Hemnet Group AB (publ) (HEM) Earnings Call Transcript & Summary

Welcome to Hemnet's Q2 Conference Call. [Operator Instructions] Now I will hand the conference over to the speakers. Please go ahead.

Good morning, everyone, and a warm welcome to this 2026 Q2 release call for Hemnet Group. My name is Jonas Gustafsson, and I'm the Group CEO of Hemnet. We are a slightly bigger group than usual today given the extended agenda. Today, I'm joined here at our Stockholm headquarters by our Chief Financial Officer, Anders Ornulf; our Chief Operating Officer, Lisa Farrar; our Chief Technology Officer, Hanna Lindqvist; and our Head of Investor Relations, Ludvig Segelmark. Today, we have called for an extended session to cover an update on some important strategic and commercial topics and will, therefore, have a slightly longer presentation than usual. With that, let's have a brief look at the agenda and what we have ahead of us for the coming 120 minutes. Please move to Slide #2. Firstly, we will start with a normal quarterly presentation where we will go through the financials and a business update from Q2. After that, we will follow up with a deep dive on Hemnet's market position, followed by a deep dive on both the commercial and product side as well on the technical and AI side. As always, there will be opportunities to ask questions at the end of the presentation. We will combine the Q2 Q&A with a deep dive Q&A into one session in the end of the presentation. Today's presentation will be moderated by our operator, so please follow the operator's instructions to ask questions through the provided dial-in details. So with that, let's get started, and let's move on to Slide #5, please. During the quarter, net sales decreased by 23%, driven by lower listing volumes paired with a timing shift in revenue recognition from Sell First, Pay Later. The financial results reflect an expected transitional phase following the nationwide rollout of Sell First, Pay Later. Anders will explain this revenue recognition effect in more detail further on in the presentation. Published listings declined by 14% and amounted to 43,300. Paid listings came in at a total of 32,900 with a difference between paid and published listings being explained by the Sell First, Pay Later listings that were published but not yet sold in the quarter. ARPU, average revenue per listing grew by 12.4% in Q2, driven by higher demand for Hemnet value-added services. The EBITDA margin amounted to 46.4% in Q2. The lower margin year-on-year is explained by lower revenues driving lower fixed cost leverage. As volumes improve and our Sell First, Pay Later listings continue to convert into sold listings, we will see a stronger margin development going forward. Now let's turn to Slide 6 for a quick look at the financial performance. Net sales amounted to SEK 372 million, down 23% compared to Q2 last year, driven by a combination of a timing effect on revenue recognition from the introduction of Sell First, Pay Later and a decline in listing volumes during the quarter. As a result of the lower net sales, EBITDA decreased by 33.9% to SEK 172 million. Lower revenues leads to lower fixed cost leverage, which explains the lower margin in the quarter. The EBITDA margin amounted to 46.4%. As per usual, Anders will break down these profitability dynamics in more detail as we move on in the presentation. Now let's turn to Page 6 for a look at the underlying property market and the listing volumes development during the second quarter. In Q2, we saw a sequential improvement of listing volumes compared to the previous quarter. On the right-hand side of this slide, you will see a combined chart showing published listings per quarter and yearly published listings as well as the year-on-year change between quarters. Published listings decreased by 14.3% year-on-year in the second quarter. The trend improvement in number of listings versus the previous quarter is driven by a number of factors. Number one, the launch and the rollout of Sell First, Pay Later; number two, the introduction of strategic partnerships; number three, new regulations implemented by the beginning of the quarter; number four, more healthy underlying market transaction dynamics. However, the market remains soft and the slow market also continues to be negatively impacted by longer selling times. And the average listing duration on Hemnet has increased by 33% year-on-year to 64 days compared to 48 days in the same period during the last year. Please remember that this is an LTM figure. This number is also impacted by the higher number of transactions that we saw in April and onwards, where more old inventory was sold, which pushed up the age of the sold listings on the platform. I will break this down in more detail on the next slide. All in all, a broader market recovery is progressing slightly slower than anticipated at the start of the year, but we do see clear signs of gradual improvement during Q2. So with that, let's turn to the next slide and have a closer look on the historically high supply levels going into 2026 and how that impacts the market recovery. From the spring of 2022 up until the fall of 2025, we had a gradual and steady buildup of supply of listings on Hemnet. That means that during this 3-year period, more listings were published on Hemnet and in the market that were sold. You can see this quite clearly on the graph on the left-hand side of the slide, where the supply graph is higher every single year from 2021 to 2025, leading up to the record levels that we saw during last year in 2025. During the second half of 2025, this trend started to reverse. And during the first half of 2026, we've seen a declining year-on-year trend for the first time since 2022, which you can see in the graph in the middle of the slide. This follows a pattern that we've seen historically, where a period of oversupply is followed by a correction where the number of transactions outweighed the number of new listings on the platform. From a short-term perspective, this has negative impact on Hemnet, but from an overarching level, this positively impacts the market dynamics in the quarters to come. In the last graph on this slide, you can see the age distribution of sold listings in 2026 and how that compares to sold listings in 2021. In 2021, almost 80% of sold listings were less than 30 days old. In 2026, the same number is 44%. On the same note, in 2021, only 5% of sold properties have been on the platform for longer than 90 days, whereas today, that number is almost 30%. As the property market is showing signs of improvement and recovery, more sellers and buyers are able to meet, which is driving more transactions and more of the existing supply being sold. This has a short-term effect on the listing days development, but it's very positive going forward as it opens up for new listings to come to the market and onto the platform. Let's move on to the next slide and break down why we expect to see a continued improvement in the Swedish property market for the second half of the year. The Swedish property market is showing signs of improvement going into the second half of this year. In Q2, transactions on the property market increased by 11% compared to the same period last year, driven by the eased credit restrictions that were put into play by 1st of April. A more active market where sellers and buyers are able to meet is very positive for Hemnet and tends to drive more listings to the platform over time. After several years of higher interest rate environment, rates have stabilized into 2026. A stable interest rate environment creates predictability for buyers and sellers, which is a good sign for the period to come. We have also seen a strong price development in Q2 2026. During the quarter, prices increased by 6.2% for apartments year-on-year, while the development for detached houses was plus 4.5%. In addition to the strong price performance, we're also seeing a continued positive market sentiment on price expectations in our monthly buyer barometer, where a significant share of prospective buyers expect prices to continue to increase in the coming 6 months, whereas the share expecting prices to come down continue to be below historical averages. All in all, there are a number of positive signs that indicate that the market will be stronger during the second half of next year and next year, which will help listing volumes on the platform as well. And with that, I will hand over to Anders for a financial update, starting with Page 9. Anders, over to you.

