Nebius Group N.V. (NBIS) Q3 2021 Earnings Report, Transcript and Summary

Ladies and gentlemen, thank you for standing by. And welcome to the Third Quarter 2021 Financial Results Call. I must advise you this conference is being recorded today, Wednesday, the 27th of October 2021. We'd now like to hand the call over to your first speaker for today, Yulia Gerasimova, Investor Relations Director.

Hello, everyone, and welcome to Yandex' Third Quarter 2021 Earnings Call. You can find our earnings release, additional prepared remarks and supplementary slides on our IR website. The key speakers on our call today are, Tigran Khudaverdyan, our Deputy Chief Executive Officer; and Svetlana Demyashkevich, our Chief Financial Officer; Vadim Marchuk, our Chief Operating Officer; Daniil Shuleyko, the Head of E-commerce and Ridetech Business Group; and Yevgeny Senderov, Chief Financial Officer of Yandex.Taxi will be available on the Q&A session. Now, I will quickly walk you through the Safe Harbor Statement. Various remarks that we make during the call regarding our financial performance and operations may be considered forward-looking, and such statements involve a number of risks and uncertainties that could cause actual results to differ materially. For more information, please refer to the Risk Factors section of our most recent Annual Report on Form 20-F filed with the SEC. During the call, we'll be referring to certain non-GAAP financial measures. You can find a reconciliation of non-GAAP to GAAP measures in the earnings release we published today. And now, I'm turning the call over to Tigran.

Thank you, Yulia. And hello, everyone. Let me give you a quick overview of the key highlights from the third quarter. Our results, again, shows strong growth momentum across key verticals, including the two largest advertising and ride-hailing. This business has continued to generate solid cash flow for us to reinvest in a number of attractive opportunities. In addition to that, both businesses continue to grow faster than their peers, further cementing our leadership in advertising and ability. Search and Portal delivered 33% revenue growth supported by our investment in ad tech, SMB products and iOS share. IOS share improved during the quarter to 43.1% in the last week of September 2021, up 2.4 percentage points year-over-year. This growth was underpinned by targeted investments in product and marketing. Ride-hailing revenues increased by almost 70% on the back of solid growth of Rite and GMV. We continue to improve our operation efficiency and optimize our cost structure with operational expenses decreasing as a percentage of GMV, both quarter-on-quarter and year-over-year. At the same time, we invested in service quality and in growing the driver base, which led to a better supply-demand balance as well as 33% increase year-over-year in the number of drivers on the platform. Our initiatives in ride-hailing position us well for further growth and improvement to profitability. During the third quarter, we also demonstrated our commitment to invest in the future growth of the Yandex ecosystem with a particular focus on expansion in the E-commerce Yandex. Plus and logistics verticals. We are pleased with the results delivered by these businesses. The number of Yandex. Plus subscribers doubled year-over-year to 10.5 million, with even faster growth in paying subscribers. In E-commerce, GMV increased more than 3x year-over-year. And delivery grew 4.5x year-over-year. This was partially driven by the low base effect of the previous year. However, quarter-on-quarter growth was also robust, with third quarter revenue up 40% plus compared to second quarter. E-commerce remains our top priority. We see substantial progress in this key vertical with the GMV growth accelerating to over 200% year-on-year, record levels of assortment reaching 21 million SKUs, further growth of seller numbers and expansion of logistics infrastructure and significant progress in terms of service quality. During the quarter, we reduced the defect rate in fully controlled logistics by 2.5x to a level comparable with our key competitors. Product-wise, one of the key priorities for us in E-commerce is further diversifying both our assortment and geography of our operations. We are actively investing in both of these areas. We have also introduced a category centric approach to organization with appointed leaders responsible for end-to-end customer experience in categories such as FMCG and Fashion. We are developing and continuing to invest in other business initiatives, such as Fintech, self driving technology and Cloud to ensure we are ready to meet future demand to create additional foundations for sustainable long-term growth. With this, let me turn the mic over to Svetlana.

