Ladies and gentlemen, thank you for standing by, and I would like to welcome you to Allegro Q3 2025 results conference call. At this point, all participants' lines are in listen-only mode. After the presentation, there will be an opportunity to ask questions. Without further ado, I will now pass the line to Tomasz Poźniak, Director of Investor Relations.
Thank you, Rafał, and welcome to all participants of our call. Let me introduce the presenters of today. Mr. Marcin Kuśmierz, the CEO of Allegro Group, who will provide you with the highlights of Allegro performance in Q3 and summarize the key takeaways. Jon Eastick, our CFO, will guide you through the financials for Q3 and update the outlook for the full year 2025. As usual, our results presentation is available for download from our investor webpage at allegro.eu. You may also download the slides from the link available on the webcast screen. As a reminder, today's presentation and discussion contain forward-looking statements. Our actual results will differ materially from the expectations expressed in such statements. Please make sure you review the full disclaimer on slide number two.
Also, please note this presentation and the Q&A session are being recorded and will be available for a replay on our website at allegro.eu. With this, I would like to hand over to our CEO, Marcin. The floor is yours.
Thank you so much for the introduction. I'm Marcin Kuśmierz, CEO of Allegro. I'm very glad to be here with you today to walk through our Q3 results, the key business developments from this period, and what we're currently working on across Allegro Group. Let's begin with the highlights of our financial and operating performance. Group GMV rose by close to 10%, and in Poland, it came very close to 10.5%. What's especially important is that GMV on Allegro in Poland grew more than twice as fast as domestic nominal retail sales. GMV per active buyer also improved strongly, up nearly 6% at the group level and almost 8% in Poland. Strong GMV dynamics, a higher take rate approaching 13%, and rapid growth in advertising contributed to an excellent top-line result. Group revenues were up by more than 12%, and in Poland, they grew by over 20%.
The number of active buyers increased by nearly 3% year-on-year to more than 21 million across the group and to over 15 million in Poland. Adjusted EBITDA at the group level grew by an impressive 24%, driven by the outstanding profitability of our Polish operations and reduced losses in our international business. In Poland, we delivered this thanks to strong performance in advertising, financial services, and improvements in co-finance. CapEx increased by 26% year-on-year to almost PLN 230 million, as we continued investing in our technology platform and expanding our logistics network and solutions. Financial leverage rose slightly following the successful share buyback in Q3 of 2025. Let me briefly recap on our medium-term strategic assumptions and pillars. We remain fully focused on enhancing our core marketplace and constantly strengthening our value proposition.
We are investing in new AI-powered services, developing projects that allow us to expand into new categories and market segments, and improving our presence across Central and Eastern Europe, giving us access to 25 million additional customers. When talking about international business, it is also worth mentioning that we have successfully completed the turnaround of the mall segment. As mentioned at the previous conference, we continue to analyze opportunities in new product and service categories and explore expansion into additional market areas. We are also evaluating the externalization of our financial, advertising, and logistics services, potentially making them available to sellers beyond our marketplace. We expect the first test of these projects to begin in the early months of next year. Our loyalty program, one of the most popular in Europe, keeps expanding. Over the past year, we've added more than 1 million new subscribers.
We've introduced a wide range of partner benefits, both Polish and international, and we recently announced a partnership with Disney, giving our customers 3 months of access to Disney+. We constantly work to ensure the best selection, prices, and convenience. We continue to attract outstanding brands to our marketplace. For example, in Q3 alone, we began partnerships with category leaders in home and garden categories, such as Homla, Hansgrohe, or Gardena. We also launched new tools for brands, including Brand Hub, a unified console that makes it easier for brands to manage their presence and activities on Allegro. For business customers, we've introduced Allegro Business, a dedicated platform for SMEs with expansive product selection and new tools such as campaign optimization and automatic currency conversion. At our last conference, we dedicated significant time to discussing artificial intelligence initiatives at Allegro.
Today, we're showcasing one of these projects, demonstrating our advanced level of AI commercialization. In our mobile app, we've launched Allegro AI Assistant at this stage for selected users. With its advisory capabilities, it helps customers to find the right products through conversational search. Buyers can specify features, price ranges, or even detailed factors such as warranty length. Using Allegro AI Assistant feels very similar to interacting with ChatGPT or Gemini. Customers chat with the platform, and it instantly responds with the best-matched products and alternatives. It's worth mentioning that Allegro is one of the first marketplaces in the world to offer such functionality, and it will be available for all customers early next year. Our advertising business continues its strong momentum. It's gaining popularity among sellers. In Q3, advertising revenues grew by nearly 32%, reaching over 2.1% of whole GMV. Our off-Allegro advertising also performed very well.
Network PPC spending increased by 60% year-on-year and continues to show strong growth potential. Additionally, in October, we hosted Retail Media Powerhouse, an event that brought together over 300 marketing leaders and demonstrated how to effectively leverage Allegro's retail media capabilities. Our financial services segment is also breaking records. Loan origination rose by nearly 30% year-on-year, reaching PLN 3.4 billion. More than 15% of GMV, or PLN 2.5 billion, was financed through Allegro Pay. This segment continues to grow rapidly, and it's strengthening its leadership position quarter after quarter. I'm also delighted to highlight a major development from just a few days ago. Last week, we announced a partnership with PKO BP, Poland's largest bank. Together, we are launching the first services for buyers and sellers. Both companies are key leaders in their segments in Poland and Central and Eastern Europe.
