I'll now like to hand the conference over to Dr. Sam Hupert, CEO. Please go ahead.
Thank you and good morning, everybody. Thanks for joining us. As most people would know, Pro Medicus is a healthcare IT company specializing in enterprise imaging. We have three jurisdictions: our head office here in Melbourne, our R&D center for Visage in Berlin, and our office in San Diego, where our biggest market is with the U.S. now, comprising roughly 90% of our revenue and growing. We have two product sets: the Visage RIS, which was our first product, largely sold here in Australia and now in Canada. And the product we sell in the U.S., which is Visage 7, which is the radiologist desktop or our clinical product, is our main revenue driver. But in the last 12 months, we had a major renewal for Visage RIS, and we've seen some growth in our Australian business as a result.
In terms of the results, definitely a record first half, not only in terms of all the financials, which we'll go into in a minute, but in pretty much every other key metric. We had a record number of new contract wins. We also had record dollar value of renewals, and we've started to see some material revenue from upgrades where clients who are existing clients take additional modules. We finished two major implementations in the half, including Baylor Scott & White and OHSU, as well as others. We continue to make good progress with other modalities and AI, and we believe the results have set a strong base for growth in FY 2025 and beyond. In terms of the financials, our revenue was up 31.1%. Profit after tax up 42.7%. EBIT margins continued to grow, so grew from about high 60% to 71.9%, half to half.
Cash and other investments increased 17.7% in the six-month period. The company remains debt-free, and as a result, we've paid out a record half-year dividend, up 38.9%, and as per usual, fully franked. In terms of the revenue splits, many of you are familiar with this graph. The pink, salmon pink at the bottom of transaction revenue, and as expected, that continues to grow materially as new contracts layer on top of our existing ones, and our existing ones continue to grow at above industry average. This is all recurring revenue. The blue is more traditional style service contracts, again recurring revenue, the green being for professional services, and the yellow at the top being for archive migrations, and as we sell more full stack, as we call it, there's more archive, there's more migration revenue coming into the mix.
So all in all, a healthy gross growth across all the sections of our revenue. In terms of our highlights, we started off with one of our smaller sales for a niche hospital called Lurie Children's. It sits next to Northwestern in Chicago but is a separate entity. And it's one of a number of children's hospitals we've won in the last two years, being a specialist market that is growing for us. We did have a renewal of a large RIS client here in Australia, which was for AUD 32 million for five years, and then followed by Mercy, one of our largest IDNs, with a renewal of eight years, which is actually longer than their original contract. And the minimums on that renewal are AUD 98 million over the eight. And then in November, just after the AGM, we announced our biggest deal to date with Trinity.
Again, at minimums, it's $330 million over 10 years, with room for growth in that contract. NYU Langone was $24 million five years. That was them purchasing Visage 7 OA and transitioning from on-premise to cloud. Then we finished off the year with Duly Health in December. Duly is a private practice and a regional group, which I'll talk more about in a minute. One last one on the next slide was Duly, again with the archive upgrade, a bit like NYU, transitioning from a third-party on-premise archive to Visage 7 OA and moving the whole thing into the cloud. Now, the next two occurred post-half-year results, but I thought I'd include them just to show that our run of new contract wins continues unabated.
In January, we announced U Kentucky , which is an academic, obviously in the state of Kentucky, and BayCare, a large IDN in Florida. We've had most probably our busiest period in terms of new contract sales over the last two months. Baylor Scott & White implementation was completed in record time, and our pipeline continues to be strong despite a slew of recent conversions from pipeline to contract. The operating model serves us well. As most of you know, it's used in the vast majority of our contracts. The model is based on minimums, usually as a percentage of their previous 12 months, volumes that they tell us about at the beginning of the contract. Each of these contracts that we announce has inherent upside in it, just based on the fact that no one commits to 100% of their volumes simply because they need a bit of headroom.
Our forward revenue has jumped materially, as you would expect, to AUD 894 million over five years. Key things are it only includes five years or roughly half or only half of Trinity because it's a 10-year contract. So that second five years is still to come. And any of the contracts that we announce beyond five years, it only includes the first five years of that. So we see a lot of potential upside that provides us annuity stream with far greater predictability. We believe in operating leverage. It's highly scalable. There's no CapEx, hardware, or cloud fees. So all of the revenue we announce in these contracts is revenue to us. We do build the professional services. As you know, they're roughly 10% of contract value. And as we've noticed, our margins have continued to grow as our footprint increases.
The segments we work in, one of the ones that's most known is the academic medical centers, where we currently do 11 out of the top 20. They used to be ranked. They're now listed alphabetically. We don't believe that anyone has got anywhere near that. I think our nearest competitor that's been in the market quite a bit longer than us would have two to three. So again, we've been able to show the product is well-suited to that high-end, leading-edge, sophisticated user. But having said that, we have an increased footprint in the IDN space. Many of these are large and sophisticated, the difference being they don't have universities and medical schools. We've noticed the sales from very large to medium to small, which is important so we can cover the biggest spread of the market, and many of our recent sales have been full stack.
If many of you have seen the slide in the AGM and other presentations, you'll notice the Trinity logo at the bottom, sort of rounding out what has been a huge segment of the market for us. The other area that we talked about previously, and it's starting to grow for us, is the private market. These are usually groups of radiologists that had regional practices. Some of them have merged into consortia where there's a practice in each state, and they may have some external funding, but most of the time, the doctors own the bulk of the equity. It was a market that was dormant for many years simply because there was a huge amount of private equity back in the 2008, and so people tended not to change their systems. Opportunities are opening up for us.
The latest one in point is Duly Health, which is a AUD 30 million seven-year deal at minimums. Cloud is a key factor for this market, so we think we have a good advantage there. We are starting to see a network effect. We had Bay Imaging Consultants a number of years ago, U.S. Radiology, which is one of the large roll-ups, and it's one of its biggest divisions. Charlotte Radiology has recently come on board off a competitive system, CRL based in Minneapolis and Duly, which we just announced just in the last half. A growing segment of the market for us, so we're really covering all bases. As mentioned during the last year, the Visage RIS, we had a major renewal. We have long-term contracts with two of the largest radiology groups in Australia.
