Bumble Aktie

Hello, everyone. Thank you for joining us, and welcome to Bumble First Quarter 2026 Financial Results Conference Call. [Operator Instructions] I will now hand the conference over to Will Taveras, Head of Investor Relations. Please go ahead.

Thank you for joining us to discuss Bumble's First Quarter 2026 Financial Results. With me today are Bumble's Founder and CEO, Whitney Wolfe Herd; and CFO, Kevin Cook.

Before we begin, I'd like to remind everyone that certain statements made on this call today are forward-looking statements. These forward-looking statements are subject to various risks and uncertainties and reflect our current expectations based on our beliefs, assumptions and information currently available to us. Although we believe these expectations are reasonable, we undertake no obligation to revise any statement to reflect changes that occur after this call. Descriptions of factors and risks that could cause actual results to differ materially from these forward-looking statements are discussed in more detail in today's earnings press release and our periodic filings with the SEC.

During the call, we also refer to certain non-GAAP financial measures. These non-GAAP measures should be considered in addition to and not as a substitute for or in isolation from our GAAP results. Reconciliations to the most comparable GAAP measures are available in our earnings press release, which is available on the Investor Relations section of our website at ir.bumble.com.

With that, I will turn the call over to Whitney.

Hello, everyone, and thank you for joining us today. This is a period of real transformation at Bumble. Over the past few quarters, we have executed a deliberate reset of our member base. We made a clear choice to prioritize quality over quantity, focusing on well-intentioned engaged members. That decision reduced overall scale, but meaningfully improved the health of our ecosystem. Importantly, this quality reset was not isolated. It was the first step in a broader strategy to reestablish Bumble as the brand that sets the pace for innovation in our category.

We focused first on strengthening the underlying supply of our platform because scale without quality degrades the experience and stifles the outcome people are seeking: high-quality, relevant connections. At the same time, we continued rebuilding our technology and product platform to better serve our members' demand for real dates and in-real-life connection. These moves required short-term trade-offs, but they were deliberate and necessary.

Now with healthier supply and stabilization in our member base, we are entering the next phase, activation. This phase is anchored by 2 innovation initiatives. First, the introduction of our new technology platform. Second, the launch of a fully reimagined experience for Bumble members, including a new interaction model and profile system. The new Bumble platform and experience will roll out over the balance of the year, beginning with the first stage of the new tech platform in the coming weeks.

Our direct member engagement and our research, including our work with author and professor, Dr. Arthur Brooks reinforces a key insight. The biggest friction in dating today is not discovery. It is the gap between online interaction and real-world connection. People get stuck in that in between. This is a central challenge faced by every scaled dating app. Everything we are building is designed to close that gap and drive real in-person dates between high-quality connections. Accelerating each member's progression towards finding that connection and getting out on a date is our priority. We've been doing foundational work on this problem ahead of introducing our new platform and reimagined experience. We have improved profiles, strengthened intent signaling, enhanced safety and built more dynamic onboarding. These changes have helped members show up better even within the limits of our legacy systems.

We are also continuing to improve the current Bumble experience by addressing core member pain points, improving recommendations and enhancing usability. Early tests are showing promising results, including improvements in matching behavior and monetization trends, but results are expected to be relatively limited on the legacy tech stack. We have more to do here in the months ahead. What comes next will go much further.

The innovation starts with our technology platform. As we shared last quarter, we have been actively rebuilding our new cloud-native AI-enabled tech stack. This modern platform will allow us to move faster, iterate more efficiently and begin to unlock entirely new product experiences.

Today, making meaningful changes to our recommendation engine or introducing new features can take months. This has been a real constraint on the rate of innovation. Our new tech platform is expected to eliminate this constraint. As the platform rebuild nears completion, we are ramping development of the next-generation Bumble Date application, a merging of the new back end and the reimagined member experience, launching in select markets in Q4 of this year.

Between now and then, elements of our new technology platform will begin powering a parallel roadmap of incremental improvements in the existing product. With our new app experience, the opportunity is not just to improve the current interaction model, but to evolve beyond it. We are designing a system that shortens the distance between intent and outcome, eliminating the friction caused by multiple steps between interest and connection. Clearer signals drive more mutual engagement and faster progression towards in-real-life connection. Early reactions to this new model have been very positive.

Our AI layer, Bee, is expected to play a key role in the reimagined experience. Testing Bee and onboarding new members has been especially encouraging, not just in Bee's effectiveness, but in members' willingness to engage deeply and share richer context about who they are and what they are looking for. Bee's ability to capture more signal and process information quickly improves our understanding of each member and will strengthen our recommendation engine.

Onboarding is just the first step in how Bee will be used in the new experience. We also expect Bee to help facilitate connection and to suggest and plan real dates among other roles. Bee is a great example of what we can accomplish on the new modern tech stack and how AI will be an important catalyst for our business. It is important to note that we built Bee separate from the legacy system.

I have said a lot here. So let me summarize. First, demand for love and human connection is as vital as ever before. We have done the heavy lifting to reset our business with healthy supply that is ready to engage. We are giving them the tools to show up authentically as their best selves. Next is our new platform, which will accelerate product innovation. Right behind that will be an entirely transformed Bumble experience, which dramatically reduces friction and gets members to in-real-life connection faster. We believe this is our path to deliver what daters are seeking today. This is the path to restoring revenue growth, and we are already at work building the monetization model behind it. That is the core of the Bumble app transformation, but it's only part of the picture.

Beyond dating, we are also investing in broader connection, which we see as both a critical need in the world and a competitive advantage for us. We have expanded groups on Bumble BFF and are seeing strong early traction with total group joins nearly doubling between December and March. This success is driven by Gen Z women who comprise the largest cohort on the platform, highlighting our opportunity with this core demographic. Overall, more than 80% of BFF members are women, reinforcing the durability of our overall brand. We will continue to expand on group connections and in-real-life meeting for platonic purposes through BFFs, but we are also bullish on the opportunity of romance beyond one-to-one in terms of how people come together and meet for love. We are testing new ways to bring people together for both platonic and romantic purposes, including a new product beta launching next month, which we are super excited about.

Across all of these efforts, our approach remains consistent: test, learn, iterate and do it quickly. We are data-driven, member-obsessed and more passionate about the opportunity and problem we are solving than ever before.

In terms of timing, members will first experience the rollout of our new platform, delivering a faster and more reliable experience starting in the coming weeks from a back-end standpoint. From there, we expect to introduce the initial features of our new interaction model and profile. This is our big thing. It will start to roll out to select markets in Q4, backed by a 360 marketing campaign. Then we'll continue to refine the experience into 2027, including adding features like group dating and expanded access to Bee. Ahead of our upcoming unveil, we are continuing to deliver innovations in the current Bumble experience that helps members show up better, more confident and ready to engage. Not all of these improvements will be immediately visible to members, but the critical signal enhancements they enable will drive more relevant connections on the back end. And the UI/UX will be on our modernized back end, which will enable the rollout of our transformed experience later this year.

As we execute this transformation, we remain disciplined. We delivered a strong Q1 compared to our expectations, and we are managing our cost structure carefully while continuing to invest in product, technology and selective marketing. Of note, we have reduced our performance marketing spend to less than 50% of pre-quality reset levels. We are starting to see the benefit of organic marketing again, including positive word of mouth now that we have improved the member base quality. Despite tech limitations, we've been able to drive meaningful improvements, which we believe signals the opportunity ahead with a modern tech stack in place.

To close, we have been hard at work rebuilding our foundation. Now we are focused on translating that into a meaningfully better product experience, which members will start seeing in the coming months. We cannot wait to reignite our brand, product and mission as we transform Bumble and our category. We look forward to sharing more in the months ahead. Thank you so much for your time. And now I will turn it over to Kevin.

Thank you, Whitney, and hello, everyone. In the first quarter, we delivered results in line with our expectations as we move past our quality reset to focus on product and technology innovation. As Whitney noted, we're seeing signs of stabilization in our member base as we enter the next phase of activation. I'll review our quarterly results before turning to our outlook. Unless otherwise noted, my comments are on a non-GAAP basis, and comparisons are year-over-year.

Total revenue for the first quarter was $212 million compared to $247 million in the year ago period. Foreign currency exchange rates contributed $9 million to revenue in the quarter. The loss of revenue from Fruitz and Official equate to approximately 1 percentage point of headwind in the quarter. Bumble App revenue was $173 million compared to $202 million a year ago. Foreign currency exchange rates contributed $6 million to Bumble App revenue.

Adjusted EBITDA was $83 million, representing a margin of 39% compared to $64 million and 26% in the prior year period. Higher adjusted EBITDA despite year-over-year revenue decline is a function of how we have executed through our reset period, most notably with more intensive operating discipline and thoughtful marketing spend.

Selling and marketing expense was approximately $26 million or 12% of revenue compared to approximately $60 million or 24% of revenue in the prior year period. In addition to the reduced overall spend, we've increased our focus on lower cost and higher return organic and targeted marketing channels. This strategy brings us back to our historical marketing strengths, which we believe also supports long-term brand health.

Product development expense was approximately $25 million or 12% of revenue compared to approximately $24 million and 10% in the prior year period. Our product development spending is focused on core product innovation and platform modernization.

General and administrative expense was approximately $24 million or 11% of revenue compared to approximately $26 million or 10% of revenue in the prior year period.

I'll now turn to the balance sheet and cash flows. For the quarter, we generated $77 million in operating cash flow, $74 million of which converted into free cash flow. We ended the quarter with $246 million of cash and cash equivalents and continue to generate substantial cash flow while maintaining a strong liquidity position.

In April, we completed the refinancing of our term loan that had been previously announced. Consistent with our plans to continue deleveraging, we paid down $114 million of debt in connection with the transaction. Pro forma for the refinancing, we had $150 million of cash and cash equivalents at the end of April.

Turning to the outlook. As we move beyond the quality reset, our focus is now on activating our higher-quality member base through product innovation and improved member experience. This transition will unfold over the balance of the year as we introduce our new tech platform and accelerate the introduction of new member experiences. While this work will take time to be reflected in our financials, we believe it best positions us to drive more durable engagement and monetization.

For the second quarter, we expect total revenue in the range of $205 million to $213 million, including Bumble App revenue of $168 million to $174 million and adjusted EBITDA of $65 million to $70 million, representing a margin of approximately 32% at the midpoint. As we move through 2026, we expect revenue headwinds to moderate as the most acute effects of the quality reset dissipate and we transition from stabilizing to rebuilding the member base.

Adjusted EBITDA margins are expected to normalize over the remainder of 2026 as we increase investment in technology and talent to modernize our platform and drive product innovation. We also plan to increase marketing spend to support our innovation initiatives, organic member growth and brand strength.

In closing, we've made meaningful progress on our transformation and are now focused on executing the next phase of the business, pairing a healthier, more engaged member base with a modernized platform that will enable faster product innovation and more effective revenue generation over time. Operator, let's take some questions, please.

[Operator Instructions] Your first question comes from the line of Eric Sheridan from Goldman Sachs.

Whitney, I want to come back to some of the comments you made in the prepared remarks and go a little bit deeper. When you think about the tech stack and how it will iterate going forward, I wanted to ask a 2-parter. One, how should we be thinking about the velocity of innovation and your speed in terms of going to market that will result from that as we continue to monitor the business from the outside in? And what do you think about your opportunity around personalization and how much of it will be either AI-driven or non-AI-driven when you think about what the tech stack might enable you to do in the years ahead?

Thank you, Eric. Great to hear from you. So I'll take this piece by piece. I think before we talk about the actual incredible opportunity we have ahead with this new tech stack, just to double down on a couple of the prepared remarks I had around what we've been dealing with.

We have had extraordinary tech debt. What do I mean by this? We have frankly not been able to make the changes that both our members are wanting, commanding, needing, demanding, but that we have wanted to roll out. So all of the results you've seen to date are done on the back end of a very legacy system, which really does inhibit the second part of your question, which I'm going to get to in a moment, the personalization of the experience.

So let's talk about velocity, my favorite word. Velocity is going to go up in such a way with this new tech stack. So as an example, if we wanted to make a change to the recommendation engine right now, which is the algorithm essentially, right, it could take us months. It's extremely clunky. It's extremely cumbersome. It's extremely difficult to navigate. On this new tech stack, we're talking we can put tests in immediately. We can be monitoring in real time. We can have A/B testing going at levels we've never been able to access before. And frankly, we can make changes in a matter of days or weeks versus months or even, frankly, years. So when you really start to wrap your head around the opportunity there, I think you can understand why I am personally so excited about this new system finally hitting members' back end here -- in the back end of the system here in the coming weeks.

Let's talk about personalization. So this is the name of the game. What's the one reason why people come to a product like ours, particularly Bumble. They're not coming for entertainment. They're not coming to use it like a social media platform. They are coming to meet people. And if you want to meet someone, the baseline is you have to be showing people you want to see and that you want to meet. And so what we're able to do with this new system and this next-gen recommendation engine, which kind of goes side by side with the new -- with the new tech infrastructure, we will be able to personalize the system in ways that we just frankly never had access to. It's not lack of innovation. It's not lack of road map. It's not lack of talent. It has been lack of technical capability. So you will see extreme personalization.

Turning to the last part of your question, AI or not AI. It's a hybrid. So I think it's important to maybe just spend a quick moment on how I look at AI for this business. AI should never replace human authenticity or human connection. And frankly, I've been saying this for a long time, but I certainly hope that the rest of the world is starting to see it the way I am in the sense that human connection is starting to matter more now than ever before and real authentic human connection. For those of you that have been following and watching people fall in love with AI bots, I mean, this is not the future we want for ourselves or the next generation. So this is why I'm at work. I'm giving it my all to make sure that we can bring people closer to real -- in-real-life, face-to-face, human, meaningful relationships and connections. So we will leverage AI to enable that, but we will not use AI to replace that.