The bank serves over 12 million consumers and nearly 650,000 companies, offering the broadest financial services portfolio in the market. These partnerships create exceptional value for customers and will help Allegro accelerate customer acquisition and GMV growth. Thanks to the partnership with PKO BP, in Q4, we were launching two innovative services that can significantly enhance the attractiveness of our marketplace. The first one is Allegro Klik, a new fully integrated payment method for buyers. It enables one-click payments, up to 3% cashback, and includes a 6-month Smart! subscription for PLN 1. For sellers, we are introducing Allegro Kapital, easy, fast, fully integrated financing. Entrepreneurs or companies can apply online for loans up to PLN 300,000, and starting next year, even higher, to PLN 500,000. Applications are processed within minutes, and payouts are made the same day. Now let's turn to logistics.
Our managed volume continues to grow and has surpassed 36%, up 2.4 percentage points compared to the previous quarter. In September, DPD, the second-largest logistics operator in Poland, joined Allegro Delivery. Since October, it has been fully integrated into our network and steadily increasing its share of Allegro shipments. The Allegro Delivery Network already includes more than 70,000 points, with over 33,000 lockers from Allegro, DHL, DPD, and ORLEN Paczka, and over 37,000 pickup points, including the chain of Żabka. Our own lockers network will exceed 8,000 machines by year-end. In Q4, we also added three new depots and a new sorting facility. We see network efficiency continues to improve and remain focused on the further development. A final word on our fast-growing international business. GMV grew by 56% in Q4, and on the LTM basis, it's nearing PLN 2.5 billion.
The number of active buyers increased by more than 50% year-on-year, reaching 4.2 million. We continue to prioritize local selection. For example, in Czechia, where local assortment grew by over 80% year-on-year. These initiatives are improving buyer satisfaction and boosting Smart! loyalty adoption. We are also very pleased to see growing cross-border sales from Czech, Slovak, and Hungarian sellers into the Polish market. Now it is time to discuss the detailed financial results for Q4. Jon, it is your time.
Thank you very much, Marcin. Good morning, everyone. It's great to be with you again today. I'm going to take you through another really strong set of financial results for the Allegro Group. As usual, I'm going to start with the Polish operations. Let's start with the key drivers of the GMV growth. This is obviously active buyers and spend per active buyer. I think the best way to summarize this is that Q3 was very much steady as she goes, in that the incremental improvement in these KPIs was very similar to the Q2 level, as you can see on the slide. On a year-on-year basis, and looking at active buyers, we're up 300,000 active buyers in Poland, 2% growth at 15,230,000 at the end of Q4. When we look at spend, this is, as usual, the stronger driver of growth.
It's up 7.9% on a year-on-year basis. The average spend per buyer on a 12-month period is PLN 4,263. That means that the GMV growth accelerated sequentially by 0.54 percentage points in Q4 to 10.4% year-on-year. That translates to PLN 16.2 billion of GMV. On a last 12-month basis, that means that the GMV and the size of the business has moved forward to PLN 64.9 billion, which is 10% up year-on-year. Looking at Q3, the marketplace itself is roughly growing at the same rate as it was in Q3, approximately 10%. We had a windfall quarter for our ticket sales subsidiary, eBillet. They were selling tickets for some of the biggest acts in Poland who are running big concerts in the summer of next year. That contributed roughly 0.4% of additional growth for Q3.
When it comes to our high-frequency categories, they continue to grow more or less twice as fast as the average category. They also contribute to the 11.2% growth in items sold, which is obviously ahead of the GMV growth. When it comes to average selling price, on a neutral mix basis, we see 1% of inflation in the average price of an item. Growth is not very strongly impacted by inflationary considerations. When we move on to revenue, as you already heard from Marcin, 20.2% year-on-year growth. This is a significant acceleration sequentially and a really excellent result. That growth translates to a revenue number of PLN 2.747 billion for the quarter. If you look at the growth bridge, you can see that the GMV, combined with the take-rate increase, produced a 14.4% increase in marketplace revenues.
If we look at the other categories of revenue, and in particular at our growth engines, you can see excellent contributions across the board. 32% growth in advertising. The logistics services revenues are up 140% year-on-year. The other category is primarily our fintech business, and that is up 128% on a year-on-year basis. Looking at take-rate itself, this has come in at 0.47 percentage points higher than a year ago at 12.98%. Sequentially, almost identical to Q2. The last time we made material changes, as a reminder, was the annual changes we put through in March of this year. That is why you are seeing such stability in the numbers here. Moving on to adjusted EBITDA, it really has been an exceptional quarter for profitability. Our adjusted EBITDA to GMV margin has come in at 6.38%, 0.37 percentage points above the last year level.
This is actually, I think, a record level of profitability since our IPO back in 2020. In absolute terms, that translates into a year-on-year increase of 17.2%. It is a PLN 1.035 billion adjusted EBITDA for the Polish business for the quarter. Looking at the bridge, you can see that those strong revenues contribute a good deal of this growth. Looking on the cost side, as usual, the most significant drag on the profitability comes from cost of delivery. Cost of delivery was up 28.5% year-on-year. Sequentially, it is a marginal increase to 5.1% of GMV in Q4. It is a fraction of just a couple of basis points higher than we reported in Q2. What is very important is the composition of that 28.5% increase. 17.3% of this is GMV-related or Smart! penetration-related, which are both good things, obviously, in themselves, and nothing to be concerned about.
An additional 8% of this growth is pure accounting. As you know, we're growing out our Allegro-managed volumes. Within the scope of operation of Allegro-managed volumes, we take a principal role. We are responsible for the delivery, which means that the revenue we earn on those deliveries is shown gross in the logistics revenues I just described, not net, as is the case when it is the agency model, which applies, for example, when InPost is doing the deliveries. There is an 8% increase coming from that, but the revenue is offsetting that to a large extent. The last element is 3.5% related to the unit price of a parcel, and that is the average outcome. Without all the efforts to move mix towards Allegro-managed volumes, it would have been 3.5 percentage points higher, so nearer to 7%.
