LivePerson - Earnings Call - Q3 2025

Executive Summary

• Q3 revenue of $60.154M was above the high end of guidance and beat S&P Global consensus by ~$3.0M (+5.3%); S&P “Primary EPS” was a large beat versus consensus, driven by a $27.7M gain on troubled debt restructuring and cost actions, while GAAP diluted EPS was -$2.76 due to share effects and non-GAAP adjustments. Revenue consensus $57.112M*; EPS consensus -$2.30*; S&P Primary EPS actual 2.6611*.
• Full-year guidance raised: revenue to $235–$240M (+$2.5M at midpoint) and adjusted EBITDA to $7.5–$12.5M (+$8M at midpoint). Q4 revenue guide set at $50.5–$55.5M with adjusted EBITDA of -$0.3–$4.7M.
• Balance sheet and cost structure transformed: strategic refinancing closed in September (deleverage $226M; $181M debt discount accretive to equity) and Q3 restructuring costs realized, with full OpEx benefits expected in Q4 and 2026.
• Commercial traction stabilizing: 28 deals signed (26 expansions/renewals, 2 new logos); ARPC up 6% YoY to $665K; net revenue retention improved sequentially to 80.4%; generative AI now ~20% of conversations.
• Product and GTM catalysts: launch of Conversation Simulator; live on Google Cloud Marketplace; expanding Google Gemini integrations (Copilot Translate), supporting improved renewals and new upsell opportunities.

What Went Well and What Went Wrong

What Went Well

• Revenue and adjusted EBITDA both exceeded high-end guidance; management attributed revenue upside largely to timing and variable revenue recognition, and EBITDA strength to cost discipline and restructuring execution.
• Financial foundation strengthened: September refinancing deleveraged by $226M, captured $181M debt discount accretive to equity, and extended maturities to 2029—addressing a key customer hesitation and enabling improved renewal outcomes.
• Product momentum: Conversation Simulator launched to de-risk GenAI deployments; early customers (e.g., Telstra) using the product; partnership with Google expanding (RCS, Gemini 2.5 Copilot Translate), and listing on Google Cloud Marketplace opening a frictionless channel.

What Went Wrong

• Top line still declining YoY: Q3 revenue fell 19.0% YoY due to cancellations and downsells; hosted services down 18% YoY; professional services down 23% YoY.
• Free cash flow remained negative (-$8.9M in Q3) and cash declined to $106.7M; recurring revenue mix remained 92%, reflecting ongoing revenue pressure and cautious services activity.
• Q4 outlook implies sequential decline vs Q3 as timing benefits reverse; management guided Q4 adjusted EBITDA to $0–$5M and flagged that full cost restructuring benefits begin in Q4 and 2026, highlighting near-term execution risk.

Transcript

Operator (participant)

Good afternoon, ladies and gentlemen. Thank you for standing by. Welcome to LivePerson's third quarter 2025 earnings conference call. My name is Irene, and I'll be your conference operator today. At this time, all participants are in a listen-only mode. After the prepared remarks, the management team from LivePerson will conduct a question-and-answer session, and conference participants will be given instructions at the time. To give everyone the opportunity to participate, please limit yourself to one question and one follow-up. As a reminder, this conference is being recorded. I would now like to turn the conference call over to Mr. Jon Perachio, Vice President, Investor Relations.

Jon Perachio (VP of Investor Relations)

Thank you, Irene. Joining me on today's call is John Sabino, CEO, and John Collins, CFO and COO. Please note that during today's call, we will make forward-looking statements, which are predictions, projections, and other statements about future results. These statements are based on our current expectations and assumptions as of today, November 10th, 2025, and are subject to risks and uncertainties. Actual results may differ materially due to various factors, including those described in today's earnings press release and the comments made during this conference call, as well as in 10-Ks, 10-Qs, and other reports we file with the SEC. We assume no obligation to update any forward-looking statements. Also, during this call, we'll discuss certain non-GAAP financial measures. A reconciliation of GAAP to non-GAAP financial measures is included in today's earnings press release.

