Podcasts
Podcasts
The Disruptive Voice
The Disruptive Voice
- 03 Dec 2018
- The Disruptive Voice
24. Butterfly Network: Innovating at the Low End in Ultrasound
Shaye Roseman: [00:43] Hello, and welcome back to The Disruptive Voice. I'm Shaye Roseman, your host for this episode and a researcher at the Forum for Growth and Innovation here at Harvard Business School. I'm joined in the studio today by Darius Shahida, a member of the Class of 2019, who's currently working full time as chief of staff at Butterfly Network, a potentially disruptive new medical imaging company and recently minted unicorn, which we'll hear more about in a moment.
Shaye Roseman: [1:10] Darius, thank you for joining us today, and welcome to the podcast.
Darius Shahida: [1:13] Thank you so much for having me.
Shaye Roseman: [1:15] So, Darius, can you tell us about Butterfly Network and the product that you've developed?
Darius Shahida: [1:21] Absolutely. So at Butterfly, we like to think of ourselves as a full stack imaging solution, and what that means is we have a device, but we also have a software as a service business. Our device is actually incredibly disruptive, in that we've created the world's lowest cost and first whole body scanner. With this device, which we priced at $2,000, we can actually image the entire body. And so last fall we received the broadest FDA clearance ever for an ultrasound device, 13 applications, and that spans everything from obstetric and cardiac deep applications to more shallow applications like a carotid or a vascular scan.
Shaye Roseman: [1:59] Talk about imaging the entire body. Is that different from how most ultrasounds today work?
Darius Shahida: [2:05] Yeah. So our technology is predicated on a fundamentally new transducer technology, and what that means is ... I think it's best to give a little history in terms of the ultrasound industry up to today and the introduction of our technology. For the last 40, 50 years, all ultrasound technology has been predicated on something called piezoelectric crystals. These are hardwired, hand-cut crystals which have very specific acoustic and bandwidth limitations, and what that means is that with traditional ultrasound technology, you need different transducers or different probes to image different applications.
Darius Shahida: [2:41] So I just mentioned cardiac and vascular applications being done on one probe. That's never been done before, and the reason why is because deeper applications with this traditional piezo technology are really limited to one transducer, whereas shallow applications are limited to another transducer because, again, the traditional technology is hand-cut, hardwired. What we did, which is fundamentally new and different, is we put ultrasound on a chip. Using MEMS we've actually created a semiconductor transducer which gives us the entire spectrum of acoustic bandwidth from one to 10 megahertz. And, what that allows us to do is ad hoc adjust the frequency and capture images in both a deep setting and also a shallow setting.
Shaye Roseman: [3:28] And can you explain what MEMS are perhaps for our listeners who are less familiar?
Darius Shahida: [3:32] Yeah. Yeah. So, at a very high level, in analogous terms, what we've done is we've created this semiconductor transducer which has 9,000 wells. These wells pulsate and they generate ultrasound, and that ultrasound is able to give us a very broad acoustic bandwidth that allows us to image the entire body. And we've seen actually over the course of history, the power of semiconductors, particularly the disruptive power of semiconductors when introduced to new imaging modalities, when introduced to new computing modalities. Couple examples worth mentioning: the introduction of digital film and what that did to traditional film. That was all catalyzed by putting film or imaging on a chip. We also saw that in the sequencing space. Next generation sequencing was actually facilitated by a chip-based solution to sequencing the human genome, and we feel that we are at kind of the third or fourth time, certainly in our lifetime, of disrupting an entire industry based on this semiconductor technology.
Shaye Roseman: [4:37] You spent some years on Wall Street prior to pursuing your MBA here at HBS and then decided not to wait until graduation to find a full-time job. Can you talk about how you got into this line of work?
Darius Shahida: [4:50] Yeah. So, as you mentioned, I worked in finance. I found it to be an interesting industry. I was running a trading desk at a large hedge fund in New York that was actually very healthcare focused, but I realized fairly early on that while the opportunity set there was great, I was really not fulfilled from the perspective of having an impact, and so one of the reasons why I came to HBS is because obviously the academic experience here is very much tailored towards building leaders who want to make a difference in this world, and so for me, that was the rationale to come here.
