Vend Marketplaces ASA (VEND) Earnings Call Transcript & Summary

Good morning, everyone, and welcome to our Q3 results presentation. My name is Malin Ebenfelt, and I'm Investor Relations Manager in Vend. As usual, our CEO, Christian; and our CFO, PC, will walk you through the key developments in the quarter. We will then finish with a Q&A session for financial analysts via Teams. [Operator Instructions] Before I hand over to Christian, let me quickly take you through the disclaimer slide. This presentation contains forward-looking statements. So please refer to this disclaimer on this slide for more information. And with this, please go ahead, Christian.

Thank you, Malin, and good morning, everyone. And big congratulations to Jann-Boje, who became a father last night. That was why Malin was stepping in this morning. So our third quarter results show further progress for a simpler , revenue growth, a lower cost base, higher margins and continued capital returns. But there are also areas that need further work, most notably the recovery of Blocket and continued improvement on cost efficiency. Over time, we see clear potential to keep growing revenues and increased margins. Group revenues in Q3 were NOK 1,641 million, up 3% year-on-year or 6% in constant currency. Our 4 verticals grew 11% in constant currency. That was partly then offset by the phaseout of transition service agreement revenues from the Schibsted separation. Group EBITDA increased 11% to NOK 711 million, with the margin expanding around 3 percentage points to 43%. Our focus on cost is delivering ahead of what we indicated earlier in the year, and we now expect the 2026 cost base, OpEx, excluding COGS, to decline by around NOK 250 million year-on-year, which is up from the NOK 150 million we indicated at Q2. And this is while we retain flexibility to invest as we do in Recommerce. Year-to-date, our EBITDA margin is 40%. That is up from the 27% that we had in the same period in 2024, but it is still below leading international marketplaces and improving this remains a key priority for us. Our capital allocation remains focused on long-term value creation. We have returned excess capital consistently since 2024. And the first NOK 2 billion tranche of our current NOK 4 billion buyback program was completed in August and the second tranche had reached around NOK 1.4 billion as of October 23. And as we announced last night, we are now taking steps to enable continued share buybacks beyond the current program. And our 14% stake in Adevinta continues to develop well operationally. PC will come back to this a little bit later. And its valuation increased by NOK 3.9 billion in the quarter to NOK 11.1 billion. And our view on AI as a meaningful opportunity for marketplaces has not changed. If anything, we are more convinced, and I will come back to this and our AI agenda and product updates a bit more in detail in a few pages. But first, I want to give an update on Blocket, where we continue to see positive momentum following the platform transition. User satisfaction reached 3.5 in September. That is up from 3.3 that we had in June and well above the post-transition low of 1.5. It's now approaching the pre-transition levels. And importantly, Blocket continued to benefit from significant scale, strong market position. We attract on average 70 million visits and more than 8 million visitors every month. And we have a stable and strong brand position with 97% brand awareness in Sweden. In Mobility, Blocket continues to hold a strong competitive position. We have a significant advantage in both traffic to car listings, but also in available inventory. At the same time, private car listings continue to recover and the year-on-year decline is narrowing from 26% in Q1 to 8% in Q3 and further to 5% in September. And this is, in fact, in line with the market, the C2C ownership changes declined by 5% in that month. Professional car listings increased by 1%, and this is actually outperforming the B2C market, which also declined by 5%. And if we look at the demand side, both visits, leads per visit and the share of ads generating leads within 14 days are all up year-on-year. And at the end of August, we launched Tryggad Bilaffär. This is our transactional C2C solution, which is based on Smidig Bilhandel from FINN. And I can say that our priorities remain strengthening the private car supply, driving traffic and growth in the subvertical categories and increasing app traffic. We are still not where we want to be on these things. In Recommerce, transacted GMV grew 17% and private listings are approaching last year's levels. And in September, we launched our seller benefit program, which aims to increase volumes further. And all of these developments increase our confidence in the progress towards bringing Blocket back to growth. And now over to AI. And of course, there's been a lot of debate whether AI agents will bypass marketplaces. But when you look at how people across the Nordics actually use our platforms, I am confident that they won't. In fact, I think that AI makes our direct user relationships, our proprietary data, our transactional services and our strong brand trust even more valuable. First, our core positions remain strong and robust. In a study we did in August '26 with nearly 4,000 respondents across the Nordics, 70% had used an AI tool in the last month. 