I'm Charles MacBain. I'm the CEO of Nordhealth, and I'm joined by my colleague Alex Cram, our CFO. We'll go through four different topics today, similar to the previous presentation, and end with a Q&A. We'll start with a general company update, then do a slight deep dive into the Veterinary BU, then the Therapy BU, and Alex will end the presentation with a financial update. Please leave all the questions for the end. I can either answer them if you add them to the chat, or you can just raise your hand to be able to ask questions. Thank you. Starting with the company updates, I always like to take a historical view so we can see where we came from. There's a lot of fluctuations year- over- year. Over the last six years, we've actually been able to grow the business at a CAGR of 49%.
As you can see, the majority of this growth has actually come from organic growth and some from M&A, but we have not made M&A acquisitions in 2025 so far, 2024, or 2023. We are still looking though for opportunities from time to time if we can find some that fit our criteria and the right price. If we look at the business year- over- year ending Q2 2025, you can see that our ARR overall only grew 9.5%. It's important to note that this 9.7% excludes the Vets4Pets and AmeriVet post-pilot rollouts. If we break down the growth and where it came from, at the end of June 2024, our ARR stood at 38.6%. We added EUR 1.8 million new ARR from new customers. Our current customers also spent EUR 3.6 million more with us, and we had a churn of - 1.6%.
Important to note is that this 9.2% net upsell is primarily driven by proven expansion within existing enterprise clients. This is not a big rollout that we've seen in 2024, and we'll see in the latter parts of 2025 and 2026. You can also see the churn rate of 4.2%, which is quite a low churn rate. The story here is that in 2024, H1, we had the CVS rollout, and in H1 2025, we've been very much working on ensuring a successful pilot of our two next big rollouts, which are AmeriVet and Vets4Pets, which will take place in H2 2025 and 2026 as well. If we go now to the next slide, here we can see the breakdown of this growth over time, not just a snapshot in time, but within a historical context.
The main points here are that our growth has slowed down in LTM Q2 2025 relative to previous years due to, first, as I mentioned in the previous slide, the large enterprise rollouts being slated for H2 2025 and 2026 versus the last 12 months. The second part to understand is that we are also driving to acquire new customers, and it's quite a healthy LTV/CAC that we've got ratio. This is especially driven by our low churn rate. If you look in the next slide, we can see that in the last 12 months, we've added EUR 2.5 million in net new spending for R&D and CAC investments. However, that's been offset 1.5%. That has been offset by improved revenue, so minus the COGS and customer service.
The additional revenue that we brought on board has offset EUR 1.5 million of that cost, resulting in a net change in EBITDA minus CapEx of EUR -1 million. Now, we'll talk more about the additional investments in the Veterinary and Therapy BU deep dives, but the two big buckets that we're investing in are, one, versus what we had predicted, is we want to be more aggressive in our localization for the DACH region for our Veterinary business units. The DACH region is the largest veterinary market in Europe, and we also want to be able to accelerate the migration of our veterinary clients on the Provet Cloud. The second is we see a massive opportunity in our potential to develop and upsell all AI features to our current customers and making this a unique value proposition for new customers which are coming from us from legacy systems.
We see that as a big potential for a catalyst for change. If we think through one of the biggest issues with growing in this business, which is the fact that the churn is very low. It's very low for us, but it's also very low for our competitors. A lot of these competitors have been in legacy software, server-based software, hosted software, and they cannot, as fast as we can, develop these new features. We see that our ability to be at the forefront here will enable us to convince a greater proportion than historically customers to migrate over. It's a reason to switch our source. Now let's go into the Veterinary update. A couple of key business topics. One, the big news is AmeriVet, which is a 200+ US corporate chain, location corporate chain, approved the pilots and full migration. This is a big, big deal for us.
It's our first flagship US customer that's a corporate, and I'm very happy with the team and our progress on this one. We're expecting that the full migration will happen in H2 and 2026 over time. It's important to note that the ARR from this decision is not included in our signed ARR that we previously put as the approval happened at the end of Q2. You will see this in the Q3 signed ARR numbers. The second is we actually did get a second smaller US enterprise customer, and we're currently rolling them out called PetVet365. We've rolled out as of the end of H1, 6 of 27 locations, and that rollout is continuing.