So I hope that answers the question. I could talk about this for 6 hours, but I want to give other folks an opportunity to jump in. But thank you again, Eric, for your question.

Your next question comes from the line of Shweta Khajuria from Wolfe Research.

As we think about the time line, could you please talk to what gives you confidence post the activation phase of the renewed tech platform in 2027 or in Q4 of 2026 into 2027? You will start seeing potentially market improvements in the refreshed tech platform. So could you point to what you saw in your test that gives you that confidence? And what should we be looking for starting in Q4 into next year?

Shweta, it's great to hear from you. So let's talk about these different kind of work streams. I want to be very clear that the back-end tech rebuild is different than what the front forward-facing member-facing interaction model and profile redesign are. So these are 2 separate things that will converge into each other. However, one comes before the other. That is the back-end technology migration and enablement and rebuild. That is coming here in the coming weeks for select members, and we will start to roll out globally and more broadly, obviously, over the weeks following and the months following.

So that is the enabler of everything. That is where we can go in and make algorithmic improvements. We can start to make matching and recommendation economics better for folks and really make sure that you are seeing who you want to see.

Now very importantly, so that's the back end, and that will start to enable everything. But very importantly, I fundamentally believe, and I feel that I am a trusted source here because I've been on the front line of this industry from its kind of mobile explosion inception, if you will. I fundamentally believe the interaction model is outdated, not just for us, I'm talking about the industry at large. And I believe it's time to leapfrog anything that currently exists and help people break through these areas of friction where these cliffs exist.

So right now, to get somebody from first sight to first date is extremely difficult. There are so many areas of drop-off opportunity where that mutuality of needing to like each other, needing to chat to each other, needing to keep the conversations going on this double-sided format, it's quite difficult to get you to a date. And frankly, Shweta, we're a dating app. We're not a matching app. We're not a swiping app. But have we really been behaving like that? And that is the impetus of the new interaction model. So we have listened to our members. We have been in the trenches with them. I personally have been on the front lines of research and deep in the data. So that forward-facing, member-touching interface interaction transition and profile redesign, that is what you will start to see in a major market in Q4 and then, of course, rolling out more broadly through the end of Q4 and early into '27.

So let me actually try to answer your precise question. When do we start to see a rebound in the numbers you're all looking for? Well, the answer is very simple. When our technology and our next-gen recommendation engine can actually help better connect people more compatibly and show people who they want to see and then get them out on great dates, that's where the magic happens. And every single thing we are doing, I'm spending every waking hour of my life right now in effort of serving that one goal: get people out on great dates. So I hope this starts to answer your question, and thank you again for taking the time.

Your next question comes from the line of Nathan Feather from Morgan Stanley.

Digging in a little bit more on that kind of pipeline from discovery to actually getting out on dates. What do you feel are the current real pinch points that cause people to maybe have a match, but not actually convert that into an in-person connection? And to what extent can you actually solve that problem? Is there any issues from a perspective of a lot of people have different preferences? There's local markets? Are there ways that you can kind of solve those? And so that's the first part of the question.

And then second, continue to see really strong performance on gross margin. Can you give an update on what you're seeing in terms of payment adoption? And do you think about the uplift that's driving EBITDA?

Thanks, Nathan, for the question. I'll take the first half. I'll kick the second part to Kevin. So the reality is, you're right, everyone has different dating preferences. But the one thing everybody can kind of agree on at this point is everyone is exhausted from this passive model of just low-effort -- like low-effort interest that there's very little follow-through. And frankly, the industry, at large, and us included, we've made it just too easy to express low-intent interest. And so we are turning that on its head. I can't say much more. I really believe that this is going to be category defining, and we want to keep it close to the chest. But what we will tell you is the early testing has come back remarkably positive. There is very little concern that this is not the right direction. But to your point, every market is different, culturally different, preferences are different. We have no issue with being really agile and making sure that we test our way into the appropriate sequencing and the appropriate rollout strategy to make sure that those nuances are accounted for.

But I really -- listen, I'm now 36. I've been doing this since I was 22. I cannot tell you how much this is needed right now for people to really feel reinvigorated with finding love. And there's a few frank realities. We are on our phones more than we've ever been on our phones before, much more so than when I started this company. The need for human connection and love is greater right now than ever. We are more disconnected. Everything is working in our favor. The only thing that has been going wrong is our ability to execute on product innovation, and that is simply due to legacy tech debt, and we are working extraordinarily hard. The teams are incredible, and they are so close on getting us to a place where we can finally innovate and deliver a modern product to our members so that they can continue to make meaningful connections in the real world. Kevin?

It's Kevin. So the improvement in gross margin is primarily a function of increased adoption of alternative billing methods and therefore, a reduction in aggregator fees. So you're right to point out that we had very strong gross margin in the quarter, about 300 basis points than the prior year period, and we continue to see strong adoption of our Apple Pay program, for example, in the U.S., and that program is slightly ahead of expectation, but we expect to see alternative billing be a tailwind to margin throughout 2026.

Your next question comes from the line of Andrew Marok from Raymond James.

This is Raj dialing in for Andrew Marok. So as it relates to the post-reset disclosures made today, could you update us through March and April and explain how the curves for registrations, retention, MAUs and payer penetration trended from October until now? Given that this is the first month -- that was the first month of post-quality reset, which metric should best predict payer recovery going forward?

Yes. Ron (sic) [ Raj ], thanks for the question. So obviously, the disclosures were provided specifically as a way for us to meet a contractual obligation to prospective lenders to cleanse data that we shared with them in connection with the refinancing. That information is all for the periods provided. You can see them outlined there in the specific disclosure on the website. They're all reflected in our current financials. They're out-of-date, stale, and have no sort of import in terms of the business today.

The only thing I can share is that the business has stabilized with respect to KPI performance. And in particular, on registrations, I think you see highlighted there the steps that we took quite intentionally to bring the member base down to what we viewed as a healthier, higher-quality ecosystem from which now we can build. So that's all I have for you on that.

Your next question comes from the line of Ken Gawrelski from Wells Fargo.

As you look out a couple of years in success as you kind of transition the business, can you talk about how you see -- how you could see the financial profile of the business just relative to [indiscernible] built up in the past. You obviously don't want that to recur. Could you just talk about any changes we might see to the financial profile of the business as you kind of get back to growth in '27, '28?

Ken, it's Kevin. So apologies, you broke up. Can you repeat the question or summarize the question quickly?

Sure. Sorry. Is that better? Can you hear me better, please?

It's still a little shaky. Try one more time.

I'm sorry. Is this better? Sorry.

So why don't you go ahead and we'll do our best.

Yes. My quick question is this, are you -- when you think about the future kind of financial profile of the business, if you go out 24 months, 36 months relative to what we've seen in the business in '22, '23 time frame, how may it look different in your view? Different tech stack? You didn't -- don't want to -- and maybe a different kind of marketing go-to-market strategy. So can you just talk a little bit about what the changes in the financial profile might look like?

Of course. Okay. So you're right to point out 2 key things. First, in the time frame you referenced, it was a marketing-led business, not a product- and technology-led business as it has been since Whitney returned as CEO. So what you'll continue to see is a much more efficient marketing spend. It will never return -- marketing should never return to the levels that you observed in '24 and '25. Marketing is used as -- in support of and as a tool to enhance product and contribute to new product introduction launch and of course, to some degree, brand.

You will see a higher rate, overall, in technology and spend or product development. We're in a period of investment now. You see us beginning to gently increase product development expense to deliver all of the innovation that Whitney was describing and is expected for the second half of the year. So overall, with steady revenue or revenue growth, there would be substantial operating margin in the business. So you should expect to see continued adjusted EBITDA margin expansion, again, so long as revenue is stable or revenue is increasing. Let me know if that answers the question.

Yes.

At this time, there are no further questions. This concludes today's call. Thank you all for attending. You may now disconnect.

Hello, everybody, and welcome to the Bumble Fourth Quarter 2025 Financial Results Conference Call. My name is Elias and I'll be your coordinator today. [Operator Instructions].

I would now like to hand over to Will Taveras, Investor Relations. Please go ahead.

Thank you for joining us to discuss Bumble's Fourth Quarter and Full Year 2025 financial results. With me today are Bumble's Founder and CEO, Whitney Wolfe Herd; and CFO, Kevin Cook.

Before we begin, I'd like to remind everyone that certain statements made on this call today are forward-looking statements. These forward-looking statements are subject to various risks and uncertainties and reflect our current expectations based on the beliefs assumptions and information currently available to us. Although we believe these expectations are reasonable, we undertake no obligation to revise any statement to reflect changes that occur after this call. Descriptions of factors and risks that could cause actual results to differ materially from these forward-looking statements are discussed in more detail in today's earnings press release and our periodic filings with the SEC.

During the call, we also refer to certain non-GAAP financial measures. These non-GAAP measures should be considered in addition to and not as a substitute for or in isolation from our GAAP results. Reconciliations to the most comparable GAAP measures are available in our earnings press release, which is available on the Investor Relations section of our website at ir.bumble.com.

With that, I will turn the call over to Whitney.

Hello, everyone, and thank you for joining us. We have a lot to cover today. I'd like to recap what has been a very productive first year in our transformation and the excitement building inside Bumble as we work towards our platform and app relaunch in just a few months.

I don't want to sugarcoat the challenging process that we are working through, but I am proud of our team for their pace of execution and early accomplishments. Starting with performance, we closed out 2025 with fourth quarter revenue and EBITDA at the high end of our guidance ranges as we continue to emphasize financial discipline, balancing investment in sustainable long-term growth with healthy margins and cash generation.

Kevin will provide a perspective around the financial strides that we've made in just a moment. Turning to our transformation. The headline today is that we believe that the heavy lift of our quality reset is behind us, and we are full steam ahead on product innovation. Before getting into the details, I want to set context. When I returned to CEO about a year ago, I came back with a clear focus to rebuild the company from the inside out and return to what originally made Bumble so successful. What people come to Bumble for is to find love and make in-person connections and they favor us for a simple but powerful reason.

We build trust with women. We believe that when women feel safe, confident and intentional about who they meet the entire ecosystem works better. You get healthier interactions, a more balanced member base and better outcomes for everyone. Our goal is to continue to lead in this area and to build the most women-centric dating product in the market across features, design and outcomes, one that solves real pain points women face in dating today.

In 2025, we accomplished our most important goals, putting trust, authenticity and member outcomes first. We also are running the company with sharper discipline so we can invest where it matters most and execute faster. We knew that doing this the right way would create near-term pressure. And it did, the choices we made were intentional. We began by raising standards across the platform and fundamentally changing how we approach acquisitions. Performance marketing was reduced by well over 80% year-over-year.

That was a deliberate shift away from volume-based acquisition and towards higher intent, organically driven growth as we return to our roots of brand and organic marketing. Despite raising the bar on new members and dramatically limiting marketing, Bumble app registrations and active users have stabilized. We proved 2 things through this exercise: first, demand for what we offer, the promise of human connection, relationships and potentially look is strong and enduring. Second, we believe the Bumble brand has tremendous affinity with our target audience as we tighten trust and authenticity standards, user metrics behaved as expected.

Although trust and safety is a competitive advantage and the work never ceases, we believe Q4 marks the completion of our quality reset. As we move into early 2026, trends suggest the rate of sequential member base decline is slowing, consistent with the expectations we shared with you on our last call. And the quality reset is working as designed. We've talked on previous calls about our internal framework for improving the mix of our member base by emphasizing approved members and removing accounts that were not here for the right reasons.

Members who are truly here to find love and connection will be more engaged, and that's what we are, in fact, seeing as our mix improves. Engagement quality is improving significantly with Bumble app week 1 retention in the U.S. up materially and monthly retention trending higher as well. These progress points demonstrate that the member experience is getting better due to the nature of the quality reset even before we innovate the product. At the same time, we've improved the quality of our monetization base.

Payer penetration for Bumble subscribers has increased and the portion of payers choosing subscriptions has risen from 80% to 89% as we reduced certain promotional activities related to consumables. We anticipate that this shift reflects stronger intent and a greater alignment between value delivered and value paid for. Taken together, these improvements likely signal that we're building a healthier funnel for registration to engagement to monetization. The data shows that our foundation is now stronger with a higher quality, higher intent member base and a more sustainable payer mix. But this essential foundational work can only take us so far.

To fully recover and return to growth, we must focus on product and technology innovation, which is where our efforts are now. Since the beginning of the year, I have personally spent 90% of my days with our tech and product teams, reimagining what finding love looks like in the era of AI. We are rearchitecting the entire Bumble experience from start to finish, and I am taking a very hands-on role in shaping the core user experience across onboarding, profiles and matching. As we rebuild the experience, one thing is clear. We cannot deliver the innovations we are creating on our legacy tech stack.

That is why we're building the new Bumble experience on our cloud-native technology stack built with AI productivity at its core. The launch is targeted for Q2 this year. What we're calling Tech Stack 2.0 is not just the back end upgrade. It is a fully new platform that will transform how we build and ship product. We expect it to allow us to deliver member experience improvements more quickly and intelligently, deep in personalization, support how our monetization model evolves over time. and serve as the foundation for more potential product launches in the future.

These are all things that we have been unable to execute effectively on with the current tech back a fact that has limited our ability to deliver the innovations that our members should expect from us and compete better with other offerings in the category. Given the significant tech stack, the fact that we've been able to execute a major member-based reset without being able to deliver the innovation that we need is a pediment to the enduring strength of our core product and our brands. This makes me excited to show what we can do once our new platform is live. The platform development is currently being executed by our newly consolidated product and technology organization led by our Chief Product and Technology Officer, Vivek Sagi and supported by the engineering-led innovation hub that we built in often over the last 6 to 9 months. The buzz in the building is strong, and our renewed energy is already accelerating our pace. We are working more tightly across teams, putting member satisfaction and product at the center of our daily work. and building with clearer accountability and execution.