25% of these AI users had used it for something related to cars or real estate. Yet, our user behavior remains unchanged. In fact, FINN grew in number of visits over the past year. 95% of our traffic is still owned, more than half of it comes through apps and less than 0.5% of our traffic comes through external AI tools. Our brands lead the Nordics when it comes to awareness, and also consistently ranks highest when it comes to trust, safety and also being fun to use. So you can say that people come to us to search, to decide, but quite many also come to us to browse for inspiration, dream about the cabin, a car, look for these interesting secondhand points. And this is an engaging habit-driven behavior that I think is very unlikely that AI will replace. Secondly, I believe that we will lead the Agentic future with our unique data. The key here is that AI agents will only be as good as the data they rely on. They can, of course, pull open listings from multiple sources. That is not different from what Google has been able to do for years, but that doesn't mean that it will be the best experience. To get the best outcome for consumers, these agents need data that cannot be scrapped. What is the right price for this exact car, which models are holding up in value and which sellers replied within the hour. This is data that we have, that we have earned transaction after transaction over many years, and it compounds every day through these 4 layers of data that we are showing here, real-time ads, aggregated outcomes, verified users and behavior and transaction history. We are now building on this advantage, scaling native vertical agents across our platforms. In addition to that, we have set up what we call Team Aura. This is a small independent unit. And in our portfolio of bets, they are tasked with thinking more radically, answering one question. If we are building a marketplace from scratch today with AI at the core, what would it look like? And third, I would say we are very excited about the opportunities we see that our markets are getting bigger because we can serve needs beyond search. In fact, search is only a small part of the deal, but AI lets us do so much more. Think about all the things that a consumer needs as part of a deal, valuation, comparison, financing, negotiation and many more things. And it's the same thing for the professionals. And I think the logic here is quite simple. The more of the transaction that we coordinate, the more we can create trust for the 2 parties. And the more unique data we can capture and in the end, also the more value that we can create. And we are fortunate to have significant transactional services already such as Fiks færdig, Smidig Bilhandel, Nettbil, Qasa and others, and we will continue to build services that help users beyond the search, so to speak. So in essence, you can say that our brand trust, our domain -- deep domain insight, our real-time transaction data, that is what gives us the best foundation in the Nordics to build native AI agents. We are certainly not sitting back waiting to see how AI impacts our marketplaces. We are using actively all the things we have built over many years to make sure that Vend stays the best place also in the future to find your next car, your next home, your next jobs or your next pair of sneakers. And that is no matter where you come from, who does the search or what start the search. So let me now take you through a few of our recent product and AI-powered features. In total, we now have around 30 AI-driven products and features, and these are some of them. In Mobility, AI is now working for both car dealers and carmakers. Dealer Hub gives dealers AI advice on price, on ad content and on visibility and usage of this tool doubled from August to September. With new car, which is now live on Blocket, it's actually AI that builds the carmakers' brand stores over a few hours, and that extends our marketplaces into the new car sales. In Real Estate, we have our AI tool we call Bytt Stil, which lets users redecorate a home virtually and to see its potential. And it's interesting to see that the users who try it are 6x more likely to sign up for a viewing. We've also launched image-led discovery on Oikotie, Meglerpromo on FINN. This latter feature is a new agent promotion feature, which is part of the large package. In Jobs, we have Agentic Apply. This is a feature that simplifies how candidates apply to jobs. Here, we use an AI agent to fetch the recruiter's application form and it fills it in with data from the candidate's profile on FINN. And this is actually solving a significant pain point for many users. We're also piloting 