Welcome to Carasent Q2 Report for 2025. For the first part of the conference call, the participants will be in listen only mode. During the questions and answers session, participants are able to ask questions by dialing pound key five on their telephone keypad. Now I will hand the conference over to CEO Daniel Öhman and CFO Svein Martin Bjørnstad. Please go ahead.
Thank you and good morning to our report for the second quarter. We'll start with me giving a company update and then Martin will give a financial update. Looking at the highlights of the second quarter, two days ago we came out with revised targets. The performance is a bit behind plan and I will get back to that. That will be the major part of my presentation. Other updates: we now have our own functional ambient AI- listening functionality built into three of our products. That means that the AI is listening to the meeting between the patient and the doctor and automatically proposes a medical record. We have that in three products: Webdoc, Webdoc X, and Metodika. It's only Webdoc that's selling it now, with roughly 30 users live at the moment. It's going forward. We continue to develop functionality and I will get back also to this.
What I think is the major highlight of the second quarter is that we sold the first primary care centers in VGR since 2018. It's in total 8 clinics, which gave us an ARR of roughly SEK 3 million with the start of January 2026. I think it's fun always to sell to new customers. It's a pretty big customer, or it's not new, they have Webdoc in other clinics already, but to win new contracts is always fun. I think what's the most important is the sign that our customers in Västra Götalandsregionen are less worried than they used to be about Millennium and that they actually go ahead with the change of systems now. I think it's a very positive sign and we have under good discussions with primary care center CDR, but we're very happy for this contract with Medt anken.
It will mean a bit of work this autumn to get them live and then they will start becoming a paying customer 1st of January next year. We have also given out new medium targets also two days ago. I will get back also to them a bit later in my presentation. Finally, with the rapidly improving cash flow we have, and also since our acquisition pipeline is very targeted, we know exactly what we want to acquire. We don't have any other discussions, we have a few discussions and we'll take them in an orderly pace. We have too much cash on hand. That means that we're now starting and initiating the buyback program of up to SEK 150 million until the next AGM, which I think I look forward to this process. Looking at growth, signed not implemented ARR now amounts to SEK 12 million.
A large part of this is Volvat and Medtanken. Also, we had 30% organic recurring revenue growth and 25% total recurring revenue growth in the quarter. Profitability continued to improve even if we're a bit behind plan this quarter and the first half of this year. I get that by going to the main reasons for the revised targets for this year. If you're looking at this for Ad Opus, Metodika, and our AI investments, starting with Ad Opus, we so far this year are SEK 3 million behind plan. We have had a long period of underperformance and high churn in this business unit. It's our smallest business unit. Most of these SEK 3 million we are behind on, it should be said, is cost that we take for the restructuring of the organization. That means costs for letting people go.
Most of those SEK 3 million we didn't plan, and Ad Opus is doing a great year. We're doing a reorganization where Ad Opus and Ad Curis are combined into one unit. Ad Curis has done a great job with sales, done a great job with support and delivery to customers, and I think that Ad Opus can learn a lot from that. They also share a lot of software already, so it's more efficient to be in one organization. It's a way of saving cash. I strongly believe in Ad Opus still. The reason I do that, even though they actually are shrinking by 25% in the second quarter, which is part of the restructuring, is that we are launching the new system Ad Opus Web, which is a completely newly built system replacing the old system.
It has taken some time, but now we have been interviewing a lot of customers within Ad Opus, and the largest one, Free Optics, is now really, really happy with the product, and they're giving out testimonials saying how good the product is. You can see them on LinkedIn and also other social media, and I think Ad Opus Web is a great product. We have to learn how to sell it, we have to learn how to implement it in a good way to make it easy for customers. Once we do, we will be able to really sell. What this reorganization does is set us in a cost structure that gives us time to succeed with Ad Opus Web, and that's the main reason. What that means is that the coming 12 months we will not try to do any new sales of Ad Opus Web.
We just try to move all customers from old Ad Opus to new Ad Opus Web. When we have happy customers in that solution, we can start selling to new customers, and then we have all these present customers talking well about a new product. It's very difficult to do both at the same time. We have to accept that, lower the pace a bit in that organization, lower costs so we can take time to move customers over to the new product and then sell to new customers. I think it would be a really beneficial structure to be in the same structure as Ad Curis. That means that the head of Ad Opus has had to leave the company and also another employee in that part of the organization. Looking at Metodika, Metodika is what I see as our enterprise solution.
It's for really big caregivers that use the same system across multiple clinics and hospitals. Those types of caregivers usually want some adaptions. If you're looking at Webdoc, which is for single clinics or groups of clinics, we don't do any adaptions at all. In Metodika, a large caregiver like Volvat, which is for all Swedish speakers, it's Tapio, Norway, they're implementing into all of their units. We're replacing roughly five different digital systems. You want as a customer to be able to have some adaptions for your business, and that's what we're doing at the moment. For Volvat, I think we had a bit of mission creep. They've been wanting a bit more function than we had planned for originally, such as a patient portal, which will drive profitability in the future.
It's increased the amount of work now and it has taken more work than we planned to get everything ready. For Volvat, it also has meant that we have not had time to work with present customers that we used to. We have a big backlog on change requests from present customers and that's what we usually get quite a lot of revenue from. In total, we have SEK 4 million banked in the first half of the year. Looking from another perspective, our consultancy revenue in total is SEK 4 million behind what it was two years ago, even though we have more employees, just because everyone is putting all the time into Volvat. On the other hand, most parts of Volvat will go live by the end of this last quarter this year and the beginning of next year.