We are seeing some upside, and that's driven not just the pricing accretion, but some organic growth within these groups that has driven some upside in our Australian business during the half. We often test our application and look at it compared to others in the market, and we're still 100% convinced we're the number one in the three key areas that differentiate systems: speed, functionality, and scalability, and in that scalability piece, cloud is a huge strategic advantage because most RFPs now coming to market will stipulate cloud as the platform because most of the large health enterprises know they need to transition to cloud from on-premise. The other thing that's driving adoption is massive data explosion. It continues to go unabated. You have new technologies like photon counting CT, where the files are absolutely massive.
Traditional systems crack in terms of trying to display and manipulate those sorts of images, which differentiates us even more so in the market. The interesting thing is wherever there is a department, the first or enterprise, the first department to crack, as we call it, is breast imaging simply because of the volume and the volume of data. In the U.S., 3D breast imaging or breast tomosynthesis has become the preferred screening methodology. So you have this massive explosion of data sets that we are now seeing, and it helps. And when we look at groups like U.S. Radiology that came to us because their breast imaging system was under stress, we see this a lot in the market as the sort of canary in the coal mine as to when systems need to change.
Legacy technology is compress and send, take the file, compress it, unpack it at a local workstation that has to be highly configured and manipulated there. The problem is the data sets are getting too big to do that in a timely manner. We moved from one system that was taking between 10- 20 minutes to do that for tomosynthesis remotely to ours, which is less than one and a half seconds. So the delta can be huge because we don't compress and send. It is a streaming technology, which means that all the manipulation and rendering in 3D is done centrally when we stream pixels to the radiologist. So no matter where they are, 99.5% of their work is left sub one second. The rest is sub two, depending on the modality.
Our solution is really a streaming platform based on a central core of cloud, one application that does everything from plain X-rays through the most sophisticated advanced imaging, 3D, 4D, fusion, and breast that used to be done by standalone systems in the past. We're now adding the other modalities to that same platform, extension of the software as a license. That concept of one core product spinning off multiple applications in cloud as a service we're building towards and achieving that. The other thing that's really important for us is fast track implementation. We believe we can do it in a quarter, sometimes a tenth of time of our competitors. We completed Baylor Scott & White.
They signed, and within 11 months, one of the largest IDNs in the country was fully live across all their regions, including the commencement of data migration and everything else, which has never been done in that timeframe before. We think it is a record implementation time. I won't go through all the sales. I'll just some of them. The key highlight in terms of volume certainly was Trinity. It is one of the top 10 IDNs in North America. It's $330 million at minimum, 10-year deal. It is a full stack, which is viewer, worklist, archive. It will be 100% cloud deployed. Implementation is scheduled to start in September of this year. Because of their size and regions, they are across 19 states. The implementation will be phased over an 18-month period.
We do believe it's one of the largest deals in radiology IT that has been announced. Duly, I mention it because we talk about the private market. It's one of our largest sales to date in that market. It confirms our platform is suited to pretty much all market segments, and we think we're unique in that regard. And sales like Duly open more opportunities in this space, which we believe will grow for us going forward. We have talked for many years about the ability to sell back, particularly archive and worklist to existing clients. And in the half, we signed two material upgrades, one with NYU Langone and the second one with Duke, both existing clients. The core of the upgrades was the adoption of Visage 7 OA from a third-party archive and moving the whole technology stack into cloud.
So AUD 24 million for NYU over five years and AUD 15 million, so AUD 39 million of additional revenue from existing clients. So we believe the Visage 7 OA will be a catalyst for transition to cloud for other clients. And I think it shows a tried and proven path for those clients with on-premise to transition seamlessly to cloud with us. The renewals, as I mentioned too in the first half, Mercy are a very large one.
We also had a large RIS client. So we still maintain our record of 100% renewals to date. All renewals, including the RIS renewal, was at a higher price point than the original contracts and many for longer terms than the original contracts. So we believe this, plus people who upgraded with us, confirmed their confidence in our product because they wouldn't do such long-term renewals if they didn't have that inherent confidence.
We are known as being the most expensive but have the best ROI. Many of you have seen the slide. It's not just in terms of financials, which we provide an unparalleled return, but also in terms of clinical capability and acumen. So we do move the needle when it comes to clinical outcomes. We do have a large number of cases where clinicians have told us that they're able to do things that they previously couldn't do, or if they could, it would take them too long, so they didn't do it. Here are some of the quotes that we get. Things like, "I've never been able to visualize vessels like this." These are the sorts of things that a neuroradiologist needs to do on one in two cases that they do every day.
The fact that they're able to do it intuitively, rapidly, and then see things that others don't see, it makes a meaningful difference to them clinically throughout their workload. The other thing that has really become very prevalent in the industry is the radiology workforce shortage. It's global. It's acute. In my 40 years in the business, I've never seen it as acute as it is at this point in time. It is worldwide, as I said. It's because of a reduced intake of radiology residents a few years ago for the fear that AI would take over. With larger data sets, there are just more images to look at. Radiology is growing at 2%-3% per year, and people are looking at work-life balance.
So all of these things together are creating a situation where some groups are even starting to cut back on existing contracts, not bidding for new work simply because they can't cover it in terms of manpower. We believe we have a surefire way of addressing this. We know that with Visage, for various reasons, radiologists are that much more efficient. We start 20%-25% as the average. We often will see even more than that in the field. So we believe it helps directly address this issue of radiologist shortage and what they call radiologist burnout, where they just can't get through the load that they have in front of them. Cloud is an important part of our offering. We see the whole transition has been quite a lot of push behind it in the U.S.