That is mainly due to the price increases that we received at the beginning of the year from our largest delivery partner. Everything is going very much according to plan on cost of delivery. Marketing spend is up 19% year-on-year. We are investing and having to pay higher cost per click. We are also diversifying our channels and getting more and more into social media. Other SG&A, the impact is much down on Q2 when it was PLN 67 million versus PLN 37 million this time around, reflecting strong cost control. The last thing to mention here, again, coming back to the high margin, is that this is an intra-year high. We are used to and are expecting a lower margin in Q4. I will tell you more about what is behind that later in the presentation. It is seasonally always the case that the Q4 margins are lower. Moving on to capital investments.
As you know, we accelerated our CapEx program at the beginning of this year. You see here, really, that that acceleration is starting to level off. The growth is down to 52% on a year-on-year basis for Q3, PLN 220 million of investment. Similar to the previous quarter, the main driver is in the physical CapEx or other CapEx. As in the previous quarters, the main component of that are the investments in logistics. This is coming on the APMs, but also middle-mile depots and work on one sorting center we put into operation in Q4 and another one that is going to come online, a big one next year. Capitalized development cost is up more moderately at 28% growth year-on-year. This is reflecting a tech team, which is about 9% bigger than a year earlier.
Obviously, salary costs are higher, but also they're spending a bigger proportion of their activity on new functionalities and projects that add value to the business rather than maintenance. That's why those costs get capitalized, and that's part of the reason for the higher CapEx. Looking at it on a year-to-date basis, the Polish CapEx is at 20.5% of the year-to-date Polish adjusted EBITDA. That's therefore well within the medium-term guardrail we set for spending up to 25% of Polish EBITDA. That's the comments on Poland. Now let me move on and start going through the international operations. I'll start with going through the progress on the new international marketplaces. As you heard from Marcin earlier in the presentation, we had a very strong quarter in terms of growth on our new marketplaces. It's up 56% year-on-year.
On this slide, you can see some of the key drivers behind that great performance. First of all, the traffic is up 18.4%. The important thing here is that the quality of this traffic is improving from month to month. We're getting more and more traffic that's coming direct, in particular because we're getting a lot of traction with people downloading the app. The conversion rates are also starting to move up as the frequency moves up. We're getting a lot of value out of this traffic compared to the situation a year ago. Active buyers is up a tremendous 50% year-on-year at 4.2 million customers. The spend per active buyer is up more moderately at a 10.1% increase to PLN 556 per customer on average for the last 12-month period.
That means that the GMV has come in at an absolute level of PLN 600 million for the Q3. The important thing here is that this is the number that we really care about. This is our strategically important GMV, the part of the international business that we're looking to grow sequentially and massively over coming years. It is a combination of the 3P GMV that these three marketplaces are doing, plus Mall North acting solely as a merchant and selling on a 1P basis to support the growth on the marketplace. Just to underline, these numbers that I've just shown you are pro forma numbers that focus only on the marketplaces. If you recall, when we met last time, I took you through the segment changes that we made with effect from Q2.
Those changes involved folding the Mall North segment, so the Mall North sales of Czechia, Slovakia, and Hungary, plus the WE|DO logistics business into the Allegro international marketplace segment. We did that because we finished closing down the old legacy store so that, in effect, it was no longer an independent, Mall North no longer an independent business with its own route to market. That, therefore, has some knock-on effects on our as-reported numbers because it means all the legacy revenues of that closed legacy business are in the comparative figures for the earlier quarters. Bear that in mind as we go through the next couple of slides. Now we'll look at the as-reported numbers for the Allegro international segment, so the marketplaces, Mall North, and WE|DO combined. Let's start with the key drivers of the GMV growth and with the active buyers.
Here you can see the 4.2 million active buyers that shop on the marketplace and a residual number of 1.1 million Mall North legacy active buyers combined make 5.3 million active buyers for the segment. The growth rate is 7.8% in total on a year-on-year basis because the Mall North segment, as you can see, is shrinking significantly. It is going to run down to zero, that 1.1 million figure, by the time we lap the closure of the front end, the legacy shop at the end of Q1 next year. It is a similar story with the GMV growth or the spend per active, sorry, for the spend per active buyer, which is at PLN 487 and is down 18% year-on-year and is reflecting the same thing with the run-off of the legacy GMV that is in this last 12-month KPI.
Again, by the end of Q1 next year and in the Q2 numbers, there will be no impact of this legacy revenue anymore. When you put all of that together into the GMV, the as-reported GMV for Q3 is exactly the same as that pro forma number I took you through two minutes ago, PLN 599 million. In the earlier quarters, you have this legacy spend, which means that the growth rate of GMV Q3 to Q3, as reported, is -3.5%. On the right-hand side here, you can see the components of that GMV. At the bottom, you can see 55.9% growth of 3P revenues from the marketplace, another PLN 25 million coming from the 1P Mall North as a merchant sales to create that PLN 599 million GMV for Q3.
If you look to the left and look at the prior quarters, the gray bars show you that legacy revenue that we're no longer generating, sorry, legacy GMV that we're no longer generating and what's generating the -3.5% growth. Again, by the time we get to Q2 next year, there won't be any of this legacy left, and we'll just be looking at the growth. Taking the same topic to revenue, you can see that that -3.5% of GMV growth actually translates to, sorry, of GMV decline actually translates to a 58.5% revenue decline. That's because all of that legacy GMV from the Mall North legacy site was done on a 1P basis, which obviously is massively over-indexed compared to the revenue you earn from take rate on a 3P model.