We start getting our revenues from the customers and that from Volvat. That will be SEK 7 million-SEK 9 million of new revenue just for the system and for different add-ons, but there's no consultancy work in those numbers. We should also have the other SEK 4 million that we're now not getting or work for other customers. It's a quite big swing of somewhere around SEK 11 million- 13 million that we should see due next year from that product alone. This is a part of the cost when we move from selling licenses, doing consultancy work to actually selling subscriptions instead. It hurts us short term, but we win long term on it. As most of you know, the majority of our products have always been subscription based, but majority has not. The final major reason is AI investments. We are, so to speak, playing around with technology.
For quite some time we have been looking at what other players do and we became very convinced of the potential for AI to lessen the administrative burden of our customers, which is the big burden which they really don't like to spend the time on. They spend a lot of time on administration. AI is really good at helping with that. We also believe strongly that to build inter systems is a great way of doing it because then you can access all the information around the patient when you do it. Not only listening to the conversation between the patient and the doctor, we can also use all the other information that are in the medical records and in the calendar and so on. We can really do a lot to burden our customers. I also think it's a technology that we shouldn't say no to.
We should invest in also that. I think that's where a lot, potentially a lot of the future lies. We decided in the beginning of the year to invest more into AI and to invest in partly new competence into the organization. I'm not very stressed about it. I think it's great potential going forward and we have no need to be first. We have the HR position. That's a difficult one. You don't need that easily as customer, you tend to stay there. As long as we can, we can be competitive. It's easy for customers to move from another competing solution to our solution. That being said, I think that we're also selling competitors' solutions for this type of support and we continue to sell also our competitors or partners, tend to put it that way, solutions.
I think for some customers one solution will be right and for other customers another solution will be the right, but I think for very many of our customers our solution will be the most appropriate. It will be completely built in, no need to work in double systems and the AI will have access to much more information than in an external system. Today we have the ambient listening, so listening on the discussion between the patient and doctor. Next we will start filling out different types of forms, referrals, sick notes and so on to really unburden our customers. We still have quite a lot of work. For example, this week we are releasing new functionality which will cut the time it takes to propose the record by roughly a third, I think. Quite a big improvement for customers that are already live coming this week.
These three are the areas which are the main reason for the revised targets. If you look at Volvat, it's really something positive for us going forward. It's of course negative that we're spending a bit more costs on it than we planned and also that it stops us from generating other types of revenue this year, but it will pass. At Ad Opus, a big part of this is a restructuring cost for moving into a more lean organization which will help us going forward and I think will be really good going forward. I think it was worth investing a bit more in this [GNRMAP] plan originally. Those are the main reasons for revised targets which now look like this. You see the revenue has been slightly lowered and the EBITDA and CapEx is a bit higher, lower by roughly SEK 10 million.
Looking into our new medium term targets, we have great products. I'm really proud of the products we have and what I think is even stronger is that we invest quite a lot in our products and we invest in the products to increase the distance between us and competitors. We said to invest more than most of our competitors and that means that it will really help driving growth going forward and it should and will really allow us to grow with an average of 15% the coming three years and together with good cost discipline, this will allow us to continue to rapidly improve profitability, reaching an EBITDA margin of 35% with the CapEx of, yeah, less than 10%.
These are our new medium targets which, yeah, we believe strongly in and I think also it's wise to guide on this horizontal when we are in this type of rapid transition from where we used to be. With those words I will hand over to Svein Martin.
Thank you. Starting off by looking at the financial highlights. As discussed, we are a bit behind the plan in Q2. In general, I would say the underlying financials continue to develop in the right direction. Our ARR grew by 27% to SEK 317 million including the backlog of signed not implemented contracts. The reported recurring revenues grew 13% organically in the quarter and net retention of 109%. As we have talked a lot about during the last year, the pace going into Q2 was not that strong. Given this churn, we have seen in Webdoc for the customers going bankruptcy. Most of that, as we spoke about last quarter, took effect at the end of Q1. This affected the pace going into Q2. The positive is that the pace going out of Q2 is stronger, which I will get back to.
On the profitability side, our EBITDA margin was 15% and EBITDA minus CapEx 3%. Improvements compared to last year, but a bit behind our targets. This slide shows a key metric for us. That is how we are able to convert our revenues into profits. Here we exclude the acquisition we did last year. Next to that, the revenue growth was around SEK 8 million and then COGS naturally increased around SEK 1 million. You see that the operating expense decreased by SEK 4 million compared to last year. Part of this is that last year was high driven by a one-off cost of SEK 2.5 million. We see that the big increase here on the cost side is the personnel expense.
It should be looked at in connection with the CapEx decrease because that's like the total cash cost for our employee base, those two combined, and that increased like around SEK 6 million. The reason for this is mainly the biggest part of that increase is this restructuring cost that we had in Norway and also the option program that we executed in Q2. In total, this cost around close to SEK 4 million. Just for this, the increase is not that big on the personnel expense side. Looking at the total effects, we converted 56% of the revenues into EBITDAC. If we had adjusted for the one-offs last year and the special costs we had in the quarter, then we would have been close to our target but still a bit behind. Our internal target on this is converting around 80%.