Both the press release and the supplemental slides, which include highlights for the quarter, are available on the investor relations section of LivePerson's website, ir.liveperson.com. With that, I'll turn the call over to LivePerson CEO, John Sabino.

John Sabino (CEO)

Thank you so much, Jon, and thank you all for joining us today. I will begin by briefly reiterating the decisive actions we took to stabilize the company this quarter. Then I will cover our results and key business updates. First, the debt refinancing agreement we discussed on our last call is now closed. This is a pivotal achievement and, most importantly, resolves a concern we heard from our customers and partners. Second, we executed a cost restructuring to reduce our cash burn. This ensures we can operate efficiently and allows LivePerson to retain cash on the balance sheet. Together, these actions address a primary headwind of renewal hesitation and slower bookings, and indeed, the tone of our customer conversations has started to change. They recognize that our cost and capital structures have been stabilized and are looking to us for continued strategic partnership.

Now, turning to our operational performance for the third quarter, we delivered results that were above the high end of our guidance ranges for both revenue and adjusted EBITDA. Revenue came in at $60.2 million, exceeding the high end of our $60 million guidance. Adjusted EBITDA was $4.8 million. This significantly exceeded our high end of our guidance range, demonstrating our continued financial discipline with the cost reductions made during the quarter. Turning to our product, we're seeing strong momentum and validation from both our customers and the market. Our customers' adoption of our generative AI suite continues to grow, with nearly 20% of all conversations on our platform now using generative AI. At the same time, Gartner recently recognized LivePerson as a niche player in their 2025 Gartner Magic Quadrant for conversational AI platforms, one of only 13—excuse me—one of only 13 vendors.

We were also recognized in the 2025 Gartner Report for digital customer service. Building on our previously announced partnership with Google, we were honored to join them on stage at their recent RCS event. LivePerson's integration with Google's RCS platform enables brands to deliver rich, interactive, and verified messaging experiences that drive higher engagement and customer trust. It also allows businesses to seamlessly transition from campaigns to two-way messaging, combining multimedia content with LivePerson's platform to create personalized, scalable customer conversations. It is an exciting development, and we expect to share more on this in the future. This partnership with Google extends even deeper. We just launched Copilot Translate, built on Google's Gemini 2.5. This capability is embedded directly within our agent workspace, eliminating language barriers by automatically translating all inbound and outbound messages.

It allows our brands to effortlessly serve customers in many languages, boosting agent productivity and delivering a truly AI-native experience. Our innovation extends beyond these powerful partner integrations. We are applying our deep conversational expertise to solve customers' most fundamental challenges. We continue to hear consistent feedback from our customers and prospective customers about the challenges they face in deploying and scaling both AI and human agent workforces. These challenges range from the complexity of safely training and validating AI models before production to the long ramp times, high training costs, and quality assurance demands of their human teams. To address these needs in the market, we are leveraging our decades of conversational expertise and deep culture of innovation to introduce Conversation Simulator.

This is a transformative product that enables brands to safely test, train, and validate AI agents in real-world conditions while simultaneously providing in-workflow training and quality management for human agents. Our fundamental differentiator is providing this dual capability in a single unified platform. It accelerates the time to value for AI deployments, improves human agent performance, and positions our customers to scale more efficiently, driving stronger business outcomes across the enterprise. This will be a standalone product with its own revenue stream, and we believe it represents a significant new opportunity for LivePerson. It has a fundamentally open architecture designed to serve as the vendor-agnostic testing and insurance hub for a business's AI and human conversational ecosystem. This means that while it integrates seamlessly with our platform, it will also work with any CCaaS platform and any third-party AI.