Thank you, Tigran. And hello, everyone. You've seen our press release and additional comments about the performance of our businesses that were published on our website. Let me focus on our updated outlook. In our Search and Portal business, we are now guiding for high 20s year-on-year growth for the full year 2021. This is the third time this year, we are upgrading growth expectations for our advertising business. Early in the year, faster than expected growth came mostly from a better macro environment, specifically a faster than anticipated recovery in business activity and the advertising market after pandemic related restrictions during 2020. In recent months, this has been driven more by our deliberate decision to reinvest part of our margin into growth. As Tigran already mentioned, these were primarily investments in ad tech, iOS market share and SMB products. With these investments, our adjusted EBITDA margin for the full year will be marginally down on a year-on-year basis, but we still expect it to be over 48%. This extra basis points have allowed us to grow faster and solidify our leading position within the digital end market. Our goal is to optimize for the high possible absolute cash EBITDA in this business. Our base case is that we should be able to keep ad margins stable going forward. However, we may consider reinvesting, provided that the opportunities we achieve, A, will help us to achieve a higher absolute adjusted EBITDA; and B, are in line with our long-term strategic priorities. Couple of words about E-commerce, which is currently the key priority for our management team and also the focus of our investments. We were pleased to see GMV growth in third quarter accelerate year-on-year to 3.1x from 2.6x in the previous quarter. This growth was underpinned by investments in expanding all logistics infrastructure and delivery channels, numerous enhancements to B2B product and B2C interfacing situation. The launch of Market Express as well as targeted marketing campaigns to strengthen our brand's recognition as a multi-category marketplace. Our guidance for E-commerce growth and investments remains unchanged. We continue to expect GMV growth to be up by 3x this year, while total cash burn for all E-commerce businesses in Russia will remain at around $650 million. As a reminder, with new total cash burn as a sum of adjusted EBITDA losses, CapEx and changes to working capital. Based on this, our spending for nine months amounted to around RUB40 billion. During the quarter, we continued to experience pressure on our unit economics on the back of front-loaded investments in logistics infrastructure as well as the success of our asset-light dropship by seller model. The latter accounted for 30% of our turnover as of the end of the third quarter and contributed to relatively low utilization of controlled infrastructure. However, this also means that we are fully prepared for continued growth towards the high season and into 2022. We have also continued to work on improving operational efficiency. For example, we have fine-tuned our pricing algorithms and assortment strategy, which helped to improve our 1P front margin. We benefited from a redesign of our pickup points compensation scheme and an improvement in utilization rates. And we made adjustments to our 3P tariffs. Finally, ride-hailing. We have raised our expectations for GMV growth to 65% to 70% from 60% previously. This primarily reflects our efforts to improve driver numbers on the platform and their utilization, which led to solid growth in rides with ride number having grown both year-on-year and quarter-on-quarter. As we said in our Q2 call in July, we have continued to invest in driver supply as well as in improving the quality of our service to support our market position and long-term growth. Assuming no material changes in consumer or competitive environment in the remaining two months of 2021, we continue to expect the ride-hailing adjusted EBITDA margin as a percentage of GMV to increase year-on-year compared to 2020. In conclusion, I would like to underscore that we have a very strong and experienced team with an established track record of finding new attractive opportunities and turning them into large and efficient businesses. Mobility is an excellent example of this. The results to date clearly demonstrate that we are making good progress on our growth strategy, which justifies future investments that we are planning. Going forward, we will adhere to our policy of financial discipline, which will remain a core principle as we continue to create value for our shareholders. With this, let me turn the microphone back to the operator for Q&A session. Thank you.

[Operator Instructions] Our first question today comes from Cesar Tiron of Bank of America.

I have two. So I'll ask the first one. First, over time, I would like to understand what do you believe is your key competitive advantage in E-commerce? Not today, but once the business scales up, would that be things like speed of delivery, SKUs, automation? Or do you think that you could use some of your competitive advantage in tech to basically gain market share? And if you can explain us what would that be?

This is Vadim speaking. Let me take this one. So when we think about the longer term, what's our key advantages could be, the way we think about those, as in any marketplace, you've got to look at the both sides, right? What you can offer to consumers and what you can actually offer to the sellers. And if you were to start with consumers, we believe that currently on the market, we are probably the only platform that has so many different components and pieces under the same roof, whether it's going to be music streaming, video streaming, whether it's going to be the ride-hailing, FoodTech, our Yandex. Plus subscription was multiple benefits, whether it's going to be the cash back on transactions or kind of the free subscription for both music and streaming. So overall, we think that by putting this kind of big complex bundle and offering together for the user, we can get him. And then obviously, another component here is the logistics, right, which allows the last-mile delivery, and we can do it quite quickly and smoothly, therefore, maximizing the user experience. And therefore, by putting those things together, rather than competing specifically on price alone, right, you will be competing both on price but also on the breadth and widths and depths of your offering. So this is the first thing. The second thing, if you look at the kind of the merchant side, right, for the merchants, we can also put together quite a -- what we believe we can put together quite a compelling bundle, which is going to help them both to improve their kind of own marketplace, but also potentially off marketplace place sales. We do believe that we have the best ad tech in the country. And therefore, this is something that we can provide the merchant as a tool to promote their goods to enhance their sales. We do believe that throwing there the kind of our logistics component can also help them get rid of all this hassle of getting the goods to the users. And then, with addition of our kind of Fintech services leader in next year and further out, we would -- we believe that we would be able to put together quite an attractive bundle that's going to cover all the needs from starting from kind of dealing with their financial questions all the way to client management systems. So, the ways to attract those clients and an ability to sell their goods and then deliver. So this is the way that we think about what's the advantages that we have and why we can put a better package than probably some of our competitors.

That was extremely helpful. Just wanted also to check on the Taxi business, specifically on the ride-hailing. Is the driver shortage situation improving? And if not, how long do you think this will take to improve and therefore, potentially positively impact tech rates and profitability for the ride-hailing?

