Salesforce, Inc. (CRM) Earnings Call Transcript & Summary

Good afternoon, everybody. Thank you so much for being here on our panel where we're going to talk about technology innovations in health care. I hope everyone's been having a great JPMorgan so far. And I know this is a lunchtime panel. So if you do have sandwiches or anything else, go for it. We want to make sure you guys have enough food. So I'm Christina Farr. I'm a health and technology reporter with CNBC, and I'm just going to let the panel introduce themselves, the 10 seconds, on who you are and who you work for.

I'm Othman Laraki. I'm Cofounder and CEO of Color. We're a health technology company that works with about 100 different institutions across -- from NIH, self-insured employers, some -- pharma as well.

Aashima Gupta. I lead Healthcare Solutions in Google Cloud. In my role, I work with broad spectrum of industry stakeholders in helping them transform, leveraging cloud, APIs and AI. Happy to be here.

Good afternoon, everyone. I'm Jujhar. I'm responsible for the Salesforce, Industry Clouds. We are focused on building industry verticals for 7 of those industries and excited to be part of this panel, look forward to it.

Ken Ehlert. I'm UnitedHealth Group's Chief Scientific Officer. We spend all of our time in health care.

Jeremy Sohn. I'm Global Head of Digital Business Development, which effectively is our partnership engine as well as our venture investment engine for working with technology, data and new business model companies. And hopefully, everyone knows what Novartis is, and we, too, spend all of our time in health care.

Here you go. So we are here to talk about health care, but we're also here to talk about technology. And everybody on this panel has a background in that, some more than others. Aashima Gupta, and then we have Salesforce and traditional health care companies also.But I wanted to kick things off with a question that kind of put you guys on the spot. And that's how do you define digital health? And is that even the right term? Should we say health tech? Is it health IT? Like what's your working definition that you use in and around kind of this conference?

Anyone want to start? I can jump in. I mean, basically, that's just health care. I think there's no version of health care with IT anymore. And so in the same way that like retail, 20 years ago, we were thinking about e-commerce versus the -- versus retail, but now it's like it seems completely inaccurate as to even think about retail without digital shopping, and I think health care is basically the same thing.

Yes. I would say it's an intersection of traditional health care with technology, and it's all coming together. To me, that's digital healthy -- melting pots of technology and health care coming together.

I think the way we define it is technology kind of ushering health care into a much more retail-centric world with patient right at the center. That's how we look at it. That's what the whole company is all about, and that's how we define digital health.

For us, it actually would depend on which part of health we're talking about. So if we're automating core business processes, that would be one version. But if we're moving into a different way of diagnosing and screening patients, that would be another way that we'd think about it. Or on the extreme, from a therapy perspective, if we're thinking about beyond conditions that we think of as body type conditions. But if we move to the mind, the ability of a molecule to cross into the brain to actually alter an underlying PTSD event, for example, that's a -- today, with current technology, that's an impossible task. But using digital mechanisms, we can actually go in and sort of mitigate the effects of that. So I think it depends on what part of health we're actually talking about. Whether it's the business or all the way to the other extreme, it's in therapy itself. Those types of tools that are there, the tools that are out there and available can be deployed against those things in different ways.

So I love the fact that you asked that question because I think what you're hearing is, is that digital health probably means something different to everyone.

Absolutely. Yes.

At the same time, what I hope is, is that we actually ultimately drop the initial adjective and we just actually talk about it as health.

Health.

And I think that that's sort of the transition, in fact, the inflection point that we're in. And a lot of what we were trying to do and the fact that we want to -- I sit within our digital office. And the reality is, is digital -- why do we even need a digital office? Ultimately, I'm trying to apply technology and innovation and new business models as a core capability inside every aspect of my core business, whether in research or development or running clinical trials or my commercial organization, or for that matter, internal finance, et cetera. All those are business units that exist and don't really need to be replaced. So what we're trying to do is to just create an agility engine inside of those around health care. And so in many respects, it's all just health and just -- it's how we run our business.

Yes. I mean, I want to kind of -- a couple of people said that they think digital health should eventually just be health, but we do see -- maybe it's because we're in a transition period, but there is certainly a difference in the culture between health care companies and technology companies. I mean, one of the great sort of mantras that's kicked about in Silicon Valley is move fast and break things. And can you imagine if I walked into a UnitedHealthcare or Novartis and said, "I want to move -- I want to break things." You guys would say, "Okay, here's the door." So how do you bring together those 2 cultures? And what do we kind of need to learn from each other to get these conversations to be more productive?

