Host Steve Krupa of Psilos Group interviews Dan Burton, CEO of Data Warehousing powerhouse Health Catalyst about fund-raising, resisting early strategic sales and how Big Data is changing health care.
Tom Salemi: HI, this is Tom Salemi. Welcome back to the Breaking Health Podcast, the podcast affiliated with the Digital Healthcare Innovation Summit, which will be taking place on November third in Boston at the Mandarin Oriental Hotel. I’m here with our host, our great host, Steve Krupa, CEO of the Psilos Group. Steve, nice to talk to you again.
Steve Krupa: Hi, Tom, how are you?
TS: I’m great. And I’m particularly excited about today’s interview with Dan Burton. I thought you both had a great exchange, just sort of VC to CEO. It was like I was sitting in on a private conversation. It’s nice to see some of the issues you addressed. And I want to get into that interview in just a second, but I want to also take a moment just to sort of get into the Breaking Health name of the Podcast and the email newsletter that we’re sending out that will be affiliated with the Healthcare Conference. I have to give credit where it’s due: it was your idea, and I resisted it at first, and you just held quietly steadfast and waited for me to come to my senses, and I did. What is the origin of Breaking Health? What does it mean to you?
SK: Well, you know, we were trying to come up with a name for the Podcasts, and I was thinking about the vernacular of venture capital, and I have a friend, Howard Morgan, who runs a very early stage VC firm called First Ground Capital. And when you go into Howard’s office, he gives away t-shirts that have various expressions on them. And one of the t-shirts said “Break, Build, Repeat,” as if that was a mantra for a startup CEO. And it occurred to me that that makes a lot of sense. You’ve got to break it and rebuild it, and that’s what we’re doing with digital technology in healthcare. So I just came to Breaking Health from thinking about that phrase and knowing I really couldn’t take Howard’s t-shirt for his company.
TS: Do you look at that, though, when you’re looking at ideas, at companies? Do you really see how disruptive they will be? And disruptive is such a cliché term, but it really is what needs to happen. I mean the healthcare system is already broken in a lot of ways, and you’re trying to find ways to fix it. But it’s broken in a way that it’s kind of functioning. It’s on life support; it’s not going to go anywhere unless something critical is done to make a change.
SK: Well, yeah. I think there’s a statement that the healthcare system is broken. I think it’s broken from the perspective of we think it’s very expensive and confusing, but actually it has more organization to it today than it’s ever had in its history, you know, going all the way back to the origins of fee for service medicine and so forth. But the truth of the matter is there’s a lot of things that are done in healthcare that do need to be broken or changed, and people have recognized that for years, but they haven’t really been able to figure out what the next step would be once it’s broken. And primarily because of the Internet and primarily because of the advanced capabilities of modern computing, the digital infrastructure is now stepping in and saying, if you make that change, we can add an overlay of technology that will make the new, more efficient process actually works. And I would expect most of the interviews that we do on this show will demonstrate that.
TS: Terrific. And today’s interview is no exception. You spoke with Dan Burton of Health Catalyst. And it’s an interesting company. Can you give us just a quick ten seconds and then we’ll hop right into the interview? Ten seconds of what Health Catalyst is trying to do?
SK: Yeah. So I would expect over the weeks that we do this, we’ll be using the term big data, data analytics, population health. They mean different things to different people, but Dan is really approaching the analytic side of this and the data warehousing side of this from a very unique perspective. That’ll come out in the interview and I’ll leave it to that for the audience to discover that. But I would point out that once we can get all of the health data available to us, then there are uses for that information, or the information that can come from the data that we haven’t probably even begun to imagine. And so Health Catalyst is at the cutting edge of that, and we are really at the very beginning of trying to do and utilize big data and healthcare right now in 2015. So it’s interesting to hear where these CEOs think that capability is going to take us.
TS: Terrific. We’re fortunate to have Dan at our Healthcare Innovation Summit, again on November third in Boston. Go to Digitalhealthcaresummit.com and let’s start the interview.
