Good data helps insurers to price and manage risks more effectively, but only if the right information is available, affordable and easy to integrate into underwriting processes.
HazardHub are providing solutions to those challenges by supplying insurers with data for perils like flood and wildfire and unique property risk information, including the location of every fire hydrant in the US.
The company was selected to join the Lloyd’s Lab last year and was named as one of the companies to watch in our new Location Intelligence report.
HazardHub Co-founders Bob Frady and John Siegman join Matthew on Podcast 135 to discuss the types of data they’re offering and how they’re making it affordable and easy for clients to use.
Talking points include:
- Accessing real-time data via APIs
- Lessons learned from the Lloyd’s Lab
- Harnessing technology to cut costs
- Pinpointing wildfire risks through modelling
- How to make clients confident about data
For more information on HazardHub's AgentRiskView, go to agentriskview.com
Sign up to the InsTech London newsletter for a fresh view on the world every Wednesday morning.
If you like what you're hearing, please leave us a review on whichever platform you use, or contact Matthew Grant on LinkedIn.
Continuing Professional Development - Learning Objectives
InsTech London is accredited by The Chartered Insurance Institute (CII). By listening to an InsTech London podcast, or reading the accompanying transcript, you can claim up to 0.5 CPD hours towards the CII member CPD scheme.
- Claim 0.5 hours for listening to Episode 135 of the InsTech London Podcast
The data you need at a price you can afford - Episode 135 highlights
Matthew: Modelling and data companies have been around for decades now. Why are customers choosing to work with HazardHub?
John: What we’re doing that is exceptionally different is we’re making it affordable and we’re listening to what customers have to say. Then we try and solve the problems that they have.
The traditional way of building a project involves looking at the size of the market, looking at internal costs and what the ROI will be. We’ve thrown that out of the window because we want the value of the API to keep increasing while the price remains constant.
Matthew: How many different types of data are you offering to the insurance market?
Bob: There are 900 variables in our API and we’ve got more than 35 distinct categories. Broadly though, we consider our data as addressing two areas: risk information and property characteristics. We will also have replacement cost data and will be adding building permit data soon.
Matthew: What are the options for insurers who are looking to get access to your data?
Bob: There are several different options. We want our clients to be able to get data in whichever way is best for them. The API is the most common as it's real-time and the insurer can connect right away. If an insurer has a database of records or even a spreadsheet, they can send that to us and we can batch process it for them. That's how a lot of our customers get started until they build up their API capabilities.
We’re looking at being able to take in a schedule of risks and append that automatically via an email submission from an Excel sheet. We’re trying to make the data available however it best serves the customer, but we stop at fax and microfiche. We won’t do that anymore.
John: We also have a report function called AgentRiskView and anybody with a credit card can use that to get a report. We've got several insurance carriers using AgentRiskView.
Bob: Carriers usually go for an interim step using a pre-printed report because it’s what they’re used to and it’s sufficient for them. We’ve built the platform to be flexible so if the customer is an advanced marketer or underwriter, we can support them. If they’re not, we’re still able to help them.
Matthew: The affordability of your data is one of your main selling points. What have you discovered to enable you to gather data at a price that is affordable for companies that couldn't afford it before?
John: We looked at all the possible ways to make machines do the job. We don’t have a lot of people and we don’t spend money on things that aren’t productive. We outsourced and we gutted every possible cost. Once the costs were down as low as they could be, we worked to get our volumes up as high as they could be so we could maintain a low price point.
Matthew: You were in the Lloyd’s Lab last year, virtually of course. How was that experience and what were the biggest learnings or highlights for you?
Bob: It was a fantastic experience, even though it was virtual. A big surprise we saw in Lloyd’s was how much risk gets bound with people having no idea of what the perils associated with properties are.
Portfolios of business come through, somebody prices it, they bind it and then they look at what the perils are. We talked to syndicates that are now in runoff and they got wiped out because they weren't aware of the perils. It’s shocking that this happens in a market as sophisticated as Lloyd’s. All of those risks should be known before binding a policy and not after. What we're trying to do is move the data higher up into the funnel so that people can make smarter decisions before they bind.
We got some good business out of the Lab. We landed a couple of new customers thanks to our involvement and we're excited to expand our footprint in the Lloyd’s market. The Lloyd’s market is the Excess & Surplus insurance market and that's where our data shines.
Matthew: The one thing that struck me looking at what your clients in Lloyd’s are doing is that they have a more dynamic way of interacting with their agents and MGAs.
