The InsureTechGeek 20: Predictive Rapid Flood Modeling with Juliette Murphy of FloodMapp
The InsureTechGeek Podcast powered by JBKnowledge is all about technology that is transforming and disrupting the insurance world. We will be interviewing guests and doing deep dives into specific technologies we see changing the industry. We are taking you on a journey through insurance tech. So, enjoy the ride and geek out!
JAMES: Oh, another day, another week. Now for us, as we record this, it is currently Friday. I think for Juliette, I think you are across the dateline, I believe it is Saturday for you. Is that correct?
JULIETTE: It is. It is Saturday morning here, 6:30 AM. So, it is a beautiful day in Australia.
JAMES: Nice. I am super thankful for you waking up so early and for coming on a weekend, and although if you are an Aussie, you are used to this very strange Tuesday through Saturday workweek that you end up having because of the international dateline. But it is really good to have you on the show from Queensland. I believe you are in Queensland, correct?
JULIETTE: Thank you so much for having me.
JAMES: And also, here, of course, the most interesting man in insurance, the illustrious Rob Galbraith Rob, how is it going today?
ROB: I am great James. Yeah, great to have you on, and of course, Juliette, great to see you again. So yeah, we are just talking a little bit off air right before we started. Juliette and I have known each other for a few years, and we had to see each other when I was in Sydney for the Anza InsureTech Conference in late February before this pandemic Corona Virus thing happened. Yeah so crazy times. So, it is great to catch up and I am so glad that we can do this virtually Juliette. Good to have you on.
JULIETTE: Yes. Thanks so much, Rob. Yeah, I was thinking it is great to see you. I have not seen you since before COVID, so it is great to catch up in this new different world and see that you are still doing okay and getting through everything. It is a strange new world we find ourselves in.
JAMES: Yeah, so we are going to avoid a few phrases today. We are going to avoid “new normal”. I do not want to hear that one.
JAMES: We are going to avoid “extraordinary times”. We are going to avoid that one. I have a few phrases. We just want to avoid them. I am tired of hearing him. But you have not said any of them, you have not triggered me yet, so it is all good.
JULIETTE: Alright. It is like reversed Bingo. I am kind of feeling nervous. I am going to say it and like “-“
JAMES: Yeah, and I am going to be “ah”
JULIETTE: And I am very competitive too, so I want to know the points. It is like, how many points do we get deducted if one of us “-“
JAMES: Five points for each major buzz phrase that you repeat. It is what it is. Look, we are going to talk about tech. Today is about tech and we are going to have a fun time geeking out. I am excited to have you on because this is a major issue. Flooding is a really big deal, all over the world. Whether you believe, well, I hope you believe the climate is changing, cause that is pretty evident that the climate has always changed. Ever since the creation of the planet earth, the climate has changed. So, if you do not believe the climate changes, then you are in complete denial of the history of the planet. It is always the causes of that climate change that are contested, not the fact that it is changing.
But the consequence has always been that places that used to not flood start flooding and places that used to flood stop flooding and water are not where we want it because it moves. I am a history nut, Juliette. I am. And history has a way of repeating itself. Human beings have a historical pattern of following water, wherever it goes. If you wind back to ancient India, the very beginnings of India, it is just a fascinating study. It all developed around the river systems. And then as the rivers went, so did society. So, water is this really deep fundamental thing to human beings, and you have built a company around water. And I think it is a worthwhile topic. Before we get into that, I want to know a little more about you because insurance is a people business. Very–very deep relationships across this industry. And so, I want to hear how somebody from Queensland, has worked their way into being CEO and co–founder of an InsureTech company. Walk me through. Where would you grow up? Was is Waka-Waka? I have a friend in, Waka-Waka. Not kidding.
JAMES: I have a friend who lives in Waka-Waka, and then I have another friend from Dubbo and, we jokingly call him Dubbo-Dubbo, which makes him very upset. So, he does not like being called Dubbo-Dubbo, but beside the point, where were you born and raised, what did you think when you were a kid, what did you think you wanted to go do, and then, what got you here?
JULIETTE: Oh my gosh. Well, love it. Okay. So yeah, I, was born in Brisbane, Southeast Queensland. But so shortly after that, my parents moved to regional, Southeast Queensland. So, I grew up in this really small town, named Bruna, and it used to flood a lot out there actually in the wet season, sometimes when I was a kid, I would be watching the school bus to come pick me up and you would see it, get to water over the road and then turn around and go back to town. And we were just like, yes, day off school is the best. So, I had always been a bit fascinated by climate weather, thunderstorms, flooding. But yeah.
By the time I was in high school, I was living in Northern New South Wales, near Byron Bay. I did a little surfing. So, I have always just been interested in waves and the coastal engineering piece as well. And, yeah, what did I think I wanted to do when I grew up? A range of things. I think my mom was always really into gardening. At one stage I thought I wanted to work in a nursery and plant trees. And then by the time I was 16, I went skiing and I was convinced I found it. I know what I am going to do with my life. I am going to be a ski instructor. And I think, you know I did a bit more research about it and people were like, yeah, it sounds fun. But I do not know if that is a really good career choice. And I was like wow, this does not pay as well as I thought. So, it came to kind of picking a course and I was good at science, really good at math’s, massive geek, and kind of talking with different career advisors, they were like, well I think you should look at engineering.