It's Yevgeny. Let me take this one. Well, we did say on the last call that we will continue to invest in the quality of the service and also in the driver base. And actually, in September, drivers on the platform exceeded 1 million, and that's up 50% versus sort of the post lockdown summer of 2020. So we don't envision, when we do our forecast that the cross-border and the border opening situation will improve significantly. But while saying all of that, we do continue to see improving profitability in the ride-hailing business. We said before, that 2021 is going to be a better year in terms of margin than 2020, and we continue to see improvements in EBITDA margin in the future consecutive years. There is some seasonality, Q2 and Q3 tend to be lower margin-wise in the Taxi business and then Q1 and Q4 expand. But the overall trend for improving margin continues. And I think, while the situation with drivers is challenging, we have found a way to deal with it.

The next question today comes from Ulyana Lenvalskaya of UBS.

I wanted to follow-up on the E-commerce business first. Could you please comment on the current competitive environment? And maybe the dynamics of the take rates for you and maybe some of the competitors?

This is Vadim. Let me take this one. So as you know, we did increase our take rates back in July of this year. And -- which I think we discussed during the last quarterly call. So far, we haven't seen any kind of negative consequences of that move. And again, as we were describing it earlier, the way to think about the take rates rather than just a pure numerical figure, probably the more correct way to think about it is kind of the overall offering that we are providing the merchant with, right? So as our kind of call it B2B side of the marketplace is improving, we do quicker onboarding. We are much better now with kind of with the pictures that we are putting there, the product cars, right? We are able to provide them with means of different delivery options, such as Yandex logistics, et cetera. As we've seen that -- as our platform improves, we believe that we can start increasing the take rates, and this is what we did last June-July. And that number -- and we haven't seen any negative response from the merchants as well. We do believe that the market is very large. We are still in relative infancy of E-commerce in Russia. We think that the take rates on the market remain fairly low overall. And we do believe that they are likely to go up in the future for the industry as a whole.

This is very helpful. And secondly, I wanted to check with regards to the upcoming kind of working holidays or whatever, way we call it, so the new potential look down, what kind of impact do you expect mostly interested in the core Search and Portal trends, advertising market trends?

It's Svetlana, and so, let me take this question. So of course, that we have been operating in a pandemic situation for over 1.5 years at the moment. And we already adopted to this situation in all our businesses. In addition to that, I think we are well hedged because of our diversification across segments. And there are, of course, some businesses that may have an adverse impact, like advertising or mobility, as we've seen in 2020. But there are also a number of businesses which will benefit from the low down, and we already see some of the trends here. In grocery, restaurant delivery E-commerce and streaming services, we should expect some improvement from lockdowns. In advertising, we do not expect material impact so far. At the moment, new lockdown measures are not very severe, while nonworking period only 5 days only. We'll see if it continues. But overall, I think that all sectors within advertising felt quite well. We do see positive dynamic and growth in most of the sectors. So, we do not expect any material impact.

It’s Yevgeny. Just – sorry, let me quickly add in quickly. Particularly to Taxi you’ve heard – to add to what Svetlana was just saying, you heard earlier, raise our guidance for the Taxi GMV from 60% to 65% to 70%. So that does include some impact from COVID in the fourth quarter. So we are planning for some impact. Of course, it’s hard to predict what the final situation is going to be, but that’s our fourth wave that we go through. So we know what to do dealing with them.

And our next question comes from Ildar Davletshin of Wood & Co.

I'd like also to say thank you for additional disclosure in the shareholder letter, which is extremely helpful. So I wanted to ask couple of questions. One is on your funding position, liquidity. So we've noticed you've been using more cash over the couple of quarters due to high investments in organic and in-organic. And I think a total net cash position is below $1 billion. So my question is, are you worried that the a long liquidity may slow down your growth trajectory later on in the next one, two years? And if so, what are your kind of priorities here? And I'm particularly interested if you're considering any partnership to maybe help with some of the verticals when it comes to self driving cars? So that will be my first question.

And let me take this question. So of course, it's very important for us to remain financially flexible because, as you know, we are considering and approaching a lot of opportunities. Of course, liquidity is an important factor when we are making decisions, including doing M&A. And as you've seen in over transaction, liquidity was also one of the factors for structuring the deal in the way it was structured. So at the same time, we understand the situation of the market, and we are very confident in our ability to attract additional funding in different instruments. We are very stable and performing, fast-growing with very attractive so the growth potential. So we're quite sure that funding is available for us in different forms. So we might consider different options going forward, subject to the market situation.

And maybe the second one, if I may, is on your Yandex. Plus business, which is a great differentiating segment. However, even though year-on-year growth looks extremely attractive, but more on a quarter basis, it's actually slowing down. So I'm curious what is it driven by this slowdown? Is it because you're raising the number of paying customers? Or is it also because of the more fierce competition with alternative subscription services available from other providers? So please comment on that.