The core purpose of this product is to make AI agents more predictable, trainable, testable, and audible. This product allows businesses to simulate, analyze, and continuously improve performance across their conversational ecosystem. This is precisely where we bring AI and human training together. For AI bots, this product validates and optimizes performance against business-critical synthetic scenarios before they ever reach a live customer. This tests end-to-end conversational orchestration of care, sales, and commerce use cases across live agents and virtual agent experiences. For human agents, it provides a new style of training. We can inject synthetic scenarios and test agents directly in their workflow. This provides real-time feedback and training without ever taking them out of their day-to-day activities. The value this provides to our customers is significant. Early data points to a 30% decrease in agent ramp time and a 60% reduction in the time to test AI bots.

Beyond these efficiencies, it is giving our customers the confidence to launch AI agents at scale for high-stakes, customer-facing use cases. Our product provides the visibility, risk management, continuous monitoring, and training necessary to bring velocity and trust and scale to AI deployments. We believe these capabilities will remove key obstacles that have prevented further generative AI adoption. This proactive, continuous approach marks a shift from reactive failure analysis. Customers can now identify and preempt errors. This is how we will deliver trust, value, and more predictable business outcomes, driven by a unified strategy for both AI and human agents. This capability unlocks a significant opportunity by extending LivePerson's reach beyond the traditional enterprise segment. Conversation Simulator has been designed to provide critical assurances to businesses of all sizes, allowing us to address an adjacent market for training, simulation, and compliance.

This market represents a $10 billion CAM today and is projected to grow to $20 billion by 2030. Best of all, this is resonating with our customers. We have several early access customers actively using the product and seeing initial results. These customers include Telstra and Open University, amongst others. Additionally, we have a strong pipeline of customers expected to begin testing in the coming months. This early adoption is validation that Conversation Simulator provides a critical new layer of trust and predictability that the industry demands, and we are uniquely positioned to lead. We look forward to updating you on the growth and success of this new product. Now, moving to go-to-market, we are seeing encouraging early signs of improvement in our commercial performance. Nowhere is this progress more evident than our renewal discussions, where the tone and confidence of our customers has shifted meaningfully.

We successfully renewed several large accounts that had previously expressed hesitation, including a major U.S. telecom company and a leading amusement park and entertainment company. This renewed confidence extends beyond renewals and into new growth opportunities. For example, a leading travel brand, which had initially raised concerns about our financial stability, recently signed a new upsell contract. In addition, a large financial services organization, which had shared similar concerns, is now expected to grow with us, including an upsell this quarter. These expansions from accounts that had previously expressed concern are powerful indicators of increasing confidence in our innovation and stability. This returning customer confidence is also beginning to appear in our numbers. We delivered a slight sequential increase in bookings this quarter, even as we continue to navigate the headwinds from renewal hesitation, longer deal cycles, and new AI-related approval processes across the industry.

Our commercial momentum is being strengthened through our key technology partnerships as well. Notably, we're now officially live on Google Cloud Marketplace, a major milestone that makes it significantly easier for organizations already operating in the Google ecosystem to discover and purchase our platform using their committed Google Cloud spend. This opens a powerful new frictionless channel for growth, and it deepens our reach across enterprise markets. In fact, we already have a deal flowing through this new channel validating the strategy. At the beginning of this call, I laid out the decisive actions we took to stabilize this company, and we're beginning to see the benefits. The tone of our customer conversations is changing, and we're seeing better momentum in key enterprise accounts. We're also seeing continued strong adoption of our generative AI capabilities.

Early traction on our new Conversation Simulator and a growing partnership with Google creating additional paths to market. With better-than-anticipated variable revenue performance in Q3 falling through to the full year, we're raising our full-year revenue guidance range to $235 million-$240 million, up $2.5 million at the midpoint, and our full-year adjusted EBITDA guidance to a range of $7.5 million-$12.5 million, up $8 million at the midpoint. With our financial foundation stabilized and commercial traction building, we're well-positioned to continue to execute our strategy. Now, let me hand our call over to John Collins. John?