Maybe I can start. So from a Novartis perspective and from a pharma perspective, certainly quality is -- and repeatability is one of those things that we absolutely believe is core to one of our strengths, right? We're best in class in science and in delivering transformative drugs for our patients because of that quality, because of that excellence. But at the same time, when you talk about digital health, for us, again, it's building a new muscle. And that muscle is that quality or being best in class actually isn't about repeatability. It's about continuous optimization. And so that's a little bit of taking that start-up engine, that technology engine and actually recreating the way that we operate. And in fact, we do that in the science, right, but we don't necessarily do that across all of our business functions. So I actually don't think about it as breaking Novartis or pharma or large companies in general. It's actually just creating a different way of operating and hopefully a more optimal way of operating.

I think at Salesforce, we actually see the perspective of both the universes, the B2B as well as the B2C, the consumer as well as the enterprise. And we don't believe that you can't be agile and still be maintaining the change management processes. We have done it successfully across multiple of our industries. Health care is no different. We believe the changes that actually you need to bring on the enterprise side of the house, they are actually going to be really pushing the digital health concept, touching the consumer. The whole patient engagement is going to be very fundamentally driven by the internal processes that change on the enterprise side of the house. So we see them mutually feeding on one another rather than being exclusive. And successfully, we've done it. We do agile releases 3 times a year. We've done it for the last 20 years, both with enterprise as well as the consumers.

I think one of the cultural gaps that has existed sometimes is that if you're purely coming from a consumer technology world, I used to be at Google for a number of years. And when you're iterating sometimes on a pure consumer product, there are very few parts of the product that are sacrosanct, in some sense, where you can iterate on the entire experience. As you can -- there's a big surface area where you can be very experimental. And I think with health care, obviously, there is some parts that are very deeply sacred in terms of ensuring that you don't cause harm, that you have certain quality standards, et cetera. And I think sometimes where there has been kind of a gap in that cultural -- in those cultural norms has been around different assumptions about where -- what parts are experimental versus not. Like, for example, just with Color's experience like we...

Can you give us an example?

Yes. I was going to say, like, for example, with Color, so we -- by now we've built a kind of an infrastructure that has been used to roll out the largest number of genomics -- population genomics programs in the world. And there are some parts of the system that we are very experimental with, like, for example, how do you get a clean consent flows and how do you get people to answer questions with the highest convenience for them, et cetera. Those are things where you actually want to increase yield and be very experimental. On the flip side, we run a clinical-grade diagnostic part of the product. And that is a part where we really cannot mess up, right? And it's very -- like the experimentation framework for that is very constrained. And so I think like that's where like sometimes if you come in with a pure consumer tech hat, it's important to realize, like, okay, where do I draw the lines and operate in these very different worlds and be kind of comfortable with them.

So you're saying you can have pieces of the business that operate more like traditional health care and then pieces where you've got R&D efforts internally that feel more like a traditional tech company?

Yes. I understand that there's just a curve of risk tolerance and impact across your product. I think we're -- which I think is much less the case in consumer products. I mean when -- if Twitter goes down, it's not great. They lose money, but they always say, no one dies, right? Like whereas in health care, that actually is not true. So I think that's really understanding -- like being sensitive to that difference, I think, is important.

Yes. And I think where I sit today, when we were working with the industry stakeholders, one comment that was made just a couple of days ago that, "Aashima, we -- yes, we move very slow, but we make much less mistakes." And so -- and now they're shifting. There's a paradigm shift. And if I go back to my old role in Kaiser Permanente, I led digital health incubations. This was a completely different team. You incubate ideas before you actually roll it out in pilot or production. So there is that discipline. Things are changing. I think the health care industry is now looking into -- another term that is overused in health care, or in general, in all the industries, digital transformation. Just like digital health, everything is about digital transformation now. It's such a generic term. Like are you transforming clinician experience, physician experience? What does...

Yes. What does it mean to transform something digitally?

Right. So it's -- and there are many aspects, just patient facing, physician facing, and there's actually operation robotic process automation. And that's an über term of umbrella people call digital transformation, yes.

To this point about culture, I mean, I've even -- I think in this space where we are still quite confused about what digital health means, I've had folks tell me Peloton is a digital health company, and you guys may agree or disagree with that. But people say that digital health companies, in general, tend to focus more on sort of the very privileged populations and often as a cash pay aspect. You see them kind of on the fringes of health care. And some might say, "Well, that's not the case because you're also seeing digital health companies actually work with traditional health insurers and pharma and others." But do you see there being any truth to that, that digital health, so far, has primarily been concerned with the wellness, the fitness, the sort of higher-income professional user base?