Thank you, Jonas. Let's turn to Page 11 and the financial summary. As Jonas mentioned earlier, we are seeing a gradually improving market environment. While the volume of new published listings fell by 14% year-on-year, this represents a significant recovery compared to the steeper declines we experienced in the first quarter. The underlying volume trend, combined with the deferred revenue effect from our Sell First, Pay Later model resulted in a 23% decline in net sales, landing at SEK 372 million. While paid listings fell by 35%, the widening gap between published and paid listings is driven by the SFPL listing that remained unsold at the end of the quarter. I want to emphasize that Q2 is the peak quarter for absorbing the financial impact of this transition. Following our nationwide rollout in April, where the service was offered to all property sellers for the first time, we naturally saw a high volume of initial listings under the new model, but with a relatively low share of listings actually sold and invoiced during the period. The vast majority of the revenue decline stems directly from this timing effect. In Q2, 40% to 45% of all published listings were under the SFPL model. At the same time, we are building a clear pipeline of around 10,000 unsold listings, which will be realized as revenue -- as these properties are sold going forward. Also on a positive note, we continue to see strong underlying performance in our paid ARPA, growing 12.4%, once again proving the sustained and increasing demand for our value-added services. Another noteworthy point is the average listing time, which on a rolling 12-month basis increased from 48 days in Q2 '25 to 57 days in Q1 '26 and now 64 days in Q2 2026. The year-on-year effect of the longer listing duration time is negative SEK 13 million in revenue and the sequential effect of the 7 additional days from Q1 to Q2 is negative SEK 5.5 million. The development of listing duration time is important even at the time when parts of the revenue are recognized in full upon invoicing. Listings sold as pay now and pay when listing is removed are recognized over the advertising period. EBITDA for the quarter amounted to SEK 172 million, corresponding to a margin of 46.4%. The margin contraction is primarily explained by the lower net sales as we maintain a large portion of fixed costs that cannot be fully adjusted in the short term to offset the drop in listings. One important component in the EBITDA margin is compensation to real estate agreements. When expressed as a percentage of property seller revenue, this ratio increased year-on-year from 30% to 31.4% in Q2 '26, driven by a further improvement in both recommendation rates and actual conversion. The higher commission reflects a substantially stronger underlying improvement in our value-added products. I will walk you through the specific cost dynamics in more detail on the following slides. The increase in leverage to 1.1x is primarily an effect of our active capital allocation, combined with a drop in paid listings during the period affecting rolling 12 months EBITDA. Notably, we expanded our share buyback program from SEK 450 million to SEK 600 million at the '25 AGM and the current mandate from '26 remains at SEK 600 million. We ended the quarter with a headcount of 184, representing a strategic increase of 19 employees compared to the same period last year. This growth was primarily driven by reinforcements within product and tech as well as new resources within the sales team to enhance engagement with the agent community. Additionally, we have strengthened our marketing capabilities. With that overview, let's turn to Page 12 to our revenues by segment to take a closer look at the Q2 figures. Our largest segment, Property sellers, which we have previously covered, generated revenue of SEK 318 million. Revenue from real estate agents decreased by 8%. While this was impacted by the weak market volumes, it was partially offset by the continued growth in our sold by product. Revenue from property developers increased by 1%. This reflects lower display revenue on reduced listings volumes, offset by our new annual subscription package launched in January '26. Revenue from other advertisers increased by 7% to SEK 17 million from continued growth from bank integration and also demonstrating our ability to improve performance with price. The B2B segment is performing well despite the fact that the lower volume of listings reduces impressions, which negatively impacted the stay sales across the B2B business, continued optimization and focus on Hemnet unique products are making a significant difference, keeping B2B revenue broadly flat, down 1.6% year-on-year. Turning to Page 13 and our EBITDA bridge. We can clearly see the dynamics at play this quarter. We start with EBITDA of SEK 261 million for the second quarter of last year. The primary impact and by far the largest, of course, comes from net sales, which had a negative effect of SEK 112 million. Once again, a majority of that negative revenue is from SFPL, where revenue is recognized only when the property is sold. Compensation to agents decreased in line with the decline in revenue from property sellers, resulting in a positive impact. Other external expenses increased by SEK 6 million, largely due to overall marketing spend. This was aimed at capturing earlier traffic and listings leading up to and alongside the launch of SFPL. Personnel costs increased by SEK 3.2 million, driven by headcount expansion and salary inflation. However, cost growth is not pacing directly with the headcount as our average FTE count remains below our total headcount. Finally, other items had a positive impact of SEK 3 million, driven by higher year-over-year capitalized development costs for our own staff, underscoring our intensified focus on product development. In total, this results in an EBITDA for the quarter of SEK 173 million. Finally, let's move to Page 14 for an update on our cash flow and financial position. Our rolling 12-month free cash flow amounted to SEK 621 million. While our lower EBITDA is reflected in this figure, we maintain a highly robust cash conversion rate. This cash generation, combined with a strong balance sheet, enabled us to consistently return capital to our shareholders. As illustrated, we repurchased shares for SEK 147 million during the second quarter, representing approximately 1.5 million shares. At the current valuation, we view the share buyback program as an exceptionally attractive capital allocation tool that allow us to deliver significant value to our shareholders alongside our dividend. Turning to the chart on the right, you can see our net debt and leverage ratio. Net debt stood at SEK 689 million, representing a leverage ratio of 1.1x. While this represents an increase compared to previous quarters, we remain comfortably below our long-term financial target of under 2x, ensuring we retain substantial financial flexibility. In line with maintaining this flexibility, we've secured a SEK 50 million extension of our revolving credit facility, bringing our total credit facility to SEK 900 million. This extension was driven by a one-off effect related to the implementation of SFPL, coupled with our ongoing commitment to our attractive share buyback program. With that, I will hand the call back to Jonas to summarize the quarter.