2 AI-powered offerings for recruiters, FINN Partner for our largest customers and Talent Search, which is a completely new business model, which I will come back to in just a moment. And finally, we have Recommerce. And here, we are building the foundation for agentic item creation. AI already helps sellers through the whole listing flow with everything from photos to titles to description to price. And what we see is that these AI-generated ads, they actually have 42% higher liquidity in the market. And as I mentioned before, we have also launched a seller benefit program, which is a key source of the volume growth that you will see in Recommerce. So all in all, we're excited about these things. These products and features make our marketplaces more valuable. And we're also monetizing these things through a number of measures, packages, upsell transactions. But Talent Search actually takes it one step further with a new outcome-based business model. And let me just go a little bit more into this product. Now as you know, historically, our recruitment business has been built around visibility. But of course, recruiters and employers, they are not looking to buy clicks or applications. They want to hire the right people. And this is where we see that AI and agentic technology actually opens up an entirely new space of opportunities for us. So with Talent Search from FINN, instead of publishing an ad, and waiting for applications, we have an AI-powered recruitment agent that helps employers find relevant candidates. It handpicks them, invite them directly through FINN. And our vision is to be able to serve interview-ready candidates in as little as 3 days. And this builds on all the strengths and advantages that we have, a proprietary candidate database, unmatched reach in Norway and, of course, a trusted brand. So I think we are still, of course, at a very early stage with this product, but the first pilots are showing some promising signals. And I think this product also signals something bigger for Vend as a whole, namely that we are focusing more on outcome-based revenue streams beyond the traditional ads. So with that, let me now turn to the verticals. And as usual, we will start with Mobility. And this time, before we go into the numbers, I would highlight some of the progress that we are making on the product offering and go-to-market. One year ago, we announced the introduction of car dealer packages in Sweden that followed a successful launch in Norway the year before. And today, around 50% of dealer adds are on our higher tier packages. And throughout the year, we have continued to introduce new features, strengthening our value proposition and improving the dealer performance. The results are reflected in the measures and the metrics you can see here. That includes higher traffic to ads, improved conversion rates and most importantly, fewer standing days. We've also completed the rollout and migration of all customers to Dealer Hub. And as I mentioned earlier, we are seeing encouraging performance and positive customer feedback on this product. Dealer Hub really strengthened our value proposition through AI-driven recommendations and ad optimization. They also add new lead management and integrated insights based on the unique insights that only we have. And we will begin the rollout in Sweden in November, which then will benefit all the Swedish dealers with these capabilities. And to align pricing with the value that we deliver, we have recently announced price increases for our car dealers. This will be effective from February 1, 2027, 10% for Blocket in Sweden, 18% for FINN in Norway, 8% for Bilbasen in Denmark. And these increases are then expected to support our financial growth in 2027 and beyond. With that, let me turn then to the financial performance. And we start with Norway, where ARPA growth remained strong, up 17% for professionals, 10% for private. This was driven by the annual price increases and also for professionals, the basic package adjustment that we did in June. And this comes alongside improved packaging and upselling. Car volumes grew in both segments, while private subvertical volumes declined. In Sweden, professional ARPA for cars grew 15% following the price increase from May 1. This was partly then offset by lower voluntary spend from dealers. Including subverticals, professional ARPA was up 12%. That is back then to double-digit growth. Private ARPA was flat, and I covered the volume development in my section of Blocket. If we then turn to Denmark, professional ARPA continued to improve. This was supported by the package split and renewals. Private ARPA declined. This was as expected because that reflects the reversal to a free-to-list model for cars below NOK 50,000 on DBA that we introduced last December. And the professional segment in Denmark continues to face headwinds from market development and also some structural shifts from the changes we have introduced. Private volumes increased 39% year-on-year. This was driven