We see a lot that's out there, and it's true. There are a lot of companies, especially from something that feels very consumer-like. I want to do a diet. I want to exercise more. And it makes sense that some -- a company that's doing that would take the perspective that I have a consumer that's purchasing something from me, I'm trying to optimize the experience, the product for that consumer, and there is a price that they're willing to exchange value on. That part is separate, though, from is it health or not health. Yes, exercising is health. When we all stand up and walk out of here, our large muscle groups are going to start producing interleukins, and they're going to have an anti-inflammatory effect. And none of you got permission from your health -- actually, probably a bunch of you have health insurance from UnitedHealthcare, but none of you got permission from your insurance company to actually engage in health like that, and yet you do. So this idea of what's health and what's not, we're human, we're alive, and every interaction that we do is part of health. When you have an interaction that's a social interaction with somebody, that actually changes what's going on in your brain. If you socially isolate, you will put your brain under enormous amounts of stress. That will cause your immune -- as it goes through all the different sort of the HPA axis all the way down, that will cause your immune system to start responding in a different way. Do I want you to start exercising when that happens? Sure, I do. But the idea that that's all in one bucket and this stuff is in a different bucket is actually not quite true. So like this morning, just on our call, we talked about -- I don't know, it's about 0.5 million lives that are in the Motion program, where they're actually earning really large incentives, about $1,000 -- up to $1,000 a year for doing something as simple as actually going for a walk, and I could take that a lot further if we wanted to. The idea that you have a molecule -- and by the way, the idea that like health care companies don't move quick or innovate, thousands of molecules will get tested to produce something that actually doesn't kill people in the process of helping them. The same thing could be said when you're looking at a genomic test. There's enormous amounts of stuff that's going on there. But when you start to narrow down and say, "Hey, we've got 0.5 million people that are actually doing things for their health, incorporating it in, reaping the benefit of that, that's a pretty cool thing to have happen." When somebody makes a molecule or some sort of a protein, whatever it might be that you need to take to improve your health, you got to remember, you still need to take that pill or you need to get that injection. That, too, is part of health. And how we make that happen, whether it's with incentives or it's with an app that actually can work with somebody to actually get that going, all of those things end up being in scope.

I would say kind of to your point, that there are 2 universes, right, in the digital health are kind of the rich and the privileged and the others. I believe that a lot of the insurance companies are actually investing. UnitedHealth, for example, acquired Vivify very recently for the remote monitoring piece. They are giving incentives for people because I think at the heart of all of this digital health is bringing the consumer into mediating and managing their own health. And all the insurance -- UnitedHealth and all other companies like that are focused on outcomes. So if the result is a better outcome, I think it's not a matter of just a subsection of the population using it, it's the whole institution is using for better outcomes. So I would say those 2 universes are very much in tandem with one another.

So I have a question for the audience, if I may. So how many digital health apps do you think exists in different app stores? There's an app for diabetes, losing weight, blood pressure. Like any guesses? It could be Play or the other app store.

Thousands.

Thousands, like...

Tens of thousands.

So the last I checked, this was like, 250,000-plus apps. Just -- there's an app for everything. Now my follow-on question, which is the crux of my point, how many of those apps actually connect to your medical record, percentage guess? Like 2%. Okay. So to me, when we are creating this app ecosystem and digital health innovations and if you don't meaningfully connect that to a patient's medical records, that all becomes a very scattered data, and I think that's my view on the digital health. It needs to be connected for patients and including the medical record.

I don't know if that's the right even metric of like defining as something being related to health because, in reality, I think there's like 3 different things that kind of get mixed when we're thinking about this. Like the first is that health that goes all the way from a basic human right to a luxury item, right? Like there's basic human right that if you're having a heart attack, someone should try to save you, all the way to going to a spa in the Swiss Alps, and that's a luxury item that's good for health. I think that's one dimension to it. I think the second dimension is that the -- there is a component of technology that tends to correlate also with luxury. So you get the newest iPhone, if you're wealthier, et cetera. And so we tend to associate access to technology with kind of affluence. But the third effect that I think is more important than the first 2 is that technology, historically, and I think, especially for health, is also a democratizer of access to services and products. And so like, I think, actually, like, over time, that's the way in which, really, we serve a lot of people much more effectively. So -- and so the -- like...

I actually want to pick on you a little bit there because you guys transitioned the business. When I first started reporting on Color, it was very much a cash business. And now it's -- you connect with a lot of insurance companies. So was that transition about trying to make the company less about serving the fortunate few?

It's actually interesting because it's -- initially, even the reason why we actually have a cash access is that we took a product that used to cost $5,000 and put a clinical-grade version of it at $250, which is still not nothing. But it made it much more accessible, right? And then -- but the -- but really, where we started working with payers, et cetera, is much more in terms of recognizing that there's not just one stakeholder in health care, right? Like it's kind of like there's individual, there's payers and the clinicians and really kind of incorporating those 3 together in the model, both from a product as well as a business model. And it's been actually pretty been part of kind of how we've been working for quite a while. So like the kind of cash access is really about an access point for people, but -- yes.

