Talkspace, Inc. (TALK) Earnings Call Transcript

Hello everyone, and thanks for being here today as we go over the results for our third quarter. We achieved record revenue amounting to $59.4 million, along with adjusted EBITDA of $5 million. I'm happy to share that our active payer members rose 8% from the prior quarter and 29% compared to the same time last year, while payer sessions climbed 12% quarter over quarter and 37% year over year. This pickup during the quarter reflects the disciplined strategy we set in motion this year across every part of the patient journey. Looking back on the quarter, I want to highlight a handful of particular areas where this approach has made a notable difference. The count of clients who activate and show up for their third session within the first 30 days on the platform has risen by more than 50%, largely thanks to enhancements in our matching algorithm, simpler scheduling, and greater provider capacity; on top of that, we keep seeing solid outcomes and better efficiency from our marketing dollars by aiming at particular new audiences, like military personnel and their dependents; we're additionally applying AI-driven methods to further refine media and try out fresh marketing investments. During the third quarter, we joined the networks of a number of new Blues plans, among them Illinois and Massachusetts, and we secured a competitive win for one of the biggest national EAPs, which went live earlier this month. We've also achieved meaningful strides in embedding Talkspace within our payer partners' ecosystems by concentrating on what matters most to them. We constantly oversee and refine our network of roughly 6,000 clinicians to guarantee that high-quality providers are accessible to our members and to make certain every clinician stays highly engaged and motivated; this stands as a crucial differentiator for Talkspace and has supported deeper integration with our payer partners, such as directory integrations that make patient sign-on a smooth experience. Our brand recognition has risen too, since we're now the most recognized insurance coverage focused brand for mental health, with more than 35% of individuals recognizing the Talkspace name based on third-party surveys. That makes our integration into payer directories even more powerful, because people are more inclined to recognize Talkspace and turn to us for care. Payer revenue this quarter also gained from momentum in our psychiatry business, which we brought back earlier this year to serve the needs of our high acuity users and those requiring medication. Thanks to these efforts, psychiatry initial session volume rose 46% during the quarter. Additionally, we expanded our psychiatry provider network by almost 50% versus Q2, and we'll keep building that portion of the network given the demand trends we're observing. We remain focused on refining the internal referral funnel connecting therapy and psychiatry services, and we anticipate further growth in the quarters ahead. To complete our comprehensive mental health offerings, earlier this month we revealed our acquisition of Wisdo Health, a clinically validated AI-powered social health platform that specializes in peer-to-peer community and coaching support. To this point, Wizzo has helped more than 500,000 grown-ups through its platform, and it applies AI to pair that group with properly trained peers plus group coaching for emotional support, companionship, and shared experiences—things that boost health outcomes, lift adherence and engagement rates, and lower the overall cost of care for the health plans it serves. Wiz fits especially well for a lot of Medicare patients, who after joining Wisdo have shown a 21% drop in loneliness and reductions as high as 10% in emergency room, urgent care, and inpatient visits. We expect plenty of Wisdo users could also gain from being referred to therapy, merely one of many paths where we see synergies in giving patients a fuller set of behavioral health solutions. I'm glad to share, too, that starting next month, Wisdo will begin backing Novo Nordisk's new WeGo Together app for patients taking Wegovy for obesity or overweight. Inside that app, Wisdo drives the group coaching experience, which assists participants in forming lasting habits, offering encouragement, and remaining emotionally supported while pursuing their health goals. Our direct-to-enterprise, or DTE, business is still strong, especially among organizations that serve youth and young adults. In particular, our 2-year agreement with the North Carolina Department of Health and Human Services to deliver Talkspace to 20,000 teenagers affected by the justice system went live first in July, and corespaces, which offers student housing choices on college campuses to 36,000 students across 32 different universities, debuted in late August. Both of those rolled out a little later in the quarter than we expected, yet they are now operating and moving along well. On top of that, contract renewals had a strong quarter, among them Baltimore County Public Schools and Colby College, which has now renewed 3 times over its 8 years with Talkspace; we also marked 2 years of our New York City Teen Space program. If you look at the results across that span, we have more than 40,000 youth signed up for the program. Significantly, 93% of participants rely on messaging, which shows how our accessibility approach lets us reach teams where they are. Our suicide risk algorithm spotted more than 500 possible elevated risk incidents, which matters enormously in a group whose most prominent mental health challenges are anxiety and depression. Turning to AI, we keep making major strides weaving our AI work into every part of the business, with several initiatives aimed at improving the customer member journey; these take in LLM search engine optimization, AI help in sharpening eligibility determination for insurance, smart insights for providers as they prepare for sessions, a comprehensive smart evaluation that gives providers a HIPAA-compliant AI draft of a biopsychosocial evaluation following intake sessions, Smart notes that supply post-session summaries for clinicians, our talkcast individualized podcast that I've mentioned before, and now our AI that assists with reviewing medical records, which has been exceptionally useful to the compliance and clinical quality teams. Evidence from our data shows how much our AI-driven advancements are worth. If a provider consults smart insights before meeting a member a second time, that member's odds of scheduling a third session within 30 days rise by over 30%, and their odds of finishing that third session within 30 days of registration rise by 31%. Fresh findings from our talkcast data show that after hearing their podcast, 21% of individuals are more inclined to schedule and finish a third session. We have also introduced 3 more proprietary risk algorithms, which join our suicide risk algorithm. Among them are risk for violence or homicidal ideation, homicidal intent or homicidal plan, risk for substance use disorder, and mouth treatment risk determination, which spots behavior or circumstances that could result in harm, neglect or ongoing abuse. Last, I'm thrilled to share that the Talkspace developed behavioral health-specific large language model I mentioned during our Q2 earnings call has been shown, in our alpha testing, to beat current AI chatbot agents. Hundreds of millions of anonymized therapy transcripts went into training our proprietary LLM, which also underwent rigorous safety and therapeutic quality testing, and we see it becoming a therapy companion and clinical support tool.

