goal of $100 million ARR from new products in 2025. We but even more importantly, ARR from customers who utilize at least 1 of our monetized AI products, which we refer to as RCAI utilizing customers, has now more than doubled year-over-year and is now approaching 10% of our overall ARR. With new logo acquisitions, AI attach rate is meaningfully higher, making it a long-term tailwind. Importantly, our RCAI utilizing customers average significantly better ARPU, and they're stickier with net retention rate exceeding 100%. This is another strong tailwind. Looking forward, I could not be more excited about 2026 and beyond. We are leveraging a scaled, cloud native, real-time communications global platform and are able to spend over $250 million on innovation annually. AI is a natural tailwind to our business. The majority of this ongoing investment is now directed towards our new AI-led product portfolio. Our investments are showing good early results. Our brand is strong. Competitive moat is wide and increasing, and our GTM is well established and differentiated. We are embedding intelligence across every interaction and creating new monetization and differentiation opportunities to further widen our moat and increase wallet share. Our financial performance is strong and improving, allowing us multiple avenues to return capital to our shareholders. With a proven team and the rapidly expanding portion of our revenue attributable to AI, we are in a unique position to revolutionize business communications yet again, now through AI. With that, I'll turn it over to Kira.
Kira Makagon: Thank you, Vlad. Let me now expand on a few points. Agentic voice AI is our strategic priority that is delivering clear ROI. RCAI utilizing customers are driving tangible value. They have higher usage, increased spend and stronger retention. As Vlad highlighted, approximately 10% of our ARR now comes from customers using at least one AI product and that adoption more than doubled over the last year. AI is driving structural improvements, making every customer more valuable. Our AI solutions, AIR, AVA and ACE, deliver measurable outcomes at every stage of a conversation, before, during and after, respectively. Each plays a distinct role in driving automation, productivity and insights. Built upon our proven mission-critical communications platform, our agentic voice AI portfolio extends a durable moat grounded in scale, reliability and over 2 decades of customer trust. Our AI Receptionist, or AIR, is a virtual receptionist that ensures businesses never miss an important call or lead. It can handle multiple calls simultaneously, is multilingual and is able to answer questions, schedule appointments and meetings, and route calls. AIR is easy to set up with no professional services required in most cases. As a matter of fact, we have proof points of AIR being set up by human receptionists who are not technically savvy. AIR is our fastest-growing agentic voice AI offering and it is helping us capture greater wallet share from our customers. In Q4, AIR customer count reached 8,300, up 44% sequentially, with customers adding usage-based minute bundles to drive more efficient front-office operations, higher call intake and ultimately, more revenue. With a usage-based model, AIR revenue scales directly with our customers' business activity and is not subject to potential reduction in seat counts. If and when a call is connected to a human, that is where our AI Virtual Assistant, or AVA, steps in to assist in real time. AVA captures notes and surfaces recommendations, accelerating workflows for our RingEX and RingCX customers. After the call, our AI Conversation Expert, or ACE, closes the loop, analyzing every recorded interaction for insights that improve coaching, quality and performance across the organization. ACE has been well received with customer count now exceeding 4,800, up 144% year-over-year. Together, AIR, AVA and ACE create a layer of intelligence at every point of interaction, automating upfront, assisting in the moment and analyzing for ongoing improvement, helping customers drive better performance, stronger customer experiences and more informed decision-making. Let me now provide some real-world examples. A large multi-specialty health care provider in Tennessee deployed AIR in Q4 to address persistent challenges with long wait times, inefficient routing and schedule appointments with integrated SMS. After a 3-month trial, which enabled them to route 100% of incoming calls properly, they expanded AIR minutes from 30,000 to 0.5 million minutes per quarter. Destination Pet, a nationwide premium pet care provider, purchased RingEX and AIR in Q2 2025, and shortly after in Q4, they added ACE. They are leveraging AIR and ACE across 180-plus locations to capture every call and monitor call quality across every site, demonstrating that tangible ROA customers look to drive as they expand adoption of our AI portfolio. PM Pediatrics, largest specialized pediatric urgent care provider in United States, is leveraging AIR, AVA and ACE to enable faster routing, higher first contact resolution and richer patient engagement across their 80-plus locations. In particular, AIR is enabling them to handle 30% more