So I actually want to -- while we're on the topic of interoperability, which I think Aashima hinted that here with the 2% of apps that are connected into the medical record. So I write about digital health companies. I talk to them all the time, and they would love to be connected to the medical record. But there's all sorts of information blocking. It's not easy within the health care industry. Are we going to see movement on that this year because every year at JPMorgan, someone says -- someone gets up and says, "I just don't have -- I'm a patient. I don't have access to my medical record." And you -- we have the same discussions, and it doesn't seem to change. So who's got some optimism from me on that front?

I didn't hear the [indiscernible] keynote, but I'm very hopeful that, in 2020, the information blocking rule will be passed. I think I'm very hopeful for that, but that's the crux of the problem that we see as a patient. I think it's more relatable as a consumer. If you go to your primary care physician and you're looking for some tests from your specialist, that information is not there, and that's the reality. And unless we fix that interoperability problem, the digital health ecosystem will just become the shells of applications with incomplete patient information. And digital health is also about behavior change, right? So that needs motivation, ability, trigger and connecting those different silos, and I think I'm hopeful. But the progress has been slow, but 2020 is the year.

So I am going to take maybe a more optimistic perspective on this. I mean if you think about some of the challenges that sort of persisted inside the ecosystem that prevented that from happening. First, the plumbing didn't exist. So the capability to actually -- first, most of the data actually wasn't even digitized. So that's only happened in the last decade. Now that the data was digitized, the second thing was that you needed to standardize and normalize. And now we have standards. Whether it's -- largely FHIR, but there are other important standards that exist that now allow that data interoperability. The thing that you need after that effectively is the company is operating in a way that they allow that data to actually flow in a liquid way, and that's happening at multiple levels. It's happening both at the commercial level, and there are meaningful discussions, both at the policy level as well, to make that much more of a -- I don't want to say a mandate, but it will be something that I think that the clarity in terms of the importance and the fluidity of that data will exist. And there are companies that do exist now to allow and enable an individual to aggregate and collect in a more meaningful way his or her own data. The piece that's, honestly, is still missing is the demand from the individual. Now there are people who absolutely want and need their data, and that's beginning to emerge. And unfortunately, those situations tend to be rare diseases and so on and so forth. And those are terrible situations, and thank god that these changes and evolutions are happening. But what we need is to create the applications that create value for me as a consumer of that data that will drive the use cases. And so I think we're just in the beginning of that inflection point, whether it's clinical decision support systems or more recreational use in some of these health and wellness apps that you're talking about. But that will certainly evolve, and it is evolving.

I'm highly optimistic that the interoperability will move big -- in a big way in this year. And the reasons are -- I actually slightly differ. I believe, actually, the whole focus on patient experience is going to be the biggest driver for change. It's going to push -- whether they are providers, whether they are payers or pharma, it's going to push the needle because they need systems which talk to one another to have a great consumer experience. So that's one part of it. Secondly, I think the regulations and the whole industry actually betting on standards. The 21st Century Cures Act is a big initiative in that regard. But even the interoperability pledge that a lot of the top tech companies have taken is also making a concerted attempt into focusing on those foundations of interoperability. I think gone are the days when people think about just products. It's much more about ecosystems. And if those are to thrive, interoperability has to happen. So consumer is going to make the change.

I'll be super optimistic. We've actually pulled in all the heterogeneous sets for all of the folks that we've served, run it through -- in ontology so that we can homogenize some of the pieces. And it is available, and it's available down to a subset of those who -- those members. We're doing pretty careful staged roll out and watching how people use it and what they use -- what those things look like. That set is also available for other outside partners, parties to come in and actually use, subject to sort of the use rules around what we should be able to use it for. The data blocking thing, it is frustrating, especially when you sit back and you don't really think about what's happening. If I said to you, "Hey, Chrissy, we're going to expose all of your notes from every conversation that you have with everybody in a public way." You'd probably retract from that quite a bit. And you might even go to the point of saying, "I think I'm not going to keep my notes very carefully anymore."

Yes. I'd be burning them in a fire.