Third is we'll talk about it and do a little bit more of a deep dive in the next slides, but we've launched our first Provet AI features, which are our Scribe, Discharge Notes, and Patient Summaries. We'll go through and show you those in the next few slides. The pilot started in August. 44 vets have already signed up even before the product was out of beta. There's a very, very strong demand for these features. Fourth is we made a decision to accelerate the veterinary migration through investments in private localizations for the DACH market, accelerating those investments relative to what we had expected before. Fifth, we were able to sign EUR 1.9 million new customer ARR. Lastly, we've migrated 109 clinics to date, and we've actually successfully sunset Provet Net, which was a platform in Finland.
They've all migrated over to Provet Cloud, so we can now focus more of our development resources on our core product. Now, let me show you a bit about these new AI features that we've launched. Let me put you in the world of a veterinary clinic. In a vet clinic, especially a first opinion clinic, these appointments are 15 minutes- 20 minutes. The history, so the patient history can be pages and pages long with lots of PDFs, lots of attachments, especially if they have referral letters and so on. The vet has minutes, if not tens of seconds, to be able to go through that whole patient history to figure out the important information.
We've come up with a solution where we take all this information and highlight the most important information in many different parts of the system so they don't have to go through the full history. They can see the most important information just in one click. If we go to the next one, it's a scribe. After they review the patient history, they will start what we call a consultation, which is an appointment where sometimes the pet owner is there, sometimes it's not, but the vet will give instructions and, for example, provide certain medicines and so on and do procedures. All these voice notes are actually recorded through an ambient recording, transcribed, and then that transcription is sent to OpenAI to be able to generate very clean clinical notes.
In the previous world, people were typing in notes, not paying attention to the pet parent which was in the room, not having their hands free, right? They had to do it after the appointment happened. Now they no longer have to type anything in the clinical notes, and they just have to focus on patient care more. The first iteration of this will be just the clinical notes, but we're looking to have a fast follow with adding treatment items. Identifying, for example, when someone says that they gave a certain amount of a certain medicine, it would identify that and add the charges. If, for example, a vet would say a new updated weight, it would add, figure out the weight, add it, and structure data. We see a huge potential in sort of converting our unstructured data into structured data for improved care analysis.
This is a really, really big feature that we've developed for therapists, and we initially trialed there. We've now brought it to vets, and there's a big interest in this one. If we go to the next one, the last one that we've developed so far in our clinical AI bundle for vets, as we call it, is the AI Discharge Instructions. After the appointment is done, the pet parent usually receives an email or a printed document in some clinics with a summary of what happened in pet parent language, right? It's not medical language. This allows you to, with one click, generate discharge instructions from the consultation, and it saves many, many minutes for the vet. A lot of the consultations actually do not have discharge instructions, so they don't have the time to do it.
The pet parent is left going home with not a real idea of what exactly happened, what are the instructions for medicine, and so on. This elevates care in addition to saving loads of time. We're very excited about this one too. Now, beyond the features, looking at the numbers, as we said, we grew 13% year- over- year, right? The net retention rate was 109%, which was primarily driven by small expansions from enterprise clients rolling out new clinics. We have not had a significant rollout of corporate clinics like we did in, for example, H1 2024 and in the future, looking at the AmeriVet and Pets at Home rollouts. Our churn remained extremely low at 3.4%. If we exclude the impact of the migrations that we've been doing, that would actually be 2.9%. Both numbers are quite spectacular.
Lastly, it's really important to note that the Vets4Pets and AmeriVet post-pilot rollout ARR is not included. That is estimated at EUR 4.5 million. That is the amount that is net needed to be implemented, not including the amount that they have of ARR based on pilots. If we go now to look at the trend over time, in 2021 and 2022, we also did not have a big enterprise customer which rolled out. You can see that impact in our numbers relative to what we saw in 2023 or 2024 where we did. We do have some seasonality based on when they decide to implement. You can see that our new customer ARR is around 4%, which is slightly lower than our average of 7.8%. That is one place where it's impacted.