Darius Shahida: [5:24] Butterfly and the opportunity that came about with 4Catalyzer, which is our larger incubator under which Butterfly sits, was very serendipitous. I was introduced to Jonathan, who's our founder and a serial entrepreneur in this space, and he actually, initially, when he presented me with this opportunity, asked me to drop out of school. I've actually worked my entire life. I worked throughout college, I worked throughout high school. It was almost second nature to just continue working, and so I told him, "Look, I'll work full time while in school. You won't know the difference." Fortunately, my teachers didn't know the difference either, so I've kept that going for the last, well, over a year now.
Shaye Roseman: [6:03] That's great, and so you've been doing it since you set foot on campus basically?
Darius Shahida: [6:07] Yeah…so, it's been over a year now. The EC year is obviously much more accommodating. I actually live in New York and I work full time in New York and I commute here to class every Monday and Tuesday, where I take all my classes.
Shaye Roseman: [6:21] Listeners who have taken our course will recall that we teach a case about a company called SonoSite and their low-end ultrasound device. Shortly after discussing that case in class, Darius, you walked into the office of my colleague and faculty member Derek Van Bever to talk about your work, and our team was so energized by what you were working on that we asked you here to join us on the podcast. I'm curious what you thought when you were reading that case and whether that's what made you want to stop by Derek's office.
Darius Shahida: [6:51] Yeah. It was very apropos that we had this case very early on in the semester. There's been a history of disruption in the ultrasound space, and more broadly in medical imaging. What I found interesting was SonoSite was indeed disruptive at the time. This whole advent of point-of-care ultrasound, or POCUS, to this day is considered to be a very disruptive and innovative paradigm shift in medical imaging. But one of the traditional limitations of all ultrasound technology, even point-of-care ultrasound technology, is the prohibitive cost. Another is this issue of image acquisition and interpretation. Until Butterfly, there really was no solution that disrupted and solved all three of those issues. We solved the issue of prohibitive costs by engineering a fundamentally new technology, leveraging the supply chain in semiconductors, hundreds of billions of dollars that poured into the semiconductor supply chain to enable us to create something which scales, and as it scales, not only does the cost go down, but also the computing capabilities and the image quality that comes with that increases exponentially as well.
Darius Shahida: [8:02] But then the other two issues we solved, actually, with artificial intelligence and machine learning, and that goes back to this whole stack imaging solution that I referenced earlier. We like to think of every technology that we drive as being a part of a virtuous cycle. What that means is the device is effectively a Trojan Horse, which enables image capture and acquisition. Once we've acquired those images, they are pushed up into our cloud in a HIPAA-compliant way, and we actually train our machine learning and artificial intelligence algorithms on these large datasets, and in turn they improve. We see ourselves and every company that we create as being a part of a virtuous cycle and one that gets better as we have more data. It also creates an interesting dynamic from a competitive perspective, in that this virtuous cycle helps us to build a network effect in a sticky ecosystem, which helps us to build a moat. And that's something which is key to us, particularly in such a competitive landscape.
Shaye Roseman: [8:59] When you say the device is a Trojan Horse, does that mean that you're thinking about the real value of the business living somewhere other than the handheld device itself?
Darius Shahida: [9:11] That's a great question. I think the device obviously from an imaging perspective enables so much, particularly if we consider developing world context. Outside of the United States, two-thirds of the world's population has no access to medical imaging whatsoever. That spans all imaging modalities. And that's largely because of these traditional barriers to entry, the cost, the image acquisition and interpretation problems. We see the device as actually helping enable a lot of those two-thirds of the world to have access to this medical imaging, and in many settings, it's a zero-to-one type step up. In those cases, the device is itself kind of the key, in that we are actually giving something to someone that has had no access whatsoever before. In other use cases, for example, in developed world context, the device and the applications that we then develop on the device enable interpretation and analysis of things that even the human eye can't detect.