So you could see why a doctor or a health system, somebody would say, "I'm not so sure I like that idea." Now there are other things that can be done. We've looked at claim and different types of disease states and do you need breadth of data, meaning everything that's going on? Or do you need depth of data? If you're in the ICU, depth of data turns out as super important. If you have a chronic condition and you're running around not hitting the hospital, turns out breadth of data is way more important. Now if you're running around, we're trying to think about sort of how to help you and how to engage, do I need to be able to have every single doctor note available and posted up for any vendor that comes to look at it? And the answer is probably no. In fact, actually, I would -- I bet all of us would be uncomfortable with that idea. Right now, I'm in a camp where I would rather that -- I don't like data blocking, but that's a really big term that captures a whole bunch of different circumstances underneath it. The idea that you can't get access to anything, that seems problematic. The idea that you can't get access to everything, that doesn't seem problematic. It feels like if I got a patient in the ICU, I don't want to have whoever is creating that note to have a restriction in their own mind about. I don't want to write something down. So -- and when we look across -- we've got about 50,000 or so physicians that are practicing in real exam rooms all over the place. So it's not -- everybody here knows that. But it's not just an insurance company there. There's all these doctors that are doing those things, too, and they have to think about those issues. And I think we actually owe those types of folks a little bit of deference in terms of what are they doing to practice and to -- and with all respect to the -- it's like a consumer industry. This little device right here doesn't change very fast either. The software on it does change. But in health care, every interaction between a physician and a patient is a unique opportunity for a unique decision to be made. There are millions, if not billions, of experiments going on because you're at the edge of what we know from a science perspective. That's why some of the stuff that's happening at Color is actually really important for where the world is going to. This idea that change happens, figuring out how to capture that and then deploy it in a way that helps is the real issue.

Yes. I think...

Go ahead.

I just wanted to comment on one thing about information blocking, right? A very pertinent point about what does get exposed. But I think when I was talking about the legislations and how they are morphing also, we are talking about information blocking, but ONC is now also chartered with the target of what are the things that are not information blocking. So that actually shows the realization that there are things which are information blocking and which are not. So the process, at least from a legislation standpoint, has evolved to a point that all these nuances are actually getting factored in.

Yes. I mean I was just going to -- I'll jump in, but we can come back to this, which would -- which is that I also talk to physicians about this idea of opening up their notes. And there's definitely fear among the physician community that they're going to write that a patient is obese or difficult and what if that patient sees the note and what would ensue there, and I think there is something extremely valid to that concern. But I want to also kind of bring up the privacy issue as well at this juncture because it's been in the press. This is -- when you mentioned any large tech company, Google, Amazon, the next word that you will often hear is privacy. And this certainly has come up for Google lately with the Ascension partnership. So I wanted to start with Aashima there and see if you had anything more to share about that and what you're thinking kind of at this point but then broaden it up to the panel to think about how we can weigh the sort of the privacy versus some of the utility that might come out of opening up access to this information.

I'm glad you brought it up, Chrissy. From the Ascension work, we're very proud to be working with them. It is a long partnership with them. And if I look into what they're doing with us, it's no different than any other health system and health care companies who are coming to cloud to really migrate their data and with the intent to really gain analytics or build AI-enabled applications on top. So they are bringing digital to the cloud. It is private and secure environment for Ascension, and it's -- we have BAs with them. It's all protected. They have a key. They control the access. The second project that we're working with them and other [videos] out there, it's intelligent layer on top. It's all EHR search, which is to really look into a patient's history. And if you're looking at the patient's notes, it's cut, copy, cut, copy. And it's -- there -- this tool would enable their clinician and nurses to really, a, define information they need from the vast history and then being able to get that at the point of care so that they can make the informed decisions. And again, the -- there is that skepticism around tech companies, but it is no different than people leveraging cloud, and we are modernizing their infrastructure. Like if you look into the TechCrunch report, which was 2 days ago, billions of DICOM images because the medical center storage was not secure. It's out there. It has PHI, right? That's news there. So moving to cloud there, that's what essentially you should realize is that we need to modernize that layer, and they are doing this on-premise analytics and warehouse onto the cloud.

So maybe just to state, hopefully, the obvious. I mean, from a pharma perspective, I mean, obviously, we hold patient identity and privacy as extremely important. And maybe the industry itself, the broader health care industry can learn from some of the models that we have, which are good and bad in many respects. They're limiting and -- but they're also empowering. So we, by nature, we have an informed consent process. We don't necessarily need to implement it in the same way, but that's the challenge, right? Because the reality is, is when you ask the vast majority of people, would they contribute their data, their personal data for research purposes, and in particular, for health care, the vast majority of people say, yes, right? And so it's really just a transparency and an ordering process of instead of finding out that your data is actually being used for research after the fact, let's just ask for that up-front. So I think it's that change, that shift that really needs to happen, and we do that, thankfully, already for good reasons inside the pharma industry. And hopefully, the rest of the health care market will catch up. I do think it's equally important for all of us to recognize that the evolution of technology and the vast amount of data that's being aggregated and the ability to actually do pattern recognition means that also the ocean of identity and privacy is also going to be changing. And so beyond even that consent, I think we have to think about those things differently. What is PHI? What is identity? What is privacy?

What's it mean to be de-identified?