This is Vadim. Let me take this one. So look, there are a couple of components in our answer. I mean, frankly, we think that we actually achieved quite a nice growth. I think it would be useful if you were to compare it to the Netflix, Enea additions from the second to third quarter. And taking into account that our geography is definitely much more limited. But there is definitely a seasonal component as well. It's a third quarter, was two summer months when people tend to spend outside of the city or go on vacations and spend less time actually consuming streaming services, which you typically consume in-house. And when you spend time outdoors, you tend to watch less. Now we do definitely focus on paying subscribers. And therefore, we've somewhat limited the proportion of trials in our subscriber base. But overall, again, let me emphasize that we are quite happy with the net additions this quarter. And then furthermore, talking about the competition, et cetera, I guess what it's -- not sure whether you've seen it, but there was recently, I would say, probably within a couple of two to three weeks time period, GFK posted kind of their analysis and survey of the streaming market in Russia. And what it actually shows that we are, I would say, probably the only one of the very few players in that market that has been added or increasing their subscriber base in percentage terms compared to the competition.

Our next question comes from Vladimir Bespalov of VTB Capital.

Congratulations on good numbers and good trends. I have a couple of questions. My first will be on E-commerce. First, could you provide maybe very broadly some outlook on how much you're going to spend on E-commerce next year in terms of your investments? And on investments, how in general do you evaluate the efficiency of those investments that you are making? What are the key metrics probably you are tracking and how you make decisions on future investments and in E-commerce? And in particular, could you also provide brief comments, for example, on underutilization of logistics infrastructure in the third quarter of this year, what was behind this? Because in general, we see the logistics as a bottleneck for the development of E-commerce. And about the closure of several direct stores for Yandex Lavka during the period? This is my first question. And the second question will be on logistics and delivery services that you are developing. Could you please provide what you see as your addressable market here? What are you -- who are your competitors? And in general, maybe what are the key drivers, the key assumptions behind the profitability level guidance that you've provided? So if you could comment on this, this will be helpful.

It's Svetlana. A lot of questions in this to be honest. So let me start, and then I think the guys will also help me to continue. So let's start with investment in E-com. For this year, let me remind you, our investment of $650 million allowed us to create $1.5 billion incremental GMV and improve our market share versus our competition. It also allowed us to expand the assortment 10x and to increase our customer base twofold to 8.4 million. We also managed to improve retention of our clients and order frequency by 26%. It's also important for us that we were able to grow the number of active merchants on our marketplace on our marketplace by 3x to 18,000 and materially improve the quality of B2B product. In terms of logistics, we added 7 new warehouses, mostly in regions and expanded our total warehouse capacity by 3x. Now it's close to 300,000 square meters. We were able to create our own logistics platform and expand share of our own delivery in orders to 89% from 16% last year. So it's also a great progress. And what is also important and was mentioned by Tigran in his speech, we were able to improve the quality of delivery by 2.5x. So we're able to decrease our defect rates. A lot of other operational improvements are also in our letter to shareholders, and you can look at them. So that's the justification of further investment for us. We do see the great progress, and it increases our confidence in the success of this investment. And overall, the fact that we will be one of the leading E-com players in the market. And of course, E-com is one of our biggest priorities. So looking at the improvements we have and understanding the financial discipline in terms of investments, including internally in E-com, we understand that we are committed for the next year also. Of course, we will not give guidance for the investments in E-com for the next year. But as we commented previously, during the second quarter call, we shouldn't expect less of investments than we did this year. So in terms of other investments, just to comment how we make decisions in that respect. Of course, we looked at the size of the opportunity, we looked at the growth profile, return investments, and we also should understand how we create additional shareholders value in the midterm. So taking into account all these factors, we, of course, prioritize, and we do maintain financial discipline. So it's very important for us to be able sometimes to deprioritize some of the businesses or delay some of the developments or sometimes even dismiss some of the projects, if we understand that we don't see improvements in operational results or we don't see the increase of the shareholders' value or returns. With that, I think I'll pass the floor to Vadim.