then by the same free-to-list model on DBA, but also 16% growth on Bilbasen. Turning then to the Mobility financials. Revenues increased 8% in constant currency. Classified revenues grew 8% in currency as well. That was driven by all markets. Transactional revenues also increased 8%. Advertising revenues declined 2%, and that was mainly driven by Sweden. OpEx, excluding COGS, decreased 2%. That follows the organizational changes we made at the end of the second quarter, while we have continued to invest in marketing and also in our C2B models. And overall, EBITDA increased 7% year-on-year to NOK 406 million with the margin expanding 2 percentage points to 59%. Moving to Real Estate. And here, we've also focused on strengthening our package offering. This autumn, we launched several product enhancements. They are all included in the large package. Pristips, which is our price estimation tool is now easier to find. Monthly price estimates have nearly doubled to 16,000 per month. Meglerpromo now generates 25% of agent profile leads. And together, these 2 send agents a lot more sellers who are ready to act. The large package continues to outperform the medium package with 31% more clicks to bid, 24% more unique visitors in the first days after a listing goes live. And -- we are also now changing how we prove our value to customers, moving from traffic and clicks to real buyer intent. And by tracking qualified leads like bid clicks and viewing sign-ups, we give agents and sellers clear proof of FINN's unmatched value in Norway. And these are some of the key messages that we took out to customers when we announced our new packages and prices for 2027 this autumn. I think agents have received them very well, and we expect these changes to improve ARPA in line with our medium-term targets. In Norway then, total ARPA grew 16% year-on-year with residential for sale as the primary driver at 17%. On volumes, residential for sale declined 4%. This is against a very strong quarter last year, and total new approved ads were down 6%. And in Finland, average revenue per office and the number of offices were stable quarter-on-quarter. That means that Real Estate delivered another strong quarter with revenues up 14% in constant currency. Classifieds revenues grew 12%, driven by the ARPA growth in residential for sale in Norway. Total revenues in Finland increased 13% or 22% in constant currency. This follows the move to the subscription-based model in Q1, but also growth in Qasa. And if we look at our transactional rental business as a whole, Qasa and HomeQ grew their revenues 10%. OpEx, excluding COGS, increased 2%, driven by higher personnel costs. This means that EBITDA reached NOK 206 million, up 23% year-on-year with the margin expanding to 53%. Then moving to Jobs. Here, our work on product and pricing changes for 2027 is not final yet. The plan is to adjust prices in line with the index across our products, but also to continue to simplify the discount model and to refine our segment model. So we continue to deliver strong ARPA growth in jobs. That's up 11% year-on-year in Q3, driven by upsell products, but also this continued refinement of the volume discount model and the index increase. And on volumes, new approved ads were down 3% year-on-year, and this is an improvement from the 6% decline that we had in the first half as a total. And then on the jobs financials, revenues increased 8% year-on-year as ARPA growth of 11% more than offset the 3% decline in volumes. And here, OpEx, excluding COGS, increased 6%, driven by higher personnel and marketing costs. And this led to an EBITDA growth of 7% year-on-year to NOK 146 million and a margin of 55%. And then finally, let's continue with the Recommerce. Here, in Q3, we stepped up marketing, shipping campaigns and seller benefits in Norway to really accelerate our transaction growth. And we see positive results. GMV growth accelerated in Norway to 28% from 17% in Q2. On Blocket GMV grew 17%, which is a clear sign of traction of the transactional model in Sweden after the platform transition. Finland grew 33%, Denmark 52% albeit from a smaller base. Take rates were stable in Sweden at 10%, Finland at 16%, Denmark at 14%. But in Norway, the take rate declined 0.7 percentage points to 15%, which reflects the shipping campaigns we had in the quarter. Recommerce then grew revenues 16% in constant currency. Transactional revenues grew 19%, driven by the record volumes, but it was partly then offset by shipping campaigns. Classifieds revenues grew 5%, driven by the private segment and advertising declined 3%, primarily in Finland. OpEx, excluding COGS, increased 11%, and that was driven by the aforementioned marketing investment in Norway. I just want to say that this is a deliberate choice that we have made. We are investing to strengthen our position to accelerate growth. And this investment has led to an EBITDA decline of NOK 8 million to minus NOK 43 million. But with that, let me hand it over to PC for the financials. Thank you.