And I had never really thought about it because I sort of perceived engineering to be about building structures and bridges. And I was like no, that is not me. But then I saw in these unique courses, there was this environmental engineering course and I was like, oh my gosh, this has me written all over it. It sorts of sets you up to go down so many different paths, model the environment, like rainfall, run off the flooding side, all of this, but also looking at systems and processes, renewable energy. It just sounded so fascinating. So, I was like, I am going to study that. And then when I graduated, it came out and I did my thesis in renewable wave energy. And as much as renewable energy excited me in Australia, there are not many jobs in that, especially when I graduated. So, I was just really interested in this.
JAMES: I am seeing a pattern here of you being interested in career paths that did not have a lot of jobs.
JULIETTE: It is true. Yeah. It is a pattern.
JAMES: Did you do a gap year? I went to Whistler and Whistler Canada was filled with Australians. There was not a Canadian in the town. It was all Australians on gap year or gap year three. I mean sometimes they do two, three, four gap years.
JULIETTE: I did not do a gap year. I was just really, pretty driven. I was getting through university to get a good job and get my paycheck. But then I saved up later and it was my goal actually to go and ski in Canada. But I just did it with the company I worked for. I kind of chose them because they were Canadian company first, they had hundreds of offices all over the world. And I got a chance to move to Canada with them and ski. I told them I was going for six months. But it ended up being four years. I loved it so much. And the advance, it was just full of Australians.
JAMES: It is all Aussies. They have invaded. It is a Commonwealth country. If you do not know much about the Commonwealth, it is the Queen‘s territories. So, you can move fairly easily among Commonwealth countries, you can get a visa fairly easily. And so that is why a lot of Australians end up going to Canada. I went to Banff.
JAMES: I spoke at a conference in Banff and I spoke at a conference in Whistler and both places, all Australians.
JULIETTE: Amazing. Yeah. Everyone who serves your coffee, dinner.
JAMES: Could I go to Sydney? Where am I?
JULIETTE: Yeah you may.
JAMES: So, let us keep talking. So, you got your degree. And then what happened?
JULIETTE: And then, so yeah.
JAMES: You worked with this company in Canada.
JULIETTE: I modeled a lot of floods. I did a lot of hydrology investigations. I mean, in Australia, let us say soil moisture, drought, and flooding rains in Canada a lot of snow mountain driven hydrology. Worked on projects for government, mining energy company is one insurance project. Super interesting stuff. I have got over a decade of experience. So, it was sort of my domain. It was my thing that I did in terms of flooding, but it was probably two personal experiences that sort of bore me on this path to starting FloodMapp. Cause I never set out to say I am going to found a company, I think as a founder sometimes to be as crazy as we are to start a company, you have got to care about the problem you are solving, and for me that was flooding.
So, in 2011, in Brisbane here, that was just this catastrophic flood that caused billions and billions in damage. I think it was our best $6 billion just in the city. It flooded kind of 20,000 homes were inundated. A hundred thousand people without power, and my friend, she was living in this super, effected place, Goodna. If this is her house, the flood went over the peak of her roof and she just had no idea that it was coming. So, she kind of moved some things up on tables but ended up losing everything that she earned. And it just broke my heart going into her house to clean up in the days after the flood receded. I was just like we had a responsibility to do better. One, it was kind of pretty criminal that her house got built in this zone for whatever reason, but we should be giving more warnings. It is crazy that we live in this world where you can order Uber eats and see the pizza coming to your door, but when there is a natural disaster, you are left in the dark.
And so, two years later, I moved to Calgary, Canada, and the same thing happened. This catastrophic level of flood. $10 billion damage at least. A hundred thousand people evacuated. All my friends texting me saying hey Juliette, am I going to be affected? Should I evacuate? Should I move my car? What should I do? Because no one is telling them this. And the reason is, I got to the heart of the problem, which made me go in this pursuit of FloodMapp is that the government agencies do an incredible job when it comes to meteorology forecasting, to forecast the rainbow. And hydrology forecasting to take that rainfall and forecast to peak river height. So, you get this message saying, okay, the Brisbane river catchment is going to reach 5.5 meters. But unless you are a hydraulic engineer, it is meaningless too, is that, to sea level, what is my house floor level? What is the Brisbane river? Do I live in the Brisbane catchment? Wait, what catchment do I live in? Everyday people do not necessarily know all these metrics.
And so, it is crazy to give this really broad warning and expect people to know what it means. And so, the next step to, if you were to give people a map, to show them what area and what properties would be impacted, you need to do the next step, which is hydraulic modeling or hydraulic forecasting to turn that vertical river height into a horizontal map, but previously this is just being so computationally intensive. It has not been feasible, and governments cannot offer it. And that is what we built at FloodMapp. We were just we are going to design a model that is built to run in real–time, so that before a flood people can see like, hey James, your house is going to be impacted or, your site or your substation or your assets. So that people can kind of prevent the loss before it happens. Some, not all. And preserve human life. I mean, ultimately.