Our new tech stack will enable us to deliver an experience built around people, not just profile. Data across the industry are dissatisfied with being reduced to images and potentially dismiss with a swipe. Bumble 2.0 introduces a chapter based structure designed to help members tell their stories more authentically and understand 1 another more deeply. This will enable them to see matches with stronger competitively signals, so confidence in the experience and get to meaningful in real life dates more quickly. Before the rollout of the new Bumble Date, we are continuing to deliver updates on our legacy 1.0 Tech Stack that improve outcomes and address the most consistent pain points that we hear from members. We are addressing global member feedback, particularly women's feedback. One example is a new really into tab, which we began testing last month. While results are early, they have been positive, showcasing the importance of strong signals of intent.

We also rolled out test of profile guidance and suggest to date. Profiled guidance delivers personalized actionable feedback on bios and prompts, guiding members step by step to help truly reflect who they are. Suggested date is a new feature designed to make expressing date intent, effortless and to get people off-line as quickly and confidently as possible. Beyond the major updates that we're making to Bumble date in the near term, we are also actively infusing AI into the core Bumble experience and recommendations algorithm. Our objective is not simply to layer AI features onto the product, but to build the underlying system that reflect how people actually meet, connect and form relationships.

So AI can operate within that structure. [ Done right ], AI helps the network function better by prioritizing fewer, more relevant matches over volume, combating flight fatigue and moving members more efficiently towards real-world conditions. What differentiates our approach is not just the AI technology with the depth of the proprietary data that we have built over more than a decade. AI on its own is neither a moat or a disruptor to online dating. Its effectiveness depends on high-quality scaled interaction data and the hardest part of building a successful dating platform is earning trust and building that member base at scale.

We believe that we have one of the largest and most nuanced data sets of real human connection in the world, leaving us uniquely positioned to apply AI in ways that are more personalized and effective than any potential new entrants. To that point, we've talked in the past about building a stand-alone AI experience. What we learned in the development process is that the best way to launch a stand-alone product is to start with enabling it on preexisting data sets.

So we formed a team to focus on a stand-alone product feature. We call it Bee that is integrated in the core member base to accelerate development and iteration. Bee is designed to become a personal dating assistant and matchmaker, learning members values, relationship goal, communication style, lifestyle and dating intentions through private conversations then using those insights to identify usual compatibility to find better dates with a higher degree of confidence and relevance. Importantly, this work is grounded in our long-standing commitment to privacy and trust, ensuring members remain in control of their data.

A pilot of these core functionality is testing internally and will be launched in beta soon. In addition to reimagining the dating experience, we continue to expand how people connect through Bumble BFF. We recently launched an important update that makes groups discoverable, allowing members to find and join communities based on shared interest. Early response has been encouraging, with a 17% increase in the number of active groups in just 2 weeks since launch. Looking ahead, we have a series of product updates planned as fast follows, including the introduction of functionality designed to bring member together in curated real-world settings.

And these are not just BFF initiatives. Ultimately, they tend into dating, and we see an exciting opportunity around group dating as we reimagine the Bumble experience. While the millennial dating experience was more one-to-one, Gen Z searches for love differently. Group socializing is a big part of that. The event capabilities that we're introducing take us firmly in this direction. They create more pathways from digital interaction to in-person connection and can further strengthen engagement. We see significant potential here, both in planning engagement with MBS and in creating new on-ramps into bundled date as members move fluidly between friendship, community and romantic connection within our broader ecosystem. Our goal with all of this work is to transform people's expectations of what a data application is. We plan to deliver our product people actively choose and want to use. One that reestablishes Bumble as the most culturally relevant and trusted dating brand.

I am highly focused on realigning with the cultural by Gist and being the dating app of choice for our core audience, which is women and men, ages 21 to 35, predominantly. To get there, we're doubling down on our unique playbook of hyperlocal organic marketing and community-driven growth rather than broad undifferentiated spend to really reposition Bumble to center of dating culture. It's already working. We are seeing Bumble's women first positioning strengthen as awareness among single women in the U.S. have increased. And Bumble continues to lead in that favorability among scaled dating apps among women. And we continue to see that influencers and other relevant people and culture want to work with tumble despite a highly competitive environment. This is the equity that we are building as we bring Bumble back to the center of dating culture. We are energized by the innovation underway and the clarity of the road map ahead.

The products that we are bringing to the market this year represent a meaningful evolution for Bumble, and we believe that they will redefine how people experience connection on and off of our platform. Brand and member experience are central to our foundation and equally so is our operating health. We've strengthened our business on both fronts. Over the past year, we have added key leaders and resized and reshaped our organization to increase execution velocity and the pace of innovation across product, engineering and go-to-market.

We've improved focus, aligned teams more tightly to outcomes and accelerated progress on our turnaround road map. We have gained efficiency across the company to fund targeted investment in our strategy, and we have streamlined and focused our marketing approach on reinforcing what our brands stands for and emphasizing organic high-quality acquisition channels. We expect the moves we've made will help preserve our attractive cash flow profile.

To summarize, 2025 was a successful year in executing our transformation. We believe we are well on our way to rebuilding Bumble around what has always made us different: listening to women and delivering a trusted, high-quality experience that drives real life, connection and love. We are clear-eyed about the work ahead, but confident in our ability to unlock a step change in the member experience beginning in 2026, which in turn should enable us to monetize more effectively over time. By resetting the foundation of business, how we acquire members, how we engage them and how we monetize.

We believe we have positioned the company to grow again as we restore Bumble to what has always been at the heart of our success.

With that, I will turn it over to Kevin to walk through the financials. Thank you, again.

Thank you, Whitney, and hello, everyone. In the fourth quarter, we delivered results at the high end of our guidance ranges. Revenue reflected the expected impact of our trust and safety initiatives and the deliberate reset of the member base, while profitability and cash flow demonstrated the underlying resilience of our model. Over the past year, we have managed the business with discipline balancing targeted investment in product with careful cost control and a focus on strengthening our financial foundation. I'll walk through our quarterly and full year results before turning to our outlook. Unless otherwise noted, my comments are on a non-GAAP basis and comparisons are year-over-year.

Total revenue for the fourth quarter was $224 million compared to $262 million in the year ago period. Bumble app revenue was $181 million compared to $212 million a year ago. Adjusted EBITDA was $72 million, representing a margin of 32% compared to $73 million and 28% in the prior year period. The quarter reflected what we believe was the most acute top-of-funnel pressure associated with our quality readout actions. For the full year, total revenue was $966 million, compared to $1.07 billion in 2024. Adjusted EBITDA was $314 million, representing a margin of 32% compared to $304 million and 28% in the prior year. Selling and marketing expense was $161 million, representing 17% of revenue compared to $259 million or 24% of revenue, reflecting a more focused and efficient approach to member acquisition with greater emphasis on targeted and organic channels.

We expect to maintain disciplined marketing spend in 2026, while incrementally increasing investment to support the rollout of our new products and in select member acquisition. Product development expense was $96 million, representing 10% of revenue compared to $84 million or 8% of revenue in 2024, consistent with our plans to increase investment in core product innovation, AI capabilities and platform modernization. General and administrative expense was $115 million, representing 12% of revenue compared to $108 million or 10% of revenue, reflecting disciplined cost management as we navigate the reset offset by indirect taxes. We believe this balanced approach containing overhead while prioritizing product investment supports both operational efficiency and long-term value creation.

I'll now turn to the balance sheet and cash flows. For the full year, we generated $250 million in operating cash flow, $239 million of which converted into free cash flow. As discussed on our last call, we are taking active steps to optimize our capital structure. We completed the buyout of all our outstanding TRA liabilities in the fourth quarter for $186 million in cash. We ended 2025 with $176 million of cash and cash equivalents and have continued to generate substantial cash flow throughout the first quarter.

Similarly, we are currently in discussions to refinance our existing debt obligations due January 2027, totaling $588 million as of December 31. Consistent with our intention to deleverage the business, we repaid $25 million of our current Term Loan B in August 2025, in addition to eliminating the full TRA liability from the balance sheet, we expect to repay a portion of our outstanding debt and refinance the balance. Turning to guidance. We believe the most challenging portion of the quality reset is behind us, and we have turned from adding incremental friction at the top of our member acquisition funnel to improving the experience with product innovation. We expect the lag between these product improvements and our reported financial performance metrics.

As such, the sequential stabilization in our business metrics, including active users and payers have not yet been reflected in our year-over-year revenue growth. For the first quarter of 2026, we expect total revenue in the range of $209 million to $213 million, including Bumble app revenue of $171 million to $174 million and adjusted EBITDA of $76 million to $80 million, representing a margin of approximately 37%.

As we move through 2026, we expect our revenue headwinds to moderate as the impacts of recent trends in our operating metrics flow through the financials. Improvements in revenue will be primarily a function of new product adoption and further retention, payer penetration and average revenue per paying user gains based on enhanced functionality as well as incremental monetization opportunities. An important contributor to our margin performance will be the continued implementation of alternatives to in-app purchases in the U.S.

In the fourth quarter, direct payments contributed approximately 1 percentage point of year-over-year gross margin expansion, and we currently expect the benefits to increase through 2026 as we are already seeing increased adoption in January and February. We believe this shift in building methods meaningfully enhances the long-term profitability of the business. In closing, the actions we have taken over the past year have been made to support our financial and operational foundation in light of our transformation work. We are entering this next phase supported by a highly efficient cash-generative business model and a cost structure better aligned to our strategic priorities.

Operator, we'll now take questions.

[Operator Instructions] First question comes from Nathan Feather with Morgan Stanley.

Congrats on the quarter. I guess, first, thinking through -- talking through a pretty ambitious product road map over the course of 2026. I guess first, when you say registrations and active users have stabilized, as we look out over the course of the next year, what's the path to get those to start to accelerate and really improve? And of all of these kind of changes you're talking about, help us think through the timing of when that might potentially happen? And then on the profitability side, 1Q EBITDA margin is showing, a really nice app up quarter-over-quarter. Can you have us think through what are the puts and takes that are leading to that improvement?

Thanks, Nathan, for the question. So I'll take the first part of the question, and we'll talk about returning to user growth. So I think it's important to just reemphasize that we just undertook something incredibly difficult without any product innovation to support us. So let me just explain what I mean that. We just underwent a membership overhaul essentially where we went in and we said quality is the goal.

Quality and safety and authenticity, which are the 3 most important things to women in dating, particularly Mayday online. And we are going to go and take this on while we're still essentially under construction because we're still operating on our legacy tech stack. So the fact that we were able to have registrations and active members stabilized during a transformation with little product innovation to support is actually quite remarkable, which is exciting for the next phase of this, which we said in the prepared remarks, which is rolling out, to your point, a very ambitious but very doable 2026 product road map. Let's talk about that for a second.

So once 2.0 back-end infrastructure is up and running. And once we have migrated the 1.0 system to the 2.0 system, innovation becomes a lot more seamless, a lot quicker, a lot more kind of the volume at which we can innovate becomes so much more plentiful than what we've seen in the past. And we have a very long road map essentially covering what I said in the prepared remarks of solving women's pain points but also delivering these really incredible innovative features, tools and ways of connecting that we believe will accelerate these cohorts of opportunities. So I think it's also important to note, the oldest Gen Z right now are 28 years old.

So we are just entering this peak intentional dating window. And when you look at our road map, it's just extremely focused on reengaging this next generation who are stepping into intentional dating. And so on the timing front, I think what we said in our prepared remarks is exactly the way to think about things. This starts as soon as 2.0 is in flight, and it will be a very consistent drumbeat of innovation. And we are super inspired, committed and we are building with innovation at the helm. So I hope that answers the question, and I will kick the profitability piece over to you, Kevin.

Thanks, Whitney. Nathan, so on profitability for Q1, we continue to be committed to operating discipline. So we're pleased to see continued adjusted EBITDA margin expansion. I will say Q1 is probably slightly elevated. We are expecting in -- as Whitney has highlighted and you heard from our prepared remarks, beginning in Q2, we are expecting to launch significant new product in market. So you'll see a slight increase in marketing to support new product innovation and some very specific product marketing around new enhancements to product.

And you'll also see product development costs increased midyear, slightly in order to support 2.1, 2.2 and the fast follows that Whitney was describing earlier. I do believe that structurally, our margin profile is higher than it has been historically. We are much more efficient today than the business has been in the past. We're not guiding the year here, obviously, but we do see the opportunity for sustained very high adjusted EBITDA margin throughout the year. And obviously, to the extent that we inflect growth in the business, there is enormous operating leverage in the model.

We now turn to Shweta Khajuria with Wolfe Research.

First is on the product revamp, so Whitney, how are you thinking about measuring progress? And what will be some of your milestones as you track that this -- the revamp is going well in how you measure it. That would be -- and is there something different that you plan to share with us in terms of metrics, whether it is either top of the funnel or some total metrics that you guys are following that would help us get a sense of how it's going. That's one because we may not see the impact on monetization just yet, but if engagement is improving, that will be a good sign. The second is on investment on the tech replatforming. So is there additional -- it doesn't sound like it, but I just want to clarify the back-end changes that you are making to make your platform better from legacy to the new tech replatforming, is there additional investments that you are expecting? Or how should we think about that?

Shweta, nice to hear from you, and thanks for the question. Sorry about that. We were having challenges with the microphone. So the way we think about outcomes is -- or KPIs, how we measure this is really about member outcomes. So what we've started to share with you today, as you saw that the deeper funnel improvements are really what measure the durability of this business and how it is performing for members. If you have really good top of funnel results but negative bottom of funnel results, you do not have a healthy consumer business. And this industry and this product is inherently dependent on the outcomes that we drive for our members.