So with that said, I'd like to turn to AI and mental health, and the crisis that's come about, which has driven those recent headlines about prominent cases reported in the press involving serious harm and deaths stemming from people turning to general purpose LLMs to deal with mental health problems. A lot of observers have pointed out that these chatbots can be risky, since they hand users instant, unlimited validation and reassurance, they're overly empathetic or sycophantic, they keep affirming poor behavior, and they stay cheerfully adaptive throughout a conversation, flattering users instead of pushing back on ongoing problems. The upshot has been social deskilling and a weakening of real-world interactive abilities. They don't challenge delusions or shore up reality, they lack real-time risk identification, they lack clinical oversight, and no HIPAA protection exists. We'd point out that even while companies try to fix some of these problems, they don't possess all the capabilities or experience needed to protect users more fully; we see these shortcomings on the part of others as a distinctive opening for us. Our major investment in an AI model, paired with our other capabilities, will give individuals a substantially better and safer experience. This new model we've created establishes a new benchmark for both therapeutic efficacy and user safety, which sets it apart from the general purpose LLMs out there today. Drawing on our database, we trained the model on hundreds of millions of tokens taken from anonymized and graded tox-based therapy transcripts. During testing, the model beat open source baselines as well as state-of-the-art models consistently when responding to high-risk mental health situations, among them self-harm, hallucinations, OCD, mania and delusion. In fact, in a recent test, compared against general purpose models lacking this specialized fine-tuning, our model produced a 50% improvement in identifying and responding to high-risk behaviors, a 47% higher therapeutic quality score on the cognitive therapy rating scale, and 3x higher user satisfaction than the base model. We anticipate these early results will improve further as time goes on, as we refine the model and run more testing cycles. Even more important, our product is being built with substantial clinical oversight, real-time risk determination, immediate referral to a live therapist, human in the loop, and HIPAA protection to safeguard patients' personal information. These are distinctive and core skills that our Talkspace members already have access to through their personal interactions with their therapists. On top of that, our model can serve as an engagement tool for intake screening across multiple different types of patients, and as an engagement tool in between sessions. We view the development of our Safe Talkspace AI agent proprietary LLM as a major new opportunity, with considerable potential upside for our existing business and a significant opening for new products; we expect to launch a full product offering in the first half of 2026, but we believe there are a number of unique and significant commercial opportunities in the near term. Our plan is to start by zeroing in on a handful of those close-in possibilities that leverage the commercial setup and brand recognition we already have — among them, serving as a budget-friendly option for consumers, and as an appealing choice for employers seeking a low-cost offering to give their workers. As time goes on, we hope to collaborate with the payer partners we already have so we can extend a reimbursable option to the members of their networks. Crucially, every one of these initial offerings will stay HIPAA compliant and deliver real-time clinical oversight, with the ability to refer someone immediately to a live therapist within our current network. We're eager to share progress on this key initiative in the months ahead. I take pride in everything our team has achieved thus far in the year, and I'm confident we're positioned for still bigger wins going forward; at this point I'll hand the call to Ian so he can walk through the financials in greater depth.