We have taken more steps on the cost side in this quarter, removed a few more roles in addition to the Ad Opus employees that we didn't see the need for. This will help us in H2 and we need to be around our target there of 80% conversion to reach the 2025 goals, the new goals. Looking at the ARR, we grew from SEK 252 million to SEK 317 million. On the organic side, we grew by SEK 36 million in upsell or 16%. Churn was SEK 9 million and new customer growth was SEK 12 million. The two main factors driving the growth were that we onboarded several new large customers and clinics in Norway for Ad Curis, in addition to a strong upsell for that product as well in the quarter.
We added around SEK 7 million in ARR towards the end of this quarter for Ad Curis alone, which is really strong for that product, partly this Frelsesarmeen main contract that we have talked about. The second effect is that we adjusted the prices for several smaller Webdoc customers that were not paying enough compared to their production, so that added a couple of million in ARR as well. On the churn side, it's a bit higher than normal for reasons we have discussed with this churning Webdoc related to the bankruptcies and also a high churn in Ad Opus as we have talked about before. In total, the net effect is that we end the quarter at an ARR growth organically at 17%, which is decent compared to what we report in the quarter. Looking at the P&L.
I t alked about most of this already, but we landed at SEK 83 million in revenues, up 26%. Gross profit margin remains stable at 85%. The big increase on the cost side is the personnel expense. Despite that, EBITDA increased from SEK 8 million to SEK 12 million within the quarter, and the EBITDA minus CapEx from -SEK 1 million to -SEK 2 million on an adjusted basis. So far this year, we have improved from -SEK 6 million in EBITDAC to +SEK 7 million. A good improvement, but still a bit behind what we had expected. On the cash flow side, we are of course helped by these improvements in profitability and lower CapEx as well. If you look at the cash flow in the quarter, it was in total positive by a couple of million for the first half of the year.
I've talked about this in the past few quarters, but we had these relisting expenses that were the costs taken in Q4, but the invoices were paid in Q1. That affects the working capital for this year. We also had a bit of currency effects when we exchanged the cash from NOK to SEK in Q1. Normally, our working capital helps us support the cash flow, but this year it will be a bit affected by these relisting expenses. That was what we were going through, and we can open up for a Q& A.
If you wish to ask a question, please dial on your telephone keypad to enter the queue. If you wish to withdraw your question, please dial pound key six on your telephone keypad. The next question comes from Elvin Rolder from DNB Carnegie. Please go ahead.
Hello Daniel and Svein Martin. I hope you're well and good morning and thank you for keeping us busy throughout this week. I have a couple of questions from my end, maybe beginning a little bit with the restructuring of Ad Opus, the merger with Ad Curis. When you look at the current organizational setup of the group, are you happy with what you have now given these changes you have made, or do you see any obvious further improvements down the line that you're looking at? How many people roughly in total did you have to let go from doing initiatives in the quarter?
Yeah. Thank you, Elvin. Restructuring of Ad Opus in general, we run quite a decentralized organization because we think the most important is to be close to our customers. That's for a number of reasons. The first one is to develop solutions that actually fit our customers. The second is that the sales and support should really, really understand their customers. In many parts, it's a decentralized organization, but in some areas we really have some clear synergies, and looking at Ad Opus and activities, they share the same software. The new Ad Opus with Ad Opus Web is actually the same software behind the scenes as our online visit tools for Ad Curis. They're more and more working in the same platform, and it makes sense to be the same organization. Also, Ad Curis is doing a great job with sales and delivery.
They've done a really, really good quarter, and as Martin mentioned in his part, we ended the quarter really strongly on new customers going online, and that's thanks a lot to the Ad Curis organization doing a good job. The Ad Opus organization, I think, can learn a lot from Ad Curis, and we gain from being part of that organization in that sense, and it also allows us to save costs. In general, I think the decentralized structure is quite good. We will always look over the organization and do what we think at each point in time for the long term, and there might be that, especially with the acquisitions we're looking at, that they might fit well together with something we already have and therefore will be combined with present structure.
For example, in Ad Opus there were only two people that go, it was the Head of IT and another person. It's our smallest business unit. It's quite a small organization, but it's always been more expensive than to let the Head or CEO of companies go. Also, we have had other roles that we have let go in the entire group in the quarter to also lower cost. The numbers that are in those SEK 3 million is only those two together with how much Ad Opus is behind channel so far this year. Did that answer your question, Elvin?
Yes, yes, more than that. Thank you. Maybe looking a bit on Medt anken and finding a customer here in Västra Götaland, very positive to see of course that customers are pulling the trigger. Was there anything specifically that made Medt anken go ahead right now? I mean they're not pretty much of a stalemate in the region so far. Has there been any changes that made them go ahead right now? You mentioned that you have several good discussions with other clients there. Do you see a similar kind of development as you've had with Medt anken with those customers? If you comment a little bit more on that.
Yeah. Medtanken is quite a big caregiver. I think they totally have the turnover for bit one being a CEC and they have a lot of primary care centers in Västra Götaland. For historical reasons they've had Webdoc in roughly half of the clinics and a CDM software in the other half of the clinics. I think that they've seen that over time the difference between those two systems has increased. Since 2018 until now there's been quite big improvements in Webdoc in the functionality and what we offer. I think they assume that that is now quite a big difference. I think that's why they pulled the trigger, so to speak, and changed. They will not be new sales in our column. They will be in growth in existing customers. Unfortunately, they sales when we look at the numbers in Q1 next year.