So the way we are thinking about 2.0 outcomes are not dissimilar from the way we think about it right now. But ultimately, it is our members satisfied. Are they getting what they came for, which is high-quality date with people they actually want to meet? Are those as safe -- are they reliable? And are they converting into what they came to look for? This is precisely how you drive top of funnel because when people come, have a good experience, they go on great days, they tell their friends, and that is the flywheel. And so that is everything we are focused on. And in order to arrive at that outcome, we had to really enhance the way in which people discover one another. We had to focus on quality. Quality in this instance, is about helping people show up better, how do we get you to basically showcase yourself in a way that is actually driving curiosity from members so that you can create matches and go out on great days.

So you will see this come to life in 2.0, but ultimately, it's all about the outcomes and the success that our members find.

All right. And -- you had a question, I think, about the tech stack and sort of the investment required on the Tech Act. So recognized 2025 was an investment year from a product development point of view. And we are continuing with that investment in 2026. Over the course of 2025, we built a new sort of modern AI-oriented engineering organization, primarily in Austin, Texas. And so Much of that investment is kind us, including both on product and on the platform. And you'll see some additional investment throughout 2026, of course. The things to recognize there's a lot of efficiency in our engineering efforts currently, we're benefiting substantially from the application of AI, not surprisingly in our product development.

And in terms of appreciating sort of the level of investment required, platform and product, there is some infrastructure costs based on the existing set of offerings, primarily data center costs that are -- that we continue to maintain, while at the same time, we're building this cloud-native AI-led tech stack that you've heard a lot about already. And so there's some duplication of costs for a portion of 2026 in that respect. And then you'll see a continued longer term, you'll see a continued modest level of product development expense increase tied to revenue just to continue to produce the sort of innovation that we're expecting. But you should see some efficiency -- some operating leverage in that in that line once we've got the duplicate costs taken out of the system.

We now turn to Stephen Ju with UBS.

This is [ Nora Robly ] in for Stephen. I appreciate the color on direct billing. And I was just curious if you could frame how big of a potential this could be for you all this year, given how much of a contributor is to the EBITDA guide? And I understand there are some trade-offs between efficiency and cause and user friction. And then a second question, how -- should we expect the new product initiatives to be rolled out globally across the bundle affiliated apps? And additionally, do you see any opportunities with respect to international expansion.

Or it's Kevin. Why don't I take the direct billing question. So you're right. So we saw in Q4 a full percentage point of gross margin expansion as a result of alternative billing. In Q4, we implemented our Apple Pay program. And what we've seen is a very, very rapid adoption by users of Apple Pay. In fact, as of today, sort of quarter-to-date, we've already got more than half of our U.S. iOS payments being made through Apple Pay.

So it is a very substantial cost benefit or improvement to gross margin. we believe it is mostly sustainable based on our understanding of the various sort of cases and settlements to date. The -- as you know, there's some changes that are occurring that are occurring in that regard. So we haven't built all of the long-term benefit into our model. And I mean, clearly, for Q1, we're far enough into the quarter that we have great confidence in our in our adjusted EBITDA guide there. But as I think about the year, we have hedged that number slightly. We have not seen, by the way, any friction. We -- when we built plans, we did think that there could be some impact on revenue. There has not been to date and we're about 3 months into the program. And we have noted too that there seems to be improvement in renewals as a consequence of using these alternative billing methods, which was somewhat unexpected, but obviously welcome.

And on the other portion of the question, so the back-end infrastructure, what we're calling Tech 2.0, that will be applied to the entire portfolio with the exception of BFF because that lives on the Geneva instructure, as you'll recall, we acquired Geneva, largely due to how great their tech and their team and that infrastructure is that they had, which was super enabled for groups and beyond 1:1, which is going to be a huge part of our focus in 2026 and beyond.

We really believe not to go on too much of a departure, but we really believe that one of the largest opportunities for us is bringing people together in groups of people. So really taking this beyond 1:1, that is inherently how a lot of the Gen Z cohort chooses to socialize and meet and date. So just to put a pin in this, tech will be rolling out to all products that are on legacy infrastructure and will be enabled globally. So that will be replacing any legacy systems.

We now turn to Eric Sheridan with Goldman Sachs.

Maybe 2 that build on some of the answers so far and just trying to take it a little bit further. When you look at the current competitive landscape of sort of investments against growth and even investments against product that you see across the dating horizon. How are you thinking about positioning yourself competitively in a 2.0 world of where you think best opportunity for both incremental growth of new users or reengagement of existing users or legacy users sits for the company when you think about that competitive locate. And then second question would be and maybe it's qualitative more than quantitative at this point. But how are you thinking about the incremental margin structure of the company in a fully deployed 2.0 world relative to 1.0 on the other side of the sort of investment cycle?

Thanks so much, Eric, for the question. I will take the first part, and then I'll make one comment on long-term investment strategy, and then I'll kick it to Kevin for the particulars. So this is really an important topic. First and foremost, let's just back up and actually look at what makes Bumble is so unique and sets us apart inherently from any of our competitors. And the most important part of our differentiator is that it's been cordless since I started this company in 2014, and that is our obsessive focus on women. .

We have become a trusted women's brand. This stands true even today on -- quite frankly, somewhat outdated technology and product offerings, we are not up to par with our product right now, but we will be -- and so our brand is so resonant and our brand carries us in such a way that this is such a strong driver for us in 2.0. And when you see the rollout of 2.0, and I know I mentioned this in the prepared remarks, but I have been in every pixel, every meeting so deep in the details, reimagine what would the Dream women's app look like in 2026 to reengage women, both Gen Z and millennial life and Gen X frankly, because this is a highly monetizable cohort. They are also equally looking for love and connection. How could we reimagine this and innovate our way to be the preferred dating platform and connection platform for women. So I just want to reemphasize that women and the trust that we have with women and the authentic design system of putting women first beyond just a function of who goes first or who doesn't. This is inherently what sets us apart. The second thing is I am a firm believer that the future is beyond 1:1 in any sense of exclusivity. Do I believe that there is still a huge demand for intentional one-to-one dating? Absolutely. Please do not misunderstand me there. I think we have not seen our potential through with one-to-one dating, and we're very excited about that opportunity globally.

But I think what's really exciting is using Bumble's brand and leveraging this brilliant new technology instructure to really go beyond just this small dynamic of 1 person seeing 1 person connecting with 1 person, bringing people together in groups of people, whether that's a small group, a midsized group or even a large group and getting people in real life, IRL, this is such an opportunity for us. And we know that we have a right to win because of the strength and the favorability that our brand brings. So I think that leads to my final point, which is the #1 thing that women want when it comes to dating, particularly online is trust and safety. This has been at the core RD&A for a year. We have been a leader in the category, even passing laws to drive safer initiatives for women online, and we are going to be doubling down on this in every sense of the word, we are going to be getting out there much more broadly, I'm going to be putting myself out there for a 2.0 relaunch in a way I've never been before.

And we are going to basically reignite this trusted, safe women's first dating product and brand so that we can bring people together in the real world as quickly and efficiently as possible.

Eric, on the operating cost point, with respect to the updated platform, I think the way to think about it is we're operating from these very, very old data centers today. When we cease reliance on the data center and have a true cloud platform, operating cost will be substantially lower. With with respect to innovation, innovation will become much less expensive, more rapid. We'll be able to iterate in a way that isn't possible today that should unlock. It's a separate question, but it should unlock some incremental monetization of revenue opportunity as well. I would say with the modern platform and what is contemplated in terms of new product introduction, you will see a greater reliance on AI, you'll see an offset perhaps in token costs associated there. But there should be a net benefit to operating margin from our move to the modern platform.

We now turn to Cory Carpenter with JPMorgan.

I wanted to ask what you alluded to the chapter a chapter based structure. Curious if you could elaborate kind of on your vision there and then the role that you see the swipe playing on the 2.0 platform? And then just bigger picture, once 2.0 is out, how radical of a change will users see and how quickly the change will they see in the app once you roll it out?

Cory, it's great to chat again. It's been a while. So yes, let's talk about the chapter based profile. So we basically took the stance of -- the last decade has sort of reduced people down to profiles. And this has happened across the Internet. And ultimately, dating only works, we really understand the story of someone. This is where chemistry and connection really happens.

It's the intersection of someone going from just a stranger that you dismissed to someone you're genuinely interested in. And so as we reimagined the profile, we thought why not bring people to life as a story because everyone has a story to tell. And this is where people become interesting. So what you will see in this chapter based approach is it won't be just your stated kind of flat noninteresting profile that everybody has become so used to across platforms. This is really an opportunity to capture the essence of who you are. so that you go from being Cory, whatever town you live in, whatever age you have listed in just some photo you'll turn into who you truly are and this will a curiosity from the people that are opposed to your story. And so this is also a gateway to allow people to interact more dynamically. So today, on the current Bumble app, you can either broadly swipe right, or broadly like left, which is a no on someone's profile.

So you're kind of saying like, yes, Cory is in or no Cory is out just in 1 swift go. And this is actually reducing the options for people to actually get better matches. So what we will be introducing are more dynamic ways for somebody to express interest in your story, rather profile, however you want to call it. And this is going to drive more dynamic engagement. It should spark better conversation, and it should ultimately drive better KPIs across the board like engagement chances to get better conversations going. You will also see us take a much more deliberate approach to getting people offline versus just in what people kind of refer to as debt and chat zones.

So we're really trying to chip away at member complaints of, hey, listen, I've got all these chats that just sit where they expire these matches that never went anywhere. So we're really trying to get people to understand who each other are and get them out real world as currently, quite and safely as possible. So that's part one. The second part of your question is a really important one, which is how big of a change will this be when members start interacting with 2.0. So this has been absolutely top of mind for the buck team. What we are going to do is we are going to piece by piece exposed members in very controlled group to different portions of the new 2.0 experience.

We are going to rigorously test every single portion of the experience, make sure that no KPIs are damaged, that monetization is strong, that we do not accidentally hurt anything. Once we have tested and if we needed to iterate and once we feel that this is safe, we will start to roll it out slowly and slowly in very controlled market expansion. Once we get to a certain place where we feel that's good, everything is going in the direction we want.

That is when it will start to kind of migrate to the entire world. You can expect that to be a multi-week thing, not a multi-month thing. And as far as our confidence in getting 2.0 out the door in our projected time line, we are very confident, teams and systems are all in a great place, and we're very excited about this next chapter.

We now turn to Robert Coolbrith with Evercore ISI.

I just wanted to ask a follow-up on Cory's question. Just to go -- I think there was a second part there about the slide mechanic. Is that still going to be a core part of the user experience? And then just yes, as you move into these more detailed chapter oriented profiles, 2 questions there. What's the strategy for getting existing users to fill those out and engage with the sort of more fulsome profile? And then secondarily, is there any trade-off or impact on user conversion versus user maybe user retention as you make some of those changes. Does it slow down people's consumption of profiles, which has, I think, historically been a way that you've driven paid user conversion?

So the good news is we've thought about all of these things, top to bottom left to right for a very long time. And we have teams making sure that revenue and monetization mechanics have been thought through and that there is a do-no-harm approach to revenue and monetization with the 2.0 experience.

So I want you to take some constellation in that we're not going to just pull something out that couples our monetization strategy. We're being very, very modest and very conservative with that. In fact, we actually have set up the -- and position the new profile to have new gateway opportunities for better monetization over time. Okay. So we'll start there. Let's go back to the swipe piece. So the way we're phrasing this internally is what does life look like beyond the spike.

And that doesn't necessarily mean this binary situation of like it's either there or it's not. It's much more nuanced than that. So we are testing several iterations of engaging opportunities of how you would -- before I go there, this slide is the way you get to a match. So if the match is the goal, the swipe is the mechanism. So what we're doing is we're testing several mechanisms to get you to a wanted and mutual match as quickly, efficiently and in the most compatible way as possible. So the mechanism of swiping may dynamically shift. And in certain markets, we may test with no swipe. In other markets, we may preserve this vibe. We have multimechanism tests that we are going to be running to make sure that we do what the members want the most, that serve our members best that is in line with the KPI that suggests we're getting people to the best outcomes. And of course, that that does not disrupt revenue and preserves the strength of the revenue business -- revenue side of the business.

So I hope that has answered the question, but happy to take a follow-up if needed.

Oh, sorry, there was a second part of that. How do we get people to fill in the more robust parts of the onboarding? I think that was your question. So this has been a long game, okay? So we've been playing a very long game here. Let me just be clear. When I came back in a year ago, I knew that this is where we would be this year. But I needed a year to get to the quality transformation. So now let's back up. What was a huge part of that quality transformation overhaul? It was trying to drive more what we call approved members what constitutes as an approved member, someone who has adequate information in their profile, enough intel that we have to work with to drive compatibility on the back end, enough photos, self verification, vacation self verification and hopefully, ID verification.

So we've already been planting the seed for this moment for the last several months by trying to drive people to fill in more information. So we've made a ton of progress there. So it's not a cold start situation, where all of a sudden, we have a brand-new profile and we have no information. The other thing that we're doing is we're being really thoughtful in how, what and where we really ask you to fill in more information. So it's not going to feel punitive, is not going to feel stressful or exhausting. It's actually going to be dynamic and fun. And then on another note, as we talked about with B or AI assistant, down the road, we are going to be able to get so much robust information about who you are, what you're looking for and really understand your story through more engaging mechanisms, whether that's voice or typing, but it's in a conversational format like an AI tool, right, like an AI product, but this is going to be the long game of how we get more dynamic information. So it's a step-by-step plan, but the good news is we've already made a lot of progress.

[Operator Instructions] Ladies and gentlemen, we have no further questions. So this concludes our Q&A and today's conference call. We'd like to thank you for your participation. You may now disconnect your lines.

Hello, everybody, and welcome to the Bumble Third Quarter 2025 Financial Results Conference Call. My name is Elliot, and I'll be coordinating your call today. [Operator Instructions]

I'd now like to hand over to Will Taveras, Investor Relations. Please go ahead.