patient calls. This integrated AI approach modernizes operations, reduces friction and enhances patient experience. The key point is that AIR, AVA and ACE are designed to automate, assist and analyze across the entire conversation journey. With RingCentral sitting at the very top of the B2C funnel and serving hundreds of thousands of businesses and millions of end users globally, we now have tangible early proof points of our ability to deliver significant customer value via agentic voice AI. The compounding flywheel of AIR, AVA and ACE is building upon the strength of our carrier-grade secure global business communications platform and sets us apart from point solutions contributing to ARPU extension and higher retention. In November, alongside our agentic voice AI suite, we introduced Customer Engagement Bundle, or CEB for EX. CEB is a purpose-built solution for businesses with non-dedicated agents who don't need the complexity of a full-scale contact center. Just months after launch, we crossed 1,000 customers, confirming strong demand. CEB is also quickly becoming another vector for RingCentral agentic voice AI growth. For customers with dedicated agents that require formal contact centers, RingCX provides an AI-powered customer experience suite, including WEM. Momentum with RingCX remains strong with adoption by more than 1,500 customers, nearly doubling year-over-year, while revenue and ARR also more than doubled. In Q4, over half of our $1 million-plus TCV deals included RingCX, and more than 50% of overall RingCX deals included AI. For example, Patient Connect, a specialized health care call center and scheduling provider, uses RingCX with AVA agent assist the surface patient insights, cutting handle times by 50%. They also use ACE quality management to replace time-consuming spot checks of call recordings, reducing escalations by 40%. Patient Connect reflects a broader pattern. Our agentic voice AI is delivering transformative results across customers of all sizes and industries, ourselves included. RingCentral customer support runs the full RingCX suite with WEM and agentic voice AI, resolving more interactions upfront, cutting queue volumes by over 50%, accelerating resolution times and elevating customer experiences. This is reflected in our latest Gartner Peer Insights ranking, where service and support scored a new high, placing us in the top tier of communications members. In summary, we're executing on our agentic voice AI vision, where AIR, AVA and ACE create an intelligence layer across every conversation. RCAI utilizing customers spend more, stay longer and represent a growing share of our business. This sets RingCentral up for durable growth, expanding profitability and meaningful long-term value creation. With that, I will hand it over to Vaibhav now.
Vaibhav Agarwal: Thank you, Kira, and good afternoon, everyone. As Vlad noted in his comments, we have a durable TAM, well-established competitive moat, a rapidly emerging agentic voice AI portfolio and a well-established GTM. Our AI, while early, is making a meaningfully positive impact on our performance and is already contributing to all key financial metrics. Let me provide more details on our performance and outlook. Q4 was a strong finish to a good year, reflecting our strong position in a growing market and disciplined execution across the board. Over the course of 2025, we meaningfully strengthened our financial profile across all key metrics. Our business is robust, growing and poised to further benefit from agentic voice AI. We believe we are well positioned to continue strengthening our balance sheet and enhancing capital returns, thus, positioning the company for sustained long-term value creation. As Vlad noted, in 2025, we surpassed $2.5 billion in revenue, achieved $100 million in ARR from new products, delivered record free cash flow of over $0.5 billion, achieved full year GAAP profitability, reduced net leverage in SBC and returned our absolute share count to 2019 levels. These milestones enabled us to drive record free cash flow per share while continuing to invest in innovation at a world-class level. Based on our strong financial performance and outlook that I will be sharing with you shortly, I am now incredibly excited to announce our first ever quarterly dividend of $0.075 per share. This strategic enhancement to our capital return strategy is reflective of our confidence in the future of our business and our ability to drive long-term cash flows. More details of this dividend are available in our press release. Turning to Q4. Subscription revenue was $622 million, up 5.5% year-over-year; and total revenue was $644 million, up 4.8%, both in line with guidance. Our core business remained durable in Q4 with stable monthly net retention rates above 99%. Within our customer cohorts, small business and global service provider business totaling over $1.1 billion in ARR, both grew in double digits with strong unit economics. As Vlad indicated, a key metric moving forward is performance from customers using at least one of our AI products. We refer to these as RCAI utilizing customers. This is currently approaching 10% of our overall ARR, more than doubling year-over-year. As these