John Collins (CFO and COO)

Thanks, John. In the third quarter, we continued to deliver on the plan we committed to at the start of the year. We closed the previously announced debt refinancing agreement and significantly reduced our cost structure. Together, these milestones give LivePerson the financial foundation needed to succeed in the market.

In addition, we began migrating customers to our public cloud infrastructure, and we launched a new product innovation, as John discussed, Conversation Simulator, for which we already have paying customers and a growing pipeline of opportunities. While it's early, we are seeing indications of meaningful demand. In terms of deals and significant wins in the quarter, we signed a total of 28 deals, including two new logos and 26 expansions and renewals, translating to a sequential increase in total deal value of 14%. Key themes for the quarter included continued traction in regulated industries, namely banking, healthcare, and telecommunications, where there is demand for compliant, centralized, and AI-agnostic orchestration to securely deploy and manage a variety of AI agents. Significant renewals and expansions included a seven-figure deal with a leading U.S. health plan provider, a leading amusement park and entertainment company, and Sanlam, a leading South African financial services group.

We also added a global industrial company as a new logo. In addition, with the debt transaction behind us, we began proactively educating customers on our improved financial foundation, which has already resulted in the renewal status of certain customers changing from cancellation or short-term extension to full renewal. As for our third-quarter financial results, total revenue was $60.2 million, above the high end of our guidance range. Note that the upside relative to guidance, which resulted in a slight sequential increase in revenue, was driven by variable over revenue and the timing of revenue recognition for certain deals. Adjusted EBITDA was $4.8 million, also above the high end of our guidance range, driven by strong cost discipline and the immediate benefits of the cost restructuring executed during the quarter. Revenue from hosted services was $51.2 million, down 18% year-over-year. Recurring revenue was $55.1 million, or 92% of total revenue.

Further segmenting revenue, professional services revenue was $9 million, down 23% year-over-year. From a geographic perspective, U.S. revenue was $37 million, and international revenue was $23.2 million, or 61% and 39% of total revenue, respectively. Average revenue per customer was $665,000, up 6% year-over-year, driven in part by expansions with our largest customers and in part by customer retention. RPO declined to $182 million, consistent with the same factors driving declines in revenue. Net revenue retention was 80.4%, up 78.2% from the second quarter. This sequential increase was driven by the same factors that caused the sequential increase in revenue. In general, we expect net revenue retention to track within period revenue and experience slight sequential declines going forward. Finally, in terms of cash, we ended the third quarter with $107 million of cash on the balance sheet.

Turning to guidance, considering the revenue upside in the third quarter, which we are flowing through, and our improved outlook on renewals, we are raising guidance for the full year. For revenue, we now expect a range of $235 million-$240 million, an increase of $2.5 million at the midpoint. For adjusted EBITDA, we expect a range of $7.5 million-$12.5 million, an increase of $8 million at the midpoint. We also expect adjusted EBITDA to exceed capital expenditures for the full year. The implication for revenue in the fourth quarter is a range of $50.5 million-$55.5 million. Note that the sequential decline is driven in part by the timing of revenue recognition that benefited the third quarter. We expect recurring revenue to be approximately 93% of total revenue in the fourth quarter.

As for adjusted EBITDA in the fourth quarter, we expect a range of $0-$5 million. Before taking questions, I'll briefly summarize by emphasizing that the third quarter demonstrated strong execution against our strategic plan. We strengthened the capital structure and rationalized costs, setting us on a path toward producing sustainable free cash flow. Simultaneously, we continue to deliver value for customers with on-schedule GCP migrations and product innovation in the form of Conversation Simulator. Collectively, we believe these milestones have positioned LivePerson to continue building commercial traction going forward. With that, we can move to Q&A.

Operator (participant)

Thank you. We will now be conducting a question and answer session. If you would like to ask a question, please press star and then one on your telephone keypad. A confirmation tone will indicate your line is in the question queue.

