Good afternoon. It's John Ransom. I've got the full complement of the Health Catalyst team with me today. Welcome to day two of our conference. We're officially over the hump now that we've had lunch and the back half of the journey. This is going to be a 100% fireside chat. I do have some questions, but if I run out of time, we'll have to talk about something else. I'm sure you guys will help me out. I think just for starters, you guys have been public since 2019. Maybe just sort of describe the ebbs and flows of being a public company from 2019- 2025 and just what some of the key milestones and developments as you see them to this point.
Yeah, absolutely, John. It's great to be with each of you. Good to see you. Thank you for hosting us at this conference. You've been there for those six years with us. When we.
My memory's not that good, though, so we need some help.
We're both getting older, aren't we?
Aging in place.
That's right. That's right. When we went public, we were about a third of our current size from a revenue perspective, with just over $100 million of revenue, and most of that was recurring. We were losing a lot of money. That first year post-IPO in 2019, I think we lost about $40 million of adjusted EBITDA. Since then, we've about tripled in size. This year we'll hit, at the midpoint of our guidance, $335 million of revenue. Still, more than 90% of that is recurring in nature. Our mix has changed a little bit. When we went public, we were closer to 50/50 tech and services. This year we'll be closer to two-thirds tech and one-third services. We have expanded in a couple of other ways as well.
Our team member base has about doubled, and then our client base has grown by about 10X. That's come through organic growth, as well as we've been a consolidator. We've acquired about a dozen point solution companies over the last six years, and that's added to our client base. Now we have just over 1,000 clients, most of whom are more app layer clients that do just one or two things with us. We've grown our largest client base, our platform clients, from about 50 when we went public to 130 as of the end of last year. It's been a true roller coaster ride. When we went public, it was seven months before the pandemic, so that was great timing to be a public company. After the pandemic, we had 40-year high inflation and kind of crushed our end market for about 18 months.
That was really tough as well. We went from being a real high-growth company with 20-plus % growth down to our growth slowing in 2023 and 2024 in particular. We are excited to see that growth coming back. It is starting to reaccelerate. We are approaching that double-digit growth. We are sort of 9% and change this year, year-over-year growth, and excited for that to continue as our end market improves and as we continue to invest in technology that makes a difference.
Anything to add, anybody? That was good.
No.
At the big competitor conference in January is when you announced the Upfront deal. The market's.
Super popular.
Yeah, the market doesn't appear to be enthralled. I think just the obvious point of what revenue multiple you're trading and where you place your bet here. Just kind of talk through those numbers and what, if anything, people might be missing there.
Yeah. No, I appreciate it. Jason, Daniel, please add as well. Yeah, we got some pretty direct investor feedback that there's a lot of sensitivity to dilution. I get it. I'm a top 15, top 20 shareholder myself. I felt that as well. With each of these acquisitions, it is always a judgment call. Often we're not in direct control of the timeline. That's better. Pivoting's a little bit better. There are sometimes just windows that open where high-quality companies become available, and then the window closes. We'd already been fairly acquisitive this last year. We'd already acquired three smaller, more tuck-in companies, and this one was a little bit larger. It wasn't the ideal timing from our perspective, but it was a really high-quality asset, a really high-quality company in one of our five areas of focus.
They were recognized as the industry leader in that area of focus around patient engagement, patient activation, very important to many of our clients. It was not a cheap acquisition, and it was one of the larger acquisitions. We made the judgment call that we believed with their growth profile, with the strength of the ROI of their offering, the quality of their offering, that we would get a return on that investment. We get it that investors expect that. With the guidance that we provided in January, there was not an immediate obvious return. There was not incremental profit. There was not incremental growth that we had yet projected. Last week at our earnings call, we made what I refer to as the first installment on a return on that investment in Upfront that we raised our EBITDA for this year by 5%. That is an installment.
We took a 10% dilution hit. We have more work to do, but we're on it. We're focused on it. I do believe we'll look back and feel good about that acquisition, but it was expensive and it did cause dilution, and we're sensitive to that as a company and as a board.
Anything, Jason, Dan, you'd add?