This is Vadim talking. So let me take part #5 for question one. The underutilization of warehouses. So look, I mean, the way I actually would think about it, right, is essentially, we've invested upfront, right, into our warehouse capacity. And this is something that will be -- the utilization will keep increasing as the time passes by. So it's going to be this quarter, i.e., the fourth quarter, next year, et cetera. The reason why we're seeing relative underutilization compared to our original plan at the beginning of the year is because, as you know, at the beginning -- starting from the beginning of this year, we started a rather massive switch from the price comparison model, the CPC to marketplace model, the CPA, right? We had tens of thousands of merchants, the small and medium-sized merchants on price comparison. And as we started moving them to kind of the marketplace, we realize that it would be much easier to facilitate that move, if we were to provide them with fulfillment option, what we called internally dropship by seller, DSBS or DBS, that some other people in this room call it. And that essentially allows the merchant to list their items on our marketplace and then fulfill that sale or the transaction from their warehouses or fulfillment centers. When it is necessary or when they actually are needed, we also provide the kind of Yandex logistics service for them and actually fulfill the last mile as well. So as we started transition from CPC to CPA, we saw a significantly higher uplift or faster uplift in dropship by seller model, which is actually good trend because it's, A, it's profitable from day one, we take commission from transaction and essentially don't carry any fulfillment or delivery expenses. And number two, it allowed us to quickly expand our assortment. Just to remind you, we went from 2 million at the end of last year, SKUs to $20 million at the end of this quarter. So this is actually quite a good model, right? At the beginning of the year, we didn't think that many merchants will actually switch to dropship by seller as opposed to move to our fulfillment centers. And therefore, for this quarter and probably some residual effect will remain in the next quarter is some of our warehouses will be somewhat relatively underutilized. However, at the same time, when you look at our growth numbers, our transactions for GMVs that go through our fulfillment, obviously growing extremely fast as well. And therefore, that utilization is going to catch up. Now moving on to logistics or I guess Part 6 is, so look, the way we think about logistics, right? So first and foremost, and I know that we mentioned it on previous calls, but I think it is still important to mention. When we talk about Yandex logistics, this is an asset-light model. Essentially, what we do, we match supply drivers in kind of in cars or couriers, the walking couriers with demand. And demand here could be C2C, that was super popular during lockdown last year, when people were locked in their apartments, and they had to send something from one apartment to another and they couldn't actually get out on the streets, so that we're doing through the Yandex logistics service. Or B2C, whereby its small and medium enterprises that trade on our platform, on Instagram or any other marketplaces, would use our service to deliver intra-city. So that business, we believe that TAM for that business are all intra-city deliveries, same-day deliveries. As a matter of fact, we're actually expanding the TAM by also focusing on the next day deliveries, and we started experimenting with those. And as such, I think there will be numbers, but basically, we already did in excess of 20 million deliveries last quarter in Q3, and that thing is growing extremely fast. The reason why we talk about the margin that we mentioned, which I think it's mid- to high single digits of GMV is because we compare -- we think the model is very similar to the ride-hailing model. We are seeing already kind of low single-digit approaching to mid-single digits. But at the same time, within ride-hailing. At the same time, we do think that logistics is somewhat lighter on the cost side as a model because you don't need to deal with customer incentives. Your customer acquisition, as a matter of fact, it's going to be somewhat lower because we fully utilize our existing properties, high-traffic generating properties like our super apps, et cetera. So overall, I mean, it's a beautiful model, we think is going to be highly marginable, and we are extremely happy with their growth profile at this date.

Let me add on Yandex Lavka, a couple of words on delivery. So your question was in regards to the seven openings. But overall, I think this is expected to happen in any retail business where a certain number of units will be closed for one reason or the other. But if you look at overall openings for the quarter, restaurants went from 362 to 395. And actually in October, they're already 400. So in Lavka, we already see that unit economic works very well in the regions with a high average check. So if we look at Moscow, approximately 60% of our dark stores on pre overheads, premarketing EBITDA level already positive. The gross margin of top 25 stores in Moscow is approaching 35%. And our dark stores in Moscow turn profitable on the pre overheads, pre-marketing level -- EBITDA level between 6 to 12 months after launch. And we actually expect Moscow to become EBITDA positive on post-overheads level within the next nine months. In Russia, overall, our pre-overheads EBITDA for top 25 stores in September is already positive. And we're seeing very encouraging dynamics with couriers, where in high-density stores, we have more than four deliveries per hour for our carriers. So overall, when we look at our Lavka dark store business that we believe our unit economics is significantly better than any other competitor out in the market. And for delivery, just to provide, I think you asked about the customers. Just a couple of questions. We’re already in a very short period of time, captured, we think, more than 10% or teens of the Express delivery market this year. Express delivery is probably way a third of sort of over the last mile delivery market with the two-thirds being NDD, and we’re going to launch NDD by the next – NDD is next-day delivery, sorry. NDD, we’re going to launch by the end of this year. And if you look at our customer makeup, we have over 22,000 BTB partners, excluding SMB, small medium business, and that’s approximately 46% of our deliveries in the third quarter. And SMB was actually another 30%. So out of the B2B customers, top 10 is probably over 70% of total deliveries, and that includes almost all large food and nonfood retailers, E-com players and so on.

And the next question comes from Kirill Panarin of Renaissance Capital.

I've got two questions. My first question is on the impact of IDFI policy changes on your advertising business, especially outside Search. It seems there's no financial impact now, but how could this changes affect the targeting capabilities, and efficiency measurement or pricing in the future? And would the potential impact be material if Android follows the same route as iOS? That's the first question. And then secondly, on Yandex. Eats, could you please comment on competitive dynamics and margin trends in the restaurant segment of Yandex, excluding groceries?

This is Vadim. Let me take the first one. So look, the kind of the gist of it, we think that the potential impact on our revenues or profitability from Apple new privacy settings or for that matter if Android decides to introduce something similar is very limited and significantly lower than it's going to be for some of our global peers. And let me explain, why? So, as you know, structurally, approximately -- well, more than 80% of our revenues are coming from the contractual advertising, i.e., this is something where a person what would actually show apps and advertising for a specific query from a from a person now would search on our search engine results page. So this is piece number one. Now when you look at the remaining, call it, like 20 or so percent of the revenue, what you need to keep in mind that unlike some of the peers that are mobile-only and single app type of companies. We operate, number one, quite a few applications on your mobile device, number one. And number two, we own a lot of desktop properties, whether it's going to be ours or what is going to be our partner network. Therefore, we actually get to know quite a lot about the user through all of the combination of our properties, whether it's going to be mobile or desktop. And therefore, we are not that heavily dependent on those limitations that Apple just introduced. On top of that, the third factor that you do need to consider, a very -- well, a majority of our users, right, they actually are logged in users into Yandex ID. And therefore is A, it's easy to track them for us; and B, they are already, to a certain extent, submitted their permission to provide us with first-party information, i.e., their information.

