Good afternoon, and welcome to the AUTO1 Q2 2022 Earnings Call. I'm Philip Reicherstorfer, our Group Treasurer, and as always, I'm joined by Christian Bertermann, our Co-founder and CEO, and Markus Boser, our CFO. We will start with a brief presentation before opening up to Q&A. If you have questions, please use the Zoom question function and submit your question. I will then ask you to ask your question directly of Christian and Markus after the presentation. If you're on a bad line, please feel free to ask me to read out your question as well. Before we start, I must remind you that management will make forward-looking statements today. These statements reflect our current expectations and are subject to these safe harbor provisions. With that, over to Christian.
Good afternoon from Berlin, everyone, and welcome to the AUTO1 Group Q2 Earnings Call. Over the course of the last quarter, we executed well on our path to becoming the largest and most profitable car dealer in the Europe. We made significant progress and reached a number of key milestones despite operating in a very tough environment. We achieved the highest revenue in the company's history and a very strong level of gross profit, while significantly outgrowing the European used car market, which shrank in Q2. We have made unit economics, especially in retail, the top focus of the group. As a result, our retail GPU surpassed our merchant GPU for the first time, while we further improved customer satisfaction levels in parallel. Across the company, we have made our goal to reach adjusted EBITDA breakeven in Q4 of next year our number one target.
Turning to Q2 performance, we increased our revenue to a new record high of EUR 1.74 billion. That is 63% more than one year ago. At the same time, we increased our gross profit to EUR 126 million, up from EUR 99 million in Q2 2021. We continued our long-term revenue and gross profit track record despite the tough used car market environment in which we operate in. To update you on that European used car market development throughout the last quarter, you can see on this slide that Q2 marks an absolute new low for the amount of units sold within our core markets. We're defining core markets as all markets that we source cars in.
With around 5 million car transactions in those markets, volumes have declined by around 1.2 million cars, representing a 19% reduction year-over-year. In Germany, our largest market, only 1.4 million cars changed hands in the last quarter. That is 380,000 cars or 21% less than one year ago. That is even undercutting the COVID low of Q2 2020. We attribute historically high prices for used cars, muted new car production, and the general negative economic climate with the war in Ukraine as key drivers behind that development. Reflecting on that market situation, I'm incredibly proud of our teams for working relentlessly hard to achieve significant year-over-year growth rates on all key KPIs, countering those market trends.
We strongly believe that the European used car market will return to its proven long-term growth trajectory once the aforementioned challenges have passed and view the current market environment as exceptional. Being able to grow in such a tough market environment proves the strength of the platform and the value propositions we have built over the last 10 years. From a strategic point of view, we are excellently positioned being Europe's used car powerhouse to further leverage the huge opportunities we see across our markets. As such, we are uniting three key business models under one strategy, wholesale, retail, and finance. We invested significant time and resources to build our market position in wholesale and are the European leader in that business today.
In retail, we have been the fastest group company in Europe ever to achieve more than 50,000 cars delivered, but are now turning to a period of more disciplined growth while improving unit economics and general profitability in that segment greatly. Thirdly, we have made first promising steps into consumer finance and will transition the segment into a key profitability driver for the group over time. Mid to long term, we see finance not limited to consumers only, given the large pool of professional dealer partners that we work with every month. Taking a look at the U.S. market, there are three specialized companies covering those different segments with a total value surpassing EUR 30 billion by our estimates.
We aim to be as ambitious and successful as these U.S. market leaders are in each of their businesses and view our ability to create digital trade systems as the key to get there. We continue to operate under the same two strategic goals. Our first aim is to create outstanding customer experiences. We believe that unparalleled customer experiences are the core driver of our long-term continued success. Our second strategic goal is to leverage our platform to gain market share. The embedded leading data set, our automatic technology and pricing, and our efficient logistics network help us to add transactions to the platform even when the overall market is challenged in the short term. Since our IPO, we have consistently improved our customer satisfaction. Over the course of the last 12 months, unit, and in parallel, NPS growth have been the top goal for Autohero.