You may press star and then two if you would like to remove your question from the queue. For participants using speaker equipment, it may be necessary to pick up your handset before pressing the star keys. One moment, please, while we poll for questions. The first question we have is from Jeff Van Ree of Craig-Hallum Capital Group. Please go ahead.

Daniel Hibshman (Equity Research Analyst)

Hey, good evening. This is Daniel Hibshman on for Jeff Van Ree. Maybe just if you could open with giving a little bit of color on the upside to the quarter and what drove that. I mean, it sounds like a few factors, perhaps customers gaining some additional confidence in the finances and a few other factors plus those. Maybe if 80/20, you could point us to sort of what drove the upside this quarter.

John Sabino (CEO)

John, do you want to talk about revenue, and I'll talk to bookings afterwards? Thank you.

John Collins (CFO and COO)

Yeah. No problem. Good to hear from you, Daniel. In terms of the upside, we characterize it as timing, which means there's some deals that would have otherwise taken place in the fourth quarter than now in the third. There's variable revenue that we recognized in the third that drove the balance of the upside in the quarter. The timing is the larger factor there for your 80/20.

John Sabino (CEO)

Okay. And then on bookings is, Daniel, as you can imagine, the conversations around financial stability and other things not only impeded renewals, but it also had to do with our ability to expand in some of those accounts. We're starting to see those conversations have some forward progress, positive progress.

Daniel Hibshman (Equity Research Analyst)

Okay.

On the competitive landscape for Conversation Simulator, maybe just walk us through, "I'm not familiar with the landscape there. What are some of the key peer products that are out there for this already? What's the differentiation that LivePerson is looking to bring to the market? What's the right to win?" Just anything about the peers there.

John Sabino (CEO)

Yeah, there's a few things with that. There are quite a few folks that are in the space, but no one that really addresses both sides of the equation, which is both human and bot. We're one of the few that we can find that do that in the market. Additionally, our experience around the verticals and the businesses we play in give us the dataset and the unique knowledge in which to train certain scenarios, personas for our customers that separates us quite a bit from some of our competition.

The interesting thing around this is that the approach that we're taking, the ability to actually inject a training scenario or evaluation of a human agent's performance right into their daily workstream or messaging queue is something that's pretty unique to us. You're not training in an outside environment. You're there doing your job, and the ROI is still there in place. When it comes to bots, we're able to do this in a way that really does allow you to simulate at scale how that full orchestration is going to perform. Because we have both sides of the equation, our product is pretty unique in that regard. Additionally, it's an open product, meaning that we can test any LLM that's out there, any CCaaS platform, in addition to activity on our own.

What I think really does separate LivePerson here is that we're not looking at one side of this in isolation. We're looking at it from a complete CX perspective, and we can do a continuous improvement and training loop, compliance and governance in a way that is pretty unique in the market. Those are the things that we think differentiate it, and we haven't really seen someone else doing exactly the same thing as us or being in a position to because we are both human and AI in how we address an agent and a CX experience for a brand.

Daniel Hibshman (Equity Research Analyst)

Okay. Thanks, John. That's helpful. Just one last one for me on the modeling. I think, John, you mentioned a little bit on restructuring and some additional costs coming out. Is that something that happened here in Q3?

Is that in reference to something that's layering more so in Q4? And then just anything on the scale of that? I see the EBITDA here guide is a few million ahead of the street for Q4. Should we take something in that scale is what of a few million, maybe sequential change in OpEx is what's driving that beat in Q4? Just any thoughts on that?

John Collins (CFO and COO)

Yeah, that's correct, Daniel. That's what's driving the beat in Q4. To answer your earlier question, the timing was Q3, so we should begin to experience the full effects of the cost restructuring during Q4 and for full year 2026. Okay. Great. Thanks, guys. Thanks, Daniel. Good to hear from you.

Operator (participant)

Ladies and gentlemen, just a final reminder, if you wish to ask a question, you may press Star and then 1 now. At this time, there are no further questions.

With that, this concludes today's teleconference. Thank you for joining us. You may now disconnect your lines.