Yeah, the only thing that I would add is I would echo Dan's comments around the growth profile of the company. We are really excited about Upfront's growth profile. They are a high-growth company. Like we've mentioned previously, we like to bring these acquisitions over at adjusted EBITDA at or near adjusted EBITDA break-even. In the case of Upfront, we do see an opportunity to reach meaningful adjusted EBITDA levels in the near future. I mean, our target as a company as a whole would be a 30% technology adjusted EBITDA margin by 2028. In the case of these acquired targets, we would expect to reach that level a lot more quickly than we expect to get there as an overall company. We're seeing that progression already with Upfront.
Just to, I think sometimes healthcare tech uses words and we kind of know what the words mean, but let's get one level. I know what patient engagement means, but in a real specific sense, when one of your clients uses them, what are they getting that they couldn't get anywhere else or not as, and maybe not the quite quality way?
Yeah. Good question. Upfront in healthcare, their real differentiation and strength is at the top of the funnel when you think about getting patients introduced and becoming a part of a health system patient community or member community. Their ability to kind of reach and activate patients in a really effective way. They use some really interesting psychographics. They have really good analytics to understand who would be an ideal patient profile, where would be the ideal place to engage, and how. Unlike some other companies in this space where they kind of stop at the point of analytics and insight, Upfront keeps going and really measures their success based on number of incremental patients that are kind of added into the system. That has a very real hard dollar revenue impact and profit impact to these health systems.
Would this be like I've scheduled a surgery and then they leap into action, or is this more missionary work before patients?
It's more missionary work. They're good at it. They're good at converting those folks that they engage with. That's where that ROI comes in. Now, lower in that patient relationship management funnel, you need to keep them engaged and make sure they have a good experience when they have a procedure and there's good follow-through. We have really good assets that go all the way through that experience. At the top of the funnel, we didn't have real strength there, and they really fill that gap.
How are they getting people excited about BayCare, just some big hospital system? What do they do to spread the word specifically?
Yeah. There is meaningful personalization and psychographics that they use in their analytics to target and understand a little bit more about you, a little bit more about what setting would be best for you, and what medium would be best for you to be reached. It is proven quite effective. When they have the right amount of data, their technology does a good job of targeting and personalizing using that data, and it has a really high conversion rate.
Psychographics sounds a little 1984-ish.
In a good way.
What are psychographics exactly?
I think understanding certainly demographic elements, but then behavior-based elements as well of how you've interfaced with the healthcare organization in the past, what has appeared to be more effective or less effective, which kind of setting works best for you.
Interesting. I think if I were to ask people the difference between DOS and Ignite, they may not make an A on that test. I get a little fuzzy myself. In layman's terms, how do we talk about Ignite versus the old DOS platform? What specific, again, sticking with specifics, a customer that migrates, what do they see out of that migration from their standpoint?
Yeah. I'll share a few thoughts on that. Daniel is leading that migration effort. DOS was heavier, and we built a lot of the componentry ourselves because six, seven, eight years ago, there was not a lot of cross-industry componentry available. If you needed it, you kind of had to build it yourself. That is never going to be the best option from an efficiency perspective. Over the last six to eight years, you have cross-industry players like Databricks and Snowflake, and even the cloud vendors have developed really meaningful, interesting componentry capabilities that allowed us to retire some proprietary elements of DOS that just made DOS more expensive and heavier. The DOS price point to get started was typically $1 million-$1.5 million.
If you think about trying to cross-sell our app layer clients who spend $60,000 a year, $100,000 a year on DOS, oh my goodness, no way, right? You're not going to jump from $80,000 a year to $1.6 million a year. Ignite is totally different. It's built on top of that cross-industry tech. We've stripped out the componentry that we can now get better, faster, and cheaper from cross-industry vendors. It's much more modular also. If a client is heavier on AWS versus Microsoft from a cloud perspective, no problem. We can kind of move modules and components to match their existing investment, those cross-industry technologies, and the price point is much lower, and it's more flexible. We can kind of choose the components of Ignite that make sense for that use case.