Thanks, Jon, and good morning to all of you. During Q3, we kept delivering against our growth and profitability goals while keeping a tight grip on operational execution. I'll begin by walking through how we performed financially this quarter, and after that I'll share where our outlook stands for the rest of the year. Let's start with Q3 results. We posted total net revenue of $59.4 million, which marks a 25% rise compared with the same period last year and a 9% gain over the prior quarter. The momentum building in our payer business stems from the product investments we've made strategically across recent quarters, along with the efficient marketing strategy we've applied all year. Payer revenue climbed 42% versus a year earlier and 12% quarter-over-quarter, reaching $45.5 million. That outcome speaks to growing adoption and our work alongside payer partners, which has driven stronger member engagement across our payer populations. During Q3, we delivered over 432,000 payer sessions, representing a 12% sequential improvement and a 37% increase from a year ago. This robust session growth came from a 29% jump in unique active payer members, surpassing 120,000 — our best quarterly number since the company began, and it highlights one advantage of the payer business line: earlier cohorts build on each other over time because members stay retained longer than out-of-pocket members do. Direct-to-enterprise revenue came in at $9.3 million, a 1% decrease year-over-year and 2% lower sequentially. As was the case last quarter, renewal rates held up well, but per John's remarks, a handful of significant launches slipped, shifting from Q3 into Q4. We have strong visibility into sequential growth for Q4, partly because those new client wins from Q3 will launch in Q4. Consumer revenue — from individuals paying out of pocket — reached $4.6 million this quarter, higher than the $4.4 million posted last quarter, yet lower than the $6 million recorded a year ago, since we now cover a greater number of Americans through in-network benefits and we've refined our checkout funnel so members are steered toward using their covered benefits. Adjusted gross profit totaled $24.6 million, rising 13% year-over-year and 5% sequentially, which works out to a 41.5% gross margin, down from 43.1% in the previous quarter. That sequential gross margin decrease came partly from the ongoing mix shift toward payer revenue overall, and partly from the timing of selective network hiring in certain areas ahead of anticipated higher demand, something we expect to return to normal in Q4. Total operating expenses were $22.4 million, falling 11% sequentially while rising 4% year-over-year. Measured as a share of revenue, OpEx — excluding stock-based comp and nonrecurring expenses — dropped to 34%, compared with 40% in Q2 and 41% a year ago, reflecting disciplined expense management and ongoing operating leverage as revenue growth exceeds a fairly fixed cost base. Adjusted EBITDA rose 111% year-over-year to $5.0 million, versus $2.3 million in Q2 and $2.4 million a year ago. Adjusted EBITDA margin widened to 8.4%, compared with 5% a year ago, once again showing the operating leverage built into the business. Now let's move to the balance sheet. We closed out the quarter with $96 million of cash and equivalents, which includes restricted cash as well as securities available for sale; that figure declined by $7 million from the previous quarter, mostly as a result of the share repurchase activity we undertook. During the quarter, we repurchased close to $9 million worth of stock, which brought our share repurchases for the year to date up to $17.2 million. And now, last of all, on to our outlook; for 2025 we are tightening the ranges in our full year guidance as outlined below. Revenue is now projected by us to land somewhere between $226 million and $230 million, reflecting growth of 20% to 23% compared with a year earlier. That's against the earlier range of $220 million to $235 million. Adjusted EBITDA is now expected by us to come in somewhere between $14 million and $16 million, compared with the earlier range of $14 million to $20 million. As was discussed today and as our improving KPIs and quickening revenue growth make clear, particularly within our payer business, the strategic investments we have made throughout 2025 in both marketing and in upgrading our technology and product platforms are now bearing fruit. Although these investments weigh on our profitability in the near term, which the updated guidance reflects, they simultaneously establish the basis for growth that can be sustained in our top and bottom lines in the near term and on into 2026. Across the company, we have achieved meaningful progress thus far this year, from a clinical as well as an operational perspective. My belief is that we are finishing the year standing on a very solid foundation. With that said, operator, the call can now be opened for questions.