Let's shift gears and look at adjusted EBITDA, where we're already seeing the growth and improvement that's coming from all the measures that we've taken to transform the Mall North business over the last couple of years that completed at the end of the first half of 2025. Here we have PLN 34 million lower loss for Q3 than we had a year earlier, which is a 23% year-on-year improvement. You can see that the Mall North loss, Mall North and WE|DO loss, has cut to PLN 32 million from PLN 60 million a year earlier. We've been able to invest more into marketing on the Allegro marketplaces but still cut the loss there by PLN 5 million from PLN 88 million to PLN 83 million. Good sequential improvement or year-on-year improvement, I should say, which is then helping us to the 24% group-level adjusted EBITDA improvement for the quarter.
Looking at the bridge, you can see where these savings are coming from. It's the marketing cost and other SG&A that's gone away on the legacy business. We've only lost PLN 11 million of margin despite shutting down all of that legacy activity. That's the Allegro international segment. Just a reminder, Mall North South still operates, and it's now a separate segment on its own. GMV is 2.5% lower, much more moderate declines than we were seeing in the northern region. This business still runs on all the legacy software systems and still essentially is an independent business. Looking at international operations in total, just as a reminder, these numbers are completely unaffected by the reshuffle of the segments. All the segments still total to the same figures. That's it for international. Let me move on to a few comments about the group consolidated results.
As previously, I'm limiting the comments really just to the leverage situation. As you know, it was a very busy quarter in the third quarter when it comes to financial transactions for the organization. In particular, we returned PLN 1.4 billion through the share buyback at the beginning of the third quarter. That had the effect of lifting the leverage from 0.72 times last 12-month adjusted EBITDA back in June to a pro forma number of 1.16, giving effect to that buyback. In the meantime, over the last 3 months, we brought the as-reported number all the way down to 1.05 due to strong cash flow generation and the strong increase in the adjusted EBITDA. We're very much on track to come back to that 1 times adjusted EBITDA level, which is what we're targeting as our long-run level in our capital allocation policy.
It's also worth mentioning that two days ago, we've announced the signing of a very successful refinancing initiative for our senior debt. We will be swapping out the old debt with PLN 5 billion of new senior debt in the next couple of days. It's being taken on better terms than we had in the legacy arrangements. Most importantly, the maturity is extended now from 2027 out to 2030, so by 3 years. That's the financials. Now, let me wrap up my section by going through the management outlook. First of all, just a quick summary of where we were at the end of Q3. We had updated the outlook, as you'll recall, slightly improving on various metrics compared to the original targets we set in March. In certain areas, we actually moved up the midpoints of the guidance.
In some areas, we just narrowed the range. Basically, going into Q4, everything is in line and on track for the full year. Now let's talk about current trading and what that does to the outlook going forward. Current trading and starting with Polish operations. First of all, the GMV. The first month of Q4 looked very much like Q3. We booked like-for-like growth of approximately or just over 10%. However, in the first two weeks or so of November, we've seen a slowdown in the growth to low single digits year-on-year growth. This is reflecting a few things. In particular, we're still waiting for the cold weather to come and bring the typical seasonal sales relating to things like cold weather or winter tires or cold weather clothing. This is still ahead of us, still to come, but obviously, we don't know when.
Secondly, the Black Week has been extended to a full month, but it got off to a fairly slow start, slower than maybe we were hoping for. The growth over the last two weeks has been a little bit lower than we were looking for. When it comes to adjusted EBITDA, very much as expected, we're seeing the EBITDA margin being lower in the fourth quarter than I was just describing for Q3. This comes from the usual factors: lower take rates, reflecting less need to make discretionary spending by merchants because of the high demand that you see in the fourth quarter. It also reflects investments into pricing refunds and sharing costs with the merchants and also additional marketing spending, which we always do in the fourth quarter.
There's also one additional element, which is a headwind that we faced this year from the volume discounts that we received and were booking mainly in the fourth quarter of 2023 a year ago. We have not repeated such an arrangement with our largest delivery partner this year because we needed the incremental volume to help build out the Allegro Delivery and the Allegro-managed volumes. Therefore, that is going to create a bit of a headwind on the average unit cost of a parcel in the fourth quarter. When it comes to international operations, the marketplace is still going very well. First half of the quarter, we are seeing above 50% year-on-year growth. The Mall North 1P sales, so just Mall North as a merchant now, is improving sequentially as they build from the restructuring that we implemented right at the beginning of the second half of the year.
Mall South is chugging along at the similar levels to what we saw in Q3, so just a very small decline. In combination, this means that the international operations are looking at a low double-digit percentage growth in GMV, not a decline, which is what we've been used to seeing for the quarter-to-date. For quarter-to-date, that means that on a consolidated basis, we're looking at a high single-digit percentage growth rate. That's the current trading. What does it mean then for the outlook? There's only one change in our outlook. Given the slow start to November, we've decided it makes sense to be cautious about our ability to catch up in the remaining weeks of the quarter. We are starting to see some recovery in the growth rate, but the big surge in spending is still ahead of us.
It makes sense to stay cautious on exactly how strongly that growth is going to come back and make up for the slowness in the first two weeks of November. With that in mind, we have cut or trimmed the GMV guidance for Poland to 9%-9.5% from around 10%, as it was previously. A marginal reduction. When it comes to revenue and adjusted EBITDA and also the CapEx, absolutely no changes. Whilst the margin is coming down, it is coming down a little bit more slowly than we had allowed for in the guidance. We are overperforming the original forecast, and we do not see any need, despite the slightly lower GMV forecast, to make reductions on revenue or adjusted EBITDA. Everything there is confirmed. That concludes my comments. I am going to hand it back to Marcin now, who will take you through the key takeaways.