SK: Welcome Dan.
Dan Burton: Thank you, Steve.
SK: You know Dan, I’ve got to say first off, as a VC, I’m sorry to have missed the opportunity to invest in your company. It’s an awesome business and it’s fitting squarely in the direction of where the future of healthcare is going. And I’ve got to say also congratulations for your success. I feel like I can’t go anywhere your company is not mentioned, where you’ve received several awards. And probably I think the most significant indicator of success for a startup is the fact that you’ve been able to attract a great group of investors and a significant amount of capital. By my count, and of course this is based on public information, you’ve raised over $150 million of capital for the business. So congratulations on all of those fronts.
DB: Thank you. Thank you, that’s very kind of you.
SK: Yeah. So I always like to go back to the beginning. Am I right, 2008 was the beginning for the company at least?
DB: That’s right.
SK: So when you first met your first venture capitalist, what did you tell them you wanted to do?
DB: Well, let me start at the very beginning, which was a couple years before any venture capital investment, and then we’ll get to the venture capital. So the beginning of the company really happened on the health system side. Our cofounders all come from InterMountain Healthcare in Salt Lake City, where they were involved on the technical side and on the clinical side in the process of using data to systemically improve outcomes. They built a data warehouse that had a different architecture from any of the prevailing, typical architectures or models for building a data warehouse that you see prevalently in other industries. So they developed this novel architecture that ended up being very effective at InterMountain. And as other health systems started to ask more and more about how InterMountain was able to harness that data and the analytics associated with it to improve care, that was really the genesis of the idea that this could be a very helpful company to a lot of health systems in this particular phase of the healthcare ecosystem’s maturation. So Steve and Tom left InterMountain, founded Health Catalyst, and begun working with their first health system client, which was Allina Health, and built a data warehouse from scratch and it proved to be flexible and agile and was been utilized by Allina Health to make some very meaningful improvements in both clinical and in some very significant financial outcomes. And that attracted attention then from some other health systems. But from the beginning from a capital perspective, Steve and Tom are both very frugal, very conservative financially. And so they didn’t raise any capital for the first three years, and the company was cash flow positive. I became involved in the company after Health Catalyst was approached by a potential acquirer, and Tom reached out to me because I had had some background in valuing companies and in mergers and acquisitions, just asking for some advice as to how they should think about this potential offer. And as I got to know the company a little bit better, it became incredible compelling from a value proposition perspective for me, and I tried to be transparent with Steve and Tom that I was running a very small private equity firm, and that I was biased, but that I would love for them not to take the offer from this other company, and instead allow us to invest and help them grow their business. So that was the beginning, when it was a very, very small investment and relationship that we started. But within about a year, as the company got more traction, one of the health systems that decided to go with Health Catalyst was Stanford. And Stanford had the Chief Medical Officer at the time, who was Kevin Taub had a relationship with a partner at Sequoia Capital. And he connected us with Sequoia Capital. And their real value proposition argument to us, which we found strategically compelling, was that this was going to be a tidal wave. The need to harness data in order to improve outcomes in a significant, systemic way, that was going to be a tidal wave. And if we were going to stay ahead of that tidal wave and really compete effectively, we needed to invest heavily in our product development so that we could really scale. So we decided that that made sense, and we’ve been on that same trajectory and with that same capital strategy really for the last 4 years. And it’s proven to be a good strategy for us.
SK: Yeah. So let me ask you sort of a personal question and you could decline to answer it if you choose. But around how much would you have sold the business for if they would have sold it quickly as opposed to going the route of building the company up?
DB: You know; it was a seven-figure offer. And at the time, Steve and Tom kind of looked at each other, looked at me and said, Boy, that’s a lot of money.
SK: Yeah, right?
SK: It is.