Or to be even de-identified, exactly. And certainly, as we look at multi-omic imprints, but not even that. I mean if I go to the same -- if you look at completely de-identified sort of data sets associated with, well, maybe not me and work because I'm in a different city every day. But like for most people, I mean, you could pretty much map out in a completely de-identified data set that, that's Jeremy Sohn, right, or that's you. And so identity means to me...

That's kind of terrifying, right, just to mention that.

Yes. It's terrifying. It's terrifying. So the trust relationship, maybe the transparency needs to shift to transparency in terms of what are you doing with that data? How do I know -- how do I actually get visibility to where my data is flowing and how it's being used. And then we can then think about, all right, how do I now sort of create limitations around the edges of that use because that's super hard.

That capacity can go both ways, right? Like, because now you also have the ability to know what is being done with your data and much more so than it ever was possible before. So I totally agree with Jeremy about the -- at least what we've seen is like over 80% in clinical care context, people actually are very willing to contribute honestly to research, especially when they -- when it's made up-front. And the -- but I think the...

Can I make a point to you? I would be willing -- just personally would be willing to share a lot of my health information with different entities up here, but I would not be willing to share it with you because of the loophole with how -- maybe -- but because of the -- because of GINA and the law against genetic discrimination doesn't cover long-term care, life insurance and disability. So for me, I just -- I don't see the regulations protecting me as a consumer to make it sort of worth my while to give you my genetic information, if it could be used against me.

So I think that's an interesting point. Like, I mean, that's -- I think there's -- and fundamentally, one thing to realize is that those are limitations in our kind of current like legal framework, not -- they're not connected to the actual technology in itself, right? When all is said and done, all that genetics does is that it's a way to generate a high amount of signal that can be used for your health. And now we're at a point where that signal can be used in a life-saving fashion, right? The ability to control the use of our data and to stop it from being misused, I think that is something that is fundamental to kind of like the trust relationship that, not just like in genetics, but through all of health care, is part of the contract, right. Like the -- that's what the ACA had a very big impact on that in terms of like preexisting conditions, et cetera. But I think that's actually a part of the long arc of data's relationship with health care.

I think that's where the...

Sorry, I just want to make one slight joke, but we should ask Ken because my guess is your genetic information is going to be interesting but less informative than probably de-identified data on what you ate, where you went, how often you sleep...

Amazon probably has that.

So on and so forth in terms of your potential cost to the health care system or who you're friends with.

I'm just going to stay silent on this one.

I was going to say the work that Kevin aligns and the different industry groups are proposing and doing around consumer-directed data exchange because to your point, you should have that right to share and maybe not of Color, maybe with the other entities. So that choice is with consumer, but that consumer-directed exchanges are still -- sometimes is growing. And good work being done by Kevin and a few others to push that ball forward.

From a Salesforce perspective, from our perspective, trust is the highest value. So privacy, as was mentioned, informed consent, I think that should be the bedrock of privacy. The consumer should know where the data is, how is it getting used. It should be nondiscriminatory. At the same time, all the time understanding and getting educated on the whole process of how it is getting used is extremely important. One point I would mention, the ability to revoke after having given data because there are certain discussions that are happening. "Hey, once we have given the data, we want to actually use it, and it's very valuable." I think the revocation piece is extremely important. I think one other part is I think technology is improving enough that there was a point about anonymized data. You still can point to things pointing back to that individual. I think with the advances in technology, we can actually build the foundation that the anonymized data at least remains, for all intents and purposes, anonymized.

Any thoughts from the insurance perspective?

There is nothing more to beat up on this topic. It's -- well, actually, there is a small piece. Trust with data, it is an important thing. Consumer trust is one thing. In health, trust takes on a new dimension that I think is important to appreciate. One of the things that you need to trust is that the information that you're getting is accurate. So if I do a search on something and I brought something back if that's ordered by who's sponsored an ad or whatever. And I'm not making a knock on a business model at all. But if the things are at the top, are those ad sponsored? Or are those the things that really are going to be most helpful to me and what I'm doing. In health, that actually takes on in very real ways, life and death type situations. So to take a really easy answer or a really easy situation. if you take out -- on an annual basis, we have about 1.6 million, 1.7 million cancer -- can really diagnose cancer cases. And in the example, I apologize because I guarantee I'm going to touch on somebody's personal situation in the room. But if we take out the top 4: breast cancer, prostate cancer, lung cancer, colorectal cancer, that leaves about half the cases left. In the rest of the cases, you max out at about 20,000 cases a year. We have 10,000 medical oncologists in this country. And all the rocket scientists in the room will really quickly go that's an average of 2 cases per oncologist per year. In reality, you have some that only see a case like that one every other year or every third year. On the other extreme, you have some people who kind of specialize in that type of case. But when cancer happens to you, almost always, it only happens once. And so the idea that I need to talk to somebody right now to figure out not just what type of cancer I have, I got to get back all the way up to who should I go see. I could do a Google search and say -- or any other search engine. But since you have 90% share, we'll go with that. You could do a Google search and say, "Okay. Show me oncologists close by me." But if the oncologist close by you has only seen a case like once last year, that's not a good oncologist to have implementing a brand new cutting-edge therapy that just came out, that might have serious consequences to it. Or do you go talk to your primary care physician? Or do you go to your social media network and say, "Hey, I have cancer. Where should I go?" And then somebody responds with, "Well, my uncle had prostate cancer, and so you should go here." Not even thinking about prostate cancer, it's almost always with the urologist. And maybe you have lung cancer, and that's, by the way, different organ systems. So the idea that, that information is being used to guide what happens next to you is an element of trust that we don't talk about very often. We talk about -- I need to trust that -- I mean, we have data with Google. It sits in there. It's protected. It would be like the world's largest lawsuit if that protective lock-and-key data became unlocked. That's not going to happen. We have stuff with Salesforce. The same thing. It's not going to happen. We have business-to-business trust that they're actually protecting the data. But that next level when the patient calls up, which, by the way, is one of the most frequent phone calls we get, I have a serious condition. Where do I go? Now there's a new level of trust that's actually necessary in that transaction about what happens next.