Darius Shahida: [10:09] I'll give you an example of this. Our first application that we're rolling out with our Butterfly device is an automated ejection fraction interpretation, and what that does is it allows us, using machine learning, to calculate heart health and it gives a metric which can then be interpreted by medical practitioners right then and there, and it can be done within 30 seconds. My colleagues who are clinicians remind me that that's a billable event at a place like MGH for almost a thousand bucks. And, really, the point of me sharing this is just to say that we are building an ecosystem on which other parties can come and develop applications, analogous to what Apple did with the App Store. These applications will enable other clinical measurements and automated measurements that we can't even foresee right now.
Darius Shahida: [10:59] An additional example of that, I should say, is that we're working with the Gates Foundation on gestational age approximation app.
Shaye Roseman: [11:05] Fantastic. Yeah. I think it's really interesting that you distinguish between what in our terms would be non-consumption and targeting low-end markets and the way you characterize building out a portfolio of products and services that may look different, depending on or where the value may be different, depending on what context you're operating in or what kind of market you're targeting.
Darius Shahida: [11:34] Yeah. No, it's really interesting and unique in that we have pursued a strategy which targets non-consumption both in a developed world context and a developing world context. In the developing world, there is no imaging, or very little, I should say. In the developed world, there's all this demand that has been unmet because the traditional technology is limited and the traditional technology is cumbersome and the traditional technology is prohibitively expensive. The important thing that I would just mention here is that our mission as a company is to maximize societal impact, and we really want to democratize this medical imaging modality. What that means is it doesn't matter developed or developing world setting. We're committed to actually this $2,000 price point when we could have priced it even 4, 5, 6,000, what have you, because it aligned with our mission and it's what we felt was the right thing to do.
Shaye Roseman: [12:23] I want to talk more about price point in just a minute.
Darius Shahida: [12:27] Sure.
Shaye Roseman: [12:28] We often talk about how technology is, in itself, just technology, but that it can be shaped to be either sustaining or disruptive. I'm curious if you can talk about what steps you've taken at Butterfly then to set yourselves on a disruptive path.
Darius Shahida: [12:45] Sure. Virtually every decision we make at Butterfly is a disruptive one, and what I mean by that is, yes, our technology has disrupted along the axes of price, image acquisition and interpretation. But even our commercial strategy, our decision to go through ecommerce as our main channel, that's something which is very controversial, very disruptive within the traditional medical device space, certainly within the capital equipment space. And as a team, we felt that this was the right thing to do and that creating a large enough funnel that allows folks, licensed medical practitioners, and there are 40 million of those globally, per the WHO, allowing them to just go organically to our website. And that being the primary source of commercial sales, that was a very disruptive, very innovative decision, but it's one that we saw unprecedented response to.
Shaye Roseman: [13:41] Talk about that controversy. What was the nature of the resistance and where was it coming from?
Darius Shahida: [13:47] Yeah. There wasn't controversy internally within our team, in that we were all aligned and we felt that this was the right decision to pursue. But when you look at the way that other ultrasound, particularly cart-based solutions, are sold, there's an extensive corporate infrastructure that helps that selling process, that facilitates that selling process. There's these complex distribution networks, regional, global, country-based. We wanted to streamline that. We wanted to streamline that because we subscribe to the lean startup mentality. We believe that every decision we make must scale, must be simple, and must be a quick one, and so our decision to go the ecommerce route was very much in line with our credo.
Shaye Roseman: [14:32] How else have the frameworks of the course helped you to bring the future of Butterfly into clear focus?
Darius Shahida: [14:37] It's interesting. We read a couple of cases on Andy Grove and the innovation that he drove and a lot of the decisions that he made. One of his famous quotes is that "Only the paranoid survive." That's something that we definitely subscribe to. We care about speed more than a lot of other folks, and that's because we know that in order to build a competitive moat, we need to scale quickly, we need to move quickly, we need to get these devices out into the hands of the people that need them the most. Another thing that Andy mentioned and he was famous for was this concept of strategic inflection points, and we feel that our industry right now is at a strategic inflection point, that going forward no one will use the traditional piezo-based technology because why would they when there is this cheaper, better, easier-to-use technology that disrupts along all these axes?