DB: And yet when you think about – it is, and yet when you think about the problem that this solution helps to solve, it’s attacking really about a trillion-dollar problem of inefficiency in healthcare. And so and it’s such an important mission to stay on. Part of what I shared with them, I’d spent a good deal of my career in and around mergers and acquisitions. And just having seen over and over the number of times where an acquisition of a company really leads to that company losing its identity and maybe losing its ability to execute against its mission, that was ultimately by far the most persuasive element to Steve and Tom as to why they should continue forward on their own is this desire to really help healthcare in general to harness data to improve outcomes.
SK: Yeah, it’s funny. There’s a lot of stories about companies that end up selling too early, right? They’ve got a great idea, and somebody grabs it, and that might have been the case here, but for your advice and then ultimately your opportunity to come in and help them run the business, which I’m sure they’re grateful for. Give me a sense of – I understand Sequoia. So they said to you, Look, this is going to be a tidal wave of basically computers beginning to dig into the operations of the healthcare system and enable efficiencies. But what did you say to them that you thought the real value that you were providing at that time to Allina and to Stanford was when they started to think about investing?
DB: Well, from a pattern recognition perspective, part of what we realized through the first four or five health system relationships that we had was there were a couple of common threads. One was that each of these health systems had invested a massive amount of money in going from paper based records to electronic records. So they’d invested a whole bunch of money, and yet outcomes had not improved. And in many ways, the physicians were a little bit more frustrated with the workflows and the process flows of delivering care afterwards versus beforehand. So they hadn’t gotten a real measureable return from that massive investment in the EMR because that investment had been necessary, but not sufficient. It wasn’t sufficient yet to marry that data up with other relevant data like costing data, financial data, patient satisfaction data, lab data, etc. to be able to bring all together in a flexible warehouse, run analytics on where the improvement opportunities actually exist, and then go and do something about it. So what we really offer to each of these health systems, in many ways, was the opportunity to realize a return on that massive investment that they had made. Now it requires a little more investment, not nearly the level of the investment in the EMR, but with a little bit more investment and a little more focus with this data warehouse as a platform, they could start producing measureable outcomes improvement on a regular basis, which then started to demonstrate that clinical and financial return on that massive investment. So that was really the critical value proposition that you unlock the value of all that investment that you’ve made so that you can start seeing the ability to bend your cost curve, the ability to improve your outcomes in a significant way.
SK: Got you. And do you think that the High Tech Act and the expansion that was in some ways government subsidized of the electronic health record adoption was a big catalyst to your business in terms of more customers recognizing that they weren’t getting what they needed out of that investment?
DB: I think so in many ways. I think one of the real successes of the High Tech Act was the massive adoption and transition from paper based records to electronic records. Clearly, that had a very direct impact on the speed with which that transition occurred. I think it created some challenges as well, but undoubtedly the success in helping health systems embrace the electronic format of data acquisition and then the ability to start harnessing that data through a data warehouse and analytics. It doesn’t go anywhere unless you’ve made that transition from paper to electronics.
SK: Yeah. So let’s talk about the product. I talked to you a little bit about how when I’m out telling my friends at parties what data analytics means in healthcare, I use the Sabermetrics analogy in baseball, that they have all this data about where balls are hit by various batters, and as a result of analyzing that data, they can place fielders in the right position, and the result has been really a significant reduction in the batting average of baseball players, and more of a defensive baseball game today than, say, five or six years ago. So that’s a sort of common man’s version of big data in terms of the analysis, the action, and then the outcome. Give me a sense of what you’re analyzing, what actions are then taking, and what outcomes occur as you begin to get your database built and your analytics rolling on some of this healthcare data.