JULIETTE: So yeah, I kind of just started, it was a hobby. I built this app in my spare time. And then yet got involved with my partner and co–founder, kind of developing the code and prototyping. And then yeah, eventually we sort of, so I think we showed it to some people from local government and they were like, you should start a business. People would be interested in this. And I was like no, that is crazy. I do not know anything about business, but then we sort of thought about it some more and ended up getting some seed funding through an accelerator program. And yeah, it has been a bit of a journey since then.
JAMES: Wow. That is awesome. Awesome. Awesome. Rob?
ROB: Yeah. So, I kind of want to pick up on that Juliette, but I kind of specifically want to talk about how you relate to insurance. I know you guys are also working a lot with emergency managers, and I imagine that is difficult particularly now. In the US certainly, some of our responses to the pandemic have been through FEMA, but yet, we just started our hurricane season, June 1st, etcetera. So maybe you can just talk about, you kind of mentioned some of the challenges with government messages, but who are some of those key stakeholders that you work with and how is insurance involved in who are some other major players outside of the insurance that you guys work closely with?
JULIETTE: Yeah. Yeah. Thanks, Rob. So, I see ultimately insurance as being a huge piece of this puzzle. Cause ultimately, that flooding causes so much damage. In the US alone Congress right now budgets $54 billion for hurricane and flood–related damage. That is a big number and it is growing. The thing is it is growing for two reasons. I am glad I am in the right place. And James wanted to check that I believe in climate change, which is good. Cause yeah, it is making flooding more frequent and more severe warmer oceans. We are seeing increased hurricane tropical stolen oppressions. But also, development is growing. Populations are growing. So, there is more infrastructure, more houses, and that means losses are climbing. And yes, this is impacting residents. It is impacting utilities, it is impacting the energy sector, but ultimately, it is impacting insurance.
Ultimately who are the people that are paying out for these losses? Insurers. And so, I think it is a super key piece of the puzzle, I think ultimately, yeah, we would love to have more and more government clients, and this be a messaging app that can also save lives. But I think with any startup, it is a journey and we see the loss prevention piece as a sort of easier piece to start with where we kind of build that trust and become the most trusted provider of predictive mapping in the world. And so, yeah, for, I mean, insurers, when you look at it, it is interesting what we have learned from so many discussions is there are multiple pieces along the insurance value chain. Where there is a lot of interest in our technology, obviously the number one is loss prevention.
And I think, it is sort of a growing area that maybe has not been a big focus traditionally, but when risk and exposure are climbing more and more, and there are these traditional methods of risk grants, but you are still getting this big exposure. The obvious thing is to sort of look at what can we do to prevent the loss and prevent the claim from happening in the first place. So particularly that I think has been a real interest to some of the big reinsurers and particularly in the commercial markets. When you have things like, and it is like a shop front and they have 24 hours warning time that their shop is going to be inundated and they can move all that electronics or mobile phones out. Maybe that just saved a $1 million claim just because they put some stuff in a car and drove it away. It is just crazy when you think how much of this damage is preventable.
People cannot move the house, but I think the world bank is saying anywhere from 35% to 65% of flood damage is preventable with early warning signs. So, there is a huge opportunity for insurers to reduce those claims and financial losses. It is also a huge opportunity for them to just streamline the claims process itself and improve their customer experience in the moments that matter. So, if you are an Insurance policyholder, you get a message from your insurer that it is going to flood, and that is an opportunity to safely leave with your family or save some stuff in your business to kind of minimize business interruption. I mean, you are with that insurer for life, as you said James, insurance is all about relationships. And if you have these two major touchpoints with an insurer when you buy the policy and when you make a claim. And if that experience in the claim is a really bad one, that is a key touchpoint to change insurer.
And we have seen that happen a lot in Australia after big events are people either love their insurer and stay for life or they will change insurers. So, insurers are seeing it as a really big value point to retain customers and even acquire customers to build that brand as this trusted insurer. As well as automate things like claims. So, you can see the flood of events coming, there has been a big hurricane and you have got this map where you can visualize your whole portfolio. How many of my policyholders have been affected? What location, how many resources do I need for claims? Where do we triage? Where are the highest volume claims? It adds a lot of efficiencies and reduces costs in the claims process as well. So, yeah, I think insurance is a huge, huge part of this problem.
JAMES: So, let us get you out for a minute. How does it work? Are you using art GIS in the background or are you using a massive GIS database and then you are taking this massive amount of water and then doing volumetric calculations? I mean, just walk me through the math and the computer science behind it.
JULIETTE: I love it. I love geeking out. Okay. So, I will preface this by saying I am not a software engineer, environmental engineer, for sure. I am not the one writing code every day. I have people much smarter in the company that is doing that. The way it works, we, yeah, as you say, it is all driven by spatial databases. It is powered by tools like PostgreSQL and prestige’s QGS we use a lot too for visualization of the results. And we are just ingesting millions and millions and millions of records of data. Across the US, we are ingesting 18,000 gauges of river height data every hour and running that through machine learning models. So, it is two steps and modeling we do. We do not forecast the weather. We ingest that data. So, we kind of ingests weather on predictions of rainfall.