Shaye Roseman: [15:33] Is the goal to wholesale replace prior ultrasound technologies?
Darius Shahida: [15:39] That's an interesting question. What I would say is when we look at the demand we've seen so far, which is unprecedented, over 50,000 inbound orders ... Just to put that in perspective, the entire point-of-care ultrasound market cumulatively since inception has sold maybe 20 to 30,000 devices, and that's across all the competitors, so to speak. We've seen this unprecedented response because really everyone is reaching out within the medical field to buy this device because we actually conceptualized it as a personal device. When you consider the fact that for the last 100, 150 years, the fundamental physical exam has not changed that much, doctors today are still using stethoscopes, that's a bit archaic to us. We feel that it's important that people move forward, the healthcare industry moves forward. We've conceptualized our device as the 21st century's visual stethoscope. It's a window into the human body, and we feel that at this price point and going forward, doctors, health practitioners are going to respond commensurately with that.
Shaye Roseman: [16:51] There are other handheld ultrasound devices on the market. SonoSite has one, Philips and GE have developed their own competing models, but a few key features make the iQ, your product, different and that contribute to its disruptive positioning. The first one I want to talk about is price. At $2,000, as you mentioned, the iQ sells for less than a third of the price of incumbents' handheld models. Talk a little bit about your chip technology and how that makes your product and your business model unique.
Darius Shahida: [17:26] So, before I do that, I just want to clarify one point. The traditional point of care ultrasound solutions that you just mentioned, that's for one probe. In order to image the entire body, you need multiple probes, in many cases, three or four probes. And so when you aggregate the cost across three or four probes, you're seeing that actually the best and cheapest solution that is an apples-to-apples comparison across these 13 applications is north of 20 and $30,000.
Darius Shahida: [17:54] Another important clarifying point is that when we introduced our device, we were very careful to show that image quality is not something we compromise on. In fact, if you look at our device and images, cardiac images, obstetric images, across these 13 applications, right next to a 60 to $80,000, cart-based solution, they are indiscernible, and that's something which is important because we are in version one of our chip, version one of our device. All of the extant technology is in the 40th, 50th iteration. If you consider the curve of marginal improvement that these guys are basically able to drive, you're seeing very minimal marginal improvements, whereas our technology, being predicated on semiconductors, gives us the power of Moore's Law, gives us the power to innovate and drive exponentially better functionality with each iteration of the device.
Shaye Roseman: [18:52] You also market the iQ as a personal ultrasound device and aim to eventually sell directly to consumers through your website, which is very different from incumbents' focus on healthcare systems. Talk about that positioning and why it's important to your strategy.
Darius Shahida: [19:11] Sure. A couple of clarifying points here as well. Right now, we can sell to medical practitioners, and as I mentioned earlier, there's 40 million of those globally. That's a huge total addressable market, and it's a market that we feel we're going to have great uptake interest, given all the kind of competitive landscape and market factors that I've elucidated. The important thing, though, is that this concept of personal ultrasound is something that is not us saying. The actual industry and the academic literature right now is very much converging on this idea.
Darius Shahida: [19:44] I'll give you a good example. Earlier this year in JAMA, one of the most prolific cardiologists published a paper where he called for insonation, or ultrasound, as the fifth pillar of the physical exam. The idea is that, as I mentioned earlier, it's a little crazy to think that we're still using stethoscopes to listen into the human body. Why not use ultrasound, or insonation, to look into the human body? When we consider the fact that north of 60%, 70% of pathologies can actually be examined using ultrasound, really it's a no-brainer to think that this device, a personal device, will be at the bedside, will be in the ambulance, will be with the nurse practitioner, will be in the battlefields. You name it - it's a personal device, fits in your pocket, and it connects to your smart phone, and that really enables a completely new paradigm, a new era in medical imaging that we feel is going to be the way of the future.
Shaye Roseman: [20:46] And you just started shipping product a few weeks ago. How's it going?