We haven't done an on-prem implementation in the U.S. now in roughly four years. Importantly, we can work and have large-scale implementations in all the three big clouds: AWS, Azure, and Google GCP. So if a client has a preference for one over the other, we're able to work within that. And we see it as a huge strategic advantage because we know a lot of our competitors talk cloud, but in reality, we believe that we're the only ones that can actually execute on that and give them a full 100% cloud implementation. Our growth strategy, we continue to deliver on it. We expand our footprints with new clients. As I said, the last six months has been a record number of new contracts in terms of dollar value that we've done in any six-month period.
We've seen strong growth in transactions from existing clients over and above industry average. We are, with our new products like Visage 7 OA and worklist, selling those not only as full stack, but in the case of NYU and Duke, as additional modules back to existing clients, and we're leveraging our R&D capability for next-generation products in cardiology and AI. The TAM in the U.S., 650 million, growing roughly 2%-3% per year. We believe we're able to address 100% of that from a product point of view, and as we've seen, we've been able to deal with the top academic medical centers in the U.S., as well as private groups, large IDNs, mid-size groups, so really up and down the spectrum across the different market segments. Our current penetration has gone up with recent sales from around 70% to just under 9% of the market and growing.
So while we've made terrific inroads, we believe there's still a huge amount of runway addressable for us. The pipeline continues to grow, so it's a dynamic thing, clearly with all the recent slew of new contracts that we moved from pipeline to contract. They've decremented the pipeline, but I'm pleased to say, particularly following RSNA, we've had a number of net new opportunities to helping replenish all of that. And there's been a number of opportunities in the pipeline for a number of throughout 2024, which is standard because sales are usually an 18-month to 2-year period. So despite having a record number of sales, we are seeing a great rate of replenishment across all the market segments and across all the market size of opportunities. Open Archive, I won't talk too much more about it. Most of you know it.
It is what we think is the lever for a lot of our existing on-prem clients to take Open Archive and then transition to cloud as part of that strategy. It's also a very integral part of our full stack offering that has become very popular in the market. The workflow is similar. We've sold it back to a number of existing clients. We see it as a key product in helping prospective clients because many of them want a full stack, regardless of size. The smaller ones almost routinely, but then look at large ones like Trinity. One of the largest IDNs have gone full stack with us as well, and a lot of the IDNs in between. The one view of cardiology, pleased to announce that we will be doing our first implementation of that. We've got a contract client. It will start in April of this year.
It will be for the full cardio echo package. Previously, we've had parts of it in test and working in various clients, but this will be the first full cardiac package that we will deploy, and that will commence and be fully deployed in April of this year. AI, we seem to make good progress. We showed some new works in progress AI at RSNA last year in December. They've been well received, and they will progress through the commercialization pipeline. We do see that the AI market has had a lot of hype in the past. There's been a lot of disappointment, but we believe the timing for when there will be some commercial reality around it is in the near future, and we're positioning ourselves to take advantage of all of that.
We have our leadership team as part of the AI, and that is currently being bolstered with some new recruits. The Visage 7 platform, as we mentioned, just from a pure platform point of view, is very well suited to AI, and we see that we not only will develop our own or develop them with joint partners, but we will be part of that third-party ecosystem. And as many of you know, we've made an investment in a company called Elucid, which has AI for cardiac CT. One of their two modules has now received FDA approval, and we will be looking to help integrate that and put it out into the market in around the six-month timeframe. So we're starting to see activity there. In terms of our newest, Visage Ease Pro, this was released just around a year ago as part of the Vision Pro from Apple.
We've just participated in the first spatial computing conference in medicine at Sharp, a health system in San Diego, where we were keynote speakers, and have had an enormous amount of positive feedback because our application is currently available, unlike a lot of others that are sort of works in progress, so we are seeing some real-world applications starting to appear, particularly in interventional radiology, surgery, and other areas where immersive or spatial computing will actually help in the diagnosis and medical treatment. As I mentioned, it was launched a year ago. We are seeing some very interesting use cases starting to emerge, and we think we'll see more and more being piloted and developed within the next 12-18 months. We think it's also very important that it underpins our belief that our technology is adding to 24 months ahead of the competition.
We don't know any of our competitors that are able to use the Vision Pro or have an application for it like we've been able to do. Just rounding out, RSNA is our biggest conference of the year. 2024 was our biggest ever, not only in terms of the booth. We built a small city, and this is just one part of it, with conference rooms, with demonstration areas. But it was also our biggest in terms of client engagement, prospective clients, existing clients. It was pretty much non-stop from Sunday to Wednesday, so four days solidly booked throughout. This was our team this year. You'll notice I'm smiling in the front in the cast. They used to do it early, but I made sure that they had to wait till I got there.
But we had a team of 52 people, and everyone was pretty much flat chat throughout the period. So our biggest and best RSNA to date, and certainly, we're looking at using those leads in RSNA to that has fed strongly back into our pipeline. We also have a new product, which is Chat. It's a joint development between Melbourne and Berlin R&D, which is a thing that we see continuing. It will be going live within the next few months within a key site in the U.S., and then will be available as an additional module for existing users. Finally, summary, and then I'll open to questions. It was a record year, revenue and EBITDA by quite a margin. We had two major contract renewals. We had our biggest record of new sales in the half with a minimum TCV of AUD 385 million.
We had two major implementations completed that are now going to contribute to the second half and onwards, and they're both material in our OHSU and Baylor Scott & White. We continue to have a strong pipeline across a broad range of opportunities. We're strategically positioned to leverage the AI and other technologies. As I mentioned, cardiology will be installed at its first site in April, and RSNA was our biggest to date. So thanks very much, and we can now open up to questions.
Thank you. If you'd like to ask a question via the phone, you need to press the star key followed by the number one on your telephone keypad, or press star two to cancel your request. For the sake of time today, we ask that you limit your questions to two per person. However, you are more than welcome to rejoin the queue if needed. If you'd like to ask a question via the webcast, please type your question into the ask a question box and click submit. Your first question today comes from Garry Sherriff from RBC. Please go ahead.