As you can see in the chart, we have constantly improved our NPS strongly while scaling volumes in Autohero substantially. This is one of the hardest things to do in growth businesses, improving customer satisfaction values while growing volumes strongly. I'm thankful and proud for the hard work our teams have put into this, and we regard these results as another key milestone for our retail business. Over the course of Q2, we achieved new highs for Autohero NPS with 74 in May and 73 in July. Those values demonstrate how important it is to us to consistently deliver world-class customer experiences. Our Trustpilot score of 4.6 across all markets confirms our internal NPS values from an external source. While we're turning our focus to unit economics in Autohero, it is our aim to keep NPS around the Q2 levels for now.
We strongly outgrew the European used car market over the course of the last 12 months. In the already described very challenging market environment, we purchased 161,000 units in Q2 of this year, consisting of 20,000 retail units and 141,000 merchant units at an average purchase price of EUR 9,160. This represents an increase of 12% more units year-over-year, while cars traded in our sourcing markets reduced by 19% in comparison to Q2 last year. As a result, our European share of cars traded, measured as the number of units sold, divided by the total amount of traded units in Europe, increased by 37% to 2.6%, showing the resiliency of the platform we have built.
As you know, we are operating under three core financial targets that represent our main focus areas. These are grow Merchant and Retail units, grow total gross profit, and three, execute our plan to Group profitability. In Q2, our Merchant business continued its solid growth track record, achieving strong results across all metrics. We have sold 150,400 units in our Merchant segment, representing 12% more units than in Q2 one year ago. Units in our remarketing business have grown from 17,900 in Q2 2021 to round about 23,000 units in Q2 of this year, representing an increase of 30% year-over-year. Revenue in Merchant grew 54% year-over-year to EUR 1.5 billion, and that is EUR 516 million more than one year ago.
Merchant gross profit came in at EUR 110 million, up from EUR 96 million the year before, representing a 14% increase. Turning to Autohero, we significantly improved its gross profit while growing units and revenue in parallel. Autohero delivered 15,800 units compared to 8,400 units a year ago. That is a year-over-year increase of 87%, and we recorded a retail revenue of EUR 270 million. That is more than twice what we had a year ago, where we recorded a retail revenue of EUR 114 million. As a result of our strong focus on improving retail gross profit, we increased retail gross profit five-fold to EUR 16.3 million, up from EUR 3.1 million a year ago.
Talking about GPU, we're very proud that with a GPU record of EUR 1,035, Autohero GPU surpassed the level of EUR 1,000 for the first time, growing EUR 672 year-over-year and representing the achievement of a key milestone for the group. Officially set out as a target at IPO for 2023, we reached the level 12 months earlier, underscoring our focus on Autohero unit economics and our overall profitability mindset. The EUR 1,000 mark is also that important to us because for the first time, a unit in Retail is generating more GPU than a unit in Merchant. The development clearly demonstrates our progress towards becoming not only the largest, but also the most profitable car dealer in the EU.
We see the strong development of our retail GPU also as an evidence that the large investments into the retail business over the course of the last twelve months are starting to pay off. Total gross profit for the group increased by 27% year-over-year to EUR 126 million, up from EUR 99 million in Q2 of last year, demonstrating the strong execution in the challenging used car market environment. Underlying this number is a strong and resilient merchant GPU coming in at EUR 732, which is pretty much in the middle of our expected medium-term range.
As already announced at IPO, we follow a clear path to our goal of adjusted EBITDA profitability by Q4 next year. In the following, I would like to give an update of key levers that we are addressing at the moment, which are leading the way towards achieving that goal. Firstly, we're speeding up our in-house production plans, given the high cost and lower reliability of our external refurbishment channel. Secondly, we will continue our track of substantially reducing marketing cost per unit. I will share more details about production and marketing costs in a second. On top of that, we're very focused on operating expenses in Autohero, including sales, operations, and logistics efficiencies. In logistics, we are focused on route efficiencies, combining units better to full truckloads and optimization of our network.