And it's Yevgeny, let me add on -- I think your question was on Eats restaurant. So the restaurant business itself, it actually grew. And if you look in terms of orders, it grew 62% year-over-year and GMV grew 67% year-over-year. And that's actually on two-year stack basis, it's actually acceleration versus the trends we saw in the second quarter. And we did indeed invest in growth in the first nine months of the year, and we believe we gained 5 percentage points of market share according to our estimates. And new users continue to grow fast and actually growing 30% quarter-over-quarter despite us actually beginning to limit certain promotion campaigns such as free delivery. So while I think we’re going to update and comment on overall profitability trends for next year in Eats on the next call. We do see an improvement in Eats in restaurants profitability, and we expect it to continue into the next year.

Our next question comes from Slava Degtyarev of Goldman Sachs.

A couple of questions, both on E-commerce. I will start with the first one. If you can comment on your strategy in the grocery or FMCG side of E-commerce in terms of the formats and also the willingness to expand beyond the capital cities? Do you see reasonable unit economics in smaller cities in the fast and specifically ultrafast delivery formats?

Look, Eats -- for us is a very sort of -- it's a very challenging business, but I think the addressable market is gigantic. So we continue to sort of invest and balance investments versus growth in our thinking. And so, I think, when we look at losses per order, they decreased 30% in September compared to June in this business. So it's a challenging business model, but we think the addressable market is gigantic, and we're going to continue to look at it, again, carefully weighting investment versus growth.

Okay. And then specifically, it was not only about the lockdown of that performance, but maybe if there is some sort of a mix shift happening between the ultra fast and to some sort of medium fast, whether you have somehow fine from the right balance already?

This is Vadim. Let me add to what Yevgeny already said. Look, it's -- I think pretty much as everybody else in the world right now, right, everybody is striving for the kind of the right balance between the ultrafast and/or immediate or something, a shorter one, we are experimenting with that. What we are seeing is overall that the presence of whether it's going to be fresh or whether it's going to be FMCG on our E-com marketplace is an important component to have because it drives frequency, it improves your retention, it allows to actually to attract customers easier. So therefore, we see it as an important component of our model. At the same time, adding to what Yevgeny has said, it's a challenging business model because you typically deal with the lower average checks and relatively painful CPR, right? And that's why we are trying to play with different kind of components of this model. So optimize the burn vis-a-vis the improvement in retention and kind of the customer frequency of transactions on the main marketplace, right? At the same time, I guess, one of the examples, what I could give you of what -- to the first question that Cesar asked, what do we think is our competitive advantage? When I mentioned the presence of many different assets under the same route, it does allow us to be much more flexible, creative in combining different components of models and assets. So for example, we mentioned in our script that we published earlier today, the fact that we launched Market Express, right? Essentially, what Market Express is, is a combination of two models that we have under the same roof. So the first one, the supply side is kind of remnants of our CPC model, right. This is a lot of partners that were present on our price comparison platform. And when we were looking to solve the kind of the quick scenario rather than trying to build all the infrastructure ourselves and deal with the inventory and SKU selection, et cetera, we figured that it would be much quicker and easier to go to our former CPC partners and offer them Yandex logistics services in order to provide hundreds of thousands of SKUs within, call it, under two hours. And again, that is something that wasn't present in our business model even three months ago. We put it together, it's flew, right, as we mentioned, it's already kind of accounts for approximately low teens of the orders in Moscow. And because we have the necessary density on the Yandex logistics side, right, the CPO on those orders for us is significantly more optimized than it would be for any other competitor. So as we think about the FMCG and food for our E-com marketplace, we will be following kind of the similar logic of trying to achieve the ultimate goal, right, with respect to the user retention and other metrics in the most efficient way by combining what we already have under the same roof.

Okay. And the second would be, if you can somehow qualitatively update where you currently stand on the process of CPC to marketplace transition in the E-commerce? Do you see much room for the further seller additions that are still on the price comparison? And maybe also related to that, if you can somehow comment on the quality of the DBS model for consumers specifically versus the experience that consumers have on your own delivery from your own warehouse?