The second primary place where it's impacted is in net upsell, where it was 12.4% versus the historical average of 19% and a high of 35% in 2023 and 27.6% in 2024. The upcoming large enterprise rollouts and revenue from AI features will support net retention in the future. Another interesting thing to look at is how we're performing in terms of our profitability. If we look at 2025, the last Q2, the last 12 months before that, we've roughly remained on the same level as last year. We've been able to add additional revenue, which has brought us an additional contribution margin of EUR 1.3 million. We spent EUR 1.2 million to accelerate tech and R&D. We have slightly improved profitability. Those investments have been in two areas that we discussed before, the first being the acceleration of AI.
The goal there is to transform Provet from a passive system of record, where the majority of the information is inputted manually by typing, clicking, into a proactive operating system that can anticipate workflows and automate admin, as we saw from the Scribe or the Discharge Notes, with the aim to boost initially clinician productivity and, in the future, other user personas, for example, the receptionist or the inventory management person. The second big bucket that we're spending on is to be able to accelerate our DACH rollouts. We've had very strong interest from enterprise customers in the DACH region that are looking to get a cloud solution to replace their legacy solutions, some of which are ours, Vetera, some of which are on Vetsy, and some of which are on other third-party softwares.
This will not only unlock this 1,500 clinics which are on Vetera today and be able to unlock efficiencies in that we don't have to have two development teams. We do not have to have as well a lot of support tickets which are generated by service-based softwares. We believe that there is going to be a big opportunity for us to upsell our current add-ons like AI, integrated payments, and so on. Go to the next slide. Now, looking at our growth split by country, we can see at the end of 2024, we had £4.5 million in the UK. That is now £5.4 million in the end of June 2025, some of which is CVS. The U.S. has grown, but there was due to currency issues that have been changed. Southern Europe has continued to grow nicely.
There's some corporates which are slowly increasing the amount of locations that they have, driving our growth. In the DACH region, we remain relatively stable with our Vetera product, but we're excited about the potential when we bring our cloud product to market. You can see in the Nordics, we've also slightly grown. If we look at the U.S., just deep diving on that, it's in addition to currencies, in Q2 2025, the PIMS ARR actually increased by 15%. However, there was also big volatility in private pay volumes. That was another one of the big drivers. Next slide. Looking at the revenue split by product, you can see that we have our SME is growing quite nicely in the last 12 months and in June 2025. We have had slower growth than previously in enterprise due to no big rollouts.
Our payments and partners and add-ons are becoming a bigger share of revenue as we have more and more partners coming on board of our marketplace. It's always good to look at, despite our focus on enterprise, our customer concentration remains low with our top three customers composing 21% of our ARR. Looking at the migration, in 2021, 40% of our ARR came from clouds, and as of June this year, it's 83%. We've made a big dent in our migration. Vetserve and Provet Net were sunset in 2024, and Provet Win and Provet Net in Q2 2025. We still have to migrate Vet Vision in Denmark, Sanimalis in Norway, and Vetera at some point in the DACH region. Now, going over to Therapy, the Therapy business has been focused on migration of the hospital users to a new unified platform.
So far, we've migrated 193 hospital users to a unified platform as of the end of June. The rollout continues at a measured pace so we can act on feedback that we get from customers and make sure that we protect our reputation, thereby protecting retention. We estimate migration volumes will increase in H2 2025. We also have had over 500 users, which is still small, less than 5% of our practitioners have activated the AI assistance, which provides an AI Scribe. We've delivered over 25,000 AI-generated summaries and over 11,000 hours of transcribed in Q2 alone. Lastly, we also signed EUR 410,000 new ARR in Q2. Looking at the growth, as we discussed before, the main focus, the primary focus for the Therapy business unit is this migration. Despite that, we've still seen some growth in current markets despite our low headroom for growth in these markets.
Given our market share, we grew at 4.3%. Our net retention rate was 98.9% and churn was 5.4%. Looking over time, our long-term average churn has been around 5%. You see it's slightly above the long-term average. Despite the fact that our revenue mix has shifted quite significantly. If you look at 2023, which is the year when we acquired Easy Practice, we've been working very hard at reducing the churn in Easy Practice. You can see that trend in 2024 and 2025. In LTM Q2 2025, our new customer ARR was impacted at 5.4%, given our focus on shifting everyone onto the new unified platform, Aspit. Secondly, the fact that we are no longer targeting actively non-therapy professionals, which were targeted before with Easy Practice. This is what has driven our churn down, as we want to focus on therapists only.