Thank you, Anders, and thanks for this part of the presentation. And given that today's presentation represents the last update to the market from your perspective, I wanted to take the opportunity, both from a personal but also from a Hemnet perspective to thank you for a great collaboration and strong contributions over the last 3.5 years. You've been a great colleague, a good friend and a true professional. Best of luck in your future endeavors, and we'll stay in touch. With that, let's move on to the summary page on Page #16. To sum up the second quarter, we do see an improvement listing performance in a continued hesitant market, and we do see a continued strong ARPU development. So the main highlights. The decline in net revenues and EBITDA reflects an accepted transitional phase following the nationwide rollout of Sell First, Pay Later. Paid ARPU growth of 12% in Q2, driven by continued high demand for Hemnet value-added services. Sign of gradual improvement in the Swedish property market in Q2, transactions are increasing and supply is coming down. Despite launching a number of new strategic initiatives such as Sell First, Pay Later and bringing live features to the strategic partnerships such as the below the radar feature, we are not satisfied with the quarter, and we're working with an accelerated pace to announce and to deliver the next wave of strategic initiatives. That sums up the first part of the presentation, and we now progress into the second part around our strategic priority areas. In this second part of the presentation, we wanted to take the opportunity to provide a deep dive on Hemnet's business. Firstly, I will go through and present Hemnet's current market position, how we -- how the market is evolving and what our main focus areas are going forward. Secondly, our Chief Operating Officer, Lisa Farrar, will provide an update on our key product initiatives and the commercial road map. Thirdly, our CTO, Hanna Lindqvist, will provide a deep dive on how we work with AI across the organization. With that, let's get started with the second leg of this presentation, and let's move on to Slide #19, please. Let me start with the backdrop. The Swedish property market looks different today than it did just a few years ago from quite a few different perspectives. Consumer behavior has changed. A slower market has driven a new kind of behavior and the way people buy and sell homes today isn't the same as before. Secondly, increased competition and an evolving industry value chain. We're seeing new players entering our space, but mostly, we're seeing new dynamics between agents, portals and sellers. Thirdly, new technology that's shifting user expectations and at the same time, enabling much faster product development on our side. I will walk through each of these in turn because understanding them is essential and crucial to understanding our strategic priorities and how we plan ahead. Next slide, please. Selling and buying a property in Sweden follows a different logic today compared to a few years ago. The 3 data sets that we showcased on this slide illustrate this clearly. Firstly, sales cycles are getting longer. Average listing days on our platform have risen steadily in the past 4 years from 20 days in Q1 2022 to 64 days in Q2 2026. The longer sales cycles is impacting how properties are being listed from a strategic perspective. Secondly, more sellers are selling before they buy. Back in March 2022, 65% of people bought their next home before selling their current one. Today, that has slipped, and we see a reverse pattern, around 64% sell first. That is a meaningful shift in how people approach and move and driving the intent level to sell throughout the extended sales cycle. Thirdly, property prices development has been weak. Both nominal and real Swedish apartment prices have been essentially flat to negative since 2022, with real prices being meaningfully below where they started. This weak price development is creating lock-in effects, especially in the high-volume apartment segment, which has impacted listing volumes negatively. All in all, and as a sum, these trends have driven a shift in market dynamics and consumer behavior, an increasingly prominent premarket characterized by lower seller intent, longer sales processes and a changed way of working among real estate agents. With that, let's move on to the next slide, please. This premarket has had a direct measurable effect on Hemnet. Sold properties published on Hemnet decreased by 5% in 2025, while total market transactions increased slightly year-on-year. That combination lowered our share of sold properties by 7 percentage points in 2025 compared to 2024. The reason is straightforward. As more of the sales cycle happens on the premarket, more properties are being sold before they reach Hemnet. This is a trend that we're clearly not happy with, and we're 100% committed to address this challenge. We will, during today's presentation, come back on this and how we attack the premarket challenge. Let's move on to Slide 22, please. These evolving market dynamics are shifting both consumer expectations and agent behavior. Premarket listings, that is properties being marketed before they are formally for sale are becoming increasingly relevant as listings are being published earlier and transactions are happening earlier in the sales cycle. The longer sales cycles and the sell before buy dynamics have [ emphasized ] this trend. At the same time, we're seeing new market entrants popping up that scrape content from real estate agents own website. These new platforms get access to a lot of supply, but they lack the traffic and reach to actually convert that supply into transactions. As a response to that, sellers and agents want to regain control of their own sales cycle, which is a strength of the Hemnet platform. That's led to more early phase under-the-radar listings being placed behind login, which has added effect on the limiting the ability of scraping sites to republish that content without consent. We've seen this trend clearly materializing during the last month and during the second quarter with the Hemnet S. Faster Mailing collaboration, Bono launching Snekit, which would be Sneakpek in English and Fasiun launching , which would be priority lane in English. Let's move on to the next slide, please. Despite changing market dynamics and a competitive dynamic, Hemnet's core position remains extremely strong. Hemnet has almost 40 million sessions per month, about 2.5x more than our closest competitor. On a per listing basis, we have roughly 4x more sessions per published listing than our nearest competitor had in 2025. So while the competitive landscape is evolving, our audience advantages remain substantial. Let's move on to the next slide, please. Hemnet's core model is strong, but changing market dynamics require adoption and change to cater for the new circumstances. Our overarching strategic ambition is simple to state. Hemnet every step of the journey for every property transaction. That ambition rests on 6 pillars that reinforce each other. Number one, we should have all the relevant listings. Number wo, we're the #1 partner for agents, property developers and banks. We are top of mind with the largest property audience in Sweden. Number four, we know our users and give them a superior experience. Number five, we hold the most comprehensive and valuable data set in the market. Number six, all of this has been built on more than 25 years of relationship with the real estate agent industry. That foundation is strong. But just as I've shown you, changing market dynamics mean that we cannot stand still. We need to actively adapt each of these pillars for where the market is heading to ensure that we continue to win, and we will win. With that, can we please move to the next slide, please? At our Q4 presentation and business update that we conducted in end of January earlier this year, we outlined 4 strategic priorities for the first part of the year. Number one, launching Sell First, Pay Later nationwide to address