Good morning, Sam and Clayton. Two questions, one on pipeline and the other on AI. If I look at the pipeline, is there any way we can get a sense of the quantum now versus six months ago? Just trying to get a level of color or quantification given you've had a really strong contract win rate in recent months and just wondering whether that pipeline refresh rate has kept up. So I guess that's question number one just in relation to the pipeline.
Yeah. Well, these things are very dynamic, so if you have a look at it, we've always said pipeline is lumpy, but they'll convert from pipeline to contract when they do. We've had an unprecedented amount in a two-month window, which is a plus, but having said that, there were plenty of opportunities that were already in the pipeline, and following RSNA, obviously, there were a number of existing ones that came to see us, but actually a material number of new ones that have subsequently followed up with us. So while we don't give an exact number, we're very pleased with the replenishment rate and how the others in the pipeline are still progressing.
Understood and in terms of AI, Sam, I mean, if I look at Harrison.ai as an example, they've just recently raised AUD 180 million for their radiology AI suite with plans to expand to the U.S. Are there any plans in terms of partnership agreements with the likes of Harrison.ai for your AI strategy, or do you see the likes of them as being more a competitor to your offering? Just trying to get an update on the AI strategy and using someone like Harrison.ai, who's a local player and attracting funding and clearly pushing to that AI radiology space.
I'll answer the second part first. Definitely not a competitor. We do see AI, the algorithms coming from three buckets. We develop ourselves, we co-develop with our research partners like NYU, Mayo, Yale, and others, and then third party. Harrison would be in that third bucket. We know the Tran brothers well because, as you say, they're local. Yes, we're looking to where they're best in breed. Look, we can put it on our platform. We can integrate it more tightly than having it as standalone. We are looking at those relationships. Elucid is one of those. You heard about it a little earlier than normal because we took an equity stake in it. We don't need to take equity stakes given that we're now somewhere roughly around 9% of the market and growing.
We think we have a compelling proposition for those third-party AI players. So yeah, you'll see our ecosystem expand in that direction, something we've flagged in the past. So yes, they certainly could be. Now that they've got FDA, they certainly could be one of those parties, but definitely not a competitor.
Thank you. Your next question comes from Annabelle Lee from Goldman Sachs. Please go ahead.
Good morning, Sam and Clayton. I've just got two questions. So first one, good visibility on your contracted revenue growth as Trinity and others come online. Just from a timing perspective, can you help us understand some of the drivers between the first half and the second half? And is it reasonable to assume that revenue growth will accelerate in the second half?
Thanks, Annabelle. Yes, it will. Implementation of Baylor Scott & White, as Sam mentioned, was completed in September. So we had three months' worth of revenue in the first half. Clearly, we will have six months' revenue for that in the second half. And also, Oregon Health & Science University was the same slight bit of revenue within the first half, which will have a full six months' worth of revenue. And as we implement new opportunities in January and February, such as Nationwide and Nicklaus and Moffitt, there'll be part revenues of that throughout the period. So there'll be some of the revenue drivers that come through. In terms of margin expansion, it's obviously been a fairly strong one throughout the period. The only cost that's not replicated in the second half is the RSNA cost. So that's usual from half to half.
Yeah. And the only other thing I'd add, Annabelle, is of all the new contracts, their revenue won't start. That's all ahead of us. That won't start till first half 2026. So we anticipate a good second half and then an even bigger step up into 2026 as the contracts we've just announced will start coming on stream.
Great. And second on the cardiology contract, are you able to provide a bit more color around what ultimately got this across the line? Maybe how long this process took, how many discussions are currently live, or what's unique about this customer that might make this package more attractive? Any color around this would be helpful. Thanks.
Sure. So with cardiology, we're not quite back where we were with radiology 10, 15 years ago. Obviously, we've got reputation. We've got the platform in there. But you're dealing with a slightly different end user, particularly around echo because it's all done by cardiologists, not radiologists. So while being in the organization and dealing with radiology and having a good reputation helps, you still have to win the hearts and the minds of the cardiologists. And I think we've been able to do this in this instance. And I think you'll see more of that coming through as people understand that the product offering not only has all the benefits of what the radiology one has, but it also has a lot of the feature function that cardiologists need over and above radiology.
I think this sort of proves that point and clearly will use that as a key reference site and a key learning site. The first one is always the biggest step, and that's now been done and contracted. Clearly, we were able to do all those things, convince the cardiologists, and show the feature function.
Thank you. Your next question comes from Josh Kannourakis from Barrenjoey. Please go ahead.
Hi, Sam and Clayton. Thanks for taking my question. Just first one, just on Trinity Health, can you give us a little bit more detail on both the split up of the contract just with regard to archive, the worklist, but also just in terms of what the migration costs are within that contract? And also just a little bit more context around the weighting of the timeframe. Is it evenly spread out over that 18-month period, or should we sort of think about some of it front-end or back-ended?
Yeah. In terms of data migration, that'll commence in the second half. I won't give specifics on how much that is because we can run through that separately. In terms of the exam split, viewer archive worklist, we've been pretty consistent that viewers around 60%-65%, archive around 25%, and worklist remainder. And once they go live throughout the period, that'll come through on all three products. The build-up over 18 months will be exam revenue will continue to build where a lot of the sites will be complete by the end of the calendar year. And then there are two hospital groups that are a bit further out. So that's where the 18-month period comes from. So exam build-up will build up over that 18-month timeframe. Data migration will start pretty much back in June, July of this period.
So there'll be virtually no revenue within the financial year, but more into the FY 2026 year.
Yeah. And the two at the end are because they're changing their EHR, and that's slated for 18 months. And so we're trying to fit in with that timeframe. But the bulk of Trinity will be relatively linear over that 18-month period. And it may be accelerated over that, but at this point in time, it's scheduled for 18 months.
So just to go back, so the eight-year period after that will be fairly even. The two years building up will be exactly that.