We put that into our hydrology models to forecast the river heights, so this is doing a simulation to work out, okay, what is the catchment characteristics with the area? How much of that rainfall is going to soak into the ground? How much is going to run off? Then how high is it going to get? We have this cool machine learning models that are learning the relationship about all the gauges throughout the stream gauge network and learning from each other so that if one gaze goes out during an event, you can kind of still predict, all the unengaged parts of the Katherine. And then that goes into, those hydrology results go into our rapid hydraulic models. So, we have just, we are very excited to say we have just developed a name for our rapid hydraulic model. It is called “Dash-2D” Dynamic Automated Scalable Hydraulics, 2D
JULIETTE: Because flat engineers have to have an acronym for everything.
JAMES: The next version will be Dash–3D. Then the next one will be Dash-4D, and the next one will be Dash-5D, and eventually, it will on Dash–12D the next time we talked to her.
JULIETTE: Pretty much, pretty much.
JAMES: Now look, the amount of water that gets absorbed into the ground is certainly highly dependent on how much impermeable, permeable surfaces there. So, are you using satellite data to calculate an impermeable surface? Or are you using just gross estimates based on population density and urbanization?
JULIETTE: Yeah. A range of different factors. It relies on a lot of aerial imagery to understand different land–use types, whether it is urban, forest, farming, etcetera, because all of those as you say, have different infiltration rates. And then a lot of kind of calibration or machine learning of models to just basically learn how relationships, so sometimes we do not necessarily tell the model exactly what the surface is and therefore exactly what the infiltration is, but just let the model learn that relationship. We are doing a lot of exciting R & D on that front at the moment.
JAMES: Could not you just measure input and output? You know the gross measure of inches of rainfall over a specific acre or hectors or acreage, and then look at your river levels and then interpolate the permeability of the surfaces between the two? Is that another way to validate your model?
JULIETTE: Exactly. Yeah. So usually, and the US does have an incredible stream gauge network. It is also pretty good in Australia compared to some other countries where they will measure streamflow at certain points and usually pretty strategic points of a catchment outlet, and that is your kind of point for calibration. So, you are essentially training the model in historical rainfall data and maybe some evapotranspiration data, like how much of the water is being taken up by the plants and going up to the sky. The physical catchment characteristics, how large, what area, the steepness of the catchment land–use type, that sort of thing. And then training it on that measured streamflow as you say.
JAMES: So, before I hand it back to Rob. Who buys this from you and what kind of company buys this from you and what do they do with it?
JULIETTE: Yeah. So, and probably to go back to your point, you mentioned OpTIS earlier, so how they buy it, it is delivered via a data feed, like a web map service. A web feature service, or kind of API. It can connect our outputs of the rapid hydraulic model, which is usually a vectorized polygon file that can then land in their database. That is something like OpTIS or even Overlay on Google Maps or Maps on Map Books. Products that we can export a shapefile to Jason or KML does not matter. And then that way they get the most power out of it because they can overlay it over their insurance portfolio or their kind of network of assets or infrastructure. So, who buys it? Insurers, we certainly worked with insurers in Australia and the US, but also power utilities have been interested. We got a few inbound leads from power utilities in Australia. And for now, we are kind of talking to pricey power utilities in the US and they are super interested in it.
I think for them, it is big, I mean, public safety, number one. You got to know which substations and power lines to switch off because you cannot have power lines in the water, a public safety issue. It is a huge problem. And then yeah, also government cities, mining energy companies. But yet it is interesting, a lot of the driving force for buying this product for those commercial entities too, it all comes back to insurance. Because they have suffered flooding in the past. And if they have had an event where there is power mixing with water or they have lost their assets, what happens is they struggle to get insurance, or their premiums take a massive hike. So, we are kind of working with all these industries together to ultimately find that solution that is win, win, win for everyone.
JAMES: Awesome. Rob?
ROB: Yeah, that is awesome. So, Juliette, I know a lot of people. So, there are a couple of things that are kind of coming together and I want you to help maybe kind of parse this out and particularly, maybe differentiate what you guys are doing, from FloodMapp versus, some other folks. So, there has been a lot of interest in flood modeling over the last five to ten years, both from startups and from some really large corporations. There is also a catastrophe model that insurers and reinsurance heavily rely on these types of events. And so, it can be very confusing to kind of know what I use for what and what is complimentary versus what is a competing product, etcetera.
I believe, when you and I were in Australia at the Anza of InsureTech Conference, during your presentation, one of the things that you were emphasizing is, knowing the catchments. Knowing when the rain comes down. Well, which catchment is it going to go into? And that is not something that is typically kind of mapped or captured. It can be very challenging to capture accurately. So maybe we can just talk about, I guess, help us sort throughout all these different types of models, and then, where would I want to use FloodMapp or what do you think kind of differentiates you from some of the other providers?
JULIETTE: Yeah. Yeah, sure. Rob. Yeah, you are right. There is a lot of solutions out there and there is a lot of modeling technologies that are being used for a long time. And there are a few different spaces, so, I feel I do not have a whiteboard here, so I am going to do lots of hand motions to help explain this, but yeah. And the insurance sector, there are these CAT modeling firms, there are AII worldwide, RMS, JBA risk–based types of players and they typically use models that are known as 2D to define different models to a model that hydraulics, which is useful for flood hazard mapping. So amazing models, and then definitely built for purpose great for flood hazard mapping, but they have a downside.