the emerging free market and lower the barriers to use Hemnet. Number two, building out strategic partnerships with agents and franchises to strengthen collaboration with key players in the industry. Number three, leveraging AI and product innovation to accelerate what we build and get better return on investments from our products and tech teams. Number four, increase our sales and marketing efforts. On an overall and on a summary level, we have executed and implemented all of these. Over the next few slides, I will go through each of these in more detail. Let's now move to Slide 26, please. Sell First, Pay Later launched countrywide in April after a gradual geographical rollout, and it had a strong uptake from day 1. The share of sellers choosing this payment option has been stable at 40% to 45% since the nationwide rollout, an adoption rate we are quite pleased with and which is very much in line with our expectations going into the launch. The new model has been well received among both sellers and agents and has successfully lowered the barriers to list on Hemnet, while generating increasing customer satisfaction. In Q2, seller NPS for Sell First, Pay Later was 7 percentage points higher than for our payment option -- other payment options. Sell First, Pay Later is also contributing to ARPU growth. The BOS conversion, the uptake of our value-added services was more than 5 percentage points higher among sellers using this option compared to other payment methods in the quarter. We're also seeing that the number of listings with Sell First, Pay Later that are sold and the time it takes for those properties to sell are very much in line with our expectations. If we then could move on to the second strategic initiative and our partnership model on the next slide, please. We launched our strategic partnerships in Q1 2026 with onboarded partners receiving increased branding and more exposure on the platform. We've seen a strong interest from across the market. More than 100 strategic partnerships were signed during the first half of 2026, representing more than 30% of the market, and we have signed 9 out of the largest 20 franchises into the partnership agreements. After being live with the partnership for a few months, we're now very happy to see how some of our largest partners are really leaning into the model, proving the value that they see in the Hemnet platform. Svensk Fastighetsformedling went live with under-the-radar or Underhand in June with full countrywide rollout from 1st of July. Lisa will provide more details on this in her section later on in the presentation, but we're very happy with the initial feedback and the data and the number of additional strategic partners are set to join below the radar or Underhand within a short period of time. Please move to the next slide, please. Beyond these 2 strategic initiatives, AI has enabled us to significantly accelerate our product development in 2026. In the first half of the year, we've added more features than ever to the platform, including Hemnet Home Potential, which shows sellers the development potential for their property. Under-the-radar listings, which enables agents and sellers to publish properties earlier while still maintaining control in the cycle. In our Home segment, a way for sellers to initially get an early introduction indications from buyers' interest before committing to a full sales process, conversational search, which I know Hanna will cover more in detail and a reimagination feature that lets users visualize a property restyle. These are just a few examples of the new features we've added to Hemnet in the past couple of months, and Lisa will go through more in detail of some of these later on in the presentation. With that, let's move on in the presentation, and let's head to Slide #29, please. So with AI changing product development and features on the platform, does that mean that AI is also changing the way Sites are searching for their next property? The honest answer, based on data is not in any material way. AI services have become part of everyday life, and the majority of Sites are now using AI services on a weekly basis. But despite that widespread usage, traffic from LLMs to Hemnet remains very low, still significantly below 0.1% of our total sessions. What AI is doing is changing what product features users expect from us. It is not, at this point, meaningful change how users find their way to Hemnet or how they search for properties once they're there. Still a very low part of our traffic comes from the LLMs. Let's move on to next slide, Slide #30, please. Over the last 18 months, we've done a lot. We've done a lot of things. Firstly, we have transformed our culture where more focus has been put on accelerating operational efficiency and output generation across the organization. We have implemented a clear strategic road map where supply and traffic are our top priorities. We have increased our sales efforts and are working more closely with agents than ever before. And we put a stronger focus on marketing and brand, which is already showing visible and tangible results in our brand metrics. This has led to increased productivity. We've launched significant strategic initiatives over the last 6 months, including Sell First, Pay Later and the strategic partnerships, and we're deploying more product updates and features than ever before, enabled by the new culture and fueled by AI adoption. But there is more to come. We are not satisfied with the development, and we're working harder than ever to bring more supply to the platform. We are and we have accelerated our development pace, and we will continue rolling out new improvements to make Hemnet as value creating as a platform possibly can be for sellers, buyers and real estate agents. With that, let's move to Slide #31, please. Despite the changing market environment that I've described, Hemnet maintains a uniquely strong market position. Number one, we have superior traffic. We have superior reach. Number two, we have an industry-leading brand built and nurtured over more than 25 years of market leadership. Number three, we have the scale to win through continued marketing and technology investments. We hold a structured and proprietary data collection that we collected over the last 25 years, which is now enabling a whole new suite of AI products for buyers, sellers and agents. We have deeply integrated industry relationships supported by an active and more frequent sales force. And we have a proven platform that significantly increases the chance of a successful outcome for sellers, meaning getting better bidding premiums, selling your properties in a shorter period of time. With that, let's move to the next slide, please, and let's move on to Slide #32. Looking ahead, we will continue to build from our strong market position. With nearly 40 million monthly sessions and exceptional brand awareness, Hemnet remains the engine for Swedish property search. Our core value proposition relies on gathering all buyers and sellers in one place, providing a complete overview for the market. We are deploying new product features that solidifies Hemnet's role throughout the homeownership journey, creating more value for all our stakeholders and further strengthening Hemnet's network effects. The strategic initiatives that we have launched during the year have helped to capture more listings at an earlier stage in the selling process, but it's also very clear for us that they are not enough. It's very clear for us that we need to do more. Securing all listings from the beginning of the transaction life cycle is our highest priority from a strategic perspective going forward, and we are working with an accelerated pace to announce and deliver the next wave of the strategic initiatives as we approach the cyclical listing peak of the autumn. With those words, I will hand over to our Chief Operating Officer, Lisa Farrar, who will take on a journey and talk a bit about some of the products that we've already taken to the market. With that, over to you, Lisa.