DB: You bet. So for us, as we think about the overarching problem being that around 30 cents out of every dollar that’s being spent, depending on which study that you read, it’s somewhere between 25% and 35%, of all of the spending in US healthcare is wasteful. The challenge is knowing which 25 cents are the wasteful 25 cents, or which 30 cents. You really can’t pinpoint that at a macro level or at a micro level without access to the full dataset that you need, and then the ability to run analytics using that data to pinpoint where the problems exist. So typically what we do with a health system is the first step is you’ve got to get the data all together in a place that you can organize it and access it in a meaningful way. That’s really what we try to help them with in installing the data warehouse platform. We pull data from the EMR or multiple EMR’s depending on the health system that we’re working with, and then up to 50 to 70 other sources of relevant data that you might need to bring to bear on your analysis to pinpoint where that 30 cents are, whether it’s in the treatment of certain populations of patients, in certain clinical processes, or in workflows and workflow processes that are commonly followed in the process of delivering care. So the first analytics app that we run is a Pareto app that takes all that data from the data warehouse and plots it on a bubble matrix, bubble chart matrix, two by two matrix, where it looks at the amount that’s being spent and the kind of variation that exists. In the upper right quadrant is the quadrant where there’s a lot of cost and there’s a lot of variation, which is a good proxy for inefficiency or waste. That’s basically a roadmap that can showcase to a health system in their particular situation what should they start with, where should they focus, where most of those 30 cents will lie. If we can improve the treatment of these 3 or 4 populations of patients, that will give us the greatest benefit, or these 3 or 4 workflows, these really matter and there’s a lot of variation in the way that we’re following those workflows. That’s what guides us then to do the heavy lifting and the hard work of diagnosing what’s causing that variation, and then fixing it at a systemic level by hospital, by provider, by location and over time. And by having the right analytics platform in place, you can see over time the effect of those activities and those areas of focus, and you can see the cost curve starting to bend.
SK: Cool. I mean you’re obviously a software company. You’re first and foremost a data warehousing company. I don’t think we have time to talk about data lakes and data warehouses and the in between, other than if you could give me a quick snapshot of why your data warehouse is particularly different than perhaps other data models that have been tried in the past.
DB: You bet. I would say the most differentiating element is the pragmatic nature of our data warehouse architecture. It’s fast from the time to value perspective to implement. It is flexible in that in our architecture, we have chosen not to significantly transform the data as it comes into the data warehouse. By not transforming the data into the data warehouse, it allows us to move more data more quickly and less expensively than traditional models, where there’s significant transformation of the data to map into a data model that’s very specific and very well defined. For us, we maintain a lot more flexibility. That gives our customers the ability to move a lot of data into the data warehouse quickly. And then once the data is inside the warehouse, that’s where we start manipulating and analyzing the data in an environment that we control. So that’s the real differentiator for us. The other piece, I think, about Health Catalyst that might be different is I would classify us not as a software company, but we keep our focus on being an outcomes improvement company. The software helps, but it’s never enough. So we bring services to bear as well. We have clinical resources, we have workflow experts that know how to use the data from the data warehouse to actually pinpoint and then improve outcomes.
SK: So most of your customers, I assume, then work with you most of the time in terms of deciding where they want to focus their energies. You help them pinpoint the action steps that would help improve the outcomes based on the data, you know, where we should place the feelers in the field, if you will, and then measuring the outcomes after the change has been made to demonstrate that it’s been worth their time and energy to focus in that area. So while you’re using software that you’ve developed, you’re really using that to deliver effectively a service to your customers in terms of spotting their specific and unique areas that could be improved. Is that a fair way to think about it?
DB: Very much so. We at Health Catalyst feel like if outcomes haven’t improved at the end of a project, and the project always involves some software installation, that if that’s the end of it and we haven’t helped the health system use the software to pinpoint a problem, work on the problem, improve it and then measure the result through the software, we haven’t done our job. So yes, we bring software to bear, both the data warehouse software, but then also the analytic software that we built specific to the most common areas that most health systems need to improve in in order to bend their cost curve. But then we also bring experts to bear to help them be successful in making the changes.