When it comes to Autohero GPU, we want to grow it further from the strong level we reached last quarter. Increased inventory turns, optimized inventory structures, auto pricing for retail, as well as higher financing and product attach rates, will enable us to reach those increased levels that we target for. On top of growing our Autohero GP base, we are aiming to continue to grow our merchant GP base. The target here is to definitely grow faster than the market while aiming for merchant GPU to be stable around current levels. Additionally, taking a look at headquarters, we will optimize overhead expenses by enhancing productivity across the different functions. Coming back to our focus on internal refurbishment, our internal facilities are making first progress in reducing total production costs.
If we take Germany as an example, production costs, including rent for our external units, was around EUR 1,350 in June of this year. Our internal production cost per unit of around EUR 900 in our workshop of Kitzingen is already today EUR 450 cheaper. Our target cost for that workshop is EUR 800 per unit until the end of the year. In a similar way, we are defining cost per workshop for all workshops across the EU with the goal of significantly reducing refurbishment cost per unit over the next four quarters. From the beginning of this year, we ramped up the in-house production share from 6%- 21% in July, measured as the percentage of units refurbished in one of our own facilities out of all refurbished units in a given month.
Heading into the same direction, we're aiming at in-house shares of 90% over the next couple of quarters. Taking a look at our in-house production footprint across Europe, I'm happy to say that all of our already announced production centers in Germany, Poland, and Spain are now in production. As you might have seen, we announced two new production centers in Italy 48 hours ago, and we are expecting them to start production in Q4 of this year. Together with the Italian centers, our total production capacity at full staffing stands at 113,800 units. In-house production remains one of the largest drivers to decrease our overall cost per unit in Autohero. It has our top attention. Now, turning to marketing cost per unit. Marketing cost per unit is another significant lever next to production costs.
After we have invested substantial amount of funds and resources into building up the Autohero brand across Europe, we are now able to benefit from those investments. We're able to reduce our marketing per unit by 32% over the course of two quarters from EUR 2,800 - EUR 1,900. Lately, again, in a further remarkable step to EUR 1,100 per car, which is another 42% reduction in June compared with the cost for Q2. We carried out those optimization while maintaining strong customer demand. There's a number of reasons making these levels of optimization possible in that short amount of time. Firstly, our teams have years of experiences optimizing marketing cost per unit from our experience in our C2B business unit. Secondly, we use our brand-building funds efficiently, the results of which you can see on the right.
We created Autohero as a brand with an extremely positive brand image driven by our strong focus on customer experiences when delivering. Our brand image is excellently perceived on every single dimension by customers. They perceive Autohero as strongly convenient, engaging, enticing, exciting, likable, safe, transparent, and trustworthy. Lastly, our conversion optimization and inventory management teams have done a great job in increasing conversion over the last couple of months, and will continue to do so. From a higher level point of view, our profitable merchant business is providing us with the headroom we need to optimize unit economics in Autohero.
We aim to match the investment into Autohero with the cash flow we generate in our merchant business by Q4 next year, giving us the opportunity to continue our level of investment in Autohero while we are developing a unique set of competitive advantages in retail that will power our long-term and by then self-funding growth. With that last remark, let me hand over to Markus, who will give you a more detailed financial overview.
Thank you, Christian. Our Q2 results reflect our continued ability to grow despite challenging market conditions. While the lack of new cars, combined with overall consumer wariness currently.
Continues to weigh on the markets in which we operate. We nonetheless were able to bring the merchant business to grow over 50% year-on-year, maintaining already strong GPUs while also reducing inventory by EUR 67 million. Most notably, we achieved a retail GPU of EUR 1,035 per car, one year ahead of our stated plan of a EUR 1,000 GPU in 2023. Together with the merchant GPU, our margin focus enabled us to achieve a total gross profit of EUR 126 million in Q2, 27% above the same period last year, and also an increase quarter-on-quarter. OPEX remained broadly flat relative to the last few quarters, driven by improvements in marketing efficiency on the one hand, but offset by an increase in employee costs.