Thank you, Jonas. Let's move to Slide 34, please. Before I take you through our product strategy, latest product launches and how I view our future direction, I wanted to start off with the very foundation of how we build our product strategy, why we do what we do. Hemnet's business is built on network effects and everything we build and are building is designed to make that flywheel turn faster. It starts with the key to the network effects, which is supply. The more listings we have on the platform, the more users we attract and the more often they come back. A larger, more engaged audience makes Hemnet a more attractive place for sellers, agents and partners, which drives value for all of our stakeholders and therefore, more opportunities to drive growth. Supply attracts demand, demand attracts supply and the loop reinforces itself. We believe everything we build should make these network effects stronger for agents, sellers and buyers alike. Let's go to the next slide, please. Our ambition holds firm and our execution at the start of 2026 has been centered on delivering against it. As Jonas covered, we've communicated several key strategic initiatives so far this year. And just as importantly, we've delivered on them. Self First, Pay Later went from announcement to full nationwide rollout within a quarter. Strategic partnerships went from 0 to 100-plus agreements, including 9 of the 20 largest franchise firms. And under-the-radar, bringing the invisible market on to Hemnet is live with our first strategic partner with others in active discussion to join. Together, these initiatives are how we hit the ground running in 2026. We're shipping more value and better products than ever before. They've laid the foundation on which we continue to build from when forming the next chapter of Hemnet. Let's move to Slide 36. Three main forces are structurally reshaping property search. And together, they're forming a new era for us, as Jonas previously referenced. First, we're operating in a market that demands more from Hemnet. Buyers are demanding to see more properties as possible and more information than ever before. Sales cycles are longer with more emphasis on the premarket stage. And given the sell before you buy shift, sellers only transact if all criteria are met, significantly impacting seller intent. Secondly, our users expect more. Users demand new ways of discovering property beyond the static list. They expect personalized, interactive and extensive property information and decision support. AI enables us to provide that elevated personalized experience, which we showcased in several of our latest product releases. Hanna, our CTO, will dive deeper into our user-centric approach to AI in a moment. Thirdly, we have been present earlier -- we have to be present earlier in the property journey. We're evolving from a platform that serves high-intent sellers at the point of selling into a lifelong customer relationship, one that partners with homeowners through every stage of the property and sales journey. With an efficient platform, rich user data, a superior product and an unmatched brand and reach, we stand prepared to build for a new era from a position of strength. Let's move to Slide 37. A new era demands a new relationship with our users. We are moving from an on-sale listing platform, one that meets users only when they've already decided to buy or sell to a lifelong property ecosystem present at every stage of the homeownership journey, as we summarize it, Hemnet Hllaergen or Hemnet all the way. That ecosystem has 3 stages: buy, own and sell, all fueled by a close partnership with real estate agents and a clear focus on marketing and sales. When we provide value in every part of that ecosystem and build relationships with Swedish property buyers, owners and sellers, we manage to strengthen all aspects that contribute to our network effects while providing more value for real estate agents. In the buy stage, we're building an interactive and personalized experience for people searching for the next home. In the own stage, we're partnering with homeowners to seek information, decision support and guidance through the largest investment of their lives while also taking position where homeowners are taking the very first step towards selling their property through leaning into the off-market, the period before a property is formally listed. In the sales stage, we're securing supply by bringing the invisible market on to Hemnet and continuously lowering the barriers to this. Underpinning all of these stages is a true partnership with real estate agents, where we are uniquely positioned to deliver value to agents through data, through leads, through intent tracking and partnering with them at every step of their journey. All of this is fueled by an increased sales and marketing focus to keep building relationships side-by-side with the industry and making sure we are top of mind for all things property related. Let's move to Slide 38, please. We're actively building and strengthening this ecosystem, continuously launching new features and services for real estate agents and our millions of weekly active users across the buy, own and sell stages. This momentum strengthened our relationships across the board, which in turn reinforces our strongest moat and unmatched position and brand. On the agent relationship side, we've expanded our strategic partnerships with under-the-radar listings that I will talk more about in a minute. A core focus has been also to elevate agency branding and generate more leads to agents. This has been done through opening up more lead generation sources and ensuring lead generation for agents sits at the very core of our products. Let me share an example of the impact of our seller lead features. We know Hemnet is a primary catalyst for an agent's next listing. In fact, half of all sellers who connect with an agent through our platform list their property with that specific same agent. And for our strategic partners, we elevate this advantage even further by ensuring premium visibility across our seller touch points. Proceeding to the buy stage, we're powering our supply with more intelligent discovery for our buyers. During H1, we delivered on this by giving our buyers property X-ray vision with Hemnet Home Potential, which is now live as a beta, whilst also launching and scaling conversational search. Earlier this year, we launched an AI-enabled personalized starting page. We're also live inside ChatGPT, and we have property sold price history live. In the own stage, we're engaging homeowners earlier. Focusing on homeowners through My Home allows us to cultivate deeper relationships with our most frequent visitors across their entire property journey. This is a key to scaling Hemnet ecosystem and driving stronger network effects. And for the sell stage, we're launching new products to secure more supply on Hemnet. By launching under-the-radar listings, we are now bringing a previously invisible market to Hemnet. And together with our launch of First Sell, Pay Later, we're continuously lowering barriers to this. Each of these products reinforces our most important and long-standing moat, an unrivaled position and brand at the very center of the Swedish property ecosystem. Let's take a closer look at a few of these releases, starting with Hemnet's Home Potential on Slide 39. For the buying stage, we're completing Swedish property market with a new product called Home Potential. Traditionally, finding out if you can invest in a property through rebuilding and expanding and knowing if it's a good investment requires weeks, if not months, of work and bureaucracy. This product using AI to ingest 240 fragmented data sources, including 30 analog sources, gives buyers those answers instantly, Digitizing 15 years of building permits, development plans, risk maps and renovation history and connecting all of it directly to a Hemnet listing creates a unique data asset that is difficult to replicate. What that means in practice? Well, buyers get immediate AI-identified answers about building rights, risks and development potential, things like lucrative investment potential or necessary renovations right inside the listing in real time. Sellers get all the relevant documentation assembled upfront and agents will they get a head start going into every intake meeting. We are launching Home Potential together with a partner. It has been live as a pilot since July 1 and scale-up is now ongoing. We see this as a way to strongly increase the value we provide in segments where we have additional room to monetize. Let's move to Slide 40 and the owning pillar. For the owning stage, we're engaging the 3.5 million homeowners in Sweden beyond registering a home through an evolution of My Home, where homeowners get more data, insights and can initiate their selling journey. Our most engaged users of Hemnet are those that have registered their own home on the platform. They generate 75% more sessions than the average user, which makes them a highly prioritized segment for us to leverage and grow. With My Home growth being a priority for us, we have driven a 73% increase in registered My Homes in the last 12 months, and we continue to see good traction. Capitalizing on data that shows that more than 1/3 of Swedish homeowners are open to selling under the right pricing and conditions, our next evolution of My Home is to provide our users the ability to understand the value of the property and potential interest. We see this working in 3 simple steps. First, activate. For My Home, a homeowner indicates that they are interested in selling. They are prompted to contact an agent to receive a valuation. Secondly, signal, a pin appears on the map and buyers can show real interest and demand that the homeowner can see building up in real time. Thirdly, agent handoff. When the homeowner is ready, they contact an agent, and that agent receives a seller lead that is already comes with a tested price and real demonstrated buyer demand attached. With this, we see the potential to increase mobility on the Swedish property market by allowing homeowners to assess real demand before they commit to a sale. Agents gain access to more and earlier and warmer leads and buyers will they get access to unique properties of Hemnet that they wouldn't otherwise see. This is a genuine win-win across the ecosystem. Let's move to Slide 41 and the sell stage. For the selling stage, we're continuing to lower the barriers to list and bringing the invisible market on to Hemnet through our newly launched product under-the-radar or in Swedish Underhand. Since going live on July 1, under-the-radar gives our strategic partners the ability to publish upcoming listings behind log-in in the upcoming result list at no cost. Under-the-radar is designed to target early listings that are in the very beginning of their sales cycle. For partners who have chosen to use Hemnet for a controlled behind log-in launch, the listing is available on Hemnet before reaching the broader market. And for buyers, it means a first look at homes that would otherwise be entirely invisible to the market. We've had a very promising start for under-the-radar listings. In the first half of July, we already saw more than 1,000 published listings. Our strategic partner, Svensk Fastighetsformedling has been our launch partner in this new initiative. Their new way of working means that all of their new listings are automatically published as under-the-radar listings at Hemnet. Together with Sell First, Pay Later, we're lowering the barriers to list to secure more supply on Hemnet and are fully committed to continuously evolving our platform to ensure that Hemnet remains the comprehensive, engaging and commercially powerful marketplace it has always been. We're continuing to have relentless focus on products that drive supply to Hemnet, and we will keep developing products in the short term designed to do just that. Now let's turn to Slide 42, where I want to zoom in on the sales and marketing engine behind everything I've just walked you through. We keep leveraging our presence to keep our flywheel running faster through our side-by-side presence with the real estate industry and continuing efforts to strengthen our brand. Our sales team continues to be a strategic pillar of how we go to market. We're in the next wave of sales transformation, ensuring our sales organization serves as genuine trusted advisers for every real estate agent in Sweden. We've strengthened the team itself with dedicated sales representatives, customer success managers and customer support agents. And we've continued to deepen our proximity to the industry, more boots on the ground, meeting agents face-to-face where they are. That proximity showed up directly in the Sell First, Pay Later rollout. Our sales team contacted more than 7,000 agents, held 700 in-person sales meetings and met with more than 2,000 agents. That's the kind of hands-on execution that turns a real launch into adoption. On the brand side, we ran a nationwide campaign earlier this year under the theme more eyes on your ad, and it moved the numbers that matter. We saw improvements across the board, awareness, top of mind, consideration and preference. And once again, we hit the top score on brand awareness amongst property sellers at 97%. Together, this is what strengthened sales and marketing engine looks like in practice. More agents met, more trust built and a brand that keeps leading. Now let's move to the next slide, please. To close, I want to bring this back to the big picture. This year, we've done more and moved faster than at any point in Hemnet's history. And where it made sense, we've been willing to take bigger and bolder steps, too. Sell First, Pay Later, strategic partnerships under-the-radar, these weren't incremental tweaks. They were structural changes to how sellers, agents and buyers use our platform. And that's really the point. We're moving from a platform that connects a buyer and a seller once to an ecosystem that stays relevant across someone's entire relationship with a property before they sell, while they own and long after they bought. That shift means more engagement with Hemnet at every stage. In practice, it means 3 things: a data advantage that deepens with every transaction, seller, buyer and agent intelligence, all in one place. It moves us closer to the actual transaction, not just the moment of discovery, and it opens up monetization opportunities that go well beyond a single listing. And none of this works without supply. Supply is still the foundation everything else here is built on. It's what makes our flywheel turn, and it's what we're focused on most. We've made real progress this year, but we're not done, and there's more to come on this front very soon. And of course, AI, both in our products and in how we work has an instrumental part to play in all of this, which Hanna will talk you through next. So with that, I will hand you over to our CTO, Hanna Lindqvist.