Thank you, Christian, and good morning, everyone. Let's move to the financials for Q3. In total, revenues on a constant currency basis increased 6% compared to Q3 last year. Vertical revenues that's a better read on the underlying performance grew 11% in constant currency, driven by double-digit growth in both Real Estate and Recommerce and 8% growth in both Mobility and Jobs. Revenues in other HQ declined 58% as expected, mainly due to the lost revenues from exiting the TSAs with Schibsted Media. Total EBITDA ended at NOK 711 million, up 11% from the same period last year. The EBITDA growth was mainly driven by Real Estate and Mobility. And Christian has covered the vertical performance already, but let me give you some comments on the other HQ segment. Despite the significant revenue drop, other HQ EBITDA improved to minus NOK 8 million in -- from last year to minus NOK 4 million this year. Lower personnel and consultancy costs following the reorganization and optimization efforts after the Schibsted Media separation more than offset the lower revenues. The low loss in the quarter also reflects seasonality as it also did in Q3 last year. The personnel costs are lower during the summer months, but the allocation that we do from other HQ to the verticals are stable throughout the year. Now let us look closer into the cost development in the quarter. As before, this slide shows OpEx, excluding COGS. In total, OpEx, excluding COGS in the quarter declined by 6% compared to last year. Personnel costs declined by 2%, where the lower FTE base were partly offset by salary inflation. Total FTEs ended at 1,565, down 83 from Q3 and down 119 from Q3 last year. And this is mainly driven by the reorganization processes that we executed on during the second quarter. During the last 2 years, we have reduced the total workforce with around 500 FTEs. At the same time, we have invested into key capabilities, but also supporting the growth in our scaling units in Mobility and Real Estate. As Christian has commented on, we have invested into marketing in the quarter. So the marketing cost is up 15% compared to last year, driven by the mentioned increase in Recommerce. Other operating costs declined 18%, following lower consultancy costs and the exit of the Schibsted TSAs. Overall, despite limited revenue growth on a reported level on the group, the cost reduction resulted in another 4 percentage point improvement in OpEx, excluding COGS over revenue. The ratio has improved from 58% in Q3 '24 to 51% Q3 last year and to 47% in Q3 2026. Working towards an OpEx to sales ratio around 40%, as stated in our last Capital Markets Day continues to be a key priority for us. Let me also say a few words on Adevinta. The carrying value of our 14% stake has been revised to NOK 11.1 billion, up NOK 3.9 billion in Q3. The main driver was a multiple expansion of around NOK 4 billion. Around NOK 3 billion of this comes from removing the 15% discount to the peer group and around NOK 1 billion comes from rolling the valuation model into '26 and '27, similar as we also did in Q3 last year. The 15% discount was originally introduced mainly due to 2 reasons. First, Adevinta has a higher share of Mobility revenues that had a structurally lower multiple compared to the peer group that had a higher share of Real Estate revenues. And then second, the Adevinta performance at the time was expected to be sort of grow less than the peer group. But as we have seen over the last 2 years, that the actual performance has been very strong and also the expected performance of Adevinta is in line or actually better than what is expected by the peer group. And therefore, we cannot really justify any discount on this stake. I cannot comment more on forward-looking performance, but the financial for the first half of 2026 shows that Adevinta's performance remains strong. Total revenues grew 11% year-on-year and the 3 key positions that in totality constitutes more than 80% of the revenues grew more than the total 11%. Adjusted EBITDA grew 22% with an EBITDA margin of 50%, up 5 percentage points from last year. Adevinta's net debt as of end of Q2 was around EUR 5.9 billion, slightly down from Q4 2025 due to underlying cash generation. On the income statement, our operating profit for the quarter increased 26% to NOK 555 million compared to NOK 414 million in Q3 last year. The positive development in operating profit reflects improved EBITDA, lower depreciation and amortization costs, but also lower net other expenses. The other expenses reported in the quarter is mainly related to headcount reductions. The NOK 3.9 billion gain related to the increased value of Adevinta that I mentioned before is included here under financial income. In totality then, net profit for the group ended at almost NOK 4.4 billion. Now let's move to cash flow from continuing operations. Cash flow from operating activities ended at NOK 590 million, an increase of around 33% compared to last year. The increase is mainly driven by improved earnings, reduced cash outflows related to separation and transaction costs. Cash flow from investment activities ended at minus NOK 73 million, where the main driver is the capitalization of our own workforce in totality of NOK 81 million, down 25% compared to Q3 last year. And finally, cash flow from financing activities ended at minus NOK 1.5 billion, but this is all driven by the share buybacks around the NOK 1.5 billion number. In accordance with our capital allocation principles, we continue to return excess cash to our shareholders. The first NOK 2 billion tranche of our NOK 4 billion share buyback program was completed in August. The second NOK 2 billion tranche was also started in August. And as of the end of the quarter, we have bought back shares of around NOK 0.8 billion. As of 23rd of October, we have bought back NOK 1.4 billion of our own share and we expect the program to complete during Q4. We ended Q3 with a strong balance sheet and a net cash position around NOK 1 billion despite buying back NOK 1.5 billion of shares during the quarter. As announced last evening, the Board has resolved to call an EGM of 23rd of November to renew the authority to buy back shares and to release 14.4 million treasury shares. And this gives us the flexibility to initiate a new share buyback after the second tranche is completed during the fourth quarter. Wrapping up, I'd like to reiterate that our strategy, our medium-term targets and our capital allocation principles as we introduced in our Capital Markets Day in '24 remain unchanged. For 2026, we expect Real Estate, Jobs and Recommerce to perform in line with the respective medium-term targets. But based on the results for the first 9 months of 2026, the current market conditions, but also evaluating the effects of our mitigation activities, full year Mobility revenue growth is expected to grow in the lower end of our mid- to high single-digit percent range. Our medium-term target for Mobility remain unchanged. We now expect our full year 2026 OpEx, excluding COGS, to decline by around NOK 250 million compared to last year. This is revised from around NOK 150 million that we communicated in Q2 this year. And this reflects an accelerated cost and efficiency measures across the whole group. At our CMD almost 2 years ago, we communicated an ambition to improve OpEx to sales from the current level at that time of 65% to around 40% in the medium term. And then despite the good progress, I have to say, on this metric over the last few years, more work is needed to deliver on this promise, and this remains a key priority for us. And with that, I'll now hand over to Malin to guide us through the Q&A.