Looking at the BU EBITDA minus CapEx, we have made our additional revenue has generated a contribution margin of an additional EUR 400,000. However, we've invested EUR 1.4 million additional in order in CAC and R&D to be able to speed up the migration of Aspit customers to our unified platform. The rationale behind that is that unlocks EUR 2.8 million in annual savings. Second, it's accelerated our investments in our AI Scribe to empower practitioners to be able to focus on patient care and reduce time on administrative tasks. Looking at the development by country, we can see that Aspit was acquired in 2021 and Easy Practice in 2022. You can see why the drivers of those growth. Growth has been slow as we've focused on R&D on the Aspit migration. Once the migration is completed, we'll resume work on add-ons and new country expansion.
As you can see, there's a decline in Norway from 9.4% to 8.6%. The Norway decline in ARR is you see seasonality. In Q1, you see high churn. It's the peak churn. Finland ARR grew after implementing price increase. Now looking at our cloud share of ARR, increased from 34% in 2021 to 51% in 2025. We have made some leeway on migration. The churn for non-cloud was 3.5%. It's quite a low churn. If we go to the next slide, over to Alex with the financial update. Thank you.
Thanks, Charles. Hello everyone. To begin, I'd like to let you all know that in addition to this Q2 update, today we also published our interim financial report for H1 2025. You can find this on the company website. It includes further details on many of the headline numbers that we'll be running through today. Looking first at our reported revenues, in Q2 2025, we did EUR 12.9 million of revenue, which is a 5.4% increase versus the same quarter last year. It's important to note that our underlying recurring revenue growth has been much higher at 13.5%, going from EUR 9.9 million in Q2 2024 to EUR 11.3 million in Q2 2025. The reason for the decline in other non-recurring revenue is due to a spike in implementation revenue that we saw in Q2 2024, because Q2 2024 was a period of rapid enterprise rollout in Veterinary with CVS.
Correspondingly, the share of recurring revenue in Q2 2025 is 87.7%, up from 81.5% in Q2 2024. For H1 reported revenues, we see a similar story. Total reported revenue grew by 13.4% to EUR 25.3 million in H1 2025. Reported recurring revenues were higher, growing 17% from EUR 19.1 million in H1 2024 to EUR 22.3 million in H1 2025. Our share of recurring revenue in H1 2025 is 88.1%, up from 85.5% in H1 2024. It's also worth noting that our core Veterinary and Therapy business units have been outperforming our much smaller other businesses. Recurring revenue for the two core business units has grown by 18.9% year-on- year. Looking now at quarterly adjusted EBITDA minus CapEx, in Q2 2025, we reduced by EUR 1 million year- on- year to an adjusted EBITDA minus CapEx of EUR -1 million. Revenue grew by EUR 0.7 million year- on- year.
COGS and customer service also grew by EUR 0.6 million. This EUR 0.6 million growth is higher than proportional to the revenue growth due to the temporary need for extra client support for the early migrated Therapy clients. The largest item impacting adjusted EBITDA minus CapEx is the product development, which is by itself a EUR 1 million increase in expenditure versus H1 2024. The increased product development expenditure is for new features, platform scalability, and enterprise client custom work, although most of the custom work is charged to clients. As per our recent update, we are stepping up investments in AI feature development and DACH localization. For H1 adjusted EBITDA minus CapEx, this was EUR -1.8 million compared to EUR -0.9 million in H1 2024. The H1 variances are mainly driven by the variance in Q2 because the variance in Q1 2025 was only EUR 100,000 versus last year.
Next, turning to adjusted cash flow, in Q2 2025, we had a cash outflow of EUR 1.2 million, which is an improvement of EUR 1.3 million compared to Q2 2024. The main driver of this increase comes from changes in deferred revenue of EUR 0.9 million. As we saw, Q2 total reported revenue had a higher share of implementation revenue, and this is typically billed later than our recurring revenue. Other profitability and working capital changes amounted to a EUR 0.4 million improvement versus Q2 last year. Now looking at H1 2025 adjusted cash flow, in H1 2025, we had a cash inflow of EUR 1.7 million, which is an improvement of EUR 3.4 million compared to H1 2024. EUR 1.3 million of this comes from the Q2 variance discussed on the previous slide. The Q1 variance was primarily driven by trade debtors, which were EUR 1.9 million more favorable in Q1 2025 than they were last year.