Thank you, Lisa. So my name is Hanna. And for the past 18 months, I have served as Chief Technology Officer at Hemnet. I've spent 25 years in tech. I lived through the dot-com boom and crash, the mobile revolution, and I worked in TV during the shift to streaming. Most recently, I spent 5 years building and scaling marketplaces and here at Hemnet. And I mentioned this not to list credentials, but to give you a perspective. I've seen several moments where technology genuinely rewrote the roots of an industry, where the question wasn't if things would change, but how fast and who will be ready. And I can tell you with conviction, 2026 is the most exciting moment I've seen in my career. What's happening with AI isn't incremental, it's transformational, and it will likely exceed what streaming did to media. But here is what those past transitions has also taught me. Excitement without discipline is how companies get hurt. The winners weren't the fastest movers. They were the ones who moved with clarity, who knew what they were building and why. And that's the lens I lead Hemnet's technology organization through, and that's what I want to share with you today. Next slide, please. So let me start with how our technology organization actually works. We are a team of around 60 engineers running the technology for the entire business. And that compact size keep us nimble and able to adapt quickly. We don't sit as a separate tech department next to the business. Instead, our engineers, designers, product managers and UX specialists work together in cross-functional teams covering mobile, dev and back end. That means that the people buying -- building the product are the same people that talk to users and our own outcomes. We also have a dedicated data and machine learning teams that have been doing serious machine learning work for nearly 10 years. They haven't just been experimenting. They have actually been building production-grade models and pipelines that run every day for millions of users. That depth of experience is a real underappreciated asset, and it puts us in a strong position for what's coming with AI. So next slide, please. Our tech platform rests on 3 pillars. First of all, we had a modular architecture. We have gradually moved away from monolithic systems to a modular setup using RAQL federation. This means that teams can build and deploy independently and the platform scales without becoming more complex. Secondly, we use a standardized and proven tech stack. We are a thoughtful follower, not a bleeding age chaser. We use well-established, widely adopted technologies, and that lets us move fast while it also makes it easier to hire great engineers. Thirdly, our decisions are data-driven. Every core decision is backed by analytics, experimentation and engineering judgment and not cut fee. And underpinning all of this, we are 100% cloud-based, no legacy infrastructure holding us back, and we have full flexibility to scale whatever we build next. And this is the foundation everything else is built on. Next slide, please. We approach AI with a focus on concrete user benefits rather than for technology's sake. I want to pause on this because it's the principle that guides everything we do when it comes to AI at Hemnet. We don't chase AI because it's fashionable. Every investment we make has to translate into something concrete. Could be a better decision for a homebuyer, could be a faster workflow for our teams or a stronger product. And with that principle in mind, let me show you how our AI capability has actually evolved and where we are today. Next slide. Our engagement with AI in property valuation spans a decade, and it has evolved in 3 distinct waves. So Wave 1, 2016 to 2024, that was all about prediction and valuation. This is when we built our first statistical models to understand the Swedish housing market. And over time, we developed this into what is called an automated valuation model, essentially our own algorithm for estimating what a home is worth. It retrains continuously on more than 1.4 million historical listings and powers our My Home feature with a price estimate, a price range and a confidence level. Wave 2, 2024 to 2025 was about visualization and tagging. We built computer vision pipelines that automatically read and classify listing images and text. This created the data structures that are now the critical infrastructure underneath both our valuation models and our newer AI features. Wave 3, 2025 to today is generative AI, natural language, conversational interfaces and personalization. This is the way everyone is talking about now. But unlike many other companies, we didn't start from 0. We were building on almost 10 years of data infrastructure. Next slide, please. So as I mentioned earlier, our approach to AI centers on concrete user benefits. Those who follow us closely may recognize this slide from 6 months ago when Lisa talked -- walked through how we work with AI. While much has changed since then, our core approach hasn't. Over the past 6 months, we have focused on building a solid AI foundation, one that position us to roll out new products and features that sharpen the user experience and help people make better and more informed decisions. That foundation comes down to 3 things: automated tagging across every listing, predictive personalized recommendations powered by historical and active listing data and AI-enabled workflows that make our own teams more efficient. With this foundation in place, we are now able to build on it, delivering continuous improvements and new features for our users. This includes intent-based search that complements our existing filters, richer and more interactive property information, better qualified leads for agents and deeper personalization throughout the experience, among other things. Next slide, please. And AI is also transforming how we work, significantly increasing operational efficiency and output. Let me give you 3 numbers that show what this means for us. We now ship more than 100 deployments a week. This is up 24% quarter-on-quarter with a very low failure rate. That means that our teams can move fast, take smaller bets and course correct quickly if needed. Nearly 70% of the code we write today is written with the help of AI. This is up 190% since January. And each developer saves an average of 6 hours a week using AI tools. This is a 40% increase since January, and the trend is continuing. These numbers are the result of a deliberate decision to become an AI-first organization, not just in our product, but in how we work day-to-day. We invested in training, built internal communities and made it safe for people to experiment. I do want to be honest, though, when you increase output this much, you create new bottlenecks. The constraint is no longer how fast we can write code, it moves upstream to how fast we can align, decide, review and test. And we are actively working on each of those. So this transformation is still very much underway, and we will continue to reap the benefits over time. Next slide, please. So let me take this opportunity and walk you through a concrete example because it provides a good illustration of how we actually work with AI. So the problem we wanted to solve in this example was that when someone looks for a home, they don't necessarily think in filters and drop-downs. They might think, I want a house near a good school with a fire place somewhere my kids can play outside. Translating that into check boxes can be difficult and sometimes even impossible. The solution for this was conversational search. You type what you want in plain Swedish, the way you say it to a friend and Hemnet does the rest. And here is the important part. Taking the easy approach of pointing a large general purpose AI model at your data wouldn't work particularly well at our scale. We have millions of searches and running a large frontier model on every single search would be both expensive and slow. So instead, we made a deliberate engineering choice to run a small and fast general model on our own infrastructure and put our efforts into closing the capability gaps ourselves. So we did this in 3 ways. First, we don't let the model guess at geography. We built our own database of almost 8,000 Swedish places, adapted to how places are categorized by the Swedish people. For example, if someone said SoHo, the model should know what a person normally defines as SoHo. And this is resolved through data and coordinates rather than through AI guesswork. Secondly, we tag every listing in advance offline and not in the live search. An AI model reads each listing text and images and assigns structured tags like features, proximity, conditions and so on. This means that we get speedy answers and that nothing is invented. If we have unclear tags, they are simply dropped. Third, we constrained what the model is allowed to say. It cannot invent the price and address or account. Every fact in the answer comes from our own search index and not from the model's imagination. This resulted in a system that is exact where exactness matters and intelligent where intelligence genuinely helps. So where are we today? We are live for users on the web behind log-in and the product compares favorably with early releases we've seen from international peers. Getting a small model to respond reliably in a structured format was our hardest engineering problem. We moved from about 70% reliability to above 95%, and that hard work is largely behind us now. So why does this matter beyond this one feature? Yes, because the infrastructure underneath it, the geo database, the tagging pipeline and the intent detection is a platform. The same tags that power a conversational query can now also be used to improve search ranking, sales alerts, recommendations and area pages and much more across the whole product. This technology is built to compound. And one last point on philosophy. AI attendance is there to support people in making informed decision not to make those decisions for them. Buying or selling a home is one of the biggest moments in a person's life, and our role is to make that process more transparent and trustworthy and not to replace human judgment. Let's wrap it up on the next slide. So let me close with where we're taking this. Deeper personalization. We hold the richest data set in the Swedish housing market, historical, behavioral, geographical and transactional. And that data lets us build an experience that feels built for you and not for everyone, and that is a competitive advantage that only gets stronger with scale and usage over time. Traditional search along intent-based discovery. As Jonas mentioned, filters and traditional search aren't going away. They are core to how people find homes and will likely continue to be for a long time still. But expectations are evolving alongside it, and we are building the infrastructure to understand intent and layer richer discovery on top of search rather than replacing it. AI as an operating model. The productivity gains we're seeing are still in the early days. As AI become embedded in how every team works, we expect those efficiency gains to multiply. And the principle I keep coming back to is simple. let AI do what AI is genuinely good at and keep every exact factual decision grounded in data we control. I believe that this is the right way to build AI into a product that people trust with the biggest financial decision of the lives. So I started by saying that the winners in major technology transitions aren't necessarily the fastest movers. They are the ones who move with clarity. At Hemnet, I want to say that we have both. We have built a small and agile team of engineers shipping over 100 deployments per week and accelerating. But we also have a platform built on nearly a decade of deliberate architecture decisions and an AI foundation that is difficult to match. And with that, I want to say thank you and hand over the floor to Jonas again.