SK: Yeah. And I mean the products and the systems that you’re offering are extraordinary. I’d encourage listeners to go to your website. It’s a fantastic place to go, even if you just want to learn about healthcare data analytics from the blog, but also be able to dig into the products that you’re putting out there in the marketplace. You do a great job on that site of explaining those products and explaining their benefits. Dan, I want to get into – Breaking Health is really about talking to CEO’s, and so let’s step aside from the specifics of the company and talk about how you’re managing the business for a little while. I’ll make an observation and then I’ll leave it open for you take it from there. I look at the website, I look at the breadth of the product, I looked at the amount of people that you have to bring in to accomplish your goals, and I say to myself this is a very well planned operation. Can you give me sense of some of the values that you ascribe to as a CEO and how you go about planning strategy for the company?
DB: Happy to. I think one of the things that I appreciate about our cofounders is that very, very early on, before we took on any meaningful institutional capital, we sat down at their behest and we started to author a document that we revised over time that has really served as the bedrock for us. And it was titled The Things That Can’t Change.
SK: I like that.
DB: And we identified some cultural attributes that we felt were very core to who we are. We identified what our mission was as a company, and what operating principles we wanted to use to guide our behavior. And we started the process of actually documenting it. One of the operating principles that we agreed to was transparency, that we needed to be up front and open with all of our constituents all the time. So that includes investors, it includes customers, it includes employees, or team members, as we refer to them at Health Catalyst. In that spirit, as we developed these cultural attributes, the mission of the company and the operating principles, it led us to certain conclusions that we felt investors needed to know before they invested in the company. For example, earlier in the Podcast we talked about the effect of being acquired and the way that it often degrades a company’s ability to accomplish its mission. We believe that very strongly, and so we have been transparent about not being open to being acquired as a company. We are not open to that, and we felt like we needed to be very candid about that with investors before they invested so they knew what they were getting. Another operating principle we decided on was to be focused on long term customer success, that no matter what we did, we wanted to hold on to that, regardless of our capital structure or capital strategy. So we put some structures in place to try to enable that to be a long term reality. One example is we decided early on we were not supportive of commission based sales reps. We had seen too many examples where that incentive structure led to a short term mentality, but it was not focused on long term customer success. So we telegraph that to investors before they invest so they know what they’re getting. A third example: we identified three cultural attributes that became our simple hiring formula and our retention formula. We hire people who are smart, hard-working, and humble, with humility being the most important of the three. We are students of Jim Collins’ work around what makes companies built to last, or Good to Great. And humility was such a central aspect of the success of those companies, we decided early on we needed a way to counterbalance the natural course that we’ve seen over and over with so many companies, where the leaders tend and all the team members tend towards drinking their own Kool-Aid and a sense of a bit more arrogance. So we tried to systematize and reinforce the attribute of humility. We want to be good learners, good listeners, we want to celebrate each other’s successes and be open to the need to change and evolve over time. So we tried to be systematic in thinking about the right operating principles the right cultural attributes, and the right mission and then build our infrastructure around that as we grew from four people to 400 people. And It’s really hard, and we haven’t always been successful, but we always go back to those things that can’t change, and that seemed to help us over the last several years.
SK: That’s interesting. I would say that I think that’s where a lot of startups get themselves into trouble. Like they touch the money, if you will. They start to think about how much wealth they’re creating more so than if they just think about building a great company, then they will create wealth as a byproduct of building a great company. And it feels that humility is part of that equation, right? I mean you’ve got to sit there and say, well, we were great today; we’ve got a lot of work to do to get better and keep going forward, or somebody is going to compete with us and do this better than us. Right?
DB: Absolutely. Absolutely. And I think you touched on another really core element that we’ve seen play out many, many times over the last several years, and that is if what motivates you is money, it will likely lead you down a path that’s much shorter term. If what’s motivating you to continue to do your work and your company is mission, and you deeply buy into the mission of transforming healthcare, which is the first goal in our mission statement, and dramatically improving outcomes, and you can keep tying yourself back to that meaningful mission of seeing patients’ lives improve, seeing outcomes improve that affect families and individuals in a very, very meaningful way, that’s far more motivating and it has much more staying power with all of our team members. We average about a fifth of the turnover of a typical Sequoia startup at Heath Catalyst. And I think a lot of the reason for that is for the people to come, buy into the mission, and yes, as you said, Steve, along the way there’ll be positive financial outcomes, but that’s not at the core of what we’re doing and what motivates us.