Let's move on to the summary of the third quarter. Summarizing, we delivered excellent financial results with strong profitability, which is confirming our strength and potential of our business model. Additionally, we saw dynamic growth in both advertising and financial services, with both areas making a significant contribution to our financial performance and driving GMV acceleration. We welcomed DPD, the second-largest logistics operator in Poland, into Allegro Delivery, expanding our network of lockers and pickup points up to 70,000 locations. We also launched Allegro AI Assistant, further proving our ability to innovate and adopt cutting-edge technologies at scale. It is worth mentioning that we are one of the first marketplaces in the world having this functionality on the marketplace. We also announced a high-potential partnership with PKO BP, the largest bank in Poland, unlocking new financial services for both buyers and sellers.
Finally, we continue to advance our strategic evolution, setting the stage for even faster growth in the future. That is all we had in our presentation. Thank you for listening. It is time for the Q&A session. Tomek, the floor is yours.
Marcin, Rafał.
Yes, yes. Thank you. Thank you so much. Thank you very much for the presentation. We are now opening the question and answer section. If you are connected via the phone and you would like to ask us a voice question, please press star two on your phone keypad and wait for your name to be prompted. If you are connected via the web, you can also request to ask a voice question or send your question as a text. We already have some participants eager to ask a question. We will start with Cesar Tiron from Bank of America. Cesar, please go ahead. Your line is now open.
Yes, hi. Good morning, everyone. Thanks for the presentation and the opportunity to ask questions. I have three questions, if that's okay. The first one relates to the slowdown that you're seeing in the GMV growth in Poland in November. Can you please give us a bit more comfort that this is really driven by the cold weather and not, for example, by macro or potentially market share loss? I would suspect that the categories that are related to cold weather would have to decline probably by 20%-30% on a year-on-year basis to trigger such a significant slowdown on the GMV. Probably give us more color on that, if that's possible. The second question relates to this partnership which you've signed with PKO. That sounds very interesting. Would you be open to share maybe more on the economics and what it means from Allegro?
More revenue, probably lower acquisition costs, etc. The third one would be on the AI agent. When would you expect this to be fully operational? Basically, when would I be able to use the agent and just tell him, "Help me buy some clothes for the winter season?" He is going to be able to buy all that with the relevant clothes for what I like and be able to do the check-in as well, the checkout, sorry, as well on the platform. Thank you so much for your help.
Thank you for all questions. Maybe I will start with some explanation how we see the first two weeks of November and our performance, but also performance of the whole market. Because, of course, we are in discussion also with other players trying to understand the reason of slowdown of the first two weeks. We see that this is nothing related specifically to Allegro. We have the same information with all major players that something really happened on the market. Probably it's related to weather, but also you should remember that we had in Poland a long weekend. This is something always affecting sales quite strongly. What we see the last couple of days, this is something positive. We see that some categories are growing faster than in the first two weeks.
We see, we think that this is something what will be changed in the next couple of weeks, and our results will be improved.
Let me just follow up and attack the same question, Cesar. You were asking about the cold weather categories. It's not quite as steep a decline as you calculated, but yeah, the categories like automotive and fashion in particular are down year-on-year, and that's essentially because we haven't had that cold weather so far. The bigger topics are the ones that were just mentioned by Marcin. Also, we've extended the Black Week, which is an experiment. We've extended it now effectively to a Black Month. I think customers are essentially still waiting to start spending their money for the season. We are starting to see some recovery in the growth rate. It's still that the big surge in the spend that surely is going to come, it's still ahead of us.
We need to be a little bit cautious as to how it's going to turn out and whether or how much we're going to recover the total growth for the full quarter. I'll take also the second question on the economics of PKO and then of the PKO deal, and then Marcin will come back on the AI agent. Yeah, the PKO deal is very exciting. It's combining the power of the biggest bank with obviously the biggest marketplace in Poland. For the bank, this is mainly a play about getting new accounts. They see it as a, for them, acquisition cost will be quite competitive if they work together with us. For us, what we see is two things.
One is a new payment method, a one-click payment method that goes straight through to the bank account of the PKO customers, making it incredibly convenient to shop on Allegro. Secondly, if you shop on Allegro versus on other sites, you're going to get cashback, right? It can be anywhere from 1% cashback to 3% cashback in various scenarios. What we know is that from talking with PKO is that they have something like PLN 25 billion of e-commerce spend going through their customers, which is not being spent on Allegro. Whilst we haven't got any specific estimates about the impact of this project, we're hopeful, obviously, that with those two enhancements, some of that PLN 25 billion will start to be spent on Allegro rather than at competing sites. Today, obviously, PKO is one of the biggest banks.
Many, many millions of their clients are also within our 15 million clients already. For us, it's less a play about acquiring new customers. It's more about getting a bigger share of spend. Marcin, the AI question?
Yeah, maybe I will add something also to our cooperation with PKO BP because you covered a new method of payment, very effective and potentially improving conversion called Allegro Klik. We have the second product called Allegro Capital helping sellers with financing and having a fully online process of getting some money from the bank. This is great information to us because this is capital helping our merchants to improve selection on the marketplace and potentially to make it even more attractive. This is a fully digital process, very convenient, very easy, and also based on some attractive rates. Of course, the cooperation with the largest Polish bank is very promising. We expect that it should boost our GMV. We just announced this cooperation.