I think the reason is that they see both softwares and they see the difference. I think that's quite a big reason. The other one is that everything coming from VGR, it seems very unlikely that any system will be pushed out anytime soon. I think that all of the caregivers see that and we have discussions with other primary care centers. We have no present discussions with a primary caregiver of the size of Medtanken switching all clinics. It will be more one clinic at a time in that case. I think it's a really positive effect and positive signal to everyone. I will try to create some media around it after the summer to try to influence the politicians locally in the region.
Okay, great. Thank you so much. Regarding the ambient AI- listening and the AI tools, I just want to clarify when we're talking about AP users. I mean, I can equal that to, I guess, 80 physicians, so it can be several in the same clinic. I guess if you can clarify on that. Roughly, how much is the revenue per user, would you say? If you can give some flavor on that. Finally, on AI, the 30 users you have live now, have these been using the other tools that you have partnered with, or are these completely new to this kind of tool?
Yeah, yeah. So it's 30 users that's live. Yeah. If I was on [Kyona] for users, and that's individuals. Yeah, so that's correct. At stage we are, we started focusing really on this product in the beginning of this year. So we have half -a- year of, I mean we're investing in it, but we're not investing very heavily in it. I think we still have some work to catch up to competitors when it comes to the quality of all parts of it. We have advantages fully built in and so on, but we're rapidly catching up. I think for example, this week we will cut the time it takes to wait by, it will become a third of what it was last week and we will probably be the fastest in the market.
I think delivering the actual result after the discussion with the patient, that's extremely important for the physician that when you say goodbye to the patient that you don't have to wait too long to have a generated note. These kind of improvements that we need to do, we'll do them. I think we have a clear roadmap on how to improve and how to move forward and add more functionality. As I said before, I'm not stressed. We don't need to be first. I think it's quite good from our position to look a bit what works for our customers and what doesn't work and then we build the things that work. There's been so many investments in health tech that has not been picked up by the customers, even if they bring good initiatives. What we have seen is that this really fits the users.
Many users like it. That's why we started investing in it. How they pay is that we charge today, we charge SEK 750 per user. We're cheaper than the competition. We still have really good minds on that. The reason why we can be cheap is that we're not running in the American clouds, which I think have other advantages too. We're running it in our own server environment and that means that we have quite high fixed cost for the AI-driven solutions and the cost you already are taking for it in terms of server capacity and GPUs. We can scale to at least 500 users without a problem, so probably quite far beyond that multiple costs. There are some cost increases, but that's where we are at the moment. We are in early phases, positive feedback.
We have some improvements we need to do and we will stop any users. The ones that are live now, many of them have tested the different solutions out there. It's still early adopters that are using this type of technology, I would say, and it's still a minority, a clear minority of all users in our system that use any such solutions. I think going forward it's likely that there will be more and more uses of this technology as it gets better and better and also as it spreads. Healthcare is really conservative. It's slow in moving into new technology, but many are big fans and they really like it. Over time I think it will really be a major point. In our plans it's still a cost entire year. It's not something that drives a lot of profitability this or next year.
Did that answer your question, Elvin?
Yeah, very clear. Very good answers. Just one clarification. SEK 750 per user, is that per month, I guess, or is.
Yeah, that's per month, per month and use it. It's.
Yeah, okay, good.
You can say it depends on how much you use Webdoc, but it's a little bit less than what most you pay for the entire Webdoc. If you compare the amount of time spent on building this compared to the amount of time spent building Webdoc, it has really good revenue per hour invested.
Good. You answered a lot of my questions during the call, but I have a couple of questions on the financial target here because you lowered the EBITDAC target. Looking at the H2 level, it doesn't seem like it's a big difference really from how it was before. It's mostly been from what has now occurred here in the first half of 2025. How confident are you in the current 2025 target? I guess you have better line of sight now than when you set out the initial target for more than a year ago, and is there some uncertainty still regarding the 2025 EBITDA target given the implementation of Volvat that can maybe delay that, or how should we think about that?
Yeah, compared to our own plan, we still have. There are some effects in the second half too. As I've said for quite some time, we felt that the targets were very little space left in them. I think I said it already in the end of last year that I was a bit worried about targets, and now we have of course much better visibility. It's not a walk in the park together, so to speak. We still have to succeed in what we're doing. It's, so to speak, our best guess of where we'll end up. I think we're on a good track to get there. We've taken a lot of measures in the first half that will help us in the second half. We'll always do a good Q3. We have the vacation effects there.
I think the autumn, I mean obviously if you look at the numbers, the autumn has to be really good. I expect autumn to be really good, as you say. There is of course unrest at all, but we're moving from six different HR systems into one. I mean, Millennium [NVIDIA] already didn't import any data from any old systems, and that's quite complex because you have the same patients in many of those systems and each system saves data in different structures, and they continue to meet patients all the time. Only 500 employees of that organization, so it's not the easiest process in the world. We've done it before in Metodika. We know how to do it. We moved entirely Norway into Metodika a couple of years ago. I think we have great experience in doing it. Obviously, we underestimated the amount of adoption work needed.
It also took longer to get access to our competitors' databases that we need in order to export the data, but we have that now so that there is a bit of risk involved, I would say. We expect to manage and reach these targets. It's not a walk in the park, but we should be able to hit that.