RCAI utilizing customers come from all cohorts, this metric better reflects how we manage our business. We plan to report on our progress with RCAI utilizing customers periodically instead of previously disclosed cohort-based metrics. Moving to profitability. Q4 subscription gross margin remained above 80%. Non-GAAP operating margin reached 22.8%, up more than 140 basis points year-over-year, driven by operating leverage and improved sales and marketing efficiency. Our disciplined approach to equity management resulted in an SBC reduction by over 300 basis points as a percentage of revenue year-over-year. This contributed to us delivering GAAP operating margin of 6.6%, up about 4 points year-over-year and GAAP EPS of $0.26. Non-GAAP EPS increased more than 20% to $1.18, above the high end of our guidance. In Q4, we generated $126 million of free cash flow, up 13% year-over-year. During the quarter, we also repurchased approximately 5 million shares for $135 million. For the full year 2025, subscription revenue grew 5.6% to $2.43 billion, and total revenue increased 4.8% to $2.52 billion. Subscription gross margin was 80.5%, and non-GAAP operating margin improved 150 basis points to 22.5% or $566 million of operating profit. Revenue growth again outpaced operating expense growth, reflecting disciplined hiring, expanded offshoring, vendor consolidation, increased internal use of AI and investments in higher return products and go-to-market [ motion ]. Our strong operating performance, combined with working capital improvements, drove a record $530 million in free cash flow, up 32% year-over-year, representing a 21% margin. New equity grants declined 36% to approximately $160 million or 6% of revenue, driving a 340 basis points reduction in SBC as a percent of revenue. As a result, we achieved a full year of GAAP operating profitability with GAAP operating margin of 4.8% and GAAP EPS of $0.48. Non-GAAP EPS grew 18% to $4.36, above the high end of guidance. Weighted average fully diluted shares were approximately 91 million. Free cash flow per share increased 36% to $5.81. Expanding free cash flow per share as well as our GAAP profitability remain core priorities. Turning to our balance sheet. We reduced debt by more than $275 million, ending the year at 1.7x net leverage. We have $955 million of undrawn credit facility, which we expect to use to address the $609 million convertible maturity in March 2026. After that, we have no maturities until 2030. We also used $334 million towards repurchase of shares in 2025. Before I get into specific guidance for Q1 and 2026, let me highlight a few key pillars that are foundational to our long-term strategy. First, we remain committed to investing in durable growth rooted in world-class ongoing innovation. We are spending over $250 million in innovation with the majority going towards our new AI-led products. Second, improving GAAP and non-GAAP profitability and free cash flow. We expect free cash flow of $590 million in 2026 at the midpoint. As Vlad noted, we also expect GAAP operating margins of 9% in 2026 at the midpoint with a goal of reaching 20% over the next 3 to 4 years. Third, we remain focused on reducing SBC to drive improvements in EPS and free cash flow per share. We expect annual grants in dollars to decline further to approximately $150 million in 2026 with further reductions over time. Our goal is to reach a steady state of 3% to 4% SBC as a percentage of revenue over the next 3 to 4 years. Fourth, continued deleveraging with a near-term goal of achieving investment-grade credit rating. To that end, we remain committed to reducing our gross debt to $1 billion by the end of 2026. Fifth, returning additional capital in the form of dividends and share buybacks. On the latter note, our Board has approved a $250 million increase in our share repurchase plan, bringing the total authorization to $500 million. With that context, let me turn over to guidance. For the full year 2026, we expect subscription revenue growth of 4.5% to 5.5%; total revenue growth of 4% to 5%, GAAP operating margin of 8.6% to 9.6%, expanding approximately 430 basis points at the midpoint; non-GAAP operating margin of 23% to 23.5%, expanding approximately 75 basis points at the midpoint; free cash flow of $580 million to $600 million, up 11% at the midpoint; SBC of $240 million to $250 million, down about 2 points to approximately 9% of revenue at the midpoint; in-year new stock grants of $145 million to $155 million; free cash flow per share of $6.67 to $6.94, up 17% at the midpoint based on 86.5 million to 87 million shares; non-GAAP EPS of $4.76 to $4.97, up 11% at the midpoint. For Q1 '26, we expect subscription revenue of $622 million to $625 million; total revenue of $640 million to $645 million; GAAP operating margin of 7.1% to 8.2%; non-GAAP operating margin of 22.8% to 22.9%, up approximately 100 basis points year-over-year; non-GAAP EPS of $1.16 to $1.19; SBC of $60 million to $65 million. In closing, I would like to thank our customers and employees for a strong 2025, and now we look forward to another strong year of execution with agentic voice AI, providing a durable tailwind to our business. With that, we will open the call for questions.