Many thanks, Hanna, and also a big thank you to Linda for these presentations. With that, let's move on to the wrap-up of today's presentation and a quick overview of the main highlights. So if we could please move on to Slide 24. Financial performance in Q2 2026 was characterized by the Sell First, Pay Later rollout negatively impacting net sales and profitability. Q2 absorbs the peak financial impact of this timing shift before the platform will benefit from a maturing pipeline of completions. Number two, market recovery has been slower than anticipated, but there are clear signs of gradual improvement in Q2 and promising signals for the coming year. Transactions are increasing, pushing down supply from historical high levels. With prices pointing up, we're in a better position to see more supply coming to the market for the second half of the year and next year. We continue to build from a position of strength. With nearly 40 million monthly sessions and exceptional brand awareness, Hemnet has a unique opportunity in an evolving market environment. Number four, our ambition is for people to use Hemnet on every step of the journey for every property transaction. Therefore, securing all listings from the beginning of the transaction life cycle is our highest priority going forward. We have announced a number of things today, but that's not going to be enough. We're working with an accelerated pace to announce and to deliver the next wave of strategic initiatives as we approach the cyclical listing peak of the autumn. With that, that concludes today's presentation, and we will open up for the Q&A.

[Operator Instructions] The next question comes from Will Packer from BNPP.

Three from me, please. Firstly, do you mind spending a bit of time talking about how you will monetize under-the-radar listings? I think it's a really interesting idea, and thanks for the presentation, but I didn't take away the precise mechanics. Will it eventually -- will they pay a typical fee if those under-the-radar listings turn on to sale? How do you stop sort of adverse selection where people take them off, et cetera? Secondly, you provided lots of interesting updates on how you plan to close the inventory gap with your peers and some of those actions have been in place for a little while now. If I do a very basic piece of analysis, which is compare the absolute number of listings on Hemnet versus Buy, the gap has actually increased in the last 12 months. So buy gone from 1.8x to 1.9x. Is that a methodological floor? Is that not a representative underlying? Or will these actions take time? And then final question for me. Consensus expects double-digit ARPU growth for 2027 and 2028. You look well set to deliver that for this year. Is that a reasonable assumption in the context of today's update?

So I'll take the first 2 questions and then hand over to Anders. So when it comes to below the radar and those listings, it's important to take one step back and think about sort of why we're doing this. And the key priority is obviously to get more listings at an earlier stage and also talking about the relative ratio that you mentioned on your second question is an important dimension. The way this model work is that this is a period of time in the very early listing stage, meaning the absolute sort of beginning. And the plan that we do have, these under-the-radar listings are not monetized from a Hemnet perspective, but they will, as they move on to the premarket or to the on sale segment, that's where we basically will monetize them in the same way. And this is sort of traditional upsell where we get a clear relationship with delisting and something that we will work on. Obviously, this is a new model, but I think we have a number of different tools to ensure that we steer this and not cannibalize on the on sales segment, which is most importantly from a monetization perspective. So early stage, but we will ensure to have business rule to ensure this becomes also very attractive from a monetization perspective. Number two, in terms of the supply, we're working on this on a number of different actions and a number of different areas. And to your point, we have launched a quite large set with both the Sell First, Pay Later. And then most recently, we have the strategic partnerships with the below the radar that obviously went live sort of in full-blown mode in the early Q3. So I don't think we've seen the full upside potential coming from this, but we also need to ensure that we continue to do more. Working on supply is absolutely a top priority, and we will have more actions to ensure that we continue to work on that dimension. Important to say as well is obviously, when you look at that and you look at the totality on the number and the 1.8 to 1.9, please keep in mind that the relationship and the development of the free market has been slightly stronger during this period of time where Hemnet has a lower listing market share. So that trend shift is also impacting it. But going back to our top priority, this is something that we are addressing across a number of different levels. Then in terms of the ARPU growth for 2027, Anders, I'll just hand over to you.

Thank you for that. So we would say that we are well equipped for double-digit growth going forward. Some of the parameters for the time being are against us, such as the very slow housing market in Sweden when it comes to new published listings for the time being. That will get better over time. So -- and we also know that we have the products and services in place to be able to deliver better when the listings come back, such as the B2B division that will also, of course, benefit from a better -- we will be better positioned with the growth in published listings on the platform. So yes.

Just to quickly come back on the first question. So a listing is under-the-radar listing. And if it sells in that initial period because the market goes back to being quite hot, Hemnet would not monetize. But in the event that it transitions to be a typical listing, you'd monetize it in the usual way. Is that a fair summary?

That is a fair summary. And in a case where we would see that -- where we see that too many listings would get sold in a very, very early phase. That's obviously something that we would steer to ensure that you don't get a too high share being sold under the -- or below the radar. So I think we have various tools, and we just went live. And based on the early indications that we have, we don't have any sort of concerns around that. But if that would materialize, we have various tools to ensure that we can convert them.

The next question comes from Georg Attling from Pareto Securities.

I also have one on the under-the-radar. So I mean, when I think of this and the impact on paid listings, it's penetration of under-the-radar, then conversion of those listings to paid listings and then the volume uplift, call it, the market share regain that we might get from it. So I'm just thinking -- or just wondering how you think of these different factors all impacting what is going to lead to in terms of paid listings...

So just one thing to clarify and also going back a bit to Will's question before. I think important to say that these very early-stage listings that we would have from below the radar are listings that typically don't come to Hemnet today. So this means an incremental volume uplift coming from this. And we think from a strategic perspective, that it is very important, and it's a clear upside for Hemnet from a -- both from a short but also from a long-term perspective to get these listings. So it has a big strategic importance. When it comes to working with partners, the sort of the concepts that the various partners have that you see out in the market, they differ a bit. So if you would compare how Svensk Fastighetsformedling are running their sales processes, that would be different to Pastiskhlson. So right now, we've launched a big partnership together with Svensk Fastighetsformedling. When we roll out the other below the radar or Underhand partnership, there might be certain tweaks. But I think it's very important for us that we continue to ensure that we monetize. It goes without saying. And I think this opens up a great opportunity to increase our market share among those listings also in the paid segment, given the fact that we do get a relationship, we do get a relationship with the agent. We do get a relationship with the seller, and we could showcase and convert these into paid listings given the strength of our paid products.

Okay. And just a follow-up on that. If a listing is under-the-radar, then removed and then sold outside of Hemnet, is there an obligation from the seller to pay you? And what would they pay you then essentially?

No, there is no such an obligation. But we definitely have the ambition to convert 100% of those into paid listings.

Okay. Second question because if I understand it correctly, you have to have the strategic partnership to be able to offer this feature or product how have the discussions evolved with additional strategic partners since you launched this under-the-radar with?

I think to answer the first part of the question, Georg, the -- you're absolutely right. This is a part of the strategic partnership slate and the strategic partnership proposition that we do have. I think there's a lot of interest from many of our other partners, and it's also a gate opener for additional discussions. I think there is a clear trend in the market that having a controlled start-up of the sales process is becoming increasingly important. And as you could imagine, this requires some technical adoption and integration. And so the tech team is -- and Hanna's team and Lisa's team are working very intensively to also tag on new partners.

Okay. Just a final question. You said sell-through on SFPL is in line with expectations. If you could just quantify that. Is that the same sort of sell-through as you see on regular listings on pay when sold? Or does it differ a lot to the sell-through of normal listings?