Okay, thank you. I should let some other people ask a question as well, but I have one more. It's a bit on the new medium term target. It's that average 15% organic growth between 2026 and 2028. I'm just wondering because that feels kind of like the organic growth that the business grows on a normalized basis given your high retention, the normalized low churn and so on. How much in that figure do you bake in from this kind of, as I see it, at least four kind of organic growth optionalities that you have with Stockholm opening up, with Germany expansion, with this AI tool, and perhaps even the surgical module? How should we think in, how do you think when you bake those four factors into this target? If you can comment a little bit on that.
Yeah, it's obviously difficult looking that far into the future. I think we will have great opportunities. If you look at Stockholm, you look at AI, you look at Germany, I think we have really great opportunities. On the other hand, things always happen as we see in this quarter and I think that the serious includes that things happen and it's dangerous to put out three years targets based on everything we see as an opportunity then. Being in my position, obviously I believe very strong in everything we do. I could see us growing much faster also. I think the sensible is to put out targets that you feel very comfortable about and these are financial targets.
Then you have other targets that we're trying to drive the business faster than that and then you have internal toys on what you want to accomplish, what you want to do. I think when you go about financial targets you need to be a bit prudent and I think those targets are a bit prudent and especially given what we just have done two days ago. It's the first profit warning I think I ever have given as a stock, as a listed company. I'm not too happy about it. I think that it makes sense to put out targets which, I mean, also this is not a walk in the park. The opportunities are much larger than that. You have to find a sense per level to communicate and make sure that the expectations out there are reasonable. That's what we try to do.
It's not our business plan that we've given out as financial targets.
Yeah, fully understand. Thank you. That is very clear. Thank you so much for taking all my questions. I'll get back in line and see if there's any other analysts.
The next question comes from Frederik Neilsen from Redeye. Please go ahead.
Thank you. Good morning, Daniel and Svein Martin. I want to start with the primary care deal in Västra Götaland. I mean, is this really the start of perhaps the positive case that you talked about last year, discussing the potential churn and so on starting to play out, or is that reading in too much in a first new customer?
I think it's reading a bit too much into it. I think it shows that customers are not that nervous about being forced into something very soon. We have not yet gotten to a stage where they see it as risk free and that's where we would obviously like to end up. We're pushing different ways of getting there. The court process is still ongoing. I think there's a lot of other good reasons why it will be free. We know that some political parties are now pushing that also and the caregivers of course are pushing it quite hard. On the other hand, the region doesn't want to be guilty. Private caregivers a lot of advantage over their own business. That's the difficult part. I think that sense will prevail, but it's not given and we're not there yet.
I think that it means what you should read into, I think, is that our customers are less worse than before. They are and that we are now quite a lot better than competition and that's why they changed despite it still being a bit of risk because you really don't want to change the HR systems unless you have to or you don't have to but unless there are really good reasons to do it. I think they obviously see those reasons and that's really positive. I would more see it as Webdoc doing a good job and that the risk is a bit less in the region but we're not there yet in the really positive case now.
Okay, great. Could you elaborate a bit on the mix between the old and new product in Ad Opus so we could get some kind of understanding when the negative trend might turn?
Yeah, so Ad Opus Web is, I mean it's a completely new product. It's really forward thinking. They don't have patients but I use that word anyway now to make it clear what I'm talking about. They have the caregiver and the patient working in the exact same system. I mean all of our systems have different types of patient portals, both our own and third-party solutions where you can interact between the clinician and the patient or the clinic and the patient and hospital and patient and so on. Here they actually work in the same system and I think that's really groundbreaking. I've never seen that before. Also, the UX is really thought out and it has a lot of optimization that really helps the caregivers. The challenge is that it's a bit too new.
The move from all that over to Ad Opus Web is, the step is really big and you can become much more efficient. If you look at the [three Attacks video], they talk about it and they have put in effort into implementing it and really making it work for them and then they are much more efficient than before. That's great. Not all customers want to do that journey. I think that every time we develop solutions, they should always allow for you to work in your old ways. They should allow for new ways of working, but also old ways because customers don't always want to change how they work and they don't want to change just because we tell them to change. That's something we'll work a bit with the system to make it also allow for the old workflows to function.
It will be less efficient to work it that way, but many customers will want to work in their old ways. That's something we're doing. It's not really development, it's more setting up the system. That's something we can do fairly quickly. We have roughly three customers, one of them being the largest, that are fully into the new solution. We have many others who have started testing it. That's also something I want to lower the pace on. I want us to take online like maybe three, four customers at a time, not more, to make sure that we do it really well, we take care of them and they are happy with the process.
If we try to take too many at the same time to the new system and it's, you know, everything looks different and if we're not close enough to that, then we will have churn, even if the new one is good. They would appreciate if they got the right education, if they got us really helping them every step along the way. That's what we need to do. That's what we do great in Webdoc and Ad Curis and all that and many of our other products. We've been trying to push it a bit too quickly. Actually, in Ad Opus, I would say trying to move too many customers at once.
I interviewed some of the ones that left us to make quite clear that the problem is we tried to move them over a bit too early, but also that we hadn't just learned how to move customers into the new system. I think we're getting there and it's really good to see the first customer being happy. It's just the last two months they've been prepared to actually be what the English word that they are out there talking positively about the system and prepared to recommend it to other users. We just about got there and the coming 12 months we will not see any growth in the system. I hope we don't see any more churn. There will be some churn that's not yet in the numbers. What really made us do the organization now was that we lost the second largest customer.