So, these models, they work by basically dividing up the land that you are simulating into millions or thousands of different grid cells and running the model over thousands of time steps that the lower you make the grid size, so, you say a one-meter cell has to have a finer time step then a 10 meter or 30-meter cell, because the calculation sort of have to be at a steady enough point that the model does not go unsteady. Anyway, the point I am getting at is they are incredibly computationally intensive. So, to run one of these models for just one catchment of 5,000 kilometers squared might take you 600 hours at a 10-meter resolution. And so, what that means is these firms often to go and have a large scale, they need to increase the grid size to have an acceptable runtime. Because you cannot just have these models running for years. So that often means for the CAT models, they are typically a 30 meter, sometimes a 90-meter resolution, and then you are kind of losing some of the accuracies. But sometimes that is okay. It does mean you have got a large–scale solution.
And then you have also got the engineering consultants typically have done a lot of flood modeling, doing flood modeling projects for FEMA and hazard mapping to cities and things. So, they also use this 2D to kind of funded different site models. It is very much what I used to do in my previous career. And so, the way kind of engineering consultants work is, because flood modeling is so computationally intensive, takes so long to set up, can take weeks, a month to set up, weeks a month of simulation time. So, they will often do just a small area, just a government, Local government flood study simulation. And these models are not feasible for running in real–time.
So, what we developed and why it is different, we are not a 2D funded different models, so we do not use the SIM vengeance equations to simulate continuity and momentum down to a physical kind of step in each cell or each time step. We sort of take a more high-level approach and yet by implementing a lot of big data automation and machine learning techniques, we are essentially able to sort of building a work surface in 3D across a whole catchment scale in sort of anywhere from seconds to three minutes where other models might take 600 hours. So, we are getting sort of run speeds of 10,000 times faster because of the new technology we are using.
And so, the interesting thing is we have made a new set of algorithms and approximations and simplifications when it comes to handling some of the physics behavior and the hydraulic behavior of the catchment, but the interesting thing in that research is that, because our model can run much faster, it means we can decrease their resolution and run a much smaller essentially equivalent cell size. We have done sort of 15 centimeters, which ultimately is showing that actually in some places, even though some people might look at the two approaches and think ours is a bit more simplified, we are showing in some ways it is outperforming a more accurate even than the traditionally considered very accurate approach.
So, we developed this model very much for emergency management, knowing that just out in the market, across the different flood models, they are all very focused on flood hazard mapping or infrastructure design. No one sort of build a model that is focused on that real–time instant stimulation. And that is what sets us apart. So, the technology we developed, it was sort of purpose–built for event-based modeling. So not modeling things the one in 10,000 or one in a 500-year flood, but to model, there is a hurricane off the coast, and it is going to hit in 24 hours. What is the flooding going to be?
JAMES: Awesome. Super cool. This is all really exciting stuff. We need it badly. Especially here at Houston is arguably one of the, you know Houston and New Orleans. I am from the New Orleans area. I live in the Houston area.
JULIETTE: Oh wow.
JULIETTE: Two of the cities that have had the biggest problems because they are at sea level. And, now they tend to, they seem to be flooding more and more and more and more. The whole real estate market talks when you buy a house or sell a house there, they talk about whether or not it was in the flood area the last time. And nowhere seems safe. Nowhere seem safe. And so, I would say, I am just curious as to like, what is the next big thing. This is a big deal, giving them a KML fee. They can bring into all their system and do their modeling. That is really what you are. You are a data subscription company, right? Yeah. And so, they are subscribing to your data feed. They pull the data feed into their software, then they use it and it is different than the hundred years, I mean, this is the crazy thing you know, the hundred–year flood plain, the five hundred–year flood plain.
People talk about that, but I do not think even understand exactly what it means as far as probabilities go. And it seems to be fairly meaningless because a hundred–year floods seem to happen more often and so; they do not seem hundred-year floods by the definition of the phrase. So, you have already made major headway in creating a technically distinct product from the market. More accurate. You are leveraging machine learning and a whole bunch of data sources and a whole bunch of math, a whole bunch of computer science. I mean, you are leveraging every tool at your resources and you are delivering far more accurate data feeds that carriers are making better decisions on probably premium and rates, right? Whether or not to underwrite it. And if so, at what rate? What is the next big thing in flood mitigation and mapping though? Like where are we going to? Are we going to a place where you can simply key an address in and hit your data feed and make a binding writing decision on the spot? Are we going to dramatically reduce the amount of time for underwriting? Are we going to start getting into prevention, using your data on making recommendations on what can be done to fix it?
Cause that seems to be the big thing. I was a city councilor in my city and city councils and flood districts do not exact our river districts for the river authorities, we call them in Texas, do not exactly know what buttons to push, to fix the problem. So, are you going to go that far where you start saying, by the way, cities, if you want to lower your, your ISO rating, you want to lower your flood rating, here are the three things you can do based on your geography and our modeling that will help fix this? Right now, a lot of it is emotional, and I have seen the presentations because they make them, they take their best guess at what would mitigate flooding, but they do not know. They do not know because a lot of the recommendations in them are not working. So, so what is, what is the next step? Are you going to start working with NASA, the municipalities, and river districts and start making recommendations on how to fix this stuff?