Yes. Great point. So I wanted to kind of raise another issue about something that seems fairly unique to me about health care, which is that I have colleagues that cover other areas like education technology, and fintech is another one. And in the course of the past 10 years of their reporting, you've seen so much infiltration of disruptive technology into the space, and it's really moved in leaps and bounds. But with my sector, it seems to be much slower, much harder. There's a lot more failure. I mean, not to pick on Google, but Google Health didn't work. We had -- Microsoft HealthVault just shut down. It hasn't been an easy road. So would anybody comment on just is health care more challenging? And is that a good thing because you're talking -- unlike fintech, I mean, it's certainly important, but lives aren't at stake in the same way.

Let me take a first crack at it. I think the point about Microsoft Health or Google Health kind of not getting the same amount of traction was fundamentally focused on the fact that they were relying on the patient actually collecting all the data. Fundamentally, that was -- now if you compare it with Apple, the way they are trying to do it, it is -- they are mediating across different EHRs getting that information, the success is much higher.

Is anybody in the room, by the way, using Apple Health records to access their health information?

It's a small number, reinforces my point. There aren't use cases yet for it.

Check this out. though. Is anybody in the room ever had a claim processed?

So I think there's a different angle to that from the complexity, Chrissy. Health care, again, having spent years in Kaiser Permanente, the health IT landscape is pretty complex. So it's not a big tech or a health care or start wave. There's an ecosystem problem to really push the needle forward. Now we talked about the trust from the consumer standpoint. Let me ask another query, again, to the audience. Let's say I'm a big health system. I'm really looking to reach out to my population for female aged 45 to 55 who have a BRCA1, BRCA2 gene , and I really want them to come back to get a mammogram done. And they have insurance, and the insurance allows for the test included. And I want to reach out to that population today. How many people think you can do that query today? Like you -- that's the internal. You have the systems, and it's not the vast information, but in the -- do you think how easy to do that query?

Health systems allow or...

As a health system, for them to make that outreach to the patient population, to reach out to folks so that they can get the mammogram done, like screening, because your health information is in EHR, your membership system is completely -- and then your DICOM images are sitting in some PACS system. So today, the problem is all data silos are very much reality of health care. And I know we talk about standards, and there's that -- it's true that HL7 standards have done a phenomenal job, and FHIR is new and up and coming. But just because a new standard exists, do you think you can map your back-end system to the new standard? Is it magic? It's hard work. It takes years, mapping around thousands of fields for millions of patients with validation. It takes years for them to take that leap. These systems have been built for many, many decades in the past. And for them to take leap forward, it's very humbling. It's hard.

Chrissy, I was trying to quote the example of Microsoft and Google, right? From a personal -- I think this circumstance was different. My belief is that the technology adoption in health care is definitely slower. If you look at telecommunications, they have digitized their whole processes. In case of health care, the silos are huge, and the focus on interoperability has been way slower. If those 2 things have changed, the adoption of all these new technologies should be as easy as in any other case. I think it's a matter of telecommunication. Their interoperability standards are equally important. You try to get a cell phone, changing it, making it easy, you can do everything online. Try doing that in health care. One silo doesn't talk to the other. That is at the heart.

Yes. They use -- I mean you used to keep your phone number hostage, and now they don't. And in health care, I think we still think that we have to trap patients inside the health system and...