Thank you for joining us to discuss Bumble's Third Quarter 2025 Financial Results. With me today are Bumble's Founder and CEO, Whitney Wolfe Herd; and CFO, Kevin Cook.

Before we begin, I'd like to remind everyone that certain statements made on this call today are forward-looking statements. These forward-looking statements are subject to various risks and uncertainties and reflect our current expectations based on our beliefs, assumptions and information currently available to us. Although we believe these expectations are reasonable, we undertake no obligation to revise any statement to reflect changes that occur after this call. Descriptions of factors and risks that could cause actual results to differ materially from these forward-looking statements are discussed in more detail in today's earnings press release and our periodic filings with the SEC.

During the call, we also refer to certain non-GAAP financial measures. These non-GAAP measures should be considered in addition to and not as a substitute for or in isolation from our GAAP results. Reconciliations to the most comparable GAAP measures are available in our earnings press release, which is available on the Investor Relations section of our website at ir.bumble.com.

With that, I will turn the call over to Whitney.

Hello, everyone, and thank you for joining us today. Helping people find love has been Bumble's focus since day 1. What made Bumble successful from the start was simple but powerful. We built trust with women. That trust became our moat. It created a healthier, more balanced member base that drove engagement, retention, growth and a global brand. Historically, most dating products skewed male, leading to an uneven experience where women often felt overwhelmed and disengaged. Bumble changed that by putting women in control, creating trust, balance and a higher quality member base that produce better outcomes for everyone. Today, we have the brand long identified with putting women first and a deep understanding of what women want from love and connection. This brand identification is perhaps our biggest strategic asset. That's why we've returned to that core, winning with women because we believe that when women feel safe, respected and confident and when they meet quality matches, everything else improves. Connection becomes more meaningful, engagement more genuine and the business more sustainable.

When I returned as CEO in March, the first thing we did was listen deeply and intentionally to women. What we heard across markets was consistent. People want to trust who they're meeting. They want better quality matches, and they want more authenticity. Those insights became the foundation of the quality over quantity reset I've been talking about over the past 2 quarters. Every step we have taken since our product updates, investments in AI and every marketing move ties back to one goal, making Bumble a better experience for women because when women are happy, the entire ecosystem thrives. That is the focus driving our transformation into the love company, expanding beyond dating to build the global platform for meaningful relationships, romantic and platonic. At its heart, what women seek on Bumble is love, and we are building the product, the technology and the brand to deliver it.

Since our last call, we've executed this reset with real focus and urgency while also delivering on our financial commitments. For the third quarter, our results are near the high end of our guidance range for both revenue and adjusted EBITDA. We are also seeing the expected short-term effects of prioritizing quality over quantity. Member registrations are lower in the near-term, but we believe that the foundation of the business is getting stronger. Early indicators suggest that retention is improving and ARPPU is growing. The entire team is heads down, laser-focused on executing this plan. The early signs we're seeing just months into our reset reinforce our confidence in our strategy, and we look forward to sharing more details as we progress further and have more data and visibility into trends.

We are working aggressively to improve the mix of engaged approved members on our platform, reduce the noise and build a higher quality community. Our Beehive Fit framework is how we are measuring our progress against these goals. We are emphasizing members who are verified, thoughtful and focused on finding love and connection while filtering out low intent profiles and bad actors. This transformation isn't about growth for growth's sake. It's about growing right. We're deliberately trading near-term volume for quality because that's what we believe builds long-term trust, stronger engagement and sustainable growth.

Consistent with this goal, in August, we launched a major update to Bumble Date focused on trust and safety. This was more than a feature release. It was a foundational step forward. These updates were built directly from what we heard from members. Men told us they weren't getting enough meaningful interactions. Women told us the profiles they were seeing didn't give them the context that they needed to take action confidently. So we've built tools to close the gap. We introduced richer profiles that let people show more of who they are. We added stronger verification features like phone number verification and selfie video verification to increase confidence. And we launched our new coaching hub filled with content from relationship experts and a help hub to reduce the time it takes to get our members to resolutions for any problems on the app. These are not cosmetic changes. They are designed to create trust to make dating more rewarding because when women feel safe and inspired to show up as their full selves, connection becomes easier for everyone.

The problem we're solving for isn't a demand problem. The desire for love and relationships is universal. To deliver on that desire, we have built a product road map that is defined by fixing our members' pain points. Let me spend a moment on exactly how we think about this. It starts with strengthening the foundation. When you ask members what their #1 complaint is with dating apps, it's really wanting more trust, safety and authenticity. The August update landed this first piece. We're building tools that guide members to put their best selves forward and engage with confidence. Between now and the spring, we'll maintain a steady drumbeat of releases, tangible improvements that reflect what we're hearing from our community. Individually, the tools and features we introduced may not sound game-changing, but taken together, they are the critical building blocks for addressing members' pain points.

The second largest complaint members have is that they're not seeing who they want to see. This is about the fundamental technology that any dating app has to get right. A major portion of the work we are doing on the UI side is building better signal capturing so that we can supercharge our matching and recommendation engine. Right now is an incredible time in technology to be doing this. AI makes so much more possible. It unlocks the ability to reinforce and enhance our matching engine in ways we could never before. These AI-driven improvements won't necessarily be visible to members, but they will deliver what they want, more quality matches.

How does that work? To get good matches, you need great signals. Great signals come from stronger profiles, which is why we are starting at this foundation. More robust dynamic profiles gives both sides of the marketplace more info to react to. This gives our matching engine the high-grade fuel it needs to deliver the best possible experience and outcomes faster. As we get these pieces right, you will see more member-facing innovation. More social tools, for instance, which are coming next year and behind the scenes, better customer service. We've already made a lot of improvements in response times and problem resolution through automation over the past year. We believe there's a lot more to accomplish here. Together, all of these components drive the experience of our members today and those yet to join our platform. We're being really thoughtful about how we're sequencing our work and what we're putting out there and when, which brings me to the final pain point.

Members want all of the above, and they want the outcomes faster. Our AI-first cloud-native platform, which we expect to launch in mid-2026 will become the engine of Bumble's future, helping us personalize at scale, enhance safety further and continuously improve the experience for our members. This new platform is expected to give us the innovation speed and flexibility to adapt faster than ever before. It's how we'll turn insights from member behavior into meaningful real-time improvement. We see this new platform as the unlock to restoring long-term product-led growth.

In parallel, we're building something entirely new, a stand-alone AI product that will become part of our portfolio. We believe it will be unlike anything else in the market, powered by our deep understanding of human connection and supported by the robust data, brand and infrastructure that already set Bumble apart. We have begun internal testing as well as getting direct feedback on the concept to ensure we get this right.

While we continue to enhance Bumble Date, we're also broadening our ecosystem through Bumble BFF, built on our modern Geneva platform, the new BFF app combines friend matching with group management and event planning. It's the next step in transforming Bumble from a dating app into a full platform for connection, and it is already performing well with increased retention compared to the old app. Our primary audience in the near-term is Gen Z and millennial women, the same demographic that powers Bumble Date. In 2026, we will expand this even further, adding group and community discovery so people can more easily find their people and form real friendship offline. BFF is both a growth opportunity and a launch pad for new social experiences that we can apply across our portfolio, further strengthening Bumble's position as the place women choose for friendship, love and belonging.

In Q3, we introduced our For the Love of Love campaign, a celebration of real success stories and real connections. It's a reminder of why people come to Bumble in the first place to find love in all of its forms. The reaction has been encouraging with a 4 percentage point improvement in awareness among single women in the U.S. aged 22 to 45. Our goal is simple: to be the most trusted, highest quality women-first platform in the world, the brand people turn to when they're ready for something real. Across Bumble Date, Badoo, BFF, we're more differentiated and AI-driven innovation, we are building a connected portfolio of love. We believe the work we are doing today is strengthening our business and setting the stage for durable, profitable growth. We are proud of the progress we are making, confident in our strategy and focus on execution. We have conviction in where we're headed, and we are so excited for the quarters ahead.

Thank you so much for your continued support. And now I will turn it over to our new CFO, Kevin Cook, who I am happy to introduce to you all today. Kevin has already brought fresh perspective and strong operational discipline to the company, and I am so excited to have him here to round out our outstanding new senior leadership team.

Thank you, Whitney, and good afternoon, everyone. In my first months at Bumble, I've been impressed by the strength of our brand, the capability of our teams and the clarity of our vision. As Whitney mentioned, our third quarter results reflect our quality-first prioritization and ongoing strategic reset. We remain focused on improving member experience, strengthening the foundation of the business and maintaining financial discipline. While some of these actions create near-term headwinds, they are designed to position Bumble for healthier growth and stronger monetization over time. Together with continued product innovation and market expansion, we believe this is the path to durable long-term revenue growth.

I will focus my comments on our third quarter performance before sharing guidance for the fourth quarter. Our third quarter results came in ahead of our expectations, but also were heavily impacted by our transformational work. So I think it's useful to start with context on the status of this work and the related puts and takes into what we are reporting today as well as our outlook for Q4.

First, with respect to revenue, during Q3, we launched our August product updates focused on trust and safety. As Whitney has explained, we are committed to improving member base quality, and we expected these updates to result in increased attrition of targeted member segments over the near-term. That attrition is reflected in our monthly active user counts with the associated reduction in paying users creating a headwind to revenue this quarter. Since the trust and safety rollout occurred relatively late in the third quarter, results for Q4 will reflect a comparatively larger full quarter impact, both from a paying user count and revenue perspective.

The second factor is marketing. We discussed last quarter how we largely paused marketing spend and in particular, stopped most performance marketing as we shifted our marketing posture to align to product launches and highly targeted user acquisition. Overall, this shift drove a significant year-over-year reduction in marketing expense and a corresponding benefit to adjusted EBITDA. This reduction is inclusive of the cost of our For the Love of Love brand campaign launched in August. At the same time, the reduction in marketing substantially contributed to the decline in registrations, active members and payers. The current performance marketing strategy has begun to show encouraging results with targeted audiences, attracting more approved-ready members into the ecosystem. While marketing spend is not expected to return to pre-transformation levels as we are focused on efficiency, we do expect some spend to return moving forward.

The third factor to discuss relates to personnel. At the end of the second quarter, we restructured our headcount to align with our product and marketing strategies. We noted at the time that we expected to reinvest much of the savings from headcount reductions, and we are already making selective headcount additions primarily in AI, product and engineering roles that support further innovation. As a result, we saw modest benefits to our Q3 expenses related to headcount. Consistent with the tech and product-led organization, this controlled hiring will continue into Q4 and beyond.

Hopefully, this discussion is helpful in shaping everyone's understanding of our performance as we execute on our strategic priorities. I'll now take you through the numbers. Unless stated otherwise, results are presented on a GAAP basis and all comparisons are year-over-year. Total revenue for the third quarter was $246 million, a 10% decline from a year ago. Foreign currency exchange rates contributed $4 million to revenue in the quarter. Bumble App revenue was $199 million, also down 10% year-over-year. Badoo App and Other revenue declined 11% to $47 million. Total expenses for Q3 were $183 million.

On a non-GAAP basis, which excludes stock-based compensation and other noncash and nonrecurring items, operating expenses were $163 million, a decline of 15%, driven primarily by a decrease in marketing activity as well as the headcount restructuring previously discussed.

Turning quickly to our key expense categories, which we report on a non-GAAP basis. Cost of revenue was $69 million, representing 28% of revenue, down approximately 1 percentage point year-over-year, with incremental improvements due in part to early testing of direct billing initiatives. Product development expense was $25 million, an increase of 14% year-over-year. Sales and marketing expense was $32 million, down 50% year-over-year. G&A was $37 million, an increase of 38% year-over-year, driven primarily by the cumulative adjustments for certain indirect tax obligations related to prior periods. Net income was $52 million. Adjusted EBITDA for the quarter was $83 million, up 1%, representing a margin of 34%, up from 30% in the year ago period. Please note that included within adjusted EBITDA is a negative impact of $12 million related to prior period indirect tax obligations. Nonetheless, adjusted EBITDA margin is temporarily elevated due to the factors I described, including the cadence of both marketing spend and our organizational realignment. We expect our margin to revert closer to historical norms as we complete technical and specialized hiring, reinstitute brand and targeted user acquisition spend and invest in updated product and our new tech platform.

Q3 cash flow from operations was $77 million compared to $93 million in the year ago period, and we ended the quarter with $308 million in cash and equivalents. As planned, we repaid $25 million of our term loan in the third quarter.

Looking ahead to the fourth quarter, as previously contemplated, our outlook reflects our expectation for continued attrition in active and paying members as we maintain higher quality standards across the platform with a full quarter of impact planned from the initiatives implemented in August. While Q4 will be challenging, we currently anticipate that the rate of sequential paying user declines will improve beginning in early 2026 as we largely complete our trust and authenticity work.

As Whitney highlighted, it is early, but these measures are showing signs of improving retention and increasing average revenue per paying user. The thesis continues to be that a better member experience will result in higher retention and drive members' perception of value, leading to increasing revenue.

For Q4, we expect total revenue in the range of $216 million to $224 million, representing a year-over-year decline of approximately 17% to 14%. We expect Bumble App revenue in the range of $176 million to $182 million, representing a year-over-year decline of approximately 17% to 14%. Direct billing tests continue to progress and nearly all members in the U.S. now have some form of direct billing available. We expect to continue to refine our direct billing offerings in Q4. We expect adjusted EBITDA in the fourth quarter of $61 million to $65 million, representing a margin of approximately 28% to 29%.

Before wrapping up, I want to call your attention to additional information we reported today in our earnings press release and accompanying 8-K pertaining to our tax receivable agreement that was created in connection with our IPO. A special committee of our Board has agreed to a transaction whereby the company will purchase all parties outstanding TRA rights for approximately $186 million. The transaction eliminates the company's TRA liability in full. We believe the transaction is a positive development for Bumble and our shareholders. It removes a large liability from our balance sheet at favorable terms, thus simplifying and creating a more efficient capital structure. By terminating payment obligations under the TRA, the transaction also improves future cash flows. And finally, it greatly improves the company's strategic flexibility moving forward as we work to create shareholder value.