SK: Well, listen, I love hearing that, and you got Sequoia to travel outside of the state of California, so congratulations for that.
DB: Thank you.
SK: I would be remiss if I didn’t mention some of your other investors, Sorenson, Norwest, Kaiser, Epic Ventures. Did I miss anybody in that list?
DB: It’s a growing list, so we have a number of other health systems as well: Allina Health is an investor. Partners Healthcare in Boston is an investor. Indiana University Health, Cedar Sinai Memorial Care in Southern California are also investors. And then as you mentioned, the financial investors Leavitt Group is another financial investor. Sands Capital in the DC area. When you raise over 150 million, you end up with a –
SK: A lot –
DB: – a good long list of investors.
SK: You gotta make a lot of friends. OK, last question, then I’ll let you off the hook here. I would say what you just described to me is what I think about as a great source of differentiation. A mission, a set of values, a set of uncompromising principles, and of course I’m a student of Michael Porter’s teaching. I know you’re an HBS guy. They must have talked to you about that while you were there. Creating differentiation of course can lead to a competitive advantage over time. And it sounds like that’s where you’re headed. So my question really is you’re not going to get to have this market all to yourself, right? There’s competitors out there today; there’ll be competitors in the future. What do you think the parameters of competition will end up being over the next 5 to ten years in the healthcare data and analytics space?
DB: Yeah, very good question. As we think about what it will take for health systems that will survive and thrive over the next 5 to 10 years, first of all it seems clear to us that they’re going to need to harness data to pinpoint where they should be making changes, and then measure the effect of those changes. The question that competitively and strategically we’ve tried to focus on is what will be the best data platform from which to run all of that analytics, to make those changes, and measure the improvement. I believe there will be 2 compelling platform alternatives. And I think both will be successful over time, and many health systems I think will wisely choose to have both at their disposals. One is using the EMR as a data platform. I think there are and there will continue to be five, 10, 15 years from now a series of use cases where the data that resides within the EMR is a sufficient data set to run the analytics, to make changes, to close the loop and measure the effect of those changes. So I believe that is one competitive platform. And we’re already seeing some of that today. And the second platform would be more of an independent platform, a data warehouse platform, whereby you can bring data from many sources. And I believe that there are a number of use cases today, and there will be 5, 10, 15 years from now use cases where you need a flexible, independent warehouse that can ingest data from multiple EMR’s ingest data across organizational lines, and ingest data from lots of different sources. That, I think, is a compelling strategic data platform, and Health Catalyst is obviously in that camp. IBM and Oracle also offer data warehousing options in that camp of an independent data warehouse that pulls from many sources. So I think those will be the two major categories of potential long term strategic data platforms that health systems will need to identify when do I use the EMR as my main data platform, and when do I use the data warehouse. And if I choose the data warehouse, which data warehouse is going to be the most compelling and the most useful for me?
SK: Very good. Well, thank you for your time. I really appreciate you speaking with me, and I’ll look forward to talking to you again in the near future.
DB: Wonderful. Thanks, Steve.
Tom Salemi: Well, thank you, Steve Krupa of the Psilos Group and Dan Burton of Health Catalyst. That’s another great conversation for the Breaking Health Podcast. I’m really enjoying the dynamic, the VC-CEO dynamic really able to dive deep into some of the issues and challenges facing digital health companies. To hear more of those challenges and the opportunities, you’ll have to come to our Digital Healthcare Innovation Summit on November third at the Mandarin Oriental Hotel in Boston, Massachusetts, my home town. Go to digitalhealthcaresummit.com to find out more information about our agenda and of course to register. So we’ll see you in Boston.