I'd like to ask about the large language models. In what way are those going to be incorporated into the Talkspace app? And what's your strategy for making money from that? Would it work as a subscription? Or does this fall under something existing patients get at no cost?

Thanks for asking. Our network is where it's going to be incorporated. There's also — we're also developing the clinical oversight component of this. That way, we can monitor anyone requiring off-ramping. So depending on which population adopts it, that forms part of the new patient journey included here. You may have noticed we have several areas where this can be monetized, be it large populations, consumers, employers; therefore the model and the patient journey will differ for each of those; so really, what I'd say in response is wait and see until we're prepared to actually launch it commercially, and that will be early 2026; however, how we commercialize each particular entity will in fact have a distinct impact on us.

All right, and as a follow-up question, perhaps you could discuss a few of the adjustments made to the matching algorithms that are producing the robust retention you're experiencing?

What matters most in that area, I believe, is that once individuals come to us, we explicitly inform them an appointment will be arranged on their behalf. After that, in certain situations, we locate a suitable match for them rather than pairing them immediately — that's one initiative. The other involves our capacity to book individuals for several sessions, or into the time slots they desire, which has gotten considerably better and helps fuel that portion of the sign-up process.

Yeah, Steve — this is Ian speaking. I don't think I could identify a single item, or even a small set of items, that we could attribute the effect to in a straightforward way. What's actually happening is a whole bunch of minor adjustments, and once you layer them on top of each other, they generate a genuinely strong cumulative benefit. So for instance, we've made our sign-up process much simpler. Then there's network management, where we constantly monitor supply and demand and maintain that supply-side planning perspective — watching things both short term, meaning 1 or 2 weeks out, and over a longer horizon with network planning that looks quarters ahead. We even drill down into particulars such as which days of the week, what times of day, and which states, and we make sure we're working with our network to get more schedules and more availability opened up, so that when a person arrives, we understand — alongside price, which we've discussed before — that scheduling plays an enormous role in whether that consumer decides to go ahead with proceeding or chooses not to.

Ethan here, filling in for Charles. It seems like payer KPIs are performing quite nicely. And do you folks feel at ease, in your view, regarding how many credentialed therapists you currently have within your network? Or do you figure you may need to expand that count before long or over the medium term?

Yes, we're absolutely at ease with our current position. I mean each — once more, we gather weekly, even among senior executives, specifically about this matter for capacity planning, since clearly it sits at the heart of our business. Broadly speaking, I'd say we feel fine about our situation; we're always trimming and bringing people on; discussing it in broad terms is tough since we drill down very precisely into which areas of expertise they hold credentials for, whether they're paid $10.99 as opposed to a W-2, what their maximum capacity looks like, and which locations they operate in. That's merely a handful of the factors we must account for; yet I actually brought this up during my prepared remarks concerning our gross margin, the quarter-over-quarter shift from Q2 to Q3. A portion of that came from — we brought on some people within our network to stay ahead of expected demand, which materialized in Q3. And honestly, what we're observing in Q4 is more of the same, that exceptionally robust payer demand. So that somewhat weighed on gross margins during Q3 as we brought people on ahead of that demand; however, as things stand now, we feel quite confident about the overall health of the network.