That will mean a churn in Q1 next year. Not a huge amount of churn. I mean, still, Ad Opus is our smallest business area, so we shouldn't put too much time into that or shouldn't guide us too much. I think that it has the potential to really grow. If you're looking at Ad Opus, since I think it was acquired back in 2018, I think it was in Amazon and then it went into Carasent a bit later. They've gone from 90 customers to 30 customers. That's because they've been building. We've been building a new solution to replace the old one and that has taken time, but now it's there and I think we'll start winning back those customers going forward and we have a much better solution than anything else out there.
I think maybe it was the wrong choice back many years ago to build such a completely new system to replace the old one. I think you could have done it more gradually, but that's in the past. There were many right decisions taken too. That's why I think we have this kind of review curve where we've been losing customers because the old one has not gotten enough love and the new one has not been ready. Now we're getting there and the churn we actually had now was leaving for another old system that's actually very similar to the one we're replacing. Just because they didn't feel comfortable enough going to our new system because we were not close enough for them, and we tried to move them a bit too early. We are slowing the pace. I'm sure we'll get people there and the entire robust organization.
I feel like lots of engine. I think they like being product curious. It's a positive vibe. The development we do is done in Moldavia, so we have low, low cost in what we're building, and it's, I think in many ways, a groundbreaking system actually. I look forward to that. It will take some time.
Great, thanks. That's clear. Regarding ambient listening products in general, there seem to be mixed feedback. Reading some news articles about larger caregivers implementing those, could you give us an update on the customer feedback regarding Medsum?
Can I ask you, Frederick, is that the report in GP about Sardinians just implementing that you're referring to?
Yeah, that's one of the examples.
Yeah, yeah. I think it's important to note that your pronounced implementation of what they call AI and which has been criticized in media by the doctors is not AI as we see it. It's old speech to text. It's Dragon, they call it. They call it AI. I don't know why. It's a Dragon Nuance product, Dragon, which is an old speech to text solution that we've been selling to our customers for at least five years. Many of our customers use it. It is a good speech to text solution, but it's just a speech to text solution, and that's what VGR and Solinsky is trying to implement. The reason I think why they get this mixed feedback is that they also take away their medical scribes, the people, completely, so the doctors do not get any support anymore with, you know, writing letters.
They move them into a completely separate building. I think that actually reasons why the doctors are upset about the new tool. I don't think it's the tool because that tool that they are selling we have sold to our customers for at least five years. I don't know exactly when that started. We have very many happy users in that. Correct. That's what we're replacing. What [sorting scale] is going live with now is what we're replacing with our AI solutions. The ones that have been criticized is not ambient AI- listening, as far as I've seen in [Västra Götalandsregionen or VGR].
Great, that's clear. Also, regarding Medsum, according to your website, you will shortly launch Medsum broadly. Could you give us some flavor on that? I mean, what time period are we talking about? Approximately, yeah.
We are selling it now. We're selling it. I've had the process where we send out information about it to all our customers. Then I just took Webdoc. The other ones have not done it yet. What we want to reach is those early doctors that are interested in being part of the journey, so to speak, that like the technology, are open to give feedback, they're open to that. There might be not everything working perfectly yet. They are happy to just be part of that process and many, many doctors, nurses, physiotherapists and so on. They like to be part of development. They are close to them, to our development people and get to be part of discussion. They get to influence what we build.
What we got done is to let everyone know that it's coming soon so that they know and they don't sign up for other products long term. We've said that if you are interested in being early, early users, let us know. We have a list of I don't remember how many that has contacted us saying that I'm interested. Out of those we have taken 30 live and we'll continue to take users live. Some of them are paying, some are not paying yet because we moved from just giving it for free to pilots to now charging users. Not all 30 are paying, but all new ones we're taking on will be paying users and all will be it eventually. We have no hurry to make them start paying.
The most important is to make them really happy with the solution and then they will be happy to pay. That's where we're at.
Great. That's all for me. Thank you very much.
There are no more phone questions at this time, so I hand the conference back to the speakers for any written questions and closing comments.
Okay, we have gotten quite a lot of questions in the chat as well. We can try to be quick in addressing them. The first ones are from [Niklas]. It's in Swedish, so I'll try to translate here. Development for Volvat has taken more time than expected. Are you getting paid for just development, going for better pay for development going forward, or do you see risk of delays further into 2026?
Yeah. Usually, historically, Metodika has been selling licenses, not renting them, and then charging for development. We're changing that around to renting it and having focus on ARR instead and the long-term value creation. That meant in Volvat that we halved the cost per hour for all deductions than what we usually take. We still run break even per hour, but we're not making a profit on the adoptions for them. Also, I mean we are the one with the best view of how many hours it should take. Not all of it, but quite a lot of it. When it gets more work than needed, we have to take the cost. It depends a bit, but in general, in most cases, we are taking the cost. That means we're actually losing money on the adoptions. As I mentioned earlier, we'll have a really big swing next year.
It's a really complex, big project. I think we're on a good track to deliver it to most of the users in November, which is the time that's set. There are some small systems that will move over in the beginning of next year. I think we're on good track to get there. It's not without risks, but it's our best guess at the moment.
Next one. When do you expect Germany to be a substantial growth driver? As we had talked about before, we will aim to have start paying users within Webdoc X this year and then it will of course be a gradual step up in terms of growth. The growth rates in percentage will be higher than the rest of the group. It will not contribute a lot to the group organic growth the next couple of years. Towards the end of the new target period we could see that Webdoc is materially contributing basically. Next one is regarding the AI solution. I think we have addressed that. The question is how big part of your current user base is within the sweet spot for Medsum?