This is a temporary increase, primarily driven by our push to insource refurbishment as we transition from predominantly external to predominantly internal refurbishment, and we expect employee payroll to decline as efficiency initiatives discussed by Christian start to take effect. We continue to maintain a strong balance sheet with no corporate debt, ending the quarter with EUR 640 million of cash before any unused inventory ABS commitments, resulting in a cash decline of only EUR 14 million. This reflected, on the one hand, positive cash flow from the sell-down of inventory, as well as the refinancing of a significant proportion of our consumer loans, offsetting our quarterly losses and CapEx investments, as well as the repayment of a portion of refinancing ABS.
Regarding inventory, as we discussed during our Q1 call, our inventory was impacted by the outbreak of the Ukraine war at the end of February, with an increase of inventory at the end of Q1. Despite the ongoing war impacting Eastern Europe and consumer sentiment, we were able to reduce our inventory by 9% and have also reduced the average value in our inventory. We think this is a testament to the strength of our inventory platform and the ability to de-risk by reacting quickly to market changes, something that we have proved during COVID, and before that Dieselgate, and demonstrates the power of our pricing technology and distribution power as the largest wholesaler in Europe.
Lastly, consumer loans, which we sell in Germany and Austria directly, increased by EUR 32 million - EUR 106 million, despite an increase in interest rates, which we instituted towards the end of Q2 to cover the current interest rate volatility. Finally, to our guidance. When we issued our guidance in the third week of March of this year, there was a lot of uncertainty as the Ukraine war had broken out four weeks earlier and the overall consumer situation was unclear. As a result, we provided a very broad range at the time, reflecting a range of outcomes, with the normalization of the situation at the top end and significant worsening at the bottom end. While there continues to be a lot of uncertainty, halfway through the year, we can now narrow our ranges.
As a result of ongoing high ASPs, we are increasing our revenue guidance to EUR 6 billion-EUR 7 billion from EUR 5.7 billion-EUR 6.8 billion, and maintain our gross profit guidance of EUR 470 million-EUR 580 million. We believe that at the moment we will likely end up just below the midpoint of the range. We maintain our EBITDA guidance of -2% to -3% provided at the beginning of the year in our annual report, and for convenience, have provided an expected range of between minus EUR 150 million and minus EUR 180 million for the full year.
With respect to units, we narrow our group unit range to 655,000-725,000 units from 650,000-770,000 units. We believe merchant units will land between 590,000-650,000 units, narrowed from 580,000-680,000 units. For Autohero units, as a result of our increased focus on GPU, shifting production in-house, and overall improving unit economics, we have decided to lower our sales targets so that we will end the year with between 65,000-75,000 Autohero units for full year 2022 versus our expectation in March of 70,000-90,000 units.
We continue to believe in our long-term unit targets stated at IPO of 600,000-700,000 retail units by 2030, and believe we will scale faster again when we have achieved more favorable unit economics. With that, I'd like to open up to Q&A.
Hi, everyone. I am Raphael, your Zoom operator. Before we get started, we would like to review a few technical items to make sure that you can interact with us today. At the bottom of your Zoom window, you will find three buttons: Audio Settings, Q&A, and Chat. Audio Settings. Clicking Audio Settings will bring up the audio preferences for this webinar. Please make sure that the most appropriate audio device is selected here. As a viewer in this session, your microphone will remain muted, as will your video. Q&A. Please submit your questions via Q&A icon. Click the Q&A icon and a window will appear where you may submit your questions at any time. Once received, Philip will moderate the questions and ask the author to address management live.
To that end, I will open the line for you, following which you will need to unmute yourself before you will be able to speak. We will address as many questions as possible live during today's session, but may respond to some questions offline after the event.
Thank you, Raphael. We will start with James Tate from Goldman Sachs. James, I think you had four questions. Maybe if you just ask them in turns rather than trying to have us remember them all.
James, your line is now open.