All right, that's really useful. If it's okay, let me ask one quick follow-up. It seems like sales and marketing came down during the quarter, which lines up pretty well with what you've been telling us. But how should we think about the run rate for marketing spend from here?

Sure. So sequentially it declined somewhat, and clearly it rose, I believe, roughly $1 million compared with a year earlier. What I'd repeat is that we hold — as I've said, the bar for marketing is extremely high. We never spend on it casually. We felt good about increasing it relative to a year ago given the efficiency and the very strong ROI that we're seeing. I think that showed up in the payer KPIs last quarter and once more this quarter, if you look at new user growth, which is a really solid leading indicator of the sessions that will arrive and, since this is a fee-for-service model, the payer revenue that will follow. So looking ahead, I'd say roughly where we sit in Q3 is a reasonable proxy for Q4. Naturally, we'll return to all of you with '26 guidance next quarter. What I'd add is that — within our narrow guidance for '25, that 20% to 23% growth range — for next year I'd anticipate another 20% plus top line growth, which of course needs some marketing behind it, though it wouldn't differ too dramatically from this year.

About the ongoing wins in the payer channel — if we separate the rise in active member utilization, could you say how much you'd credit to the core commercial population versus, clearly, the ongoing expansion of lives within the Medicare population so far?

Sure. As I'm sure you'd expect, the meaningful expansion is still coming from the commercial part of the business; we're watching Medicare keep growing. And I'd add that we're seeing considerable adoption of the military initiatives we have underway at this time. So everything is advancing at nearly an identical rate. Still, without a doubt, the commercial payer orbit is a major driver behind the gains in both registration and sessions.

That helps. Now, when you consider the Wisdo Health purchase and begin bringing it together, how might that — what role could it play in unlocking a larger portion of the Medicare chance for you, or in generating more active participation there? Also, from a marketing standpoint, is there anything you're doing right now during open enrollment to genuinely increase those efforts this year?

Yeah, good question. So — so without a doubt, folding Wisdo into our Medicare approach, we believe, will deliver a meaningful lift thanks to the group coaching they provide. What we're seeing is that a large share of seniors respond well to the concept of group coaching and peer-to-peer support, and to connecting with individuals who've been through the same things they have, especially when it comes to loneliness. So yes, we do expect a meaningful lift when it comes to the Medicare population. I'm not — I follow what you're asking about open enrollment. I'm not certain what effect that would have; it's definitely uncertain when it comes to the Medicare Advantage population. So I don't — that one I can't respond to. It remains to be seen what unfolds during open enrollment. But keep in mind, our central focus is marketing straight to patients after they've enrolled. I'd also — I simply want to restate a bit regarding Wisdo. The Novo Nordisk announcement we issued alongside it is truly a meaningful favorable development for the Wisdo go-to-market strategy. It was quite evident that those using GLPs for weight management are genuinely seeking a group coaching answer to help them make it through by observing and engaging with others who are on that same journey.

Ryan, just a fast bit of modeling regarding that. So revenue from Wizz is going to appear for, say, the GLP-1 program alongside Novo and their Wego app, and that is going to appear within DTE revenue. Then, regarding what you asked before, bringing our current payers on board — I'd say that up to now, the receptivity and interest coming from Talkspace's existing payer partners has genuinely been a pleasant surprise to us. Their interest in perhaps adding Wisdo on as an additional benefit. That clearly is going to appear in payer revenue; therefore it won't be broken out separately, as such.

Nice work on the numbers and everything you’ve shared today. Maybe we can dig a bit deeper into Wisdo and what’s possible with pharma. Could you walk us through how you expect that to unfold? Is Novo the only company involved, or are you thinking about how to make real inroads across that sector?