Operator: [Operator Instructions] The first question today is from Brian Peterson with Raymond James.
Brian Peterson: Congrats on that above consensus free cash flow outlook for '26. Just maybe double-clicking on that and with that cash flow, I'd love to understand what are your capital allocation priorities as we think about 2026 and beyond and maybe the longer-term strategy with the opportunity, both for debt payback or the dividends. Would love to get more perspective there.
Vaibhav Agarwal: Thanks, Brian, for the question. So yes, on free cash flows, we are super proud of what we've accomplished over the last 5 years -- sorry, the last 3 years. If you look at our trajectory, we've gone from $100 million to $500 million in free cash flow, and this year, at the midpoint, we are guiding to $590 million, so up 11% year-over-year. So we are super happy with the progress there. Now with those levels of free cash flows and the consistency with how we are producing them, we have a lot of optionality in terms of our capital allocation priorities. So clearly, the first priority is investing in the growth of the business. So as you read probably in the transcript, we are spending over $0.25 billion of R&D spend, majority of which is going in our AI-led products. So think of that as 4% to 5% of margin that's getting invested in growth. From there, we are looking to strengthen the balance sheet by reducing our leverage to being investment grade. So we remain committed to bringing our gross debt down to $1 billion by the end of 2026. And also as a reminder for people, we have a $609 million convert that's coming due in March, and we expect to refi that with the undrawn Term Loan A facilities. So once we pay down the debt, there is no debt maturities until 2030. So net-net, we've taken care of any near-term debt maturities. After deleveraging, we are returning additional capital through a balanced combination of buybacks and super excited to announce our first quarterly dividend. So we repurchased about $300 million of stock. Our Board has authorized an incremental $250 million, which takes our available share repurchase balance to $500 million, and we've lowered share count to 2019 levels. And dividends will just complement share buybacks. It's an incremental way of returning capital to our shareholders. And the reason we are very excited about that and the reason we initiated it now is because of the confidence that we have in our business as well as the strong free cash flows that we are generating. So overall, look, with our recurring revenue model as well as growing portfolio of AI products and improving profitability profile, we feel comfortable in our capital allocation strategy.
Brian Peterson: Great to hear. And maybe just following up. I would love to understand how you would characterize the demand environment versus enterprise, mid-market, SMB. Any color you can kind of share on the various customer segments?
Vladimir Shmunis: Repeat the question, please.
Brian Peterson: Yes, Vlad. I'm just going to get some perspective on what you're seeing in terms of demand. Kind of enterprise, mid-market, SMB, just by customer size, what are you seeing in the environment out there?
Vladimir Shmunis: Yes. You know what, great question. So demand actually continues very strong across all segments. And we're doing well with new logos, and we're actually doing pretty well with upsells as well across all segments. We are seeing more pricing pressure in the enterprise than in SMB and in particular, in small business, where contracts are shorter duration, and we don't have any COVID lapping contracts at this point in those segments. Because of that, they are doing -- they are -- specifically the small business, is growing in double digits and have actually accelerated year-over-year. By the way, this was not always the case. Okay? So that seems to be very much an area of strength. And as a reminder, between small business and global service providers, in total, there is a bit over $1 billion, closer to $1.1 billion of combined revenue, and that's growing in double digits and performing well above the Rule of 40, if it were standalone. So we are seeing that. With enterprise, there are still pricing pressures, still mostly having to do with COVID lapping contracts. And I think as we've indicated on some of the past calls, we expect for that headwind to subside over this current year, so entering '27 with a clean slate from that perspective.