Hi, good afternoon. It's actually Lisa Yang from Goldman. I'll just have three. The first question is on the outlook for Autohero units and the relationship between the units sold and unit purchased. I think historically you said you were targeting 80%-90% of your unit purchase to be converted into units sold. I think in H1 you were a bit below that at probably 70%-75%. But I think to hit the midpoint of your full year guidance of 70,000, obviously we need to see a bit of acceleration. I'm just wondering what gives you that confidence that this conversion from basically unit purchase into units sold would improve into H2?
Any, you know, color behind that would be helpful. That's the first question.
Yeah. Hi, Lisa. Maybe I take this question. Indeed, is this like a key relation, the number of units purchased versus the number of units sold? The main driver behind that is the structure of inventory and how fast we're turning it. If we're turning inventory not fast enough, it will be sold off via the merchant business as a return. We have made good progress over the course of the last two quarters to speed that inventory turns up and to reduce the amount of units that we return as wholesale units into the merchant business. That gives us the confidence that if we continue this trend, then we should get the two numbers much closer together.
Okay, thank you. The second question is on your guidance. I mean, clearly, the retail, the revenue per unit continues to suppress the upside. You're raising your revenue target. Any reason why we shouldn't see a better gross profit or just maybe that number for the full year as well? Like, you know, surprised to see like, you know, maybe a lack of operating leverage here, or maybe I'm missing something. Could you maybe talk about the rationale here?
Yeah. I think the relationship between revenue and GPU potential is not linear. Obviously we're targeting for a certain amount of absolute euros per unit. Yes, if ASPs are increasing, then there's like a slight increase potential for gross profit per unit, but it's not linear. If like revenue is increasing 10%, your gross profit will not increase 10%. To that extent, it comes down to a little bit of what we are expecting for the market to do over the second half of the year. There are factors that might come in positively, like new cars expected to go up, and that could result in used car prices to be stable or also go down, create more liquidity on the market itself.
Therefore we are not thinking at the moment that this trend of increased ASPs that we have seen over the last two years will continue. This is, yeah, there's really like a high level of uncertainty around projecting what will happen with the market.
Okay, great. Yeah, I just wanted to check whether you're seeing any sort of pressure on costs anywhere in the business. It doesn't look like that's the case. Even my last question, we've seen obviously a few developments on the competitive landscape. I know you don't operate in U.K., but we saw Cazoo, I think, you know, going to administration. I think Cazoo announced also a review of their European assets. Wondering, you know, have you seen any other major change in terms of competitive dynamics, and at what point that could potentially benefit, you know, your financial or your market share?
In terms of the Cazoo's assets in Europe, like any asset that you think could fit your portfolio? Anything you might be interested in? Thank you.
Yeah. I mean, if you look at the numbers that we talked about, I mean, I think we're benefiting already today from the market and competitive situation. Having now AUTO1 standing at a 2.6% market share versus last year, 1.9%. I think this is like a very strong achievement given the market condition and just shows that we are operating differently than the one or the other competitor. Yeah, regarding to your last point, I think, you know, I think we have all the ingredients that we need to realize our vision of Autohero. That is not only the largest but also the most profitable car retailer in Europe. Yeah, therefore I think we are very well equipped with everything that we need.
Great. Thank you.
Thanks, Lisa. We actually got a question from Nizla Naizer from Deutsche Bank, turning to the merchant business. How is the health of your dealer customer base? Are some markets more resilient than others in the current environment? Please provide some color.
Yeah. Yeah, it's a good question. Overall, our dealer base is stable, but of course, they're also feeling the reduced amount of units traded across Europe. I think if you are looking at the total dealer base, then the health indicators that we are looking at are broadly in line with where we want them to be. Obviously there's additional pressure in our Eastern European dealer market base, which are just impacted by more than I think other parts of Europe by high ASPs inflation and then also being that close to the war, general economic situation, refugees and so on. I think if there's any color, then I think our Eastern European network after the war is under more pressure than, for instance, our Southern European network.
Okay, great. With that, let's go over to Will Packer from BNP. I think, Will, you had two questions.
Will, your line is now open.
Hi there. Sorry. I have three questions. Maybe you kind of touched on one of them already, but it'd be great to just have a quick refresher. Firstly, could you just remind us of the retail GPU drivers, and could you confirm the proportion you view as sustainable and underlying versus cyclical? Secondly, or maybe fire away on that, and then I'll come back.