Slava, both are very good questions. I think the second question is actually, I would say, somewhat better than the first one. But let me start with the first one nonetheless. So look, where do we currently stand? I would say, this year, we added approximately 10,000 active merchants. And just to align on the definitions, when we say active merchants, this is somebody who has sold or who did at least one transaction through our marketplace in the past month. So, we added 10,000 of those, approximately two-thirds of those additions came from CPC. The rest are actually newcomers to our platform. We do see there is still room for growth, both kind of intensive and extensive as other merchants are doing the shift as merchants that originally didn't switch from CPC to CPA are coming back. And as so -- and then there is obviously the increase in sales in terms of GMV volume of those merchants that actually are on our platform. So we still think there is a significant potential in the conversion itself. Now when we talk -- the second question or the second part of your question, which relates to dropship by seller, you are right. Overall, like all else is equal, we do think that, that model is somewhat inferior to a situation where we fulfill the transaction and the delivery by ourselves. At the same time, as I mentioned before, this is something that allowed us to convert CPC to CPA easier and grow faster in that conversion. But the way we think about it, we mitigate that experience or the user experience by inserting the Yandex logistics last leg, if you will, because then the user does know exactly what’s happening with the delivery because we control that leg. At the same time, be focused on this and the way we think about it is also, if we look at the kind of the matrix of the categories versus the frequency versus the margin that we can earn by keeping those SKUs in our warehouses, et cetera. And I do think that some of – it’s probably correct to say that some of the goods that are currently being fulfilled by dropships by seller model will be moving to other models, fulfillment models where we will be taking more and more kind of involvement. That’s pretty much it.

Our next question comes from Luke Holbrook of Morgan Stanley.

I've got a couple of questions on the self-driving group. Just wondered if you have any -- if you can kind of provide more color on the plans that you have to accelerate maybe investment in this part of the business, given what competitors are doing in the space? And can you provide a bit of an update on the progress that you're making in terms of launching the Robotaxi service in Moscow by the end of the year? And also on the importance of the deal you struck with the Russian Post earlier this week on the Robo side?

This is Vadim. So, let me take this one. And let me start with the last one. So the kind of the deal that we announced with the Russian Post earlier this week, I would say this is something similar to what we did or done with Grubhub in the U.S., whereby we're doing the kind of the last segment delivery with our Rovers and therefore, replacing essentially the mailman, if you will. We are seeing kind of -- the way it will work, we're testing the different kind of user cases in this particular model. And we will see whether -- and we will see whether this is the pilot that we will start kind of convert it to a much, much more kind of a mass or larger project. Now with respect to the SDG or the self-driving group, the autonomous vehicles, we do think that our investments next year will be kind of somewhat similar this year was some modifications. We think, we don't need to kind of increase, in absolute terms, our investment as much as our competitors because historically, we've been extremely cost-efficient in achieving the same results as any other self-driving autonomous company. And therefore, just kind of -- while there might be some increase, it will be nothing compared to what others are spending. And then finally, I think it was the second part of your question with respect to the Robotaxi services in Moscow. We are already operating self-service, limited self service, self-driving taxi service in Moscow in one of the districts. It's -- will be kind of more officially launched a little bit later this fall. And the passengers will be able to kind of order a Robotaxi via our Yandex taxi app and travel between different pickup and drop off points. We think this is actually pretty cool, given that our technology is only four years old.

Great. And just a quick follow-up. Is there any update on your search for an OEM partner at this stage?

Look, this is a process that we’re kind of still evaluating and when we have something to report, we will.

Our next question comes from Maria Sukhanova of BCS.

I have two questions. So first one on revenue growth in Search and Portal segment. So of this 33%, you've mentioned the factors, but if you could put them into numbers, like the say the contribution of CPA strategies this new tools for SME and increase in share of iOS or for instance tell us what the growth would have been if you didn't have these extra factors? So that's first one. And second in car sharing with Delima competitor about the go public. There's stronger risk of that they do get more aggressive. So do you think -- do you plan to chase them a number of cars or you would rather like evolving gradually in this area? That's it from me.

This is Yevgeny. Let me start with your last question with drive. Just sort of give you a couple of words in our strategy overall. So with drive -- we've been successful in B2C, and it continues to be extremely important, and it's also very synergistic for the ride-hailing business, but it also turned out to be a great R&D, an idea platform and vehicle sharing, and that technology we can use it also in B2B space fleet management vehicle sharing. And we think this space besides B2C has huge opportunities for growth for drive business, especially with the rapidly evolving E-commerce delivery markets. And this business has significantly better vehicle utilization, which leads to better revenue per car and ultimately to better margins. So if we just look at the third quarter of this year, B2B was approximately 12% of drive revenues and 30% of EBITDA -- adjusted EBITDA. So -- and we continue to accelerate and grow in both areas of the business. Our EBITDA margin reached 11% of GMV in the third quarter, but I think it's important to know there is operating in financial leasing. So it's, I think, important to look at EBITDA margin post finance lease costs or kind of all in. And our EBITDA margin there was 7% of GMV accounting for sort of taking into financial lease costs if they were accounted in EBITDA. And operationally, the business continues to perform very well. We have 535 miles in September. And that's where the user base, we tend to have thresholds, which are higher for registration in the service versus our competitors. So it's 21 years old plus with at least two year driving experience, usually, these restrictions are more loose with our competitors. And if you look at our revenues, sort of the so-called incidental fees, which essentially finance paid by the users for various violations. There's half as much as they are with our closest competitor. So the business continues to perform very well. And we plan to both grow the B2C and the B2B segment of it.