I think basically as I see it over time all users that do a lot of documentation are in the sweet spot. In the beginning it's mostly doctors, all doctors. I think it will add to. More and more. Users and I mean we, as I see it, we will. For example, if you want to do, say you want to do a referral, you can just, you might not have to say anything at all and you don't. You might not have to have the recording from the patient visit. You can push a button and we'll take data from different parts of the system and write the proposal. I think that it will become, I think I said before, roughly half of all our users do notes in any larger extent. I think that's where the future users will be. It will take time. Healthcare is really slow at note technologies, but this has been one of those that's been growing the fastest I would say of new technologies. It's a real time saver and that's what everyone's looking for.
One, and next one is can you sell it standalone, and how do you resonate around that?
Yeah, maybe we do not have an interface to melt some. I mean it's completely built into different systems, and I think that's the advantage we have and also the access to a lot of data in our systems. It is a standalone that we're working to of our systems, so potentially we could sell it to HR providers that are not ourselves. That's not something we're taking decision on and not something we have in our plans. That's something we have thought a bit about.
Next couple of questions from Richard. Do you do any NPS with customers, and if so, have you considered publishing them?
We don't today have it in a format that's the same across our products. In general, I think that each product we have should focus on getting a feedback structure that fits them and not what fits us centrally or if it's external parties. I still think it's worth thinking about. We don't yet have one structure for all products. Also, something I would like to have implemented actually in all our system is that it uses every month. Get the question of how was your experience with the system last month? Because then we get a much more real-time data on how they like our systems and how they experience, if you have a period of time, maybe some issues with the system, how they really experience it and so on. We're not there yet. We have not been able to prioritize that development yet.
I think we get there, but it's not a prioritized question for us.
Right, next one. Excluding Ad Opus, what do the churn levels look like? Are you happy with the levels? Yeah, so close to 1% of the churn, so it would be like approximately around 3%. It's higher than we have had the last few years, and I think most of it is related to these bankruptcies and basically churn that it's difficult for us to impact, basically involuntary churn. If we look at the voluntary churn, the customers leaving us because they changed the system is still very, very low, below 1%, so that level is good. Of course, the churn that is similarly hurts us this year and we.
We don't really have a good explanation why we had a lot of that. To be honest, it's a bit surprising. I've been working in the private healthcare sector for a long time and I haven't seen this much bankruptcies, and there's nothing real in the market that I see that would drive it. I think part of it is that some of the tech startups are not getting funding anymore. That's a part of the shift that we can say. There's even more than that, and I think it might also be a bit adjusted. It happened to be a couple of them at the same time.
Okay, next one is relating to primary care and VGR. I think we covered it quite well, but there's one question regarding the interest from primary care centers if they will need two systems, Millennium or yours, or are they simply not believing in the rollout of Millennium at all?
I would say that if you change systems now to Webdoc as this customer's done, I don't want to put too many words in their mouth, so to speak, but I would say it's because we believe that it will not roll out and that there's nothing else being forced on them at least for five, six years. I would say otherwise you don't do that change. It's not that you're planning for it being better when you have two systems that you might have to have for a few of them, but you wouldn't change systems for that scenario.
Great. Next one from [Erasmus]. It would be interesting now how you build up the 15% average growth. Given the business current growth rate in the low teens, do you expect that to decrease as AI and Webdoc X drive the growth or what are the different growth drivers you see? As I mentioned, I think Webdoc will build up gradually. It will be a very small part of the group during this forecast period. Basically, we see that the 15% is mainly driven by the current business we have and tools such as AI will drive the net retention naturally. In general, the core business, selling to new customers and selling all types of functionality including AI, will be the main growth driver as we see it.
Yeah, we have improved sales quite a lot. We're selling more and more. We have this high churn, but we're also selling much more. This salesperson is selling more and actually we're selling on and a bit about plan so far this year. New sales are quite good and sales to existing customers are quite good. I think we can get better at both. This is more of a, it's much more of a top down target than a bottom up because the bottom up can end up at crazy places.
Yeah, I would say that Webdoc will be a very important driver of the growth, which will grow probably above these targets, and then the rest of the business a bit lower. Can you provide an update on the operations, the surgery module?
Yeah. We have, I think it's now two pilots running it. We have two new clinics, quite large clinics, starting after summer, still existing customers, and actually three after summer. They are all paying customers. They will not pay from day one, but they will be paying quite quickly and at a quite good level. We will now start to really target the surgery clinic that are not Webdoc users. Today it takes time to convince a clinic of 50- 100 users to change software. From the point that they start thinking about it after meeting and we're showing them things, I think it generally takes probably a year before they actually commit to changing. It's a very slow process. We're doing the best we can to really push it. We have really good feedback from the users of the system.
We still have some functionality to make it even better. I think this period will be a really important part of the growth for a clinic, especially Stockholm. If you have insurance patients, which many of them have, and all of them have it in, the strategy goes to improve on that. If you have surgery, then Webdoc is clearly a better choice than what's out there for you, even more than other cases. If you also want to have different types of online tools connected directly into the system, then it's really clear that it's the best by far, I would say. That's really the sweet spot to grow. It's also when we talk about big customers and quite long processes.
Next one. It would be interesting to hear your thoughts about Skane now that the smoke from VGR's launch of Millennium has settled down somewhat.