Operator: The next question is from Siti Panigrahi with Mizuho.
Sitikantha Panigrahi: Great. I want to dig into a little bit on your profitability. That's impressive, seeing the GAAP profitability and the target for further expansion. So Vaibhav, could you talk about the levers that you're seeing to get there? Is it more like gross margin expansion with your own product mix or you're expecting some kind of operating leverage on any particular line? And also in the same context, where do you see this stock-based compensation to come down in next few years?
Vaibhav Agarwal: Yes. Thank you, Siti, for the question. So let me unpack maybe the three-part question that you had. So relative to operating margins, again, if you look at our trajectory over the last 3 years, we've doubled operating margins from 12% to 24%. So -- and this year, we are also guiding to a 75 basis point expansion, and we are expanding margins while we are continuing to invest in innovation. So where that expansion is coming from is a few areas. Number one, as you mentioned, our gross margins continue to be strong at above 80%. Number two, we are very disciplined in terms of spend. And we also have operating leverage in our business, so we've been consistently driving revenue growth, which is outpacing expense growth. And then in terms of our spend, we are really disciplined in terms of our hiring practices. We are offshoring. There is a lot of vendor consolidation. And then Kira talked about this in her script, that we have increasing use of AI internally. So all of that is driving operating margin expansion. Now operating margin for us has a broader definition. We also look at it in the context of SBC reduction as well as conversion into free cash flow and free cash flow per share. So we are continuously driving reductions in SBC. We are disciplined in terms of our brand practices. So if you'll see from our guide, we are going from almost 11% of SBC to 9%. So there's a 200 basis points improvement and over the longer term -- in the medium term, I apologize, we have laid out a target of SBC being 3% to 4% in the next 3 to 4 years. So I think that's point #1. Point #2 is we also look at operating margin in the context of free cash flow conversion. So if you go back a few years, there was a delta between our operating margins and free cash flow, and that has come down pretty significantly. So now the quality of the operating margin conversion is pretty high, and we've guided to $590 million this year or up 11%. So net-net, overall, look, we have a lot of operating leverage in our model, and we can always drive higher margins, but we are balancing that expansion in margins with reinvestment in innovation and growth. Overall, from a long-term perspective, we feel comfortable with the long-term sustainability of both operating margins and free cash flows because there are a number of structural drivers. And we have a scaled revenue model, which produces recurring revenue and as high net retention rates as Vlad indicated. We have embedded operating leverage in the model, and we are being very disciplined in terms of how we are deploying that cash. So overall, we believe we have a strong foundation to continue to keep improving both operating margin as well as free cash flows.
Sitikantha Panigrahi: Okay. That's helpful. And then a quick follow-up on AIR that grew 8,000 customer plus. So that's pretty good. But what's the average contract value for those AIR customers? Are you seeing the ARPU for AI-related customer different? I mean, anything, changes? How does compare that versus non-AI customer? Basically, I'm trying to understand if you're seeing any kind of meaningful dollar expense and [ per seat ] with AI.
Vladimir Shmunis: Yes. I'll take that. Vaibhav can add additional details. Okay. I'll start with your last question. Are we seeing lift? Absolutely. We already said that -- so here are a few things I really want people to appreciate. So one is we have achieved a $100 million ARR exit rate with new products, which we said 2 years ago, we had 0 revenue. We said would at the $100 million, we have achieved that. So that's check. We also said that AI -- pure AI comprises a meaningful portion of that. Okay? And just as a reminder, our new product initiatives include our 3 AI products, AIR, AVA and ACE, as well as contact center, which is RingCX. Okay? So AIR, AVA and ACE together are contributing to that $100 million. But even more importantly -- and this is a new metric that we have put out there and that would urge people to be judging us on that moving forward. And that is a percentage of overall revenue that comes from customers that utilize at least one of our paid AI products. Now why say paid? Because almost all of our customers are utilizing some AI in their portfolio. Okay? But a subset is paying for AI. So this is the dollars, AI dollars, paid AI dollars that comprised the $100 million. But they pull together, at this point, almost 10% of our ARR, so in the $250 million range to date. Okay? So it has a direct impact but a much more stronger and $0.25 billion like indirect impact, both in ARR and also very importantly, it is showing significantly better retention. And our overall is pretty good, world class, we think. But now it is showing net retention substantially above 100% across the board. That includes small business. So I hope I answered that question.