And also at a different phase of the digitization of the industry. Like, I mean, this actually, like right now, health care feels like the mobile industry felt in the early 2000s. And like, for example, now Vodafone was the product of, I think, like done dozens of acquisitions. And they still were running literally like 20 different billing systems internally. And so if you moved states, they could not like reconcile your bills, et cetera. So you have like very similar problems actually. And so I think just the -- I think health care is maybe just at a different part of the cycle. It's probably actually not even that more complex than telecoms in practice. It's probably just that we are kind of still in the earlier phase of ingesting technology and as a...

I think we need interoperability to be real. There's no dearth of ideas in health care. These ideas need execution, and those execution need doers, and those doers need tech to really make it happen. Like we talked about air mapping from [FHIR]. It's a pretty massive undertaking by itself. So what can we do to bring those tools together? The -- one other example I was going to use was we talk about interoperability, and I've been saying that both as been before joining Google as Apigee in launching interoperability with Project Argonaut and then [indiscernible] with Kaiser. Each year, we go to conferences. We say interoperability, and we sound innovative, but it has not moved. Right. We keep saying this word, but it really needs actions, doers, and doers need tech. And I think we need to be building those tools to help the health care industry to move forward.

So I should preface this by saying I spend every day -- every minute of the day arguing that we can be faster, and we could be better, and it's just too slow. So I preface by saying that. But in a reality check, the statement that health care is not innovating in speed, I don't think is a fair comment for a whole host of reasons. I mean we could look at many examples. I mean we are now in an age where we are curing disease. Started in hep C, and now we're doing it in cancer. And thank god for cell and gene therapy and many other next-generation technologies, technologies, not health care, technologies. We are literally eradicating certain forms of disease, unheard of. 15, 20 years ago, the health care system was all run on paper. You would go to the doctor's office, okay? You would -- they would take the file. They had carbon copies of your documentation. I mean that's how we ran clinical trials. And so the entire industry is now digitized. It hasn't been perfect because, as everyone has described, in some sense, we are in a 3-sided marketplace, a 4-sided marketplace, maybe even a 6-sided marketplace, and that absolutely creates challenges, right, challenges that are super frustrating, right, for -- I'm an entrepreneur, actually, before I came to Novartis. But the reality is that there's a tremendous amount of change, and I can keep going on. In the third -- you made a point earlier,that may be sort of less second-world, third-world countries don't necessarily benefit from digital health. In fact, some of the most exciting opportunities are happening there. And we do a lot of our innovation work I think actually wrongly. We should be doing it in third-world countries or in smaller countries, but we should be doing it in both, in first-world countries and in developing countries. But why? Because we're in Ghana, and we are distributing life-saving medicine by drones. Think about that for a second. We're not doing that here yet, right, but we are absolutely beginning to do that in third-world countries because we can, right? In third-world countries, the infrastructure, it doesn't need to be replaced because it's being built. And so a lot of the frustration is the 6-sided marketplace and the established institutions. And that's a good thing, and it's a bad thing. But it's absolutely happening, and we should celebrate it, and we should continue to struggle through it and fight for making it faster.

I agree.

I just have one element. The health care industry is innovating, right? Precision medicine. Color sitting right here is there. The fact is all the innovation is happening in the silos. The innovation from a patient experience is way behind as compared to other industries. I think that's the fundamental point. And those innovations are happening in those silos. As a result, the end consumer is actually not seeing the results of the connected system. That's at the heart of what fintech is doing much better. Because as part of my portfolio, I see them evolving across 7 different industries. That's a fundamental difference between health care and other industries.

I don't think that's true.

That's a good discussion then.

The outcome -- when you think about the outcome and experiencing the outcome of innovation, today, we have people that live that yesterday, they died. That is an outcome that is pretty hard to put a dollar amount or an experience or anything else around. That's a pretty darn good experience. There are babies that are born today that, 15 years ago, didn't stand a chance of making it a week. Jeremy is right. Innovation is happening everywhere at a pace that is absolutely incredible. This idea around information and how it flows, though, I think people who are outside look at it through a lens of the wrong analogy. I understand the analogy of like the telecom industry. That's sort of true but mostly not. It's more like -- because, see, in telecom, we all use our phones every single day, lots of times. We don't all go to the hospital and go to the ICU and hang out there every single day with a new rare disease. It happens once. It happens twice. So it happens whatever it is. It's a lot more like show me the really cool digital app that tracks all of your home purchases that you've made this year. There isn't an app that tracks all of your home purchases because there are only like 5 people on the planet that buy lots of homes in the same year. The reality in health is that these things are so skewed, you have a few people that have a very few -- I'll stop right there. I'll stop mid-sentence.

I was going to let you wrap it up with...

We'll bring it out in the hall and have a different conversation.

We'll continue the conversation. Thank you so much to our panel. This is so much fun, and I love the discussion between all the different stakeholders today. So thank you very much.

Thank you.

Thanks.