So let me quickly add to what Yevgeny said. Look, we actually welcome that listing. We think that it would allow all of you to finally get a benchmarking car sharing and incorporate the relevant metrics into our sum of the parts for Yandex. Drive, which we actually think again, in our humble opinion, is a better business. But going back to your first question with respect to the kind of advertising on Search and Portal growth. Look, overall, it's very difficult to say the impact of any particular component or any one of the ad revenues because they're all interconnected. Probably some of the useful numbers here would be talking about our CPA, right, it’s the share, our CPA conversion strategies that the clients actually be seeing really accepting rather widely, it’s already 30% of our revenues today. I think it was in approximately 20% range in spring. We are seeing that the kind of on the numbers that we run together with the clients, the return on their marketing investments, it’s much higher than when compared to the CPC strategies. Overall similar kind of exercise for the Yandex business subscription for the smaller businesses. We are seeing a pretty high kind of reception or well, we see that kind of subscription model as well received by our customers because for the clients on a small budget, it is definitely more difficult to optimize our typical tools, such as Yandex. Direct, and that essentially be giving them kind of a tailored product just for them, whereby we actually optimize their return on investments, on marketing investment. And that’s pretty much it. I mean, it’s really difficult to kind of separate all the different factors and their impact.

Next question comes from Anna Kupriyanova of Gazprombank.

I will ask a couple of very quick follow-ups. When you discussed your logistics business, you gave us breakdown of B2C, B2B. But maybe you can also mention which part of your logistics revenues is coming from internal group deliveries? This is my first question. Second quick question will be on your Fintech update. You mentioned that you will be prepared like to start the Taxi operations next year, but maybe just for our understanding some comments where you are at this point? And finally, maybe you can say a couple of thoughts on your other businesses including HF or Cloud business which of them do you see most potential is maybe next driver of one, two year period? That's it.

It's Yevgeny. I'll quickly answer your question. Probably I would say, low double digits. I think that's the area we're working with.

Low double digits. It's a share of internal deliveries, right?

Correct. Yes

This is Vadim. So let me take the Fintech question. So it's essentially what we've been saying before, right, we are building the infrastructure to launch both the client-facing products as well as kind of the -- as well as the, kind of, call it, banking as a service to our internal services. And we'll be ready to report more, call it, next year, I would say, not early in the year. Other than that, we reported before that we launched a limited experiment with Buy Now Pay later, called Split. We talked in the beginning of September that we opened that product for 10% of Yandex market users. Since then we expanded it to 90% of the audience. And frankly, we're seeing quite encouraging results.

Okay. And on other businesses, if possible?

I'm sorry. Can you repeat the question?

So, my question was regarding your other bets and experiments, where you have Zen Cloud and some other things. Which of them do you see as the next potential may be driver of your revenues over the next two, three years' period, which develops is the best way, do you see most potential? Where would you like to focus your operations in future for example maybe it's has a next potential target for Yandex as a Group in 2021, 2022?

Well, honestly we like all our other bets and experiments because otherwise we wouldn't be doing them. But probably out of many favorite children, we would like to highlight our Cloud initiative. We are seeing extremely encouraging results. It's 3 times year-on-year revenue growth of this quarter and 25% quarter-on-quarter. We continue on-boarding new clients. We have approximately 30,000 active clients at the end of Q3, which is an increase of 60% year-over-year. We do believe there is a number of kind of key advantages, why we think we would take or be kind of the leading player in this market. And number one, it's 100% in top -- in-house technology platform, i.e., we are the only Cloud provider in Russia. Hyperscale was a fully-fledged scalable in-house infrastructure which combines our own data centers, our own hardware and our own software capabilities, we are locally present. And then finally, this is kind of interesting title. But I think we are the most certified Cloud in Russian market, i.e., we received all the required certification from different regulators and authorities that allows us to work with personal data of Russian clients and we can work with clients and financial industries and state on enterprises, et cetera.

That's great. And do you plan to actually develop your ad tech maybe to push some more effort into this area or you don't see it as priority for your business development at this stage?

Sorry, ad tech as an advertising technology?

No, sorry, it education?

Education. Well, look, we've been kind of saying I think for the past number of years that we see it as an important kind of national wide projects, important for the state. We believe there is -- we have a lot of in-house kind of knowledge and abilities to provide cutting-edge and top of the line education, whether it's going to be in mathematics or whether it's going to be in programing et cetera. And this is what we're doing and investing quite a lot in. And our practicum, which is the professional, or this is the online courses for people that are willing -- that are looking at changing professions, is gaining superb traction. And we think this is probably the best product on the market, and we are very excited about it.

Okay. But you don't plan any other investment specific like M&A in this segment, given its very good momentum for HS development currently in Russia?

As usual, we do not comment on M&A.

And ladies and gentlemen, that concludes our question-and-answer session. I'd like to hand the call back for any additional or closing remarks.

Yes. It’s Yulia Gerasimova, Head of Investor Relations. Thank you very much for all dialed in participants and all your questions. We hope that the answers were helpful. If anything that we haven’t discussed, please feel free to reach out the IR team. Thank you so much and goodbye.

Ladies and gentlemen, that concludes today's conference call. We thank you for your participation. You may now disconnect.