Thank you. Yeah. Retail GPU, I mean, I think main drivers are increased, or optimization of the inventory structure. If you compare, for instance, the inventory that we assembled for our retail customers in Q4 with the inventory that we assembled in Q1, and then the inventory that we assembled in Q2, this is just fitting far better. Based on the data and based on the supports that we see per unit, per cluster, per segment, we have optimized pricing, we have optimized the assortment. That, I would say, is one of the main drivers behind the GPU increase. On top of that, we are very focused on external refurbishment costs, have optimized them further, but I think it's not the main driver. I would say it's like the number two driver.
All of that GPU improvement is sustainable because we are trading units at market price. Even if ASPs are coming down, there might be, to my earlier comments, a slight pressure maybe on the absolute GPU, but then on the other hand you are getting faster inventory turns because your overall inventory is just priced less. We have the ambition, as we also said in that presentation, to grow it from here, but we think that for the next two quarters there will be not so much growth. There will be more and higher levels of growth kicking in at the beginning of next year.
Thanks. I suppose moving on to the EBITDA margin. The GPU, as we just discussed, increased significantly. Marketing cost per unit fell, but the margin deteriorated year-over-year. Why is that? Is it mix or other areas of investment?
I mean, the margin for a used car dealer is always something that if you have a stable ASP environment, I think you can look at it. If you have an ASP environment where since two years, you know, prices are increasing, depending on the category, between 60%-120%, so a 2013 Vauxhall Corsa, 80,000 km will now be EUR 8,000 or EUR 9,000. Used to be like kind of EUR 3,500. So if you see that, then the percentage EBITDA will or the EBIT will naturally go down. I think when ASPs are normalizing, then you will see the reverse effect. What is really important is that the P&L per unit works, right?
That the GPU and then also the unit costs are coming in line over the next couple of quarters, and that's what we're working on. We're really looking at absolute euro levels, on the one hand on intake, so car trading margin and also gross profit per unit. And then on the other hand, total unit cost per unit in Autohero, but also in the rest of the business. And those two just have to match.
Maybe Will, I think just as it relates to the EBITDA specifically. I mean, the key differences between our OPEX from a year ago to our OPEX today is of course on the payroll on the one side and marketing costs, absolute marketing costs for Autohero on the other side. We've obviously been building up the business significantly since, I mean, a year ago we had just finished the IPO, and I think we're putting, you know, that investment to work. I think where we're now is, you know, we've continued to invest in across tech and across retail production in particular.
Now, you know, having created a product that has the NPS scores that we have, the fast delivery, all the other items that go into a great product, now looking to optimize that more versus a year ago.
Thanks. That's helpful. I suppose leads naturally on to my final question, which is we've got guidance for the margin to improve by the end of the back end of next year. Could you remind us of the drivers of that improvement? Further GPU gains, improved efficiency or a mixture of both? Thanks.
Yeah. It's definitely on both sides. We actually think that unit cost is by now the even more important driver. By the way, I mean, that was exactly where we wanted to be. We wanted to be in a spot where we can see all the data, where we have businesses in Autohero across all the markets that we wanted to be in. Now that we have a scale which is significant. By now, we are definitely among the top 10 retailers of Europe, not online retailers, but real, like, retailers, also offline.
Now there is a phase of optimization coming, which is just, I think, a very good and healthy thing for the business to do, because we can grow this business a lot bigger, but we wanna grow it at favorable unit economics. Or in other words, step by step, this Autohero itself should become profitable, depending on how successful we are, and how quick we are with optimizing unit costs. We think that until the end of next year, there is like, a 50%-60% lever to gain on the Autohero GPU, which is mainly coming from growing the in-house refurbishment share to 90%.
Other than that, we are very focused on optimizing the marketing cost per unit further, which are already, like, at a pretty good level, given that there's, like, a lot of branding funds in those investments as well, but especially then also refurbishment, logistics and then also cost of sales, customer service efficiencies as well. At the moment, I think that the cost per unit level is even larger than the GPU improvement level.