We will implement both services in the next couple of weeks and, of course, carefully observing the reaction of the market. It looks right now very promising. Following your question about the AI agent, we are excited to implement this functionality on the marketplace, and we see first very positive reactions from our customers. We should remember that this is still on test with, let's call it, friends and family, so with a limited number of users. We see the first results that it's boosting conversion. This is great because thanks to this technology, and this is comparable to what we see on the market using applications like ChatGPT or Gemini, quite the same functionality, this is something helping our customers to find and to choose the proper products. This functionality, AI Assistant, will be available for all customers early next year.
It will be full implementation. As I said before, we are expecting something good because we see high demand, especially thinking about the young generation. They look for such functionalities. This is something great. This is something exciting. This is something showing different than the new way of shopping and for sure supporting or improving even the experience on our marketplace.
Thank you so much. That was very comprehensive and helpful. Thank you.
Okay. Thank you. Thank you very much. We will now move to our next question that comes from Luke Holbrook from Morgan Stanley. Luke, please go ahead. Your line is now open.
Yeah, good morning, everyone, and thanks for taking my questions. My first is actually back to the weather because it does look like it has got colder in the last couple of days, and you're suggesting that trading's picking up. It looks like it's going to remain relatively cold over the next week to 10 days. I am just wondering if the weather patterns do go in your favor, is there a possibility that you think you might have been a bit conservative in the guide into Q4? Could you actually outperform the guidance on where you're suggesting? My second question is just if you can kind of give us an indication on where the cost of delivery could head for your largest logistics provider into 2026, or if you can't comment explicitly on that, where we should regard the cost of delivery increase into next year.
Finally, just more of a conceptual one here on Agentic AI, is it possible just to comment on how you see your proposition evolving if consumers do end up going to LLM more to search for products or items, or perhaps you can comment there on any partnerships that you have in the space? Thank you.
Yeah, thank you very much for those questions. Yes, you've obviously got your eye on the weather very closely. It has started to get colder, and we are indeed seeing some improvement in the categories that I mentioned, and that's contributing to the recovery and the growth rate that we're seeing. The weather is not the only part of the story. As I said, the initial part of Black Week has not gone so strongly. As I said, that was an experiment to try and extend it. We do think that the growth and the Christmas season spending is going to come through strongly over the next weeks. We're just being somewhat cautious that having lost essentially 2 weeks out of the 8-week season in terms of the growth rate that we've seen, getting back all of this over that remaining 6 weeks is very optimistic.
We want to be a little bit on the cautious side, right? We do expect recovery, but in the circumstances, we need to be a little bit more cautious given what has happened in the first 2 weeks of November. The second question was about cost of delivery. In particular, I think it was related to the pricing for InPost for 2026. There is a binding contract through to 2027, as you all know. Within that contract, there is provision for indexation increases on an annual basis. They kick in on the 1st of January. It is totally at InPost's discretion if they want to implement those increases or not. Inflation this year has come down considerably. Right now, it is running at just under 3%. These increases are going to be, if they are implemented, low single-digit increases this year for 2026 compared to 2025.
If they choose to implement them, then obviously they're already more expensive than the alternatives, as we've said many times. If they implement those increases, they're going to be even more expensive than they have been up until now. We will see what happens. The simple answer is that they have the right to do that should they choose to use it. Agentic AI, you want to take that one?
Yeah. Thank you for this question because this is an extremely interesting one because especially in the past, we heard many times that Agentic shopping mainly or some applications like ChatGPT, they can be quite competitive to marketplaces.
We see it completely on the opposite side because we are in discussion with all major players about cooperation, what we can do together because we see as an opportunity to just generate higher GMV thanks to accessing through these applications to a new group of customers and, of course, changing a bit our way of shopping.
You see that we implemented our own technology or our solution to the marketplace, but of course, you can expect that we will be also visible with our offering in different types of applications on the market, major ones, because cooperating or discussing with all major players, we see almost unlimited opportunity because we can do a lot for them, giving a wide selection of products, giving some value-added services, and of course, using these applications, using this opportunity for us to, again, generate additional GMV or to be closer to a new group of customers. This is something promising to us. Of course, we will be announcing, I guess, next couple of quarters, many new initiatives related to AI because this is really technology changing the industry and helping marketplaces as Allegro to even boost GMV and the whole business.
Understood. Thank you very much.
Okay. Thank you. Thank you very much. We will move to the next voice question that comes from Mia Strauss from BNP Paribas. Please go ahead. Your line is now open.
Hi, good morning. Thanks for taking my questions. I just got two. I know you talked about a Q2 end today that you're going to be looking at offering services outside of Allegro. Could you maybe just get a bit more color on maybe the delivery sort of services you could do and what timing that could be? Secondly, just on the Polish adjusted EBITDA , where you talked about that 8% or 8 percentage point impact from your principal cost, can you maybe just explain this? I think I missed a little bit of that and how we should think about this going forward. Thanks.
Thank you. I will take the first section of your questions. Yes, we do analyze externalization of our services, some of them, like for example, some of financial services. You should know that our buy now, pay later is, I guess, the largest or most popular service on the market, of course, very competitive to others and existing on other marketplaces or online stores. Of course, we see huge demand from our merchants for potentially to use this service also of Allegro because many merchants, they use mainly Allegro as a key sales channel for them, but also having additional sales channels, like for example, online store of them. We think that some financial services, some logistics capabilities, or maybe some advertising services can be also sold outside Allegro. Of course, we analyze attractiveness.