Bob: Yes, people can be much more dynamic using our data, but it's a mentality change for people. They have to want to, and that's just something that comes over time.
Matthew: You have some data that didn’t exist before. One of my favourite stories to recount is what you’re doing with fire hydrants in the US. Can you talk about that?
Bob: Understanding the location of fire hydrants helps identify the risk to properties from fire. California is great because the counties release the data and we’re able to put it into a geospatial format, but most places in the country aren't that good at it. We petition the municipality directly and some of them say the data can’t be released due to security concerns, so we have to drive the streets. John and I have probably done at least 250,000 locations ourselves. We've gone in and tracked where the hydrants are located because it had to be done.
Now we have a team that does that work for us, but we built the data sets ourselves and we're not above doing the work. We've personally encoded 55,000 fire stations for positional accuracy, and we examine how many bay doors are on there so we can determine what the power of that fire station is.
Matthew: How do insurance companies use that information about fire hydrants in a business application?
Bob: There are two ways. The first is that some insurance applications require information from the customer on where the nearest fire hydrant is. The insurer asks the customer, and they have to wander around to try and find the nearest hydrant. Now they don’t have to ask that question anymore because we know where the nearest hydrant is.
The second is that our hydrant network is an input to our fire classification model, which we call Property Fire Score (PFS). We take the network of roads, the network of water, the network and types of stations that are nearby to assess the fire risk of a property.
What we see is that the higher PFS in our model are the ones where the increased fire losses are occurring. They're far away from a fire station, they don't have access to pressurised water and they don't have streams nearby that can infill for the pressurised water. We’ve just released a case study that shows that these examples are where people see the most increased amount of risk from fire.
Matthew: Losses from wildfires and floods are becoming more prevalent. They’re difficult to model comprehensively, so how do you get enough information to make your data attractive for clients?
Bob: Those risks aren’t hard to pinpoint if people have the right tools. It's relatively straightforward with HazardHub to pinpoint where these risks are because we've done the work to put that information together.
John: We've been building wildfire models since the start of hazard modelling and we know where they're going to happen. Wildfire is not just a California or a Western States event, wildfires happen everywhere and the worst wildfire that's ever happened in the US took place in Wisconsin. Our model's a 50-state model and we look at the factors that contribute. Everybody looks at slope and aspects but in a catastrophic wildfire event, there is a 65mph wind blowing the fire in whatever direction that wind happens to go.
We test the model every day. If we get information that a wildfire broke out, we run our model on that location. Did we get it right? Fortunately for us, the answer is almost always yes. Occasionally we miss it, so we go back and figure out how to make the model even better. We are on our fourth wildfire model because we’re constantly improving what we do. Our customers understand that if they want to keep our data forever, they can, so they can build a time series history. All we are concerned about is what's going to happen today.
Matthew: How do you give customers confidence that your data is correct?
Bob: There are two ways. John and I have a saying: "We don't want to sell you anything, but we want you to buy a lot." The way we do that is to take a client's data, append our data onto it and let them compare their loss history to what our data tells them.
That's usually the most direct method of selling that we have. We want them to be confident in what they’re getting, rather than just relying on somebody else's confidence.
The second way is we do have actuarial studies. Milliman did a study for us and we've got a couple of other studies that show what the data says. Different customers have different needs. There are innovators and there are the followers. The innovators will take the data and test it themselves; the followers will rely on the work of other people. I would much rather companies test the data themselves to prove what the value is, rather than relying on the words of others, but we offer both.
Matthew: Once life returns to normal, hopefully later on this year, are you going to be out and about seeing people face to face? Will we see you back in the UK?
Bob: I can't wait to get out and start to talk to people again for two reasons. The first is it's a great opportunity to sell, of course, but it's really important for learning more about what the issues are. The more we understand what the issues are, the more we can help derive a product that solves a need. I absolutely can't wait to get to London. It'll be sometime in the fall, I hope.
Matthew: Are there any companies that you’re working with that you can tell us about?
John: Well, in the UK, Atrium got to know us when we were in the Lloyd’s Lab and they announced that they were working with us. We've always protected our customers because they know what they’re doing with our data is their secret sauce. Small insurers and big insurers get the same data, but what they do to make it their own is what we want to keep private.
Our partners, on the other hand, we'll talk about them all day long because they want us to. The list includes Intellect SEEC, Betterview, JMI Reports, Signature, and a tonne of other inspection