Thank you, PC. I see quite a few raised hands already. And first in line is Andrew from Barclays.

I've got a couple, please. First one is on the extra OpEx savings you've announced now up to NOK 250 million this year. PC, can you help us understand how much of that fell in Q3, how much will come in Q4? And then how we think about the annualization of that into 2027? And then when we think about the OpEx ex COGS into 2027, appreciating you're not guiding at this point, but what do you think is a likely direction? Like can that continue to be down year-on-year as we think into 2027? And then my other question is to get a bit more color as to the investment that's going into Recommerce in Norway and to better understand the timing of that? Is it in reaction to something that's changed in the competitive environment? Just trying to understand why now? And the third question builds on the first 2, which is, are you kind of assuming that, that Recommerce investment will continue in Q4, and therefore, we should think about these extra OpEx savings as a cushion to that extra investment in Q4? Or should we think about the extra OpEx savings as being something that's incremental to how consensus was thinking about EBITDA previously?

Thank you. So I'll start on the first one, and then we can -- maybe Christian can start on the second, and I can fill in if needed. I think year-to-date, 9 months in '26, we are a bit, let's say, higher than NOK 250 million year-on-year cost reduction. So that implies that Q4 will be broadly stable year-on-year. Remember that during last year, we winded down both the revenue and the cost significantly from the CSA. Some of that happened in Q4, but you got full effect in Q3 and then full effect in Q4. Looking into '27, I'm not going to give sort of a cost outlook for '27. But I think what you can read into our message -- what the actual development as you have seen over the last 2 years and also now in the first 9 months of this year and also including the whole year, will continue into next year. And also then combined with the midterm ambition to work towards 40% OpEx to sales, that will require us to continue to work on cost and efficiencies also going forward.

Yes. And on the Recommerce side, I think it's important to remember that Recommerce is a scale game where in the end, it will be volume that will be driving success. We have done a lot in the past year or so to improve our unit economics, and we now saw the opportunity to further strengthen our position, further grow volume by investing then in marketing, but also in seller benefit program and campaigns and such. So -- and we've seen great success with that. And I think you should expect that to continue also going forward.

Okay. Next in line is Yulia from UBS.

I have a couple, if that's okay. So on Adevinta, I believe previously, your plan was to disclose like revenues and EBITDA once a year, whilst now we have this 1H update. So what's your plan going forward? Are you planning to give us this update 2 times a year or like that would be helpful to understand. And also in the beginning of the presentation, you mentioned that you continue to see a lower voluntary spend in Swedish Mobility in Q3 in the Professional division. Can you please elaborate on this? And what's driving the dynamics? Why is it still not recovered at this stage?

I can start on Adevinta. Yes, you're right. We disclosed the full year '25 numbers when we reported Q4, I think -- Q1, sorry. So what you see now, we intend to also report on a quarterly basis going forward.

Yes. And on the voluntary spend in Sweden, you are absolutely right. What we're seeing is that the price changes that we implemented in May, they are kind of holding up while the, let's say, the bumps, they are somewhat reduced. That is also something we have seen at the earlier price changes and something that tends to come back over time. This is also what we expect here. But there is also an added element of electric vehicles selling very quickly in Sweden so that you don't need to bump your ad as often because they will sell quickly anyway. So it's kind of both those 2 things. We think it will be a gradual recovery of that over time.

Next in line, we have Olav from Pareto.

I was just wondering if you could give some color on the development of IT expenses year-to-date as it was a big part of your other operating costs in 2025. And I was just wondering if IT costs have come down year-to-date given the platform migrations or if that's a potential lever for 2027? And then I'll ask my second question afterwards.

Yes. On IT expenses, so that goes in other -- in the bucket other costs. And as you have seen, we have a quite sort of big reduction in that. Part of that is coming when we can reduce IT expenses as well when we exit the Schibsted TSAs. And then we have a multiple set of initiatives within the IT domain addressing that cost as well because we carry with us quite a lot of complexity from our own Schibsted base that we are working quite proactively to address. When it comes to the platform transition, that will also help us to address IT expenses going forward. But the biggest effect on that will come more on the resourcing side. So a number of people needed to operate, maintain and develop and a little bit less on the IT expenses. The exception of that is on cloud costs, where we will see a reduction in cloud costs as we migrate into one common IT stack.