I think on an overall level, the sort of -- it's a very fair comparison comparing Sell First, Pay Later and Pay Later if removed. You see similar levels. Then you have when it comes to the Pay now, that is -- we expected to see some differences there given the fact that there is a different sales dynamic in those, but it's very much in line with our expectations and actually slightly better.

The next question comes from Yulia Kazakovtseva from UBS.

Firstly, Anders, I wanted to say thank you and I wish you all the very best in your future. With regards to my questions, I have 2, that's okay. So my first question is about Max package penetration. So have you seen any consistent and sustainable improvements in penetration rates in 1H and in 2Q specifically? It would be helpful if you could provide some numbers and update on where the penetration stands today? And my second question would be about the 2-day listing rule for Sell First, Pay Later. So you recently announced that you will not reintroduce the rule following the investigation by the Swedish Competition Authority. Could you please elaborate a little bit on this? How you view the medium- to long-term impact of this decision compared to your initial plans? Do you think this may somewhat limit the Sell First, Pay Later this generation in the future?

Thank you, Yulia. And I see Anders just across the table and he's very grateful for your words. To go into the specific questions, so number one, in terms of the Max penetration, we don't comment or disclose specific penetration numbers on our individual packages. But what I can say is that -- we've seen slight and gradual improvement throughout the year, both in Q1 and Q2, but we're still at low levels. We've done a few product changes over the last 2 quarters that we have seen is having a positive impact on the penetration levels, clearly sort of clearly interlinked to the new feature coming in. We've seen a slight uplift. I think also the market is now becoming a bit warmer and as we've spoken about the outlook for 2026, the remainder of 2026 and also into 2027, it looks like the underlying property market is more healthy. And I think that would also benefit the Max penetration as we move ahead. Secondly, when it comes to the 2-day rule, I just wanted to clarify there, Yulia, that the main rationale and the main decision was not driven by the competition authority initial investigation. The main rationale was that we looked at this from a holistic perspective, including a few different dimensions. I mean, first of all, one of the rationales for having a 2-day rule when it comes to Sell First, Pay Later was linked to the fact that, as you know, this had quite some effect on our short-term financials, and we wanted to ensure that we have a viable and sustainable model. And given the findings from the first 4 months or the first 5 months since we launched Sell First, Pay Later in early February in Stockholm, we do see that the model is sustainable. Second to that, and maybe that's where you're coming from, Yulia, as part of the SCA investigation, it was pretty clear that there is -- there were concerns from the agent industry around especially the 2-day rule. And this feedback, obviously, we think that this is a fantastic product. We don't only think we know it's a fantastic product, and we want to have the sort of the positive reception from that specific product. So that's also baked in. But then obviously, a discussion with FCA, there's one component, but it was definitely not that one being the needle mover. I think from an overall level, this will actually have a positive impact on Sell First, Pay Later because if we remove the rule where we basically had quite some negative feedback, that negative feedback is obviously being removed. And anecdotally, we've heard from a few of the chains that this is sort of improving the likelihood of them using Sell First, Pay Later. And if we take a step back, what's most important with Sell First, Pay Later, it is to get more listings. And with that from the industry, I think that's definitely helpful. Hopefully, that clarifies, Yulia.

The next question comes from Ed Young from Morgan Stanley.

I've got 2 as well, please. First is on package, a bit of a follow-up on the Max question, I guess. To us, it looks like it's a very stable and low level. And I guess with sell now pay later increasing VAS by 5 points, premium listings already representing the majority of inventory. Just sort of wondering, does that not strengthen the case for you to revisit the product structure, particularly that it wasn't really talked about today? Or is it not the right time from a market context perspective? Or is it not the right time for the business when you've got a focus on a lot of different areas? So that's the first question. And the second is on development of the homeowner services on My Home. Just wondering if you could give us an idea of the proportion of homeowners that are on the platform right now. And you made it clear that this is under development with input from agents. Just wonder if you could talk about the tensions there or what are the areas that are key to get right in that respect?

So first, when it comes to Hemnet Max, I think we've seen improvement, as I also indicated based on Yulia's questions before. But per your point, we're still at a low level. I think going back to one important thing that's been hindering the penetration of Max has also been the underlying market conditions where we do expect to see an improvement going forward. When in terms of the package structure that we do have, I think it is something that we are very much working on as we speak. We currently have 3 packages right with -- or 4 packages with BOSS, with Plus, with Premium and with Max. And the relative feature differences is something that we -- that's a lever to pull. And I think, I mean, the most obvious one is obviously the relative price difference between premium and Max, but also the relative feature difference between premium and Max. And I think one of the key strengths that we do have with Hemnet Premium is obviously the renewal where it gets renewed every 30th day. You could think about exploring alternatives with Hemnet Max being renewed more frequently. And in terms of the increased BOSS penetration coming from Sell First, Pay Later, I think that would also benefit Max as we move ahead. On the second question, Lisa, do you want to take that one and just elaborate a bit?

Sure. Thank you for the question. We don't disclose the absolute numbers for My Home registrations. But as I mentioned on the call, we are very focused on it. We see they are incredibly engaged on the platform. So I think I mentioned 75% more sessions than our average user. And so we've spent a lot of focus these past 12 months just to increase the number of My Homes registered, but also increasing the features that we make available to them. This specific one around putting your home on the market and the My Home is a feature we are developing together with the reagent industry. And our focus is to increase the mobility on the market. So actually seeing this early supply grow. And there, we're very focused on how can we make that mobility into specific leads for agents, so more business for our agents. And that's what we meant by working together with agents to launch this feature.

The next question comes from Andrew Ross from Barclays.

I've got 2, please. First one is back to on the radar. Can you give us a sense as to how many listings you may expect from that product if you fast forward, a year or 18 months to try and understand how differentiated this may be in terms of your presale content versus Booli, first question. And then second one builds on that and also on your comments you announced in the opening remarks about your biggest priority being securing all listings from the beginning of the transaction life cycle. So I guess, philosophically, why not go much harder in terms of unmonetized content and have some kind of free tier or presale inventory to get it all on your platform? And when you think about the balance between inventory and monetization and network effects, can you just give us a glimpse in terms of how you as a management team think about it? And why not go much harder in initiatives to get inventory on your platform even if it means sacrificing ARPU growth into '27?

So on the first one, in terms of below the radar, I think right now, we're going live with Svensk Fastighetsformedling. They represent some 12% to 14% of the overall market volumes. I think as they are really leaning in and have the ambition to start every single sales process with this below the radar, this would have an impact on the overall numbers. I don't want to sort of put any overall number out there given the fact that we're in discussions with so many other partners and how they roll out and when they roll up would obviously impact the numbers. But I think it is a very strong future from a strategic perspective to ensure that you get all early listings for a certain partner in a certain way. In terms of looking at our overall strategic priority of ensuring all listings, that is something that we are actively working on. There's many different levers and many different ways of getting there, and we look forward to sort of revert back on how specifically we plan to do that. But that's definitely something that we are working on. I think that's all we can say on that specific topic right now.

And maybe just to follow up on that. Is it right to think that, that could involve a move into more unmonetized content in the presale market? Or is that not something you're considering as part of the suite of options?

I mean we're looking at all various options. Commenting on specific ones, we don't want to do at this point in time, unfortunately.

The next question comes from Eirik Rafdal from DNB Carnegie.