Darius Shahida: [20:49] Yeah, we started shipping last month. As we mentioned, right now the demand is incredible. We've seen an order of magnitude greater demand than really the entire industry's aggregated demand to date, and that's very telling. Right now, our team is focused on scaling and fulfilling orders as quickly as possible, but as you can imagine, every day that goes by, we get all these other organic orders coming in, and so scaling and fulfilling these orders is of the utmost importance to the team right now.
Shaye Roseman: [21:21] Where do you think that demand is coming from if you've seen it in numbers larger than historical, aggregate demand across the industry? Who's reaching out to you?
Darius Shahida: [21:33] Yeah. That's a great question. We've seen everything from specialists to emergency medicine practitioners to cardiologists, sonographers, radiologists. I'll give you a good example. Last year and actually this year we presented at ACEP, which is an emergency medicine conference in San Diego, and within, I think, something like 90 days of presenting at ACEP last year, we saw almost a 10% penetration of all emergency medicine physicians in this country, which, just to put things in perspective, is pretty telling, I would say, of the demand that exists out there. We'll be at RSNA, which is the largest radiology conference, later this year. We just actually won the award, the AuntMinnie Award, for Best New Radiology Device of 2018, which is pretty exciting, considering the fact that we were up against the major incumbents with much more expensive technology that they were introducing.
Shaye Roseman: [22:31] You've taken this technology that was previously expensive, inconvenient, and that required certain expertise to operate, and you've made it cheap, portable, and accessible. What would the theory of disruption predict about the future of your business?
Darius Shahida: [22:48] It's interesting, I think about this a lot. I think right now the most important consideration that we have as a team is just continuing to be very thoughtful and I would say deliberate with our strategy going forward. As I mentioned earlier, only the paranoid survive. We subscribe to that. We know that the incumbents are going to respond. In fact, the incumbents have already responded by changing their pricing on some of their traditional point-of-care offerings. The important thing is we're not worried right now about the incumbents. We're focused on our core strategy, which is getting as many of these devices out there as possible.
Darius Shahida: [23:23] As we do that, we build this ecosystem which is really edified by our virtuous cycle, and this ecosystem builds a network effect which hopefully in the next year to even six months will allow us to build the largest labeled ultrasound imaging data repository in the world. That's pretty exciting in this new day and age of the importance of data and analytics, and when you consider the machine learning and artificial intelligence applications that we're building, the suite of applications that are going to have real, tangible clinical impact, that's something which makes this even more exciting for us and which just edifies our competitive position greatly.
Shaye Roseman: [24:06] What's been the biggest takeaway from the course for you in the context of your work?
Darius Shahida: [24:10] Yeah. The course has been great in the sense that we've kind of chronicled a lot of instances of disruption over the last 50, 60 years, and what's interesting is history does repeat itself. A lot of the lessons of corporations in the '60s and '70s are being seen today in new industries. What we see is that at these strategic inflection points, it's really important that you're able to grasp the bigger picture, not to lose the forest in the trees, so to speak. We're very cognizant of that and we're acutely focused on that, which is to say we feel we're at a strategic inflection point; the medical imaging space more broadly is at a strategic inflection point. Everything we do, every decision we make is, therefore, predicated on us being able to own this strategic inflection point so that we don't become a case about what should have been done, but we're really the case on the appropriate playbook to capture and own a market.
Shaye Roseman: [25:11] Are there any key lessons for you in putting together that playbook?
Darius Shahida: [25:16] One of the interesting lessons of the course recently was on good money and bad money, and one thing which we feel very strongly about is we want to align ourselves with investors who want to do well by doing good. What we mean by that is because our mission comes first and foremost, we have found, and we've been very fortunate to find investors who are aligned with us strategically on this mission.
Shaye Roseman: [25:42] Well, Darius, thanks again for being with us today. Where can listeners go if they want to learn more about Butterfly iQ and what you're up to?
Darius Shahida: [25:50] Yeah, sure. Be sure to go to our website, butterflynetwork.com. You can order a device if you are a healthcare practitioner and follow us on social media. There's going to be some really exciting updates in the coming months, so I'm excited to share it with you, and thank you for having me.
Shaye Roseman: [26:06] Thanks again.