The largest individual item here is a payment that we received from one of our large enterprise clients in Q1 2025 for a backlog of their invoices, which totaled about EUR 1.1 million. Finally, looking at the June 2025 balance sheet, cash as at June 2025 is EUR 20.5 million, of which EUR 13.5 million is in money market funds. There were no changes to goodwill in Q2 2025 except amortization and changes to FX. There was no external financing taken in Q2 2025. There were movements in treasury shares. In Q2 2025, we transferred approximately 48,000 treasury shares to participants of our performance share plan. We also completed a share buyback after the balance sheet date in July, which I'll talk about in the next slide. Our balance sheet remains healthy with equity at EUR 68.2 million, and the company has no interest-bearing debt.
The full detailed financial statements for Q2 2025, including P&L, balance sheet, and cash flow, are in the appendices. Next, looking at the share buyback, in July, we settled the transaction to purchase 300,000 shares at a price of NOK 36 per share. As at the settlement of that transaction, the company owns 1,377,793 shares in the company. A key use of these shares is the performance share plan. The company is moving towards equity as its primary form of bonus compensation. In 2025, we expanded our PSP scheme from 16 participants in 2024 to 55 participants in 2025. For our 2025 guidance update, we're reiterating our full-year guidance on Vet plus Therapy recurring revenue based on December 2024 constant currency and excluding acquisitions of 12%- 17% growth.
As discussed during this presentation, our focus in H1 2025 has been on preparing for the large enterprise rollouts in Veterinary, and these will lay the foundation for strong revenue growth in H2 2025 and 2026. On adjusted EBITDA minus CapEx, as we announced earlier this week and discussed in the presentation today, we're accelerating our investments in DACH localization for our Veterinary business unit and in AI development across both business units. We believe the timing is right for both of these investments to optimize growth in the coming years. The adjusted EBITDA minus CapEx guidance for 2025 is being updated to between EUR -4 million and EUR -2 million, excluding acquisitions. Previously, it was EUR ±2 million, excluding acquisitions. Lastly, a reminder on our financial calendar that the Q3 2025 presentation will take place on November 11th, 2025.
The calendar can be seen on the company website, as can our history of presentations. I'll now turn back over to Charles for Q&A.
Thanks, Alex. Just as a reminder, feel free to add questions in the Q&A functionality on Zoom. I've already received one question, or you can also raise your hand if you would like. Going through the questions, we had three questions in the chat. The first was, how do you currently perceive the competitive landscape in the DACH region, the UK, and the Nordics? Are you seeing new startups emerging with a modern tech stack and strong customer proposition? I believe that the question was aimed for Veterinary, given the three of them. We'll talk about each of the different regions first, both in terms of emerging startups and the competitive landscape. In the DACH region, we are the number two player with Vetera. All of our clinics, except for a handful which are on Provet Cloud , are on Vetera.
The number one player is easyVET or VetZ, which is not to be confused with the New Zealand EzyVet, which was bought by IDEXX in 2021. That company was independently owned, then was purchased by a diagnostics manufacturer, which was then purchased by Mars. Mars is the ultimate owner of this company. There are a few low functionality cloud players in the DACH market. For now, as we can see from our Vetera, people are not looking for a new solution, and churn has been relatively flat, especially from the SME. However, we have been receiving inbound interest from corporates which are on our solution and also our competitor solution. That is why we're actually accelerating that opportunity, in that the current cloud pages of the market do not currently accommodate functionality to accommodate these large enterprises in the cloud.
That's the primary reason why we want to go after it. There are no other entrants in the PMS space in that region which are a significant threat. In the UK, there are some upstarts such as Lupa in the market, which do not have a full-fledged PMS yet, but have an AI Scribe. There will be more competition from these AI Scribes trying to go into the PMS markets. I think that once they enter the PMS market, they realize how big of an endeavor it is to build a full RERP system. I'm a strong believer in only the paranoid survive. The reason why we're investing heavily in AI is to make sure that you're seen as the front runner in these markets. As if you've got the PMS plus the integrated AI, that is way better than having an unintegrated AI value proposition.