I'd also like to note that we are funding the termination agreement with available cash given our solid balance sheet and cash-generative business. As a result of the TRA transaction, which substantially reduces our liabilities and deleverages the business, we no longer plan to pay down $25 million of our term loan as discussed last quarter.

In closing, there's a lot of work ahead, but early indicators suggest that we're on the path to reshaping the core business and positioning the company for future revenue growth. From a financial perspective, we're prioritizing disciplined expense management, solid cash flows and the flexibility to invest in our strategic priorities while preserving profitability. We believe we're setting the foundation for a healthier, higher-quality business that will monetize more effectively over time.

Operator, we'll now take some questions.

[Operator Instructions] First question comes from Nathan Feather with Morgan Stanley.

Two from me. First, Whitney, now that you've gotten back in the CEO seat and are really starting to make some real product changes, if we zoom forward 2, 3 years, what's your key vision for how Bumble will really work differently from today and be able to drive the successful outcomes? And then a little bit more short-term, but what visibility do you have into the timing and magnitude of when revenue growth might bottom and then hopefully start to improve, especially given the talk of potentially paying user growth bottoming in the first?

Thank you so much for the questions. I appreciate it. So let's start with the 2- to 3-year outlook. I think there's a slight misconception that exists with folks looking at the dating industry, thinking that it's the swipes that people are dissatisfied with or it's this or it's that functionality. But the reality is, and I just want to put this extremely bluntly, our product road map for the foreseeable future, for the years to come is solving for people's pain points, particularly women's pain points. And frankly, that is how we got here. That is truly what differentiated Bumble from any other dating product that had ever existed. If you think about it quite simply, you don't have a balanced member base or ecosystem or dynamic when we're talking about heteronormative dating if you don't have a place that women want to be. And frankly, that's what Bumble has been synonymous with for all of these years.

So when we look to the future, I don't think it's all that complex that you have to reinvent the wheel. The wheel works. And frankly, the demand has not changed through our history. As human beings, we just want love. We want relationships. And frankly, we need it to survive. How we deliver that needs to be modern. It needs to be current, and it has to feel up to par with where people are today with what they expect from technology. So what's so exciting about right now, and I cannot overemphasize this.

What AI gives us the ability to do? It gives us the ability to solve our members' pain points to hear women and say, "Oh, this is what you want from a dating product. Oh, you don't want this to ever happen to you again. Oh, you want to meet this type of person and you want more control over that experience." What's phenomenal about our opportunity right now is, when we have this modern tech stack that we are operating on top of we will be able to deliver these changes to our members in a matter of, hopefully, days to weeks versus months. And frankly, even historically, it's taken companies years to build certain things. So we have a superpower because we listen to women. We have women that believe in us globally, and this gives us market expansion opportunity. This gives us a dynamic opportunity to really rescale once we have this quality approach really under our belt, and we're making great headway.

So as you know, and I'm so grateful to all of our shareholders, we're very early in this transformation. I stepped back into the seat in March. We actively got to work. We have a brand-new executive team. They are superstars, and we are very bullish on going all in on having the highest quality platform. And that's not just from a member base standpoint. That's from an experience standpoint. You should come to Bumble no matter who you are, and you should be able to be very deliberate in what you're looking for and you should be able to get the great high-quality match as soon as possible. And hopefully, that should lead you to love. So that's the 2- to 3-year plan is really just build for the demand and answer the wishes and the wants and solve the problems of the dating market around the world. AI is going to be a huge part of this. But ultimately, we're here, and we're so excited about it.

So as far as when do we see a return to revenue growth. So I know this is a tricky one for people to follow along with in such tight time lines as far as earnings calls go. But we've got to complete this reset. Frankly, we have to complete the trust and safety efforts. You've seen that the Q4 numbers really reflect most of the impact of that. And it's largely what drove the payer decline along with us really pulling back on the performance marketing. But one point I'd like to make quickly, and then I'll wrap this up is, we've actually -- because of all this work we've done around quality and the Beehive Fit framework, we've actually found ways now to go do targeted performance marketing that brings in high-quality, what we call approved or likely to be approved members. So we can go back to that here in a very precise and targeted and measured way. So overall, that is on the horizon. We do see an end in sight if that makes sense, and we're very fully committed.

Kevin, is there anything you'd like to add to that?

Yes. Nathan, it's Kevin. Thanks, Whitney. So as you might expect, we're not forecasting revenue beyond Q4, but I can give you an idea of the arc, right? So just thinking about the Q3 decline in payers, for example, the decline was driven primarily by 2 intentional strategies designed to improve member base quality and, of course, member experience.

You're familiar with some of these, but let me just repeat them for clarity. So trust and safety work and a reduction in our marketing spend together contributed approximately 80% of the decline in paying users on a year-over-year basis. So it's important to recognize that we control these reductions to a large extent, and that this strategy is all consistent with the reset that Whitney outlined.

In terms of sort of progress in executing that strategy, just looking at some of the things that we know, and as Whitney highlighted already, it's extremely early. We've only been at it for about a quarter, but we are seeing some signs of encouragement. We saw a modest improvement, and you might have observed this in our prepared remarks, but we saw a modest improvement in retention, for example, over the quarter. We saw an 11% improvement in ARPPU in the quarter for Bumble. We, in addition, saw very strong uplift in brand awareness and brand perception among women in the quarter, mostly related to our brand campaign. And we are -- we continue to make progress on this internal framework you've heard us refer to before, the Beehive Fit framework, where we are attempting to lift to members from what we call internally the improved category into the approved category. And there, we are seeing all of the categories moving in the right direction. And so, that incremental progress, it's going to take time, of course, but that incremental progress is encouraging as well.

Remember that the approved members show significantly higher engagement and they monetize at more than twice the rate of improved members. So apart from understanding, as Whitney was suggesting that the functioning of the ecosystem, it's essential to have high-quality members as proxy approved members, it also has an obvious impact on business performance.

So I'll pause there, Will, and see whatever questions we have or Nathan's got a follow-up.

[Operator Instructions] We now turn to Andrew Marok with Raymond James.

Maybe digging into that last point that Kevin made there on the improved bucket. I know that was going to be a major initiative in kind of bringing along the improved members to ARPPU. And I understand it's probably a pretty wide spectrum, but is there anything specific that you're seeing out of those improved members that are giving you signs for encouragement or perhaps caution? Or anything else that you'd like to call out within those that you think would be important for us to look for as we go into 4Q and beyond?

Thanks for the question. So I'll take that. And if Kevin wants to add on, we'll do that after. So I think it's really important to understand that a lot of folks have the capability to be what we would call approved members in just a couple of tweaks. So I'll give you a quick example. You might have a great person who is looking for love, but they have no clue how to write a bio. They only have one photo. So while they could be going out on lots of dates and getting a lot of interaction, because they have limited knowledge on how to set up a dating app or how to show up at their best self, they actually get stuck in what we call this improved category. And so, what we're really seeing is the more profiles photos someone adds going through just a couple of these quick steps with the trust and the safety tooling that we've introduced, they get more matches. They get more right swipes, they get more engagement. An approved member has better outcomes. And as Kevin just stated, they have higher retention rates. They monetize much better. And frankly, it just improves that entire flywheel, and it really gives people more to operate with.

And so, when we focus on improving the improved, not to be redundant there, there's huge opportunity here because the vast middle is somewhere stuck in the middle. And so, this is what we're focused on with this road map that we keep talking about. Building an exceptional product experience doesn't have to be some fancy flashy new feature. It's frankly, "Hey, how do we get onboarding tools in the hands of our members so that they can set up a remarkably robust and authentic and great profile in a really short amount of time, and then they're out in the dating pool." And so, these are just a couple of examples of how really just enabling the system to provide easier tooling to get out there and to move into the proved category, it pulls the tide up for everyone, and it really enhances the experience for everyone.

So while we're early and while the metrics at math aren't suggesting this huge monumental moment, bearing with us and having some patience is critical here because this is the strategy that wins in dating. And I believe that I can speak to this with a lot of conviction. I've been at the forefront of this industry, frankly, on the front lines of modern dating technology since I was 22 years old. And I have seen this front and center when you have a healthy balance of women on a product, when people fill out their profiles with high-quality information, when they are not flooded with removed profiles and they see who they want to see. They get good matches, they get into great chats, they go on dates, and that's why I meet Bumble babies every day when I go out into the world. This works and this changes lives. We just need to land the strategy with returning to quality, and then we will reaccelerate on all of our growth opportunities in new markets and core markets, and we're here for it. We're very excited.

Great. Really helpful color. And then maybe one quick one for Kevin. I'm not sure if you've seen the proposed transaction or the -- excuse me, the proposed settlement details between Google and Epic today. But just wondering if you have an opinion on how that kind of contrasts with what you have assumed into your guidance for the cost of revenue line.

Okay. So no, I haven't seen the details. So I don't have any insight to share there. I can comment briefly on sort of our direct billing initiatives. As you might expect, we -- as soon as permitted, we set up alternative payments, and we've been testing those strategies throughout Q3, and we'll continue to do that in Q4. You saw in our cost of sales a meaningful improvement in Q3. And we would expect to continue to -- while we will refine strategies around alternative billing throughout Q4, we'd still expect that benefit to persist into Q4, and we should have a full quarter effect of cost of revenue benefits.

We now turn to Ygal Arounian with Citi.

I wanted to ask specifically about the stand-alone AI product, Whitney and sort of your thoughts and visions around that, but something that will sort of always be stand-alone. What's the sort of AI-first dating app experience like? And how does that impact how you view the core Bumble experience and how that overlaps with that vision and strategy?

Thanks for the question. So I think before I talk about our upcoming stand-alone AI product, I actually did want to speak about 2 things surrounding AI. So first and foremost, I'm sure you and everybody else tuning in have been reading a lot and hearing a lot about the new AI-focused dating apps. I really want to make an important point here. Throughout my career, I have seen hundreds of dating apps, dating products, dating matchmakers, you name it, hit the market and fizzle and fumble. There's a reason why there's only a couple of us that stand as strong as we do on a global level. And that's because building critical mass and building a trusted double-sided marketplace is incredibly difficult to do.

So what really gives us a unique opportunity here and a right to win is, we have extraordinary data. We have unbelievable sets of groups of people around the world that are looking for love that are actively searching for dating and us being able to lean into this moment with AI and provide them a modern experience that doesn't collapse or change the current experience. So you've got to separate these in your mind. You've got Bumble Date today, and we're going to talk about how AI affects that in a moment. But the stand-alone product is something that, in my opinion, has never been done before. It is going to be very unique. I am extremely excited about it, and our team is very excited about it.

The way it ultimately -- we can't disclose the details of how it will work. But what I will leave you with is, if dating apps have predominantly been discovery oriented, right? You get on and you kind of just discover people. This is really the first time that there will be precise search involved in this. And so, when you look at a lot of our consumer products we all use on a daily basis, they're powered by great search and great algorithms. So this AI product is going to lean into a new way of thinking about dating. And how it really will flow in our category -- in our portfolio over time is very exciting because, yes, it can be stand-alone. However, we can take a lot of those learnings, and we can take a lot of that modern technology and layer it in to the core products that we already have.

You're seeing us do a bit of this already with BFF because BFF, we've migrated on to the Geneva tech stack, and I'll give you a quick example, what would have taken us a few months at Bumble Date to really update. So Date 1.0. That's the tech stack we're on right now at Bumble Date. It might take us months to build a certain feature or product change for our members. We had a piece of feedback from members at BFF, for example, and we had that problem fully changed and involved within a week where that would have taken months on Date 1.0. So we are already integrating some of these AI changes into 1.0 to solve customer needs. But what's very exciting is you'll have this tandem approach to AI in our group because the upcoming 2.0 infrastructure for Bumble Date, which we just said will be mid-'26, that will give us the capability to build tooling and we were speaking earlier about how certain features can be engineered in hours now due to this technology, it's revolutionary. So we're going to be able to move so fast.

So the answer is yes, it will be stand-alone this new product. However, it will have overlap crossover function just like we're doing with BFF, where we're learning about how groups behave and how communities behave so that we can build that into the core dating product.

So I hope that answered the question. If it didn't, please let me know.

It did and very helpful. And I guess maybe a little bit more near-term, just as you get through to the spring product release and you're talking about kind of the, I guess, steady releases through spring and solving pain points. Are there a few things that you see as being sort of the biggest drivers? I know we're doing all the cleanup work and enhancing the user experience on one side, but in terms of like new products that are coming out, are there a few things that are sort of excite you the most?

Yes, that's a great question. So this might be a slightly unpredicted response that the least exciting features on paper are actually the ones that move the needle the most for us sometimes. So when you go and invest time and energy and optimization, for example, into customer service, that has a meaningful impact on our members' lives. They are feeling heard. They're getting their problems resolved. Now, on paper, that's not some flashy new release.

So what we're really focused on right now between the launch of 2.0, and I'm using that as a reference point. 2.0 is the new AI cloud-native tech stack that Bumble will live on in the future, which is the mid-'26 technology. But if you look at everything we're doing between now and then, it's just listening to our members, right? Our product road map is our customer pain points. And as I said, you don't ask members, "Hey, what's not working for you and they say, "Oh, wow, I really wish you had this flashy crazy feature." They just say, I want a better profile. I want to be able to see more information about someone. I want a safer experience. I want the algorithm to be more personalized to me. These are at face value, quite basic asks, but this is what builds an incredible experience. And frankly, this is what got us to being the great company we are. And yes, the technology has been innovating and iterating over time. And obviously, we have to evolve with where the category is. But ultimately, the strategy doesn't change. Just listen to women and give them what they want. It's very simple, and that's exactly what we're doing, and that's how we win with the category.