JULIETTE: Hey, I love this. I feel we could just talk all day about this. There is just so much I want to talk about. That is a great question and I do not know that there is an easy answer. It is a complicated problem. I think ultimately building more flood resilience. It is so multifaceted. Some things need to happen where, I am a big believer that there is a real physical kind of infrastructure problem and, consultants or solutions like FloodMapp that can kind of offer that rapid, real–time and scenario–based kind of flood modeling can help inform better development planning and things. But it is complicated. In some places, you can build a flood levy, but then, that might protect one side of the river, but make it worse on the other side or increased velocities.
And so, I think it is just this complicated thing where so much of humanity, we have settled on flood plans next to rivers because we need water. Humans need water, but by nature, we are always going to be exposed to flooding from the river, or maybe from the coast, living near water. It is just, it is sort of going to be a part of life. And I think it is the smarter development choices that can happen. Things like better planning where, rather than having these fancy- houses on the river, that is perceived to be this status and beautiful thing to have. Like we need to have the golf course on the river, the parks on the river, and the houses on the hill. Like there was this town that flooded in Lismore and new South Wales.
And I saw someone do some drone footage after the flood event and they took off and you could see the drone going up and I am, like, oh, they are on the golf course. And I was like, oh wow. The golf course is at the top of the hill. And then they fly over the central business district where all these, the McDonald’s the petrol station, the whole city town is underwater. But the golf course is high and dry, and it just seems logical planning. I think that needs to change a hundred percent. I am passionate about that, where I think historically, we maybe just did not have enough data and this planning has been done, but schemes property buybacks or where in Australia, some insurers are doing some really exciting stuff in North Queensland where homes are affected by cyclones hurricanes. If they make a claim, the insurer is kind of helping them invest more money if they will make resilient infrastructure changes to their homes “-“
JAMES: Or raise the house, raise the house.
JAMES: So, these physical things, I think is one a key part of the problem.
JAMES: But if it just flies in the face of what Aussies want. I mean, I took a boat ride up the river in Brisbane, Vegas. I went to Brisbane, Vegas, got on the boat. And we are cruising up the river and all the houses, all the Aussies are just lining the river. And I did find out something interesting. I asked the boat captain, why is no one in the water? In Texas, everybody would be in the water. Water skiing and playing around and swimming and he goes, “Oh no mate, there are bull sharks here, that is the key. They are really fast”. And I am like, wait, wait, wait, wait. There are so many things that if it is not the floods if it is not the crocodiles, it is the bull sharks that will rip you to shreds. So, no one is in the river anyway, so they cannot go swim in the water. So, you might as well.
JULIETTE: What is the point?
JAMES: What is the point? Everyone was on the water and they cannot get in the water because they will die anyway.
JULIETTE: Yeah it is very funny. I actually, I have been wakeboarding and the Brisbane River. Touch with luckily did not get eaten by sharks so people do-do it, but yeah, the bull sharks are a real thing. They are aggressive.
JAMES: Oh yeah.
JULIETTE: I have seen a lot of shark attacks.
JAMES: Especially, and he goes: “and right now it is the mating season” and I was like, oh, crikey. I was like, where is Steve Irwin when you need him? “Now you will see here, the male bull shark swims up on the female” and then they are all words like Roy too much! Alright, Rob, last quick, last question. Bring us home. Let us wrap it up, Rob. What have you got?
ROB: Yeah. I want to save everybody from your “-“
JAMES: Shane Martin Queensland.
ROB: So, Juliette, I am going to kind of share with our listeners a little in secret. And that is that we’ve kind of made a resolution that every time we get together at InsureTech at Las Vegas every year, we are going to go out and grab notches. It is an annual tradition now that we are going to do with you and the FloodMapp team. So, I do not know if we are going to be in Vegas this year or not, but I kind of think about where you have been. I know you have won a ton of pitch competitions like you said. You guys just started, you and Ryan and you built this great company.
And so maybe you just, I guess, take a moment to reflect on the journey so far, but I am interested in kind of then what is next for FloodMapp. Are you talking about the kind of future flood modeling, but I am kind of interested in where you think you are going to go as a company? You have been growing, and you are doing tremendous. It is so wonderful to see your progress over these years. So yeah, just take a moment to reflect on where you have been and what is next for FloodMapp.
JULIETTE: Yeah, such a man. I get so inspired thinking out this sort of stuff. It has been a journey actually, I think back to this super key part of our journey was, I think it was way back in, it must have been the end of 2017 maybe. And so, I connected with Rob on Twitter and I am like, hey, we do not know you. We are these random people from Australia, we are working on this flood thing. And we want to understand if the insurance industry would be interested in this. Like, do you have some time to chat? And Rob was so kind, he was like, yeah, I am doing a road trip. I will get you on the phone in the car and I will tell you what I know about the insurance industry, which it turns out with a great deal. And we learned so much from him. But we just think back to when it was just the two of us, just trying to validate if there was even a market for a solution this to now.