Yep. No, that's perfect. Thanks, guys. And then just the second question, just with regard to cardiology, I know Annabelle just mentioned it before, but you've got a couple of other, I think, trial sites still going. Are they still underway? And what has been the feedback from those sites? And I guess the possibility for those converting into commercial contracts over time?
Yeah. The feedback's good. As I said, you've got to win the hearts and minds of cardiologists. And unlike radiologists, cardiologists are not in front of imaging all day, every day. It's a part of what they do. Whereas with radiologists, it's everything that they do. So getting their attention takes a little bit more effort. I think they've been pleasantly surprised at the speed of the application because some of them hadn't been exposed to it before. And cardiac ultrasound now, the file sizes regularly get larger simply because of a video that they put in there. So I think we're getting very strong feedback. I think people are realizing how to fit in to their existing reporting systems because you have to report all the measurements and things to the government.
And we've done a lot of work in that to fit it in with Epic's product called Cupid and other bits and pieces. So I think it's all progressing really well. I think having our first one, the full kit, is a very important first step, and we're hopeful to have more.
Thank you. Your next question comes from David Low from JPMorgan. Please go ahead.
Thanks very much for taking my questions. Sam, I noticed in the interview that was published, you talked about pipeline conversion improving. Can I get you to elaborate a little bit on that? I mean, how much of the pipeline have you been able to convert in the last six or 12 months, and how does that compare to previous periods, please?
Well, yeah, the last two months sort of eclipsed everything because we had unprecedented. And some of that is already in the first half. So U Kentucky and BayCare occurred in January, February. But if you look at that window, I think there was about, what, AUD 485 million conversion at minimum between pipeline and contract. So if you look at it, we always say the pipeline is lumpy because it's not up to us when clients sign. I wish they'd just sign one every month and it'd be easy. But even if you look at any six-month period, let's say six months, we've never converted as high dollar value from pipeline and contract as we have now. And if you look back historically, we've always had these steps. This may have been one of the biggest.
I guess the other metric that I would be thinking about is if you had 10 customers that the sales team had put in the pipeline, are more of them being converted and not choosing a competitor as well? I mean, I get the dollar value has stepped up in that two-month period was extraordinary. But do you think you're winning over those who are prospective, high proportion than in the past?
I think we're winning at least the same proportion, which is 80% +. I think the other thing that's interesting is we're winning some in a market that we weren't really playing in before, which was the private radiology space. So not only are we winning in academics and IDNs, we're starting to take market share in this third bucket. And it's an interesting space because a lot of the groups have their own imaging centers, but that's a much smaller part of their business. They read third-party hospitals. So it's a different sort of market to what we have here where an I-MED or a Sonic have their own imaging centers, read their own work primarily. In the U.S., it's very different. So we think we're equal, if not better, in the academics and IDNs.
And then we're starting to take market share in this area that we really didn't have much before, and we're starting to see some more activity.
Thank you. Your next final question comes from Andrew Paine from CLSA. Please go ahead.
Yeah. Morning. Thanks for taking my question. Just coming back to EBIT margins that you touched on before, just getting to understand the outlook here, just especially around the OpEx needed to ramp up in other ologies or support recent and maybe anticipated contract wins. Really just trying to understand the top-line opportunity and how that flows through to earnings, especially with your previous comments of accelerated revenue in the second half.
Yeah. So Baylor Scott & White being on board for six months versus three months won't change our OpEx because it's already been done. So that won't change it. So clearly, margin would expand there. But do we need more people for cardiology AI? We are continually increasing our resources, and you've seen our operating expenses and headcount increase over time. I think, well, clearly, the revenue is outstripping that. So part of, Sam, and my job and the leadership team is to make sure we are not too thin on the ground and that we can actually satisfy our customers both from implementations, which I think we've done and showcased in the past, but also supporting them once they're fully live. So that's an ongoing daily activity of ours.
Yeah. But just on top of that, if we're not looking to hire a whole team of people to do Trinity, for instance, so our hiring pattern will be consistent as it has been in the past. We're building the bench, as they say, but we're not looking to all of a sudden we've won these contracts and go out and hire another 30 people. We won't need that. It's our standard hiring pattern.
Yeah. Okay. So it sounds like margins wouldn't be expected to go down, but if anything, directionally positive, which I think is what you said on the last call.
Yes.
Yep.
Yeah. Okay. Thanks. And then just another one. Just coming back to your comments about hospitals cutting contracts, I think that's what you're referring to. I'm just trying to understand how material that is and what percentage of these contracts are being reduced. And if possible, can you talk about your work around a little bit more and I guess the phasing of those headwinds, if they are headwinds?
I know there's tailwinds. It's not the hospital that's cutting the contract. It's the radiology group that. So in the past, the hospital would put out a contract for a radiology group to read all of their work. And there might be three or four groups bidding for that work, right, saying, "Choose us, choose us." Now you see sometimes we're not even one group bids because they can't take on any extra work. They're already flooded as a manpower entity. So the hospitals are finding it harder to get people to read their work. Whereas with our clients, what we're allowing them to do is do 20%-25% + more work with the same manpower. So we're shifting the dynamic.
That's why you're seeing places like some of these reading groups like CRL and Julian, people starting to realize that, yeah, you pay more for visits, but hey, it lets us really improve the efficiency of our manpower, which is our stock in trade, and that's really important now with this shortage because if you can do work when others can't, it's a big plus for you.
Thank you. Your next question comes from Julian Mulcahy from E&P . Please go ahead.
Sam, first, two questions. The first one on AI. When it first sort of became a hot topic in imaging, it was all about reading scans. But it seems to have shifted now more towards workflow and automated reporting just to sort of save a lot of time. So do you think that's kind of a risk and it'll allow some of the incumbents to actually hold on to contracts because they can provide that sort of time-saving that way?