Think about the, there's no need to copy-paste. We know both the history of the clients and what happens in that certain consultation. We also can add in clinical notes as V1 of it, but you'll also be able to add structured data if we are the ones who actually provide that as a service. We have a right to win in terms of the value problem product perspective. Also, the second is this is a paying add-on, which we can afford to sell at a lower price than the competitors because we don't have to offset the full CAC given that they're already our customer. In the Nordics, we have not seen any major competition on the Vet side. There are some upstarts which have created an add-on for scribes and so on, but they have not actually launched a PMS solution.
Similar to in all markets, we'll keep vigilant in that by having a great AI solution, there's no reason for our customers to go to a third party. You mentioned a more modern tech stack. Provet is a Python backend and a Vue.js frontend. We're hosting on AWS. We've got a very modern tech stack. We don't see any competitors with a more modern tech stack, especially at scale, right? We've done a lot of work to migrate our frontend over the last few years, and we're continuing on that work. We don't see a big risk from a more modern tech stack coming on board. In terms of value propositions, there is one theory from this great investor, a VC fund called Tidemark. I highly recommend that you take a look at their website.
They've got this vertical market software project where they detail a lot of the ways that vertical market software can grow. There's a strategy that our competitors can use called integrate and surround, which is basically at first you integrate the PMS, and then you surround them by slowly adding feature after feature. That's what we're trying to protect against because we have the core PMS. If we don't develop these additional features, right, at some point our customers will ask us to integrate with these startups which are providing the same feature, even if they're slightly worse. Over time, they can add more and more features and slowly surround fully in order to become the PMS. That's what we have to prevent against. The second question was, what is your view on AI in terms of both threats and opportunities for Nordhealth?
Do you believe you can monetize new AI-powered features? How do you see AI impacting the efficiency of your software development processes? This question is twofold. One, it's about how do we see leveraging AI to be able to improve our products and our business model? Then secondly, internally, how we can use those. On the value proposition, I'm so excited by the potential. If we think through the original value proposition that we thought we had with cloud, it was that people don't want to have to have the hassle of servers and servers burning out and so on. They want to move to cloud as well to access their data from anywhere. Even though that value proposition did cause migration, it didn't dramatically change the way people work. The catalyst to change was less intense, right? AI dramatically changes how work is done.
If you think about a clinician, the amount of time that that clinician can spend on admin, like typing clinical notes, can be reduced by 95% with these features. There's a huge opportunity to dramatically transform how work is done. With cloud, it was a nuisance for the practice manager, but not so much for every single practitioner. This is quite a unique time in that I believe it will be a big catalyst to change, much more than it was for the shift with the cloud. From a business model perspective, prices are a bit all over the place. Startups are trying to figure out what pricing they should charge for their AI solutions. For the scribe alone, we've seen anywhere between EUR 69 on the low end to all the way to EUR 129 per user for the scribes only in both the vet side and on the human side.
This is substantial, to almost double our average revenue per vet or per therapist if we are looking at the higher end of that. There's a big, big opportunity to be able to increase our average revenue per account. Internally, there's a couple of places where AI is particularly adept at being able to provide improved efficiencies. The first, if we go to the top for P&L, it's looking at the cost of customer service, where as we grow, we've got more and more customer support tickets. The ability of us to, we use both Zendesk and Intercom, so we're trialing two different softwares in two different business units to be able to experiment. We have seen that AI is particularly adept at answering a lot of the how-to questions. That will be a big cost saving if we look at per cent of a revenue basis over time.
The second is on our development, where if you think about the development software development life cycle, there is a discovery process, which is done by a Product Manager to figure out which problems should we solve. AI is really good at figuring out all the information that you get from your customers internally, from support tickets, collating that to figure out what are the most pressing problems to solve. The second one is design, where a discovery process was a Product Manager would create what is called a document outlining the problem and so on. Right now, they do not have to do that. They actually can go all the way to create a prototype, which in the end, a document can be confusing to some. A prototype, everyone understands it very easily, and it gets people to align much easier.