Operator: Your next question is from Elizabeth Porter with Morgan Stanley.
James Faucette: This is Jamie on for Elizabeth. Would be great to just get a sense on how you're seeing the different uptake of AI across different go-to-market channels, maybe like thinking about the GSP space or sort of verticals, whether it's enterprise versus SMB.
Vladimir Shmunis: Kira, you want to take that?
Kira Makagon: Yes. Sure. So Jamie, the uptake has been good across segments on direct and channel. I would say that AIR -- where we launched our -- so we have 3 products, AIR, AVA, ACE, and AIR is having particularly a good uptake in the smaller customers as it's really easy to set up and especially in the small business, where it's -- they're pressed for just pure resources to be able to take incoming calls, for example. That's been like a lifesaver for customers, being able to go from anywhere in improved lead taking to all the way that translates to real revenue. Our AVA product particularly sells well with our mid-sized customers and ACE across the board. This is the product that does a post-call analysis. In terms of -- and a similar direct end channel, and I would say that on the GSP side, we announced last quarter AT&T is taking it to market. Now we've got TELUS taking it to markets. We've got other GSPs taking our AI products to market, so seeing very consistent and repeated uptake on AI products with the GSP constituency as well. And in terms of the results, we spoke a little bit on the -- during the script with real numbers that support why customers are using it, and it goes from essentially improving their revenue profile to be able to expand, to be able to monitor quality, to be able to achieve their strategic goals such as, for example, one of the customers that we quoted achieving their rating classification, so they can take in and attract more providers.
Operator: Next question is from Andrew King with Rosenblatt.
Andrew King: Just wanted to get some extra color on how you might have adjusted your partner program in order to reflect the company's new AI priorities?
Vladimir Shmunis: Really, really good question. Look, we have a well-established and a well-differentiated partner network. We have a pretty good understanding of which partners cater to what audiences. We also know what our golden verticals are, and a couple that really stick out at this point is health care and financial services. And then there is also SLED, which is doing very well for us as well. So at a high level, we are going with those partners, and I would say that, that would be tip of the spear for us. Over time, we believe that AI will be utilized across the board and adding values in all verticals, but this is the one that come to mind first. Also, GSPs generally tend to -- their user bases tend to be SMB in their own right, and this is another testament how well our AI and AIR, in particular, is playing in SMB, is that not only is it doing well for us but also most of our GSP partners have now lined up and is deployed or will be deploying shortly. So it's [indiscernible].
Andrew King: Congrats on the strong performance.
Operator: [Operator Instructions] The next question is from Ryan MacWilliams with Wells Fargo.
Unknown Analyst: This is [ Cyrus ] on for Ryan. With the recent announced integration with OpenAI's like 5.2 voice model, what are OpenAI's models bringing specifically to the Ring platform that are enhancing your voice offering?
Kira Makagon: Well, we -- in the press release, we clearly spoke about 5.2 model. Generally, I would say that we are -- we utilize the models that make the most sense for the transaction that is being analyzed and most cost effective. So we always test all the models. 5.2 is the latest model and shows very good results. So we look for accuracy. We look for latency. We look for things to optimize during essentially processing and apply what's best in that particular scenario. And I should add to that as well that we -- the platform generally is model agnostic, and so we also utilize other models as they're applicable to specific needs of what's being processed at the time, whether it's a real-time transaction or post processing transaction. So different models apply and there -- we always do arbitrage between the models.
Operator: This concludes our question-and-answer session. I would like to turn the conference back over to Steven Horwitz for any closing remarks.
Steven Horwitz: Thank you, everyone, for joining us today. We look forward to seeing you next quarter and also seeing you at Enterprise Connect in March. Please contact the Investor Relations at ir@ringcentral.com if you'd like to attend. I will be also sending out an invitation soon. Thank you very much.
Operator: The conference has now concluded. Thank you for attending today's presentation. You may now disconnect.