No, because a lot of the workflow things where they tell you which CT to read because there's blood in the head, we integrate to. So our worklist already can manage all of that. So there's always been multiple uses for AI. The biggest use, actually, is the one you don't see. It's the actual manufacturers build it into their equipment. That's where most of the FDA-cleared algorithms are in that group. You're right. Some are workflow prioritization. Some are reporting. But we think the biggest bucket will be as an aid to diagnosis, and that's just starting to be commercial. And that's why we're ramping up our efforts in there. But having said that, we work in the other spaces as well. So we don't see it as a disadvantage, a bit like the question about Harrison earlier. We see it as a plus.
Right, and secondly, any update on military tender work?
They've still progressed, but nothing new has come out. They're still talking about RFI, RFP. We are progressing with our ability within that cloud. But like most things with government, I think we've preempted this. It moves slowly and moves in fits and starts, and this is no different. But we have someone on board we've had since Harrison.ai, who's an expert in government and government sales and speaking government, as we call it, and navigating all their various ordering and contractual systems. So we have bulked up there and have someone with a lot of experience. It's now joined our sales team as of RSNA.
Thank you. Your next question comes from Wei Sim from Jefferies. Please go ahead.
Hey, Sam. Hey, Clayton. Thanks for hosting us at RSNA last year. Coffee was really great, and thanks for today. A couple of questions. The first one is just related to Trinity. My understanding is that at the time, Trinity was looking for kind of like a SaaS deal where just the economics of the cloud was also borne by the software provider. So I just wanted to understand just in terms of how the accounting for Trinity works. Is that any different from any of the other contracts that we've done, or would the economics and the accounting be the same?
Yeah. So we have two models and have always had two models. One is the true SaaS model where the client says, "I don't want to deal with a cloud provider. I don't have that wherewithal. I don't have enterprise agreements." So we charge them, and that charge is based on number of exams and number of gigabytes because one exam could be 35 Mb. It could be 4 GB . So we make sure in that that's all done below the line for us. So when we turn around and say the contract's worth AUD 30 million, and Duly will be a case in point, it's AUD 30 million in revenue to us. Cloud fees are not included in that. For the bigger guys, a lot of them have already done what they call enterprise commitments to one cloud vendor.
So they've gone all in and said, "We're going to spend X over five years." And for that, they get a discount. Now, if they don't spend that X, they still pay for it. So in that model, we use their cloud agreement with their provider. We call it their tenant. And so the cloud fees go between the cloud provider and the client. We don't even get to see them. We don't do anything. Trinity is that second model.
So it's similar.
Yeah, which is no different. So we don't have any liability for the cloud fees in terms of an expense for us. That deal is purely software, so the same as the other ones.
Okay. Got it. And then my second question is just regarding this Cardio client that we've won. Can you just talk about give us an idea as to kind of like how the unit economics and how we're charging for the Cardio product versus I'm assuming they already have a radiology product on a per-scan basis.
Yes, yes. They're already per-scan.
Yeah. So, what kind of uplift should we be seeing and how should we think about the unit economics? I guess an extension to that, I think we've kind of like talked about where we think cardiology could go over the medium term in terms of revenue contribution and as to how we're thinking about that currently. Thanks.
So we've always said cardiology volume is much less than radiology, but the amount we can charge per unit per transaction is more. And think that we've always said roughly double and that it's within that realm. So the pricing structure we always had in our minds and the volume structure, it has been correct. So we charge materially more, but there are fewer of them. So we think the cardiology market changes because cardiac CT is growing rapidly. So it's not fixed. Somewhere around 20% of radiology in dollar value, not in exact numbers. And that seems to hold true with what we're seeing with our product in these early stages in the market. So look at it as roughly double cost, maybe touch more. Look at it as much smaller volumes in radiology, which is what you'd expect, but cardiology volume's growing, particularly with CT.
Thank you. Your next question comes from Sarah Mann from Moelis Australia. Please go ahead.
Morning, guys. Just a first question on competition. So you've obviously come back from RSNA. You've confirmed that you're still kind of 12-24 months out of competition. Just wondering, have there been any, I guess, newcomers coming into the market, any changes from the incumbents who are trying to catch up, or do you actually think you're broadening the gap versus the competitors at the moment?
Yeah. Technologically, I'm sure we're broadening the gap. So I think for a number of reasons, cloud, even as I mentioned, the spatial computing, there were a whole lot of people from all these different hospitals that came to Sharp as a health system, but they had this whole spatial computing lab. And you could just start to see where a year ago you thought, "Where's this going to be used?"
We're starting to see some real-world applications for it in healthcare, as I mentioned, starting to come out. And I think as that becomes bigger and bigger within an organization, if you can't do it, it's sort of a box that you can't tick. And I think cloud used to be like that. So I think from a technological point of view, if anything, I think we've gone even further ahead. There have been some new entrants into the market.
One is owned by a publicly listed imaging center provider, made a big splash, but we don't know anyone that's gone for it. There's some others that have been born in the cloud that were showing their wares. There'll always be that. There'll always be that dynamic because it's a big market. But did we see anything that really changed the landscape in our minds? No, and we still see cloud as a big differentiator.
Okay. Thank you. And then just on pricing, so it seems like you haven't really received that much pushback. Obviously, we've been through a period of kind of high inflation. That's starting to come back. Are you seeing any signs of sensitivity to your normal and your price increases from either new clients or some of the renewals you're doing, or is it the inverse just because you've got a reputation now it's actually easier to get better pricing?
Well, we always see pushback. So let me start there. No one wants they all want to pay less. That's a normal dynamic. We are seeing a greater understanding of the ROI. And again, this is particularly in the more price-sensitive private market where people are saying, "Okay, it's not really what I pay. My biggest cost is radiologists. And if I can make them X%, which is only a few% more efficient, right, I'm home and hosed if I can make them 20% or 25% more efficient." And that's out of the gate. You don't need extra AI and all this other stuff. That's all on top. Then they're starting to see the value. But there are some. There are some that turn around, and all they see is the sticker price, and they will buy something cheaper. Now, there are two buckets to those.