On the design side, you can leverage AI to create designs way faster than you did before. On the development side, we use tools such as Cursor or Claude and so on to be able to accelerate developments, and also other tools to be able to test and improve our safety and our quality of our code to make sure it matches our standard. There is big potential internally. The third question is Provet Cloud and DACH. When do you expect to have a fully market-ready Provet Cloud offering for the DACH region? Although we mentioned the DACH region, there are three countries, and in those countries, they have three specialties: Small Animal or Companion Animals, you call it dog, cats, then exotics, Equine, and Production animal, cows and sheep, and so on.
We are first going to focus, we have got a few clinics in Austria, and we have got an initial focus for Small Animal in Austria, which we are almost localized for. We will go after Small Animal in Germany. After that, we will see. We are going to segment by specialty and by country. The when is really hard to say because the way we go about localization is we do have a good understanding of the gaps that we have got in terms of features, but ways of working are slightly different. It is a bit like an iceberg in that it is hard to set out an exact date because we only move on to the next segment when the first segment has happened. The definition of satisfaction can be different.
We'll continue to update you on progress as we continue rolling out more multilingual in the DACH region. The next question we had was preparing the rollouts for AmeriVet and Vets4Pets to take place in H2 2025 and 2026. When will we see the full effect of the EUR 4.5 million in reported ARR numbers? Could you help us think about phasing on both ARR growth and revenue growth in H2 2025 and 2026? The total amount of ARR, which is not rolled out as of the end of Q2, is around EUR 4.5 million, right? That's an estimate because there are transaction fees and so on for those two. When they roll out is really hard for us to predict because the enterprises have certain requirements that they want, and those requirements change. They have to upgrade their internet sometimes, upgrade their equipment.
It's really hard for us to predict exactly when that will happen in that phasing. As we've guided, this will happen in H2 2025 and 2026. You should see the full EUR 4.5 million by the end of 2026. The next question is beyond headcount and other OpEx investments in the AI and DACH launch. Should we see any impact on gross margin in H2 and 2026? We don't provide the forward-looking guidance on this, but Alex, any comments?
Yeah, I mean, I think as we've discussed, one of our biggest projects is in the Therapy business, migrating clients from the legacy Aspit platform to the unified platform. That should help us realize the savings that Charles mentioned on that rollout. In the meantime, it costs a little bit more to migrate, but then the savings from the migration will help support gross margin and other cost lines into H2 and certainly into 2026 when the larger volumes of migrations will start to come through.
Thanks, Alex. The next was, could you share some more insight into why the time is right to accelerate your DACH efforts now? As I mentioned in the presentation, the main reason we're doing it now and accelerating it versus what we had initially planned is the strong inbound requests we've been getting from enterprise. They want to migrate, and a lot of them are on both systems, partly on our system, but also on our competitor system. We see a big opportunity to be able to do work in parallel. We onboarded a new CTO, James, this year, and I'm very confident in his ability to be able to do products in parallel. Our capability internally to be able to do it, ramp up net new teams. The second one is the strong demand from enterprise. There was another question from Eric.
The EUR 2.8 million in annual savings in the Therapy from migration of Aspit. Did you see any of those gains in Q2, or is that an HQ 2025-2026 story? The savings from the migration on the Therapy side come up from probably a few different areas. First, it's the license fees that we pay for. Basically, we pay for license fees to be able to provide the Aspit legacy products to Microsoft and to remote desktop providers, Citrix. Once they migrate, those are no longer needed. That's one big area, right? The second is the efficiency. A lot of the changes that are required to be made are made manually. Someone has to contact support to be able to maintain. The number of support tickets on the Aspit platform is much higher than it would be on a new cloud platform. As a result, we should see a decrease.
The third one is that we don't have to support two different platforms. We only have to support one. Those are the three categories of savings. The first one is we get, and the second one we get as people migrate. The third one comes only when we shut off the system properly. That should happen. We don't have an exact end of life, right? We should see this improvement in mostly 2026. The next question was, are the financial targets communicated at the capital markets day still valid, or should you assume they have effectively been in band? Nope. In the capital markets day, we had communicated our growth targets, and those are still valid, right? As we see that there are some fluctuations over time, right? We've seen that historically in our revenue numbers as well, for example. These are still valid as of today. Great.
Any other final questions? Perfect. Thank you very much, everyone, for joining.
Thanks a lot, everyone.
Nice day. Bye.