Of course, we analyze potential impact on attractiveness of the value proposition on Allegro as well. We are in discussion internally. We are in discussion with the board. If we decide to go further, I guess next couple of quarters, next year, something will be visible on the market. Of course, we should test firstly. We should get some feedback from customers, how attractive our services are in comparison to others in new places, in a digital environment outside the marketplace. We are sure that we can potentially enter some new markets and to create some attractive, very attractive value proposition and be competitive against other players.
Thanks, Marcin. I think the second question, if I caught it correctly, was about the 8% of the increase in cost of delivery that related to proprietary accounting. Let me try and walk you through that.
Historically, the delivery services have been done on an agent model in the sense that we would arrange a delivery for the Smart! customer, and we would cover the cost of it. Therefore, we were acting as an agent. In those situations where there is revenue, that revenue gets netted off against the cost. The revenue in those cases is the Smart! subscription revenue. Historically, essentially 100% of the cost of delivery was agent accounting, and the Smart! costs were being netted off. More recently, we started to do deliveries using the framework provided by the Allegro-managed volumes. Within that, there is Allegro Delivery, which particularly relates to APMs. In this case, it becomes a proprietary method of delivery where we are taking responsibility as the provider of the service.
We are either doing it ourselves with Allegro One, or we are working with our partners like DPD and DHL as subcontractors to us, but we are ultimately responsible. In proprietary accounting, it means that the revenue gets recognized gross. You probably recall that the logistics revenues were up 140% year-on-year when I was describing the revenue side. The costs are also shown gross. What you have going into cost of delivery is essentially that switch in mix from agent accounting using the old methods, i.e., InPost, for example, to the 36% that is now going through Allegro-managed volumes, and the costs are therefore recognized gross, and any revenue is shown also gross. The revenue is obviously those Smart subscriptions, but we are also starting to handle some of the non-Smart! deliveries ourselves as well, where the consumer is paying for the delivery.
That means additional cost that was not recorded in the P&L previously. On the other hand, we make money on that because the consumer pays more than the cost of the delivery. Hopefully, that answers the question. If you're still a little confused, we're more than happy to take it after the call with the IR team.
No, thank you. That's super helpful. I appreciate it.
Okay. Thank you. Thank you very much. Our next voice question comes from Michał Potyra from UBS. Please go ahead. Your line is now open.
Thank you. Thank you for the opportunity to ask questions. I have three as well. The first one is a follow-up on your PKO partnership. It seems that the cashback, it's supposed to be one of the main, let's say, incentives for the shoppers. I am wondering if you could share who would be covering the cost of the cashback, please. That is the first question. The second question, I noticed that your balance sheet shows a growing amount of consumer loans on your books. I am wondering if that is something you want to continue doing, like a change in the policy, or is it more, I understand, like was commented last quarter, more of an opportunistic choice given high cash balances? The third question, a little bit of a complicated one.
I'm wondering if you can give us a little bit more color on your delivery cost changes. What I'm actually trying to understand is that once you onboard new partners like DHL or DPD, I try to understand how much of that increase in your kind of Allegro-managed deliveries is incremental and how much is still done by DHL, but just under a different agreement. If it's under a different agreement, I'm wondering if there is a difference in the unit cost for you, DHL now under Allegro Delivery and previously. Any color would be appreciated. Thank you.
I'll take these. Michał, thank you very much for those questions. Let me start with I'll take all three of those. Let's start with the PKO and the cashback. This is an important contract, but it's not considered to be material. Therefore, we don't need to disclose the specific terms and conditions. The financial impact, in short, is not expected to be material. Exactly who is covering the various elements, the Smart! discount or the cashback, is something that stays confidential. As I already outlined previously, both sides have got a lot to gain from this cooperation. We're very comfortable with the agreement that we've got. When it comes to the consumer lending, you were asking about our intentions in terms of using our own balance sheet.
is certainly the shift to using more of our own balance sheet this year is opportunistic in that we do want to run with significant liquidity to give us a lot of financial flexibility. This is a much better way to invest that money than simply sticking it on deposit. We are also able to very quickly access that money by going back to selling the loans in larger amounts from one week to the next, literally. If we decided to, we could recover this money very, very quickly with the partners that we have at the moment. We also have other potential financing partners in the pipeline. Many banks here would like to be part of the Allegro Pay story. The last question, yes, around delivery.
Yeah, building on what I said to the previous question, as parcel deliveries have moved from the agent model to the Allegro-managed delivery model, we move essentially, it's the same parcel, but we go from a netting of any revenue with the cost of delivery to showing the revenue gross and the cost gross. Yeah. That's accounting for that 8% increase that I was describing in answering the previous question. If we talk about Allegro Delivery, prior to joining Allegro Delivery, neither DHL nor DPD were doing any APM deliveries for us. They have existing courier relationships with us, but not on the APM side. The APM side is purely incremental, but it's basically then part of that 8% where we're shifting from the agency model and using somebody else to using these new partners. Hopefully, that answers your question.
If I've not covered all of it, then again, we'll follow up with a detailed answer.
If I may, just a very quick follow-up on the delivery. I get the accounting and I get the APM part, but for Allegro Delivery door-to-door, physically delivered by DHL, is the price you pay, not the accounting, the actual price different versus what it was before they joined? Thank you.
Yes, the pricing, it's a different price. Generally speaking, the prices are better than the prices we had in the previous model.
Thank you.
Okay. Thank you. Thank you very much. We are moving to the next voice question that comes from Aaron from Ashmore. Please go ahead. Your line is now open.