I'm sorry, just to confirm, did you suggest that GPU could be 1,500 by the end of 2023 or did I mishear?
Yeah. I think 50%-60% from current levels toward the end of the year, I think is something that is definitely an aim for the group.
Thanks, Christian. I appreciate it.
Thanks, Will. Over to Andrew Ross from Barclays.
Hi, Andrew. Your line is now open.
Hi there. Can you hear me okay?
Yes.
Yeah.
Good afternoon, everyone. I've got two more, if that's okay. The first one is to come back on the marketing cost per unit you gave in that helpful slide for June and ask you basically about the sustainability of that. I don't know to what degree there's any kind of seasonality that impacts June versus other months. There's no sports, for example. Do we kinda think that that actually could continue to improve from here at that June level or not? That would be helpful or at least to have a sense.
Yeah.
of where the marketing cost per unit may land, let's say, for the second half of this year and into next year.
Yeah. Our ambition is, of course, to continue to decrease. While I think the level of improvement that we have seen here over the last three to four months has been much faster than we actually had planned it and anticipated it. You're right in the comment that June is a favorable month when it comes to the share of sponsorships that are in that overall budget. Also if we see latest trends for July, this is like a route where we feel confident. It will be a little bit harder per your comment when all the sports sponsorships on a monthly budgeting are kicking in again. Nevertheless, our target is to continuously reduce it now quarter-over-quarter. Yeah.
I don't think we can expect it to improve at the same rate on an absolute base. We are really happy with the outcome that we see from the latest optimization because the number of leads and customer demand that we're seeing has actually gone up despite those cuts.
Just to be clear, you're saying you can improve it each quarter relative to the Q2 level or relative to the June level?
Definitely to the Q2 level, and we aim to also be better than June.
That's helpful. The other one is just a comment to follow up on Will's question there around the GPU on Autohero into next year. Can you just give us a sense as to kind of the profile of the ramp and how we think about the GPU in, let's say, Q3, Q4 this year? Is it gonna be front-end loaded, back-end loaded, linear? How to kind of think about it on a quarter-on-quarter basis from here.
I think, having achieved the EUR 1,000 Autohero GPU effectively a year earlier, I don't think we're gonna see dramatic increases over the course of this year, as we're really pushing hard to, you know, increase the penetration of the in-house refurbishment. Therefore, I think we should expect to see the ramp to the increase that Christian was talking about, the 50%-60% happening more towards next year rather than this year.
Okay, cool. That's helpful. Maybe if I could just follow up with a third one, if that's okay. I think at the IPO, you talked about the number of Autohero cars more than quadrupling in 2023 against 2021. Is that still an assumption you're comfortable with? Or maybe looking a bit aggressive, what's your best guess in terms of how to think about growth into next year for Autohero whilst balancing that GPU improvement?
Yeah. We're not giving specific guidance to 2023. I think which is why in my prepared remarks we continue to see this being a massive growth opportunity and reiterated the 2030 target of 600,000-700,000 units. Having said that, I think our near-term focus is very much going to be around improving our unit economics, and then we'll give more guidance towards the end of this year as we see the progress that we've made towards that and the ability then to really scale much more going forward.
Cool. Thank you both.
Thanks, Andrew. We've got Adam Berlin from UBS.
Adam, your line is now open.
Hi. Good afternoon, everyone. My first question was, can you give us any color on trading so far in Q3? Should we be expecting another two on two improvement in Autohero units in Q3?
I think broadly we're very happy with the progress. Clearly, you know, if you look at the midpoint of our revised unit range, that would imply an uptick in the second half of the year.
Thanks. You're seeing that coming through, so that's good. The second question I wanted to ask was, you've given a target to reach EBITDA breakeven by Q4 2023. Do you have a date in mind for when you'll get to cash flow breakeven?
We've always stated that we would hit cash flow breakeven one year thereafter.
Q4 2024?
Yeah.
Great. Thank you very much.
Sure.