Now when we're cross-selling one of those 900 app layer clients that may be spending $100,000 on VitalWare, we want to talk to them about Power Costing or Power Labor. Instead of going from $100,000- $1.5 million, we can go to $300,000 or $400,000. We can just use enough Ignite glue to kind of make that adjacent use case work. That's really fundamentally different. 2024 was the first year where we could start selling Ignite. It was ready. That's where we saw this increase in the number of new clients, net new clients that we were able to add. Two-thirds of them kind of came from our app client base. That's informing why we think we can add a lot more new logos this next year. Ignite just makes it so much easier.
Dan, what would you add?
Just one of the innovations that's been incredibly helpful is the elastic compute capabilities of these Snowflake, Databricks, cross-industry experts, meaning that you pay for what you use. You do not just buy a bunch of storage and compute capability, and you pay for that. Whether you are using it or not, you are able to sort of adapt to the use cases and pay as you go. That becomes an incredible efficiency savings opportunity. That is where a lot of the cost savings come from when we leverage those things. We can plug in at the next layer on top of those infrastructure capabilities and create the mappings to healthcare data sources and the healthcare-specific content and terminology mappings that we are differentiated at. We do not have to be building that componentry down at the lower level.
If you've got theoretical 100 hours to go market or go to market, how are you spending your time now versus, say, a couple of years ago when you were still out there trying to slug away at a million-dollar DOS sales? How are we spending time hunting new customers versus trying to convert app layer clients to Ignite and then just continue? How much of it is like going back to your old friends with the, "Let's try these two more modules we think would be great," versus, "We want to just keep bringing in new clients into the door"? How has the focus shifted, in other words?
Yeah. Ignite changes our focus, certainly, and it opens up real meaningful possibilities. If you think of our historical, there are two growth building blocks. There is what are we doing with existing clients? We typically have talked about our existing platform clients as our largest clients and what kind of expansion is possible there. The second building block has been, how can we add new clients? In the past, with DOS, it was such a big price tag that the best we had ever done before last year was, I think, 15, 16 new logos. They were at a seven-figure price point, but we just could not add more than that. With Ignite now, we are projecting 40 new logos this year, and I believe we can do a lot more than that in the future. That is where our sales focus has shifted.
It is focused mostly on those 900 app layer clients that we have measured and found we have about 2x the conversion rate when we are selling into an existing client relationship, even a small one. The sales cycle is shorter. Our number one focus is on those existing app layer clients, converting them to platform. I think that will be the single biggest growth driver, growth building block in the next five years. We still have a second focus on our 130 platform clients. Now, those clients spend $1.5 million a year with us on average, $1 million of tech, about $500,000 of TEMS. They are big clients, and they have been loyal to us. There is still growth opportunity. We can sell them on more apps, like you said. They provide a really good foundation of our growth.
We've projected a 103% dollar-based retention for that platform client base. They'll contribute to our overall growth. I think the biggest growth driver will be really converting those app layer clients.
The typical app layer client, the next thing they buy, is there like a pattern there, or does it just all depend on what they're looking for?
Yeah, we do see a pattern that we're really excited about. It is the next adjacent opportunity. We're investing in that growth motion where if you're a VitalWare client, what's the next adjacent thing for you to logically buy? For VitalWare clients, they're in the finance function. That's a revenue cycle charge master management solution. What else can we sell within that same decision-making group? It's Power Costing and Power Labor, which is on the cost side of the finance team. Mapping out that next adjacent opportunity, if it's a Twistle client in the clinical improvement category, patient engagement category, now we've got Upfront as a next adjacency, as an example, or Lumeon as the next adjacency that helps us convert an existing client into a deeper relationship. It's a logical next step.
We're spending a lot of time mapping all of those next adjacencies and then doing a really effective job at reaching out to those 900 clients and sharing that invitation with them.
Are you seeing clients migrate more to revenue cycle, just put simply, trying to get more dollars, or are they attacking the cost side, or does it depend? Where are the hospitals focused these days?