There are those that actually think they're getting a cloud product and haven't even tried it before they buy. And usually, they're the ones that become very unhappy, or the implementations get stalled and take forever. Or there are those where they look at price and they do a pilot. And wherever they do the pilot, we pretty much get that business because the difference, the delta is so huge that you can't help but see it. So everybody's still worried about price. We have been able to get price accretion. A lot of that is because of reputation. We'll always have some that will look at price and will do a pilot and then begrudgingly say, "Yeah, I'll go to you," or, "I'll go somewhere else at least in the interim." Now, some of them, as you know, Trinity went elsewhere a number of years ago.
That project failed, and then they had to stop it, go back out to market, which no one likes to do. But we're going to see more and more of that, I think.
Thank you. Once again, if you'd like to ask a question, please press star one on your telephone and wait for your name to be announced. Your next question comes from Peter Meichelboeck from Select Equities. Please go ahead.
Hi guys. Just a couple of questions. First one around the pipeline, more around the timing rather than the size. Given that you've got virtually all the contracts in recent times have been full stack and with obvious benefits around higher value and greater sort of embedded customer relationship, I'm just wondering, has that shift towards full stack meant that you've now, has there been some contracts or pipeline opportunities that have perhaps taken a bit longer to come through because customers are sort of holding off a bit longer on replacing one module in order to align it with the timing of a replacement of a latter module?
Not really. So not every client will take full stack. So if you look at a remote reading group like CRL that reads for Allina, which is a client of ours, so they knew about Visage. They will only ever take the viewer because they don't store the images. The images are sent to them. They make the diagnosis, and the report is then sent back. So they don't need the archive because they don't store them. They only read them. So it really depends on who the client is. But yes, we haven't noticed that. A lot of them, actually, full stack makes it quicker for them because they don't have to go back out and do three separate RFPs and assess three separate products.
Right. Right. Okay. And looking at the second one, second question's probably more of a hypothetical one, but I'm just wondering with some of the larger groups that you've now sort of contracting with, and obviously, M&A is sort of a part of some of their businesses, I mean, in terms of acquiring hospitals. Just wonder what would happen to your contract minimums if they sort of hypothetically sold a bunch of their hospitals that you'd sort of contracted for?
Yeah. So these really big groups are interesting because they're always buying and selling. They're selling off three here, buying six there. It's almost like someone with a real estate portfolio. Not that I'm trying to make it that basic, but they're buying and selling all the time. But the general thrust of all the buying and selling is that they're growing. And obviously, with Trinity, that was factored in. Growth was. We factored in minimums that moved up over time. And so the general trend of our clients is growth, not shrinkage. We haven't had anyone shrink, quite the opposite. So while they may sell a little bit here and there, they'll buy more than what they sell.
Yeah. The minimums in those contracts have been negotiated on that basis. So if they sell, the minimums are still in play.
Yeah.
Right. Great. Thank you.
Thank you. Your next question comes from Christine Trinh from Macquarie Bank. Please go ahead.
Hey, Sam. Hey, Clayton. Congratulations on another strong result. Just a quick one from me on Elucid. With the FDA approval of PlaqueIQ and commercialization expected in the next six months, I mean, how should we think about this and other future AI offerings in terms of revenue generation? Thanks.
Yeah, so remember we said that there's three buckets of where we'll get our algorithms from. Third party will be one and most probably the biggest because there's just so many of them, so with Elucid out, the commercial agreement is that we get a certain markup on what's sold. And we will be their route to market for part of that markup, so we will get a pass-through on everything sold. And we will also get a benefit because we're a minor equity player if they make money on all of that. But leaving that aside, we'll get a pass-through. That's the way we see it with most of the third parties where we will get a percentage of the sale for integrating it and doing all of that. They'll still be responsible for the algorithm. They have to be.
We'll be responsible for the integration and the client-facing side.
Thank you. That does conclude our time for phone questions today. I'll now pause for a brief moment and continue on to webcast questions. The first question from the webcast today comes from Kay Kim. Kay Kim asks, "Are there any future plans for splitting of shares for more retail participation?
This was a question that was raised at our AGM in late November, which our chairman answered that would be something that we would look into. Given the timing of that RSNA and now our half years, this hasn't been raised yet at the board, but it would be something that will be discussed in the future.
Thank you. The next question from the webcast comes from Anthony Corrigan. Anthony asks, "Do you think you should start to increase the dividend into the future?
At the moment, our guidelines for dividends is roughly 50% of retained earnings. And obviously, we reassess that every half before we announce the dividend. I think the board is comfortable with that. And I think we're comfortable retaining a growing pot of retained earnings to give us flexibility, particularly around investment, M&A, and of course, most importantly, investing in the business as we need it. So we do assess it every half. At the moment, the guideline's roughly 50% of profits per period.
Thank you. The next question comes from Curtis Larsen. Curtis asks, "Can you comment on your exposure to the AUD/USD currency and to what extent of hedging, if any, do you contemplate?
90% of our revenue is in North America. So we have a large exposure on an ongoing basis to US dollar. We bill on a quarterly basis, and we hedge on a quarterly basis. So it's a very active hedging policy that we undertake. We have increased the amount of revenue over the years in US dollars, but clearly, the gains and losses that we see go through our P&L have decreased, if anything, because of that hedging policy.
Thank you. Your next question comes from Claude Walker. Claude asks, "With these results, Pro Medicus has grown half-year profit by more than 30 x in the last 10 years. Many senior employees and leadership are extremely long-serving and also own shares. Can you share any insight into staff retention and how people are staying grounded and cohesive in the face of so much success?
Thanks, Claude. I appreciate that question. I think a few things. One, money is only part of someone's employment, and we try and make sure people aren't pigeonholed in multi-skill. We do expose everybody in our organization to the clients, which they find quite stimulating and invigorating. As you know, some of our clients are some of the top radiologists in the world. I think there's that. I think there's also the part that we do good. We move the needle clinically. We don't do anything in other areas that people deem to be not within the ethical framework. We have a lot of people that are really pleased to be working for us because of what we do. Yes, we have an LTI scheme and an STI scheme, and the LTI does give equity.