Just recently last year we closed the seed funding round and we are now kind of a company of 10 people, and trading with, selling our solution in the US and Australia. And I think what is next is really growing that steadily and sustainably, but really, staying kind of niche with the problem of flooding. People ask us a lot, are you going to do bush fire, earthquake? What is next? And we say, no, we want to pawn with those people, but the flood problem is so technically specific, and I could no doubt about it all day. It requires a lot of resources. So, we are going to say specific, but go broad. So, we want to take the solution into Europe, to Southeast Asia, to a lot of countries that, we get a lot of queries and especially around insurance interest out of Singapore for countries, like there is a big interest for this in Japan, India, Bangladesh.
There are so many places that suffer from flooding that need our solution. So slowly growing the company and ultimately having kind of some global offices like North American headquarters, maybe a Singapore kind of headquarters. Maybe a Zurich headquarters or something that, so that we can ultimately serve global markets. And then kind of more, essentially more products and more features. So right now, our big focus is on this sort of like, event-based modeling, really specific on loss prevention. Specific on claims. But ultimately, we think there is this whole other opportunity to go into that flood hazard modeling space, which may have way bigger applications on getting more accurate kind of underwriting pricing, but also for local governments for that development planning for reducing future decisions that are putting people at risk.
And I think, in Australia, that is a big conversation that goes on all the time. And I think the same in the US between the insurance industry and FEMA or the government agencies that are doing this mapping where sort of insurers are just having to take so much of the financial hit, even though it is not their decision, where the buildings are built. They are not land planners. They cannot decide where someone builds a house. They just have to price the risk. But in some regions, because flood is getting more frequent and more severe, flood insurance or just insurance, in general, is becoming more and more expensive. So, people feel they cannot have the coverage. And so, there is this big discussion about collaborating more on saying it is not one person’s responsibility. Governments have to work with insurers to lower the risk to make more affordable premiums for people. And I think we fit into that solution to help both those industries come together and give them the data to collaborate and ultimately find a better kind of insurance and a better, more resilient outcome for everyone.
ROB: I love that vision Juliette and one of the things that give me a lot of hope is, in Florida. I was caring in cases where developers wanted to build new houses, right on the Gulf of Mexico, but people would not be going to buy them because the insurance is going to be 20, 30,000 a year. So, they have moved them inland, a half–mile. They are leaving the wetlands there to beat that natural barrier from the hurricane and flooding. And so, to your point, insurers can actually can, drive proper valuating of your green infrastructure in kind of keeping some of those areas that have always been that natural protection for centuries. And people know it is valuable, but it has always been really difficult to put a price on that. Well, now you can, and what you’re talking kind of talking about, maybe you can say and exactly to James’s point right now, you can go to those local planning authorities and say, you don’t want to zone this for residential, right? You want to keep this here, but you can put it here, etcetera. So, I just love the vision and where you guys are driving towards.
JULIETTE: Thank you so much, Rob.
JAMES: Yeah. Insurance drives behavior all the time. Does not it? I am a pilot. I can tell you this, the federal aviation administrative may set the rules, but the insurance company tells me what I am going to do. And so, it is very interesting. I am always concerned about the FAA, but I am concerned about what my insurer thinks. And so, I spend a lot more time focused because they react instantly. It takes years for municipalities and federal agencies to react to changes in reality. It takes ideally days or weeks for insurers to, they cut off riding in the market. They re-rate the market, they re-rate the individual, they re-rate the risk, they do it quickly because if they do not, they will get in really bad trouble. And so that is why it is always interesting Juliette, and I love looking at when I am looking at historical documents and Eric Larson‘s one of my favorite history writers. Have you ever read them? Devil in the White City was his first, really big book that hit “-“
JULIETTE: I have not read it yet.
JAMES: Oh, read it, read it, read it. He wrote a great book called, that you would love, called Isaac’s Storm. And it is about the great hurricane flood that wiped out Galveston, Texas in 1901.
JULIETTE: Oh my gosh. I need to read this. I am so curious. Wow!
JAMES: So, for all of you out there, if you have never read a book by Eric Larson, he is the best history writer I have ever read. And I am probably number one. And then number two is Dan Carlin for me. I love Dan Carlin and his hardcore history podcast. But if you look at insurance records, they always tell you what was going on. I do not know if I mentioned this before, Rob on another show. The rebels that threw the tea in the Harbor in Boston? Remember that?
ROB: Yeah right.
JAMES: They were smugglers and they were not hardcore Patriots. They were ticked that their smuggling gig was up because the king was going to waive his stamp tax. And so, they went and dumped tea in the Harbor. That was the Boston tea party. It was a bunch of smugglers. And the only reason they know they were smugglers is Malcolm Gladwell went along with some others and pulled the insurance records. Cause Lloyd’s was ensuring the shipments that were coming across. So, they compared the customer’s bills to the insurance register. And they were completely different, which means they were lying about what they are bringing over and that is what they say. You will lie to the government all day long, but you will never lie to your insurer because if you have a loss, you want it covered. So, insurance records are some of the best ways to make policy decisions and you could not be in a better sweet spot right now, Juliette. So, I hope everyone will go check out your company and consider, rethinking the way they look at flooding and flood data. It is FloodMapp.com. And this has been Juliette Murphy, our delightful, Queensland resident from Australia. Juliette, thank you for being on the show today.