We have no further questions. So this concludes our Q&A and today's conference call. We'd like to thank you for your participation. You may now disconnect your lines.

Good evening. Thank you for attending the Bumble Second Quarter 2025 Financial Results Conference Call. My name is Megan, and I'll be your moderator for today. [Operator Instructions]

I would now like to pass the conference over to Will Taveras, Bumble's Investor Relations. Please go ahead.

Thank you for joining us to discuss Bumble's second quarter 2025 financial results. With me today are Bumble's Founder and CEO, Whitney Wolfe Herd; and Interim CFO, Ron Fior.

Before we begin, I'd like to remind everyone that certain statements made on the call today are forward-looking statements. These forward-looking statements are subject to various risks and uncertainties and reflect our current expectations based on our beliefs, assumptions and information currently available to us. Although we believe these expectations are reasonable, we undertake no obligation to revise any statement to reflect changes that occur after this call. descriptions of factors and risks that could cause actual results to differ materially from these forward-looking statements are discussed in more detail in today's earnings press release and our periodic filings with the SEC.

During the call, we also refer to certain non-GAAP financial measures. These non-GAAP measures should be considered in addition to and not as a substitute for or in isolation from our GAAP results. Reconciliations to the most comparable GAAP measures are available in our earnings press release, which is available on the Investor Relations section of our website at ir.bumble.com.

With that, I will turn the call over to Whitney.

Hello everyone, and thank you for joining us for our Q2 earnings call. Four months ago, I returned as CEO of Bumble and reset our strategy for quality over quantity across the whole business, including how we operate our revenue and payers and importantly, our member base. We have taken decisive actions over the last quarter. We've removed over $100 million from our cost base by streamlining operations, restructuring head count and shifting to a more efficient organic marketing engine. These changes have sharpened our operational discipline and positioned us for a return to growth.

We've reorganized for speed and focus, rebalancing engineering to the U.S. unifying our product road map and bringing in exceptional leadership, all while building a leaner, higher-performing team. These efforts contributed to record EBITDA margins in Q2. And while this level of performance shouldn't be viewed as a steady state baseline profitability margin expansion and cash flow will remain core priorities even as we redeploy some of our cost base savings into selective high-impact investments across product, AI, UX and trust and safety. We believe we can do both, run a healthy, efficient business and strategically invest in the long-term future of bumbles with discipline, intention and a relentless focus on quality. That quality extends to the type of revenue and payers we are prioritizing.

The majority of this quarter's payer decline came from the phase out of a legacy strategy related to promotions. -- which were skewed towards certain of our lower monetizing noncore markets. As of late April, we have been reemphasizing our core of sustainable full-price subscriptions. Encouragingly, our full-price payer base increased quarter-over-quarter and subscriptions now represent approximately 80% of total payers, up from 70% in Q1. This is reflected in our ARPPU increases. We believe this may be an encouraging signal that removing lower quality revenue moves us towards a stronger, more sustainable baseline for our payer base as we look to return to growth.

Now let's talk about the progress we're making on improving the quality of our member base. At Bumble, our product is people. The quality of someone's experience how they engage, find what they're looking for and monetize depends on the quality of who and what they encounter on the platform. We are executing this work through our BeeHive framework which is designed to create the right conditions for meaningful connections, authentic, robust profiles, clear intent signals and respectful behavior. It's not about exclusion. It's about how well people show up. To assess our members using our BeeHive framework, we are using 3 categories. The first is approved. These are members showing high intention, effort and alignment with the experience we are building. They are verified. They engage well on the product. They have robust profiles. These people are great members of the community, follow our community guidelines and are seen as high quality and compelling data by the other members on the platform.

The second and largest group is improve members who have the right intentions, but may need clear guidance better tools or a stronger sense of what a high-quality profile and experience should look like. We are focused on supporting them. And this represents one of our biggest areas of opportunity. For example, someone who follows all community guidelines, but has only one blurry photo and no bio. That's a simple high impact improvement. By helping these members show up more fully, we believe we can move more of them to the approved category. This will enhance their experience and drive better engagement for them. And in turn, unlock a healthier member base with increased retention and stronger monetization. Importantly, approved members monetize at approximately double the rate of our improved for.

The third category is removed. We have always prioritized trust and safety. The last quarter, we significantly deepened our focus. In March, we accelerated efforts to identify and ultimately remove bad actors and low intent members, a strategic step aligned with our quality-first approach. It includes removing duplicate accounts, members who break trust, but, scammers or those that don't follow our community guidelines, but also members with low quality and complete profiles and no willingness to improve. It improved members don't take the step to engage meaningfully they transition to remove. We believe that we're moving these profiles will create a stronger, safer environment for our high-intent community, making the overall experience much better, even if it contributes to a near-term headwind to payers.

Our goal with BeeHive is to move as many members into the approved category as possible. And this will be an always on focus for us over the long term as new members continue to join us. We remain laser-focused on cleanup, discipline and resetting the bar, but we're already shifting into build mode as we set the stage for renewed sustainable growth. Our August Love launch marks the first major step in this next phase. It's a trust first product update that includes phone and ID verification, mandatory selfie checks and richer profile building tools, all designed to help high-intent members show up more clearly and confidently, while naturally filtering out those less willing to engage meaningfully. We are also introducing our human and AI-powered coaching hub. This launch is foundational to strengthening the platform and improving outcomes for the people who are here for the right reasons. From there, will maintain a steady drumbeat of improvements across product and tech with another milestone in February when we debut what we believe will be the safest and most innovative version of bumble yet. This next phase is about reactivating our community and personalizing every part of the journey from onboarding to matching, to meeting in real life.

On the BFF side, we will also launch the all-new Bumble BFF app this month built on Geneva's group tech and bumble safety infrastructure. It combines one-on-one matching and events with community features to quickly follow designed to help people build real friendships off-line. With minimal investment, BFF is already a top friend-finding app in the United States, especially among Gen Z and younger millennial women, and we see it as one of our most exciting long-term growth opportunities, especially as demand for friendships, real-world connection and belonging continues to grow. Bumble and Bumble BFF are both powered by a clear technology vision, build proprietary AI capabilities to create more human connection.

Under our new CTO, we've rebuilt our core tech work with a single mandate, help people find love safely and efficiently. AI is being embedded responsibly and ethically across the entire product ecosystem from matchmaking and personalization to member support and internal operations. We have data-rich insights on what our members are looking for and how they achieve successful outcomes. For bumble date, we are building a fully modern AI-first cloud-native tech stack from the ground up with a new infrastructure set to be completed by next spring. In parallel, we'll continue optimizing the current stock to support ongoing performance and stability.

Bumble BFF, meanwhile, is migrating to an already modern platform through our acquisition of Geneva, giving us speed and flexibility in scaling group and community features. Everything we built is grounded in real world outcomes, not endless engagement. Our AI systems will be trained and continuously refined by real matchmakers and relationship experts, ensuring that insights, prompts and recommendations are shaped by emotional intelligence and modern relationship science.

Now let's talk about marketing. At Bumble, marketing isn't here to cover for an average product. It's here to amplify a product people need and love. We are focused on building the best experience in the category because a great product is what keeps people coming back. Historically, our performance marketing lacked clear guardrails. We've now realigned marketing with our focus on organic growth high-quality acquisitions and durable engagement. We are already seeing early signs of progress. retention and organic registrations are up. We're doing all of this while preserving one of our strongest assets, our brand. Among scaled dating apps in the U.S., Bumble holds the highest favorability among women and the general population with a competitive standing among Gen Z with improvement in each of these key demos in Q2.

Coming alongside the product update and the Love launch later this month is a brand moment rooted in real stories of love, connection and family that Bumble has helped to create. It is a powerful reminder for the impact we help create in the world. Behind all of this work is a team that's getting stronger and stronger. In just a few months, we have both shrunk the team to a much leaner organization and added exceptional leaders across products, tech, marketing and legal and now finance. We are thrilled to welcome Kevin Cook as our new CFO this month. Our leadership transformation has helped us attract top talent with purpose and passion. We are being selective with how we invest in talent with a very high bar for any additions, including considering whether AI can do all or part of the work of each potential new head, prioritizing a lean, high-performing team.

Across the company, we've reorganized to move faster with more focus and ownership. We are building a team designed for scale, speed, and with heart. We are just a few months in, but the energy is real, the clarity, the focus, the urgency. It's all here. We are not building for where the world was or is. We are building for where it's going. We have committed to bringing bumbled back to being a truly member-first company. Everything we're building is rooted in what our members want, need and deserve when they speak, we listen. Our primary goal is to deliver an experience that helps them find real love friendship and connection.

Now I'll hand it over to Ron. Thank you so much.

Thank you, Whitney, and good afternoon, everyone. As Whitney mentioned, Q2 reflects early results of the decisive steps we've taken to recenter bumble around quality and to deliver the best member experience possible. reshaping bumble to prioritize quality and the member experience has, by design, weighed on our revenue and paying user count, while also improving ARPPU. We are making this trade-off intentionally in order to prioritize high-quality, relevant matches and a better member experience to drive Bumble's organic, sustainable growth in the years ahead.

Today, I will share updates on how our member base improvement strategy may influence our performance near term, but I will start with a summary of our second quarter financial results. Our Q2 performance exceeded the company's late June upwardly revised expectations. Bumble Inc. Q2 revenue was $248 million, including a favorable impact from foreign exchange of approximately $2 million. Mobile app revenue was $201 million, including a foreign exchange tailwind of approximately $1 million. This brings year-to-date total revenue to $495 million. Total paying users in the quarter were $3.8 million, and Bumble app paying users were $2.5 million. Badoo App and other revenue was $47 million, including a $1 million foreign exchange tailwind during the quarter and benefited from fruits and official revenue of approximately $3 million in total. Before each was sold and shut down, respectively. Badoo App and other paying users totaled $1.3 million.

Turning now to expenses. Total Q2 GAAP operating costs were $587 million, and we reported GAAP net loss of $367 million, primarily driven by an impairment loss of $405 million. On a non-GAAP basis, which excludes stock-based compensation and other noncash or nonrecurring items, operating expenses were $154 million. The year-over-year decline of approximately 21%, was primarily driven by reduced marketing expenditure and lower headcount as a result of the restructuring plan executed in 2024. Adjusted EBITDA for Q2 was $95 million, representing 38% of revenue, benefiting primarily from a lower-than-anticipated impact from the reduced performance marketing spend, more selective spend in brand marketing, head count savings and the impact of the additional fruits revenues. Year-to-date, adjusted EBITDA totaled $159 million, representing a 32% margin. We reported strong cash flow of $71 million in Q2. Year-to-date, we have generated over $114 million from operating activities and we ended the quarter with $262 million in cash and cash equivalents.

I'd like to highlight a few notable updates from the quarter. As Whitney indicated, we have moved decisively to streamline our operations and optimize our execution on our strategy. We have removed approximately $100 million in annualized costs including approximately $40 million related to the June reduction in our workforce, which affected approximately 240 roles with the balance in marketing spend as we reorient our strategy in support of our organic growth focus. Profitability, margin expansion and cash flow will remain our priorities as we redeploy some of our cost base savings into selective high-impact investments across product, AI, UX, trust and safety over the coming quarters. With respect to performance marketing, we cut even deeper than initially planned, as we saw that the cuts we were making had a smaller-than-expected impact on our near-term revenue. This outcome has reinforced our conviction that Performance Sense was not efficiently acquiring high-quality members and validate our strategic shift away from performance-focused marketing. Overall, you should anticipate that marketing spend will increase from Q2 levels in the second half of the year in support of brand initiatives including our August product moment as well as spend later in the year in support of the BFF initiatives and our regular drumbeat of product updates.

On the talent front, we are being highly disciplined on head count with our focus primarily on product and AI-related engineering roles as we refocused these teams in the United States. The initiatives are factored into the Q3 guidance we've issued today. I also want to call out that in June, we began testing direct billing options for some purchases on iOS in select U.S. markets, offering customers discounted pricing. Early results have been positive with up to 30% of targeted members opting for direct billing. We will continue refining this process, but we already expect we could see a modest positive impact to our gross profit dollars in fiscal year '25 as a result of this change if it rolls up more broadly. It is worth noting that these tests could result in headwinds to our reported revenue as we offer lower prices but should result in improved gross margins as we keep more of that revenue compared to a full price purchase after accounting for App Store fees.

Turning to guidance for Q3. We expect total Bumble linked revenue between $240 million and $248 million, representing a year-over-year decrease of 12% to 9%. Excluding foreign exchange and Fruitz and Official, this translates to a decrease of between 12% and 9%. We expect Bumble App revenue to be between $194 million and $200 million, representing a year-over-year decrease of between 12% and 9%. Additionally, we estimate adjusted EBITDA in Q3 to between $79 million and $84 million, representing a margin of approximately 33% at the midpoint of the range. The sequential comparison reflects primarily a return to brand spending in support of our August product moment. In each of Q3 and Q4, we expect to pay down $25 million of our term loan. On that note, I will add that we did not repurchase shares of our common stock during Q2 and continue to have $50 million remaining on our current authorization.

To add context to our outlook, we've shared today that we have accelerated execution of our BeeHive fit framework to drive improvement in our member base. We expect these efforts, combined with the reduction in performance marketing spend that continue to impact year-over-year revenue comparisons. Later this month, we will be launching our August product moment focused on trust and safety, which is designed to improve the quality of the member base overall, but it is expected to drive further near-term attrition from lower intent members. You will see the beginning of this impact in the remainder of Q3, then more fully in Q4. We believe that as early testing has already shown as overall quality improves, we will begin seeing positive impacts on the retention and ARPPU from our higher intent users, ultimately setting the stage for a positive inflection in members.