For those that most of the people that got it have been keeping it, that becomes a meaningful contribution. Some use it for buying houses and paying off mortgages and all the good things that make them feel good about the job. Then we work together as a team. At RSNA, we try and bring as many of the people to be part of it as possible. It's a combination of all those things that has worked. If you saw our AGM presentation, which I believe you would have, 30% of our staff have been with us for long periods of time. We have one that's been with us over 30 years, a number over 25. We even now have second generation where parents and a child is working for us in different roles.
So all of that, we think, helps build that culture too.
Thank you. The next question is a follow-up from Claude. Claude asks, "Has Pro Medicus lost any tenders to competitors in the last 18 months? And if so, what was the reason that the competitor won the contract?
Yeah. So always there are two things. Some, where we particularly smaller ones, where you may not even get a look in. The ones where we've competed on RFPs, I'd say 99% of the time, it's on price and the perception that the sticker price of another product is cheaper, and second, it's not as good, but it'll do, but having said that, our win rate has been consistent, and we win far, far more than we lose, but you're not going to win everyone, at least not first time round. We are seeing some, and Trinity being a case in point, but by no means the only one where we lost them originally, and then a short time later, they've come back to the market, and we've been able to pick up that business, so yeah, our hit rate's as good as it's ever been.
Thank you. Your next question comes from Curtis Larsen. Curtis asks, "Do any of the module updates, for example chat, entail additional pricing on top of the existing contracts?
Yes. All the module updates. So we do update the program, and that's part of the existing contract. So in my interview, I talk about we do two or three major updates to the Visage 7 viewer every year, and everybody gets those. But worklist, archive, and chat are all additional modules. And you saw in the half, we had two big archive sales, and they can be quite material in terms of revenue.
The next question from the webcast comes from John Hester. John asks, "Are there any top 25 IDNs in the pipeline?
We don't actually answer that question other than saying we have opportunities of all sizes, so that should give you some hint that we've got some medium-sized regional groups, some smaller groups, and a fair smattering of quite large ones.
Thank you. Your next question comes from Sarah Syed. Sarah asks, "Implementing the Baylor Scott & White contract in just 11 months sounds quite commendable. Could you provide some color on how this timeline compares to the industry standard for rolling out similar enterprise imaging solutions?
Yeah. So the 11 months was from date of signing. So there's a lot of work that gets done in the background from signing to the first go live. And that's actually the biggest window. Now, all of that is working out which sites are going, when they're going to go live, things like data migration that we start, things like testing all the interfaces and everything else. So Baylor got signed in October, November previous year. The first sites went live in June. So between October, November, and June, it was all that under the bonnet work. And then they had what we call three mini go-lives and one big one. So a go-live means where they switch off their previous system on Sunday night and start with Visage in that region for everything Monday morning. We never run two systems together.
So the actual cut across for a radiologist is stop the system one night, get the new system the next day. So in terms of windows, it was very, very quick. It went from June, and we did our final one in September. Now, normally, that would take 3 years + for some of our competitors. So from signing to complete implementation was 11 months, including all the preliminary work before go-live.
Thank you. The next question is also from Sarah. Sarah asks, "Could you quantify the gap between the minimum contract value and actual revenue, highlighting any upside achieved? What has been the general trend in this regard?
Yeah. I mean, the minimum contract value that we normally announce is around that 80% of their volume. So there's 20%-25% inherent upside in all the contracts. And that's what we're generally seeing within their first year. We do see then our customers grow between 8% and 10%. So we do see the upside. So if the minimums are 80%, two or three years in, they could be doing 120% of that volume. But we have seen upside of double their volume of what they originally signed for us. So we do that is purely what we announce is what we say, which is the minimum. We always see an inherent upside in all the contracts.
Thank you. Your next question comes from Peter Kempen. Peter asks, "With Sam and Anthony selling 1 million shares in each of the last two years, at least, is there any organization building up a substantial stake in the company? And do you see any natural acquirers for the business?
So in terms of the first question, no, there's no one that's built up a substantial stake. We don't have any substantial shareholders. That is 5% or more except for Anthony and myself. And so we've seen, as far as we can tell, when we have sold that there's been a spread of institutions that have taken those parcels. So we don't have anyone substantial or close to substantial. It's a pretty well-spread register. And the other thing we're noticing is more and more of the register being offshore, as you would expect as we become more and more international.
Your next question comes from Daniel Arnold. Daniel asks, "What is your view on OpenAI or Meta AI's recent investment into scaling into Pro Medicus's markets?
I think it's complementary. A lot of that generative AI is something one of the analysts before alluded to it, I think it was Julian, about generating the report. So a lot of the generative AI, what it does is it will read the voice recognition file and then generate conclusions and do things like that. That's all additive to what we do. It's not competitive to what we do. So I don't think it's really changed the core AI that we use. It's sort of an extra area that radiologists will use rather than replace what we're looking to do.
Thank you. Your next question is a follow-up from Daniel. Daniel asks, "What is your view on the Trump administration cutting HHS funding to hospitals?" He also asks, "Is there any feedback on channel spend?
We haven't noticed any. People talked to us a few years ago about interest rate rises. As I said, as that increased, as it made it harder to sell, we didn't notice any delta at all. We didn't notice any slowdown before the election. Although it's not been a huge difference in timing, we haven't noticed any change from any of the Trump being in and any of his new policies, certainly not at this point. We don't believe it will impact us materially from what we understand about what's been released in the market to this point. Certainly, we haven't felt anything there.
There are no further questions at this time. I'll now hand the conference back to Dr. Sam Hupert, CEO. Thank you.
Thank you, and really, I wanted to say thanks to everybody for participating, and we appreciate you listening in.
Thank you.
That does conclude our conference for today. Thank you for participating. You may now disconnect.