All right. And then secondly, I was just wondering if you could give an update on how Swedish dealers are now distributed across the different package tiers and how this has developed into Q3? And then just some commentary on how this is tracking compared to Norway at the same time when launching the similar packages.

So I don't have the exact split between the packages. I'm not sure if you do, Malin. But I think in general, they're tracking in a positive way. I don't believe we are still at the place where we are in Norway. So there's still room to grow and improve, let's say, the package distribution.

And let us come back to the packet distribution.

Yes, we have. And just remember that the market dynamic is a bit different in Norway and Sweden. So I don't think you should expect the sort of exact same distribution in the 2 markets. In Norway, we have a lot of customers on the higher tier, while in Sweden is more around the mid-tier, but we also have a higher degree of monetary spend that Christian commented on in the previous question.

Next up, we have Ole Martin from DNB Carnegie.

I'll start with a follow-up on the professional ARPA in Sweden. It was up 12%. Can you give any comments on how it evolved during the quarter? How was the exit rate in September would be interesting. And also on the price changes you're implementing, have you made any comments with regards to private for next year in Sweden?

So I don't have any comments on the development in the quarter on the pro ARPA. And I think we are still -- on the private side, we are still focusing on, let's say, the Blocket back to growth and the Blocket recovery. And as we have said before, we want to be very confident in that before we kind of start thinking about price increases in the private segment. Volume is more important for the time being.

But if you were to sort of where is the voluntary spend now compared to what is a normal level? Are we at 20%, 50%? Where are we?

I don't think I will go into detail of where we are on the voluntary spend. We are at a lower level than what we have been before. And as I said, it's partly driven by, let's say, the price increases that we introduced, that's a pattern we've seen before, but it is also partly driven by the market situation where EVs are selling very quickly.

But just to comment on it. This was a big topic in Q2 when we saw quite a big hit on ARPA compared to expectations given the price increase in 1st of May. I think what you see now in Q3 is exactly what we expected and also guided at that point in time. So we're very much sort of in line with the recovery momentum that we have outlined before.

And then a second question. There's this press focus around this new potential entrant in Norway supported by Polaris Media, Amedia and Schibsted. Were you aware of this? And what do you know about this initiative considering that one of your largest shareholders looks to back this new entrant?

We were not aware of this before it was announced. It's not -- it doesn't have any connection with us at all. We don't know much about it either. But also, I will say that we are concentrating all our efforts on, let's say, the Agentic future and the opportunities that come with that and less on this specific competitor.

Okay. And on that note, when do you plan to launch your own agents?

Well, we already have agents, right? Many of the services that I went through before are agents working on behalf of both pro customers and private people. So we do have them. If you're talking about having a chatbot on the site, that's another question that's just another, let's say, incarnation of an agent that is also coming, and we have a team working on that.

And let's move to Giles.

I wanted to come back to Recommerce and get a better understanding, please, Christian, of what your long-term plan is to solving the operating leverage problem that you've got in that segment. You mentioned scale earlier, but if you wanted to expand on that. And then the second question was on the 2 senior leadership appointments to the AI ventures unit. I think you call it Christian, the AI Aura unit during the presentation. And it would be interesting to hear what do you think success would look like for that unit over the next 2 years?

All right. It was a little bit difficult to hear you, Giles, but let me try. First one was about the Recommerce and the long-term plan and how we think about scale and leverage. I think what we have seen with the Recommerce in the past year is that it's a strong improvement in profitability, like 10 percentage points improvement in EBITDA. So we are really seeing that scale kicking in. But I also think it is important to build a robust, strong position in the market, and this is what we are now seeing the opportunity to invest in with the seller benefit program with campaigns and with marketing. So that we are doing. But we are balancing these investments with the goal that we have clearly communicated before to reach profitability in the medium term. On AI and this separate independent Team Aura, as I said, this is one best out of many in our AI portfolio. We've set this team up to operate outside of the line organization to have more freedom to innovate radically. And I think we will see over time what comes out of this, but I hope it will be user experiences that are well received by many people in the Nordics. So it's primarily looking at, let's say, that part of the space, the consumer adoption.