Our defined and focused member-based strategy is well underway and showing early promise. We believe we can continue to execute with pace against our strategy while maintaining a strong financial profile including solid adjusted EBITDA margins and cash generation. By driving an improved member experience today, we're advancing towards a bundle with stronger retention richer monetization and a sustainable long-term growth trajectory.

We appreciate your ongoing support, and we will now turn to Q&A.

[Operator Instructions] Our first question goes to the line of Shweta Khajuria with Wolfe Research. .

This is Andrew Rob for Shweta. I want to follow up on the alternative payments comment. I guess could you please provide updates on kind of where you are in terms of the alternative payment options? What specifically are you testing? And I guess what kind of range of discounts are you offering? And then what you're seeing towards the adoption rates?

Yes. So what we're doing on -- we've done some testing, and it's been very limited at this point in time. And we're looking at different discount rates. And one of the things that we had seen very early on in this testing is that we tested a reasonable group of opportunity in people in that Apple iOS system and giving them a discount. And we found that the take-up on direct billing was actually 30%, which is very, very solid. We are continuing to do testing on that at different discount rates. And we've included a little bit of that in our guidance for the rest of this year or for Q3, but we are looking at -- if we -- there is an opportunity for upside in our numbers if we continue to do that through the whole rest of the year.

Our next question comes from the line of Andrew Marok with Raymond James.

Wanted to touch on ARPPU growth at Bumble App really quickly. So was the growth there solely driven by that promotional strategy, the exit that you had mentioned? Was there anything going on with maybe the underlying ARPPU growth that's worth calling out? And then I have a follow-up after.

Yes. So on ARPPU growth, I mean, I think it was a result of some deliberate changes that we have made in the monetization and some pricing optimization. It's an early signal of some of the impact of the changes that we're making, as we focus on providing fewer, better, deeper relationships. I guess I would remind you that it's really just an output. It's really not a number that we're managing to at this point in time.

Appreciate it. And then maybe on the user front, Whitney talked about the 3 buckets of users, the approve, improve and remove. You mentioned that improve was the largest, but can you give any guardrails or quantitative sense on the relative size of the buckets? And how many of those improved users actually want to improve?

Yes. Thanks so much for the question. So improve does make up a vast majority of the member base. And perhaps reassuring here remove actually accounts for under 10% of the member base. So here's the thesis. What we're really excited about is that we see real data and real green shoots with approved members. They have higher [indiscernible], they place higher [indiscernible], better retention. They get more matches. They have a double willingness to pay. So if you take this strategy which I said in the prepared remarks, which is to essentially move as many of those improved up to approve, you can imagine the output and the results that this has on the overall ecosystem and member base. And so this is exactly what we're focused on. Our product road map and our recommendation engine is very focused on making sure that approved, see the proof and that we are giving the tools through coaching, through AI, through other product functionality to move that improved base up because when you do that, everything improves across the board, and you are left with a much higher quality member base, and this is exactly the strategy of quality over quantity.

Our next question comes from the line of Nathan Feather with Morgan Stanley.

As you work through the turnaround, we know what metrics you're tracking internally as you make this push for quality and for investors looking into outside, what do you think the best way to track your progress given the noise in terms of the reported KPIs?

I appreciate the question. I'll take that. So you can imagine that with this reset, to quality first. I have been extremely obsessed and the team has been extremely obsessed with tracking deeper signals, deeper member inputs. What are the inputs across the board to get our members to success? If you think about why people are here, they come to the product, to find real high-quality interactions that lead to real quality date. So every single thing that we are watching is all in the effort of getting our members to high-quality compatible relevant matches as quickly, efficiently, safely and effectively as possible. And so while we're not willing to share those deeper metrics or those deeper signals with the Street right now, rest assured that the team is very focused on this. And we will keep everybody posted as we expand our metrics in the future.

Great. That's helpful. And then on the reinvestments you're making, just any way to think about the timing of those reinvestments and especially given some of the usual costs and AI talent associated if those might be rolled in over a longer time frame?

Yes. Thanks. That's also a great question. So we're being extremely precise and surgical with reinvestments. We are not cutting costs to just spend again. As I said in the prepared remarks, every single dollar that is being spent at this company is being rigorously evaluated through the lens of is it a must-have for the quality strategy. Is this helping get our members closer to offline durable love? And is this going to make us the leader in the space as rapidly, effectively as safely as possible. So as we evaluate that, we are putting every headcount member, every new headcount through that lens, every technology update for that lens, and we are only spending where absolutely necessary. Some of that spend will start hitting here in the near term. And some of it will be deferred to longer lead when we do things like strategic brand marketing around product launches. So just rest assured that we are being extremely diligent and we are managing to have an incredibly efficient cash flow positive and healthy business. At the same time, we are running for the long term and optimizing for long-term success.

Our next question comes from the line of Eric Sheridan with Goldman Sachs.

Maybe following up on that last question and just sticking with this theme of sort of marrying priorities with investments. When you look out over the next 12 to 18 months, what do you see as your biggest priorities to move the needle for the platform? And how should we think about the depth and duration of investments lining up against some of those key priorities, just to put a finer point on some of the messages from the prepared remarks?

Thank you. I appreciate the question. So the main priority is our products and technologies, hands-down. As we said in the prepared remarks, we are actively rebuilding a Bumble 2.0, if you will. This is AI first, cloud native, extremely sophisticated tech infrastructure. This is really meant to enable innovation at speed, at scale and very personalized. And while we do that, we are maintaining a steady state of productivity and support to our members while we continuously make these micro enhancements that really count on the current technology. So for example, over the last quarter, we enhanced our recommendation system and we continue to leverage deep learning and rich profile signals and implicit signals from these in-app actions. And those improvements alone on our legacy system actually drove a greater than 10% increase in women's yes boats. So we are able to continuously move our metrics upstream while we focus on these longer-term swings that will require time focus and reinvestment. So the priorities are really essentially very simple, deliver a quality experience that is both product but tech, recommendations, matching. Our priority is always trust and safety. This is what particularly women, but everyone wants, People want to feel safe. They want to know that who they're meeting, is who they say they are meeting. They want to feel a trust and safety backbone when they're using these products to find friends. Another big priority is Bumble For Friends. I cannot tell you how excited and how convicted we are in the future. So organic demand for Bumble For Friends, particularly from Gen Z women and younger millennial women is extremely exciting. And so we're really putting a lot of time, energy and focus in this modern technology to support front lining and communities. And I think you'll notice that I didn't mention marketing. And I think that is important because our product is going to speak for itself and the marketing that we do from both an investment standpoint but also from a team resourcing standpoint, is there to amplify a great product. there to amplify very innovative, safe quality technology. And so you will see us spread the most important message that we have to offer, which is the amazing success stories we put into the world. So those are really the high-level priorities over the next 12 to 18 months.

Our next question will go to the line of Ygal Arounian with Citi.

So maybe just first, Whitney, on the bundled users and the 10% or less than 10% in that remove category, talked about some of the improved users that don't want to improve potentially getting down to that removed bucket. Some of the trust and safety features that could cause dislocation, you've also less focused on some of the lower-value users. And so you've kind of like naturally aiming to tear down the user base before you build it up. So I'm just trying to understand what that level might be to the extent that you could help on where we kind of bottom out? And then as you begin to try to build up, you're going to be targeting a smaller base, right, to not track some of these users that you're electing not to bring back. And so how you do that and how you start to build back up and kind of get back to a user base that's higher than your previous peak?

Great question. So I think this is actually slightly different than what meets the eye. You would be surprised by how many phenomenal members we have that technically classify as lower improved. They're not bad people. They're not nefarious members. They have no clue how to build a profile. These could be extraordinary people that when you meet them in real life for like how are you single. But when you look at their Bumble profile, they have one photo, they're wearing a mask, and they have no buyout. There's no chance for them on our product in that construct. So this is really not about us needing to remove just the whole member base because everyone is a bad member, absolutely not. Frankly, the quality of our members are really good, but the quality of their profiles are not so good sometimes. And so this is really an effort to really improve the quality of how they show up, how they put themselves out there and how they express themselves so that they can move up that ranking. Now there is a subset of members that does fall into that removed category. And you're right, we will be taking them off of our platform. However, fewer better is always going to win when it comes to connection and relationships. Just more profile. If you were to swipe through 100 people, you never wanted to meet, you would walk away feeling very, very disappointed. But if you were to go through even just 5 or 10 or 15 profiles, of very high-quality profiles, and everyone was actually quite interesting to you would feel very, very compelled to return. So this is actually not so much of having to kick everyone out. It's really about just removing that less than 10% of folks that shouldn't be here, if they're box, they're scammers, the duplet hit accounts. They have no intention of behaving well. And then finding those folks that are lingering around the improved category that maybe have no willingness was bring them down within everybody is up or out. I mean, it's amazing to see what happens. When you just help people out a little bit, they really do have a much better experience and then they have positive impact on the member base. So high level, I just don't want anybody to walk out of here thinking this is a onetime strategy. This is always on. We get new members every day, and we will have to consistently sort our profile, improve our profile, and this is an always on approach, but we are very confident that as this strategy continues to be executed against with speed, with focus that we are going to have an extremely high-quality member base that will create all of the outputs that matter for a great business.

And maybe just a follow-up on the investments -- on the investments that you're planning, including marketing. It doesn't sound like you're targeting a specific margin target. But as we think about our models and as marketing comes back in and you start to roll out some of these investments, and any way to help frame the right way for us to think about how margins might flow through next year and beyond and how you think about that?

So maybe I can answer some of that. And I think -- well, I'm not going to go out into next year because we're not giving guidance that far, but I think I can give you some color towards the rest of the balance of this year. And looking essentially on the income statement, starting with the revenue side, we effect that the trust and safety, which we've talked about in our script today, will have an impact, obviously, in Q3, and we'll have a greater impact in Q4 and that includes everything from mandatory phone numbers and self identification and those kind of things. So that will have a -- it will have a near-term attrition to our payers.

From a gross margin point of view, we talked about the alternative billings, the impact of the reduced app store fees from there. And so that will be could be a potential upside from a gross margin point of view, but that could also be a little bit of an impact -- negative impact on the revenue because obviously, less revenue. But at the end of the day, we'll end up having a greater margin in dollars at the bottom line. When you look at the marketing, we've talked -- we talked a lot about this already, but we're going to have -- we will have very minimal performance marketing in the second half of this year. What we did say is that we will have more marketing, brand marketing, more organic marketing that will occur in the second half of the year than we incurred in the -- if you think about the second quarter level, it will be at a higher level than that. Again, it's going to be very focused on the organic side. So product and engineering side, again, the head count, we're going to continue to -- we've just done this large workforce reduction. But now what we're going to do is we do have to build up our presence in the United States and not merely in the engineering area. In all other areas, it's going to be very, very limited increase in our headcount. And that's really the key things for the balance.

Yes. The only thing I'd like to layer on to what Ron said is when it comes to marketing, again, I want to be extremely clear, we are not going to market for marketing's sake. We are only going to be doing marketing that amplifies the strength of the quality and the power of our product. And we are only going to be reinvesting into the halo of our brand. I cannot -- I cannot enforce enough how important our brand is, how it really is a huge strength for us. People choose us over competitors because of the way Bumble makes them feel from the brand perspective. So rest assured extremely measured and precise and surgical when it comes to marketing and only there to enhance the power of the brand.

Our last question will go to the line of John Blackledge with TD Cowen.

It's Logan on for John. As you release new features in the August launch and looking forward, could you talk about how you are appealing to Gen Z users who maybe have grown weary of the traditional online dating experience. And then following up on the CFO appointment. Could you just talk about equality that led to Kevin retire?

Thanks for the question and good to hear from you. So let's talk about Gen Z. I think there's a bit of a misconception that Gen Z is some completely different species that doesn't think about love and connection the same way most of humanity does. The reality is this Gen Z entered the dating market rather the online dating market at a time where most of these products were already very saturated. So a lot of the exact product solves that we are so maniacally focused on right now or specifically the issues Gen Z has with online data. For example, they don't want to feel like they slide evenly through people that are not interested at. They don't want to feel judged. They don't want to feel rejected. If you don't really feel like they're talking to someone that is not actually who they say they are. These are the same issues that everyone has struggled with, with all mine love. And this is precisely what has driven the strategy back to quality first. When I came back in as CEO, I showed up with a list of 10 of the top customer pain points that I had researched extensively. And I basically came back and said, until we fix all of these problems, we're not going to just roll out random features. Feature fatigue is not solving the problem for anyone. This is about solving the real needs of dating and connection, and that is precisely what we're doing. That's what you'll see in the August launch. It's all about test and safety. It's all about efficiency. It's all about feeling relevant and compatible again. It's all about helping people enjoy dating again and build confidence back up because that's what made this category special to begin with. And so this is really how we win back ends.

Now I'd like to touch on something else because it's very important. Gen Z has shown us through the trends and the data that we are tracking on BFF that they are actively seeking off-line connection, friendships and groups. They want to be a park community. And this is one of the brightest thoughts in our entire group to go and lean into this with Bumble For Friends. And the great news is we are already there. We are a leading friendship app in this space. And frankly, we're the only one in the dating space that has a friend-finding feature at scale. So this gives us a real competitive edge. And we're really excited about that.

And now to the CFO. Yes. So Kevin Cook, what an exciting hire. Listen, he checked every box. We wanted someone that understood technology first. We wanted someone that was a product thinker. We wanted a finance leader that understood the demand that members have from a product and to really lead for the customers, to lead for the members first because that's how you drive a great business. And that's the strategy that we are operating against right now. He comes with deep expertise, deep experience, deep understanding of many technical components. He's going to be a great partner to us as we navigate this next era of AI and as we navigate a return to growth, and we're very, very excited to have them. And we're deeply appreciative to Ron, who has been extraordinary.

Thank you. And with that being our last question, we will now conclude today's conference call. Thank you for your participation, and enjoy the rest of your day.