JULIETTE: Hey, thank you so much, James. Thanks, Rob. I appreciated that you had me on and great to speak with you guys this morning.
JAMES: Rob, always good to see you my friend, thank you for joining and for continuing to introduce us to a lot of your great friends around the world.
ROB: Absolutely James and I know we are running out of time, but a quick news note this week, Lemonade came out with their S-1 Filing, big IPO. I have seen a lot of people saying they are going to short the stock when it comes out, but, any quick thoughts guys on that?
JAMES: You know when classic SoftBank company when you bleed cash, the public markets are not responding very well. And so, it will be interesting to see how the public markets respond to a cash–burning machine that has an insurance carrier. Cause this will be a new one. They are used to seeing these claims of being lower than, gross loss ratios and then net loss. They are used to seeing a much different number than what they are seeing in that S-1 and so, I think, personally, Rob does not think the market is ready for rapid growth, high loss companies. But I do not think that Lemonade or other companies that want to go public, there is one in the construction space, Procore, they do not have a lot of time left because they have raised so much money. And SoftBank is not handing out any more money. They have pulled back on their commitments. So, I think they have to go public. They are going to sell shares. The question is at what price. And certainly, I am not buying any IPO stock of a company that is not profitable when they go public. Just saying for myself, that is my own opinion. Juliette, I welcome your opinion on that.
JULIETTE: Oh, that is so interesting. And I think it is just going to be interesting to just watch this one. Especially after the We Work IPO, I think Lemonade, what they are doing 67 million in revenue, raised 408 million to get there or something. And that 2 million, 2 billion valuation. I sort of think the market right now, it is people are looking at things a bit more conservatively right now. And it is a bit of going back to basics as you say, moving at you know, are they profitable? What is their revenue, what is their expenses? How much do I think these shares are worth? And that is what I am going to pay. It is less about that speculation and more kind of going back to basics. I, probably myself, I am not rushing to go and buy Lemonade shares as soon as it hits the market. It is just my personal, I am just going to kind of sit back and watch. I think they are really brave and good on them for going to IPO right now because it is yet interesting time.
JAMES: Yeah. And they have got a good, just keep in mind, I love the user experience of Lemonade. I have friends and employees that use them as their insurance carrier. I have been through the process of applying for, I love what they have done. I love what they have done. I am just saying, I am not sure the market is ready right now. And that is the challenge. They did publish an article on May 6th of 2019 saying nearly there, this was on the Lemonade website. Why is Lemonade steadily improving loss ratio is important, you can go read that. That is their literature they published. And then, of course, you can read, INS Nerds, insurancenerds.com. Title AI from Nick Lamparelli who you and I both know, said do not believe the hype, the Lemonade story.
And so, there is a lot of writing out there on both sides of this, the beautiful thing about the market is that it is brutally honest. It is brutally honest. So the day they go public, the market will tell you, and the day after you are going to find out what the market thinks about high growth, high cash burn companies, and in particular, if SoftBank is going to have any more runaway hits. That is the really big question because their model has been called into serious question. So, Rob, thank you. I appreciate you bringing it up because it is impossible not to mention that one and that one almost slipped by me. So, thank you for mentioning that. And, look, I am rooting for the best for them. Remember I want InsureTech companies to be successful. I want their IPO to go well. I am just worried about the pilot markets and how they are reacting to these. So, we will see, Rob.
ROB: Absolutely. Yeah. I might dip my toe in the water a week after it goes public. We will see if it gets big enough.
JULIETTE: I like that.
JAMES: Yeah, I dipped my toes in the water on Lyft the day it went public. So. That was not a fun swim. And so luckily, I kept my investment small, and then when Uber went, I was like, ah, I am going to wait. And that went badly pretty quickly and then I went okay, it is just not a time to buy IPO stock right now. I just need to; I need to stay out of this.
ROB: So, I have to make my bad dad joke at this appropriate time and say, are you underwater on those investments, dude?
JULIETTE: Love it.
JAMES: So, my 13-year-old, I just got, I just made two sweet, awesome dad jokes with my 13-year-old, like an hour ago, and she told me that I was super on my dad joke game.
JULIETTE: Nice! I love that. What was your dad‘s joke?
JAMES: Oh, I do not remember what it was. It was something, it was something horrifically punny, it always involves a pun and she thought it was a, she laughed and then patted me on the shoulder and said it was cute. She tried to make me feel better about myself. She is a sweetheart. So, all right, well, onwards and upwards!
This has been the InsureTech Geek Podcast, powered by JBKnowledge at JBKnowledge is all about technology transforming and disrupting the insurance world. I have been your host, James Benham, jamesbenham.com and my cohost, Rob Galbraith, endofinsurance.com Thank you to Jim Greenly, our podcast producer, Kara Dalton-Arro, our creative producer. And thank you for joining us today.