They're definitely still focused on hard dollar ROI. That tends to skew a little bit more towards the finance function than, say, clinical improvement. We have seen clinical improvement come back as operating margins have gotten better, but they're still very much a focus. I would say, in our experience, they're equally interested in the revenue side and the cost side. We have a lot of activity on the revenue side with VitalWare, a lot of activity on the cost side with Power Costing. There's meaningful operational interest as well. How do you improve throughput? We've got some good assets there and good traction there. Where there's a better, clearer ROI, like what Upfront brings to the patient engagement equation.
Before Upfront, we had a great asset in Twistle that's really down the funnel, more about when you've had your procedure, let's have great patient communication and make sure you had a really good experience. There's not necessarily a hard dollar ROI always associated with that. With the addition of Upfront, now you've got a real ROI that the CFO will agree with. We are seeing more activity over there.
If you were just to sample all your clients, what are the common things that you see that hospitals all could do better, that do maybe not as well as they could?
Yeah. Everyone's trying to figure out AI, right, in healthcare and everywhere else. What we're seeing as the most promising use cases for AI kind of attack, I think, one of the biggest areas for improvement and opportunity in healthcare, which is more efficiency in processes, more efficiency in administration, more efficiency in the way that we deliver healthcare. It's meat and potatoes. It's not really glamorous, but it's really important work. We have several AI use cases that are in that meat and potatoes category, but we're seeing some meaningful efficiency gains. I think healthcare needs to get better, faster, and cheaper. We need more efficiency in the care that we deliver. I think technology, in general, has a really useful role to play there. AI can play a role there. We still see a lot of science projects as well that aren't producing an ROI.
We get really excited about the ones we can see an ROI.
What's sort of a live bullet example of a process that used to be manual that's now been taken into the AI realm on the provider side?
Yeah. For us, chart abstraction is something every health system has to do, right? You've got to build registries, then submit them to various bodies if you want to get your Medicare payments, incentive payments, be certified or recognized as a trauma center or in lots of other ways. That's been historically a very manual process that required clinically trained resources, expensive resources, nurses, and other clinicians to do that manual work of hunting for the data in the charts, putting them in the registry, making sure that it's accurate. It was actually error-prone as well as being expensive. We've invested in and acquired some meaningful technologies that have automated several steps in that process. The latest example is specific AI.
It's probably our most promising use case of AI where we have seen an additional about 25% efficiency gain, where with many of our clients, they've asked us to manage that process end-to-end, not just the tech, but the team. We're managing those clinical resources that are doing the chart abstraction. What we found is, as AI gets trained better and better, AI's ability to predict where the data is that we need, where that data goes within the registry, interpret free-form data correctly. We're not perfect, but we're about 91% accurate. You still need that final check with a clinically trained resource to make sure that the data is going to the right place. It makes sense. It's in the right registry definition as well. With that 91% accuracy, we're seeing 25% improvement in efficiency. That's probably the best real result.
One of our companies has gotten pretty down the road with the AI reads for imaging scans. I mean, they charge $40, and the scan finds the breast cancer two years sooner. I mean, it's a pretty good use case. Are you guys seeing any opportunities with stuff like pathology, radiology scans, and getting an AI interface, or is that something kind of outside your purview?
It isn't one of the five use case areas of focus for us right now, but we're excited to see the progress there and applaud others who are pursuing that path. We do have some support that we provide from an infrastructure layer perspective, but those aren't specific use cases that we're pursuing.
Yeah. Okay. What about just the large language models and taking dictation and reading contracts and using the language models in ways that were not available three or four years ago? Is there anything you are doing there?
Yeah. We've got a couple of other use cases that are interesting to us. One of our recent acquisitions was in the cybersecurity space. There's a meaningful AI use case there where, in the certification process, you need to understand how each vendor is performing. There's quite a bit of language that can be leveraged by AI models to really speed up the efficiency of getting the certification, maintaining the certification, anticipating where there may be problems or vulnerabilities, where AI is making humans much more efficient in that regard. That's another use case.
I don't know, Daniel, anything you'd add there?