Hi. Thank you very much for taking the questions and for the call. Two questions from my side. One would be perhaps just to summarize on the delivery cost to GMV on the outlook there. Obviously, it's ticked up a little bit this quarter. I appreciate all the nuance that you've given on what the individual drivers are behind that, but kind of bringing it all together, looking for the next few quarters and into next year, do you think the delivery cost line as a percent of GMV, as it's currently reported in the P&L, do you think that will start to trend down over the next few quarters? That's question one.
I think the company has spoken in quite general terms about being excited about offering new categories, being a kind of broader lifestyle or ecosystem proposition, whether that may include grocery or other kind of new categories that Allegro has not historically operated in. Any kind of progress on that or anything that you would like to comment on there, please?
I'll take the delivery one, and then Marcin will follow up on the second question. Obviously, we're not giving any specific guidance about 2026 at this point in time. I'll answer the question more in terms of the drivers that we're looking at. The pure volume driver, obviously, the more GMV and the more Smart! customers we have, because Smart! customers spend more when they convert from being non-Smart!, that's a good thing, right? The GMV goes up, also the cost of delivery goes up as the percentage of customers doing their deliveries via Smart! increases. Essentially, the 17% out of the 28%, if that keeps moving up, we're more than happy, basically, because it means GMV growth. The 8% that relates to this structural shift towards the managed volumes, this is something that we intend to keep growing.
This is giving us access to lower unit prices than we're currently paying with the biggest delivery provider that we have, who's still in the agency model. If we manage to sign at some point a new arrangement that's satisfactory to both parties that will supersede the current contract, that's obviously going to have a material impact, presumably favorable to us on the cost of delivery line. That is still ahead of us and subject to negotiation. That's basically the key elements. I mean, at the moment, the more we do ourselves, at the margin, the lower the average unit cost.
Marcin, the second.
Yeah. The second question was related to potential new product categories. We shared with the market a couple of weeks ago about our considerations or plans, what we want to do on the market, how we want to enter in some parts of the market, and how we see potential. Of course, we're analyzing many of them, trying to identify some new cool product categories, of course, giving us material impact on our GMV. Because, of course, we should remember that we want to give our customers the right selection with the right value proposition, but finally, to have material impact on our business. We can expect that something in our selection will be visible in Q1, so next couple of quarters. We are in discussion with several partners.
We have some ideas what kind of products can be helpful or can be perceived as attractive ones by our customers. The second thing, even much more important to us, is because we're also talking about services. This is something new potentially to us. I'm saying about also how to expand offering related to value-added services, helping to boost sales of products we have on the marketplace, but also services, what we shared, like ideas in the past. It can be potentially like travel. It can be potentially like healthcare or these types of services. Yes, we are excited doing these projects or identifying some new opportunities. The first test should be visible quite soon, next couple of months.
That's great. Thank you.
Okay. Thank you. Thank you very much. We will now move to the next questions. I will hand over to Mr. Tomasz Poźniak to read them out to the audience.
Thank you, Rafał. Let's take the first question from Marek Szymański, IPOPEMA Securities. Could you please talk about the path and the perspective for a bit of profitability for the APM lockers?
Can you just repeat the question? Did not catch it.
Could you please talk about the path and perspective to a bit of profitability for the APM lockers of Allegro One Box?
Yeah, certainly. At the EBITDA level, as we've already mentioned, it's clearly a lot cheaper than paying for the alternative, which is presumably, certainly in the case of InPost, more than covering their full cost down to EBIT level. In our case, we're actually making really significant progress in terms of filling the lockers and in terms of utilization, which means that we're getting closer and closer to the break-even. The important thing about having our own locker capability is the optionality that it's giving us. If we hadn't started this project back in 2021, then the progress that we're making at the moment wouldn't have been possible. Also, the discussions around the long-term deal that we have with InPost would have been more difficult than they are in the current situation.
We're heading towards profitability at the current price level that we have in the market is the short answer to the question because we're making tremendous progress in terms of the volumes we're moving.
Thank you, Jon. Next question is from Roman Reshetnev from Goldman Sachs. Could you provide an update on discussions with authorities regarding regulatory changes to ensure a level playing field with international marketplaces? What would be the update on the timing and nature of potential legislative action? Would you see this action different compared to other EU markets?
Maybe I will take this question. Of course, the answer is that this is something that is supported by us because we want to see fair competition on the market. This is great that we have some new players sometimes pushing us to be even faster with development of some capabilities and to increase the attractiveness of the offering we have. We should play using the same rules. We should play having the same conditions on the market. Of course, we are quite close to local authorities doing some research, sharing some analysis to show the broader perspective, how it is impacting also the Polish economy, also on the negative side.
Of course, we are fully supportive to all initiatives, the local ones, but also on the European Union level to fix this problem because, again, this is something affecting badly European companies, European marketplaces because of unfair competition.
Thank you, Marcin. Thank you again for the time, which is running out right now. The last question would be from Harry Welton from Virgin Task Management. You guided for margins of just below 6% for full year 2025. If we assume 5.9%, that implies 5.2% for Q4 2025. Is that in line with expectations given the aforementioned impact?
Yeah, I'll take that question. I do have a little cheat sheet here with me, and I think you're slightly low. I think I heard you say 5.2%. The low end of the range would definitely be in the mid-5s and the high end of the high end of the range, yeah, around 5.7%, something like that for the quarter, yeah, because your question was about the quarter itself. I think 5.2% was too low.
Thank you, Jon. It is 11:20, the scheduled time to close the conference. I would like to thank all the participants. Thank you for all the questions. Jon, Marcin, thanks for the presentation. See you, everyone, in March when we will be reporting Q4. Rafał, the floor is yours to conclude the presentation.
Thank you. This concludes the call. We'll be now closing all the lines. Thank you and goodbye.