Okay. With that, over to Catherine O'Neill from Citi.
Catherine, your line is now open.
Great. Thank you. Actually, quite a few of my questions have been answered. The one I just wanted to get a bit more detail on, or a couple actually, was if you do start to see a meaningful decline in used car pricing, the latter half of this year or 2023, if we hit a recession, and certainly at Citi, I think our economists are expecting quite a deep recession for Germany, how confident are you on sustaining that GPU, both, I guess, Autohero and the merchant side? The other question I had was on the cash flow. I think you said, you know, the cash burn in Q2 is pretty limited 'cause of some specific factors.
Can you just give us an idea of whether any of those factors will continue into the second half, or were they quite specific to the quarter?
We would actually welcome used car prices to come down, because it provides additional liquidity, and it would unlock a lot of pent-up demand for customers right now that are not available, or not able to afford, certain used cars. I think it would also strengthen the position of our Eastern European dealer network again, because for them, the absolute euro level is also just a high barrier out of the perspective of their customers. Let me first say, like, we will welcome that little bit of normalization. I think it would help our overall unit trajectory. It would create here and there pressure on the GPUs, but not to an extent that we would find worrying.
With respect to your second question, we'll be providing much or full disclosure on the balance sheet when we provide the full half-year report on the fourteenth of September. Specifically, there was one particular one-off, being the consumer finance ABS, which contributed to that. Though of course, having achieved or signed the consumer finance ABS, we expect that it will, of course, continue to finance the growth of consumer loans.
Great. Thank you.
Thanks, Catherine. With that, last but not least, Sherri Malek from RBC.
Sherri, your line is now open.
Hi. Thank you very much. You've touched on both my questions. I guess I'm gonna try and get out some additional color. The first one is around the breakdown of Autohero GPU, and if you could provide any more color. I know you said refurbishment cost isn't a main driver right now of the improvement, but last disclosed it was about EUR 900 per unit. I was wondering how that's developed since then. If you could provide any color on the financing income piece as well, that would be helpful. I'll just weigh in for the second question.
I think the EUR 900 was something that was at the IPO in Q3 of that year of 2020, when I think we had significantly fewer units than we have today. As we've ramped up, I think certainly in Q&A, I think we've been reasonably consistent with the numbers that we showed here in the presentation of around, you know, EUR 1,300 or so, or I should say EUR 1,100-EUR 1,200 in COGS. I think the overall breakdown we don't provide, although I think I can say that in financing, GPU is probably, you know, around about the EUR 150 or so, a little bit north of that overall.
That's really helpful. Thank you. My second question was on used car pricing. Again, this was touched on, but just curious in terms of what are your thoughts going into next year and beyond how much of the stronger pricing can be maintained?
Yeah. I mean, as I said, Sherri, I think there's, like, a lot of uncertainty. If we see used car prices, or if we see new costs coming up and we're getting, like, clear signs that at least OEMs are trying to book more logistics, trying to rent spaces and so on, so that would mean that ASPs begin to taper off or start normalizing towards the end of the year, and that might then continue also, throughout the rest of the year. I would also, on the other hand, see that units are going back to a more normalized level, so going closer back to the 2019 levels that we had. This means, like, more affordable used cars.
This will return us, like, then into an environment that we have been operating on already in 2019. Therefore, yeah, I think we're waiting for this to happen. On the other hand, in the meantime, we have taken a lot of improvements on the GPU itself when it comes to merchant on the GPU, also for Autohero, as you have seen. We are remaining confident on that midterm guidance that we gave on the merchant GPU, and as well as the indication that we just gave for the Autohero GPU, irrespective of the ASP development for now.
Okay. Thank you.
Thank you, Sherri. Thank you everybody for dialing into the earnings call. Thank you, Christian and Markus, for your time. Hopefully we'll see most, if not all of you, at some of the conferences in September we are attending or otherwise, at the next earnings call on the second of November. Thank you.
Great. Philip, everyone, thank you very much. Looking forward to next time.
Thank you.
Cheers. Bye-bye.
Bye.