The other use case that we're excited about is our healthcare.ai capability, which is basically an automated way to interpret analytic insights. We find if we had 100 people in this room and we sort of surveyed, gave them a basic run chart, and then asked certain questions about how to interpret that data, there would be a pretty low just general capability in interpreting that data correctly, providing predictive insights, saying if something is out of control or not within just the basic run chart. This capability sort of adds on that automated AI use case to interpret that data to deliver insights without someone having to study that chart and identify those different insights. We're really excited about sort of empowering those analytic insights that are actionable to make better decisions.
This is kind of an in-the-weeds question, but I know some of the more forward-thinking payers are telling us that, hey, this predictive, taking multiple data sets and then using predictive analytics to catch something before it happens. Yeah. Like a classic example would be somebody that just got discharged. They got arrested. They missed a script refill, and they showed up. They're having a life break event that's kind of evident, and so you try to get that person some help. Is there something like the, how far out are we able to kind of overlay different data sets and start getting people getting a little more into the prevention? I know hospitals like are some of them have these pop-out contracts, and anything like that would be helpful. Is there something that?
I love those use cases. One use case that we found great success with, even prior to ChatGPT and a lot of this innovation, is surgical site infections or hospital-acquired infections. The ability to evaluate the conditions of a patient, their background, their profile, the wound site, etc., and then predict their propensity to develop an infection and intervene before that infection happens. We did a bunch of case studies with IU Health and other clients where we demonstrated incredible savings in that preventive type of care. Those are the types of uses.
Hospitals got very excited about that because they got dinged and then also readmissions.
Exactly.
Those are the two things I saw them actually react to because there was a marker out there. They got.
Totally.
Yeah. Okay. What I'll kind of leave with is, as we look out, obviously, your stock's trading under one time's revenue. It's historically pretty inexpensive. At the same time, things are inflecting. As we think about the out year, let's just pretend like you can grow 10% a year for the next three years. If we were to adopt that theoretical construct, what's the EBITDA drop-through look like? How do the variable costs grow versus the revenue? Are those moving independently, or is there a certain underlying level of fixed costs that need to go up? Just trying to think about the long-term growth algo here, at a kind of low double-digit revenue growth.
Yeah, absolutely. I think we're excited about the next few years as a company. This has been a painful transition. It feels like when we went public, we were an adolescent. Man, I love those revenue multiples. 11, 12. That was fun. Every company has to grow up. Now, to your point, we're valued as an EBITDA multiple and a fairly low EBITDA multiple, right? Eight, nine times at today's current valuation for this year's EBITDA. That's been a challenging transition, no question. As we look forward, I think there's a couple of tailwinds that we feel really good about that we're excited about. First, we're ahead of schedule a bit as it relates to our EBITDA progression. We have swung meaningfully from a negative $40 million of EBITDA when we went public to positive $41 million this year.
That is on a trajectory where we can continue that. We have shared a 2028 EBITDA goal of $100 million, and we are on schedule to a little bit ahead of schedule there. That implies about 40% year-over-year growth for the next few years. This year, we are growing our EBITDA 57%. We feel good about the leverage. That technology mix shift, where we were more like 50/50 tech and services six years ago, and this year, we are more two-thirds, one-third. A lot of the growth is actually at the most profitable layer, that apps layer. That all is a tailwind as it relates to our EBITDA profitability. As Jason mentioned, even as we have acquired companies, they have almost all been at the apps layer. Our ability to turn them into really profitable assets is another tailwind.
I'm excited, and it's one of the reasons I've been a buyer of shares in the open market at a significant level. I'm excited to be a top 15, top 20 shareholder because of the trajectory that we have with regards to EBITDA to grow from where we are today. We just raised our EBITDA by $2 million for this year to $41 million to get to that $100 million level in the 2028 timeframe. As we do that, there should be meaningful shareholder value creation. Even if we're never going to get back to those 11 times revenue multiples, I think, even at a kind of a reasonable EBITDA multiple, with that kind of EBITDA expansion, there should be a great opportunity. Not a lot of companies are able to grow their EBITDA by 40% a year the next few years.
Right. We are at time. Let's get a breakout. Thank you.
Okay.
Thank you, gentlemen.
Thank you, John.