Yeah. Skane are saying that they will implement Millennium still, I think. I always thought that Skåne is a much better project around Millennium and they have a better version of it too, I would say. I mean, both regions have spent since 2018 roughly to actually do adoptions and changes in the system. What VGR went live with, extremely many users, was something that was not even tested in a pilot. To me, that's really strange. We could never ever do something like that. Skåne has taken a much more sensible approach and they had test versions and demo versions up and running way before VGR, and they still haven't complied with any in the clinic, I think, so very much better. What Skåne has done is that they say now that Millennium will replace much fewer systems than they planned from the beginning.
I think that's quite a good process for us. There might be, and there seems to be, openings that Webdoc could be used for certain parts of care in Stockholm that in Skane that we had not planned for. It's actually not still in our plans, but there seems to be some openings that are positive. I can't say more about it yet. I think that it's not bad for us if Millennium eventually goes live in VGR and it doesn't have to be. For the hospitals, they don't have that much to choose from. Why not go live with Millennium? There's nothing really, really good for them to choose. Let the primary care use Webdoc. I think that's a sensible choice, or let the primary care choose between all the systems that are available for primary care. We'll see if we get to that point.
That's not what's in our plans, not in our targets, but it's something we're working towards.
Next one, a couple of questions from [me and Nick. You, Daniel Öhman], could you comment on the development in Stockholm? Do you see clients shifting away from CGM already or not yet? Secondly, have you seen private clinics complaining about CGM software given their decision not to invest further? Do you hear more about bugs, etc.
To answer second, first, we heard from customers that they don't really understand how that software is going to live until the end of the decade as it's in the dance. It's not too much happening there and that's understandable. If we had a system that we knew was going to be closed down in five years, I think we would invest very little in it. It makes sense. The difference between that system and ours will increase over time even more. Quite a lot of our new sales are in Stockholm. It's a clear focus for us and we're working hard to make sure that everyone knows about us. I think for every new customer we get to Stockholm, more and more get to know about the system, more and more talk about it and it helps us in the new segments.
Next one, how is the development in Germany? Could you comment on that?
Yeah, I think the data business is running really well since the acquisition. Actually, a bit better than we had planned for. We're winning customers. We have quite good results in the business. We have now combined the R&D departments of the three products we have on German markets to one organization so that we can healthcare presentation, and obviously we have to maintain the softwares we have there already in the market. We have to do some development for the data software that's going to live for quite some many years still for some of the users. We focus mostly, of course, on Webdoc X. Whereas before, we have the aim to get paying customers now during the second half of the year. It will not be a big part of the revenues or growth in the autumn.
It's anyway extremely important to start getting paying customers and start getting customers that can talk to other customers about our products.
Next question from [Christopher Gastberg]. In regards to the profit warning, when did you realize an update before Q2 was necessary, and what is the main learning from this process for the management?
It's difficult to say an exact point in time when you realize I've been saying on this call and on other calls that there's little space. There's less space in toys than I would like to compared to our own budget. There should be. There should always be a bit of positive difference. I think the main learning is our medium targets are a bit more. They have more space in it. I think that's an important part. There is more space for things not going as planned. I think that's probably the main learning to make sure to have that type of space. From our perspective, there are so many opportunities, so much good we can do, so much value we can add for our customers that it's very easy to see all those opportunities.
I think that to have some space in the targets and what you actually see as doable is an important learning.
Next, one rise target for EBITDA means CapEx is lower significantly from SEK 44 million- SEK 49 million to $35 million. Could you elaborate a bit more about this? I think, yeah, we had talked a lot about it in the call, but the main reasons here is that the result in H1 is a bit behind what we expected, and that is partly driven by this extra cost we have talked about, but also this development for Metodika and Volvat and the AI investments, and that those will naturally continue throughout the year.
Yeah, just to mention there's been a lot of questions today and I have to go with an [Intuit Dockers industry] here, but Svein Martin can I see the next couple of questions he can answer? I'm sorry for being a bit double- booked. Usually this course doesn't take this long and maybe I've given a bit too long answers. I have to go for that interview. Thank you for me so far. Martin will answer the final questions. Thank you.
Yeah, we don't have a lot more questions, but a couple. So SEK 750 for the AI solutions, is that for ambient AI- listening only? The AI for referrals is a separate product to that, which can help expand the SEK 750 going forward. That's for the base product, the ambient AI- listening, and then when we add functionality, it has potential to either increase the revenue base or reach more users, depending a bit on how we position it. Definitely potential as we add more functionality. Next one, can you give an update on the surgery module and when the market potential of SEK 150 million that you have previously communicated is plausible? Daniel mentioned this in quite detail, but the market is still there, the product is close to being final.
We think that this will be a very important growth driver for Webdoc in the coming years, particularly these types of clinics and also the Stockholm market with these types of clinics. Final question we have gotten here from [Victor Bellak]. Can you give an update on Germany, and to clarify, selective M&A would be primary in Sweden if it happens? The update on Germany, Daniel mentioned on the M&A side. I would say the selective M&A that we're looking at now, it's not a lot of companies, it's a handful. All of it would basically be to strengthen our position in Sweden as we see now, given that we believe that we have enough to work on in Germany at the moment with data and web and getting that commercially viable. We don't really have capacity there to do more M&A at the moment.
In a couple years' time, it might be an opportunity if we are able to succeed. At the moment, we exclusively focus on the local market, basically. With that, that was all the questions we had. Thanks a lot to all for dialing in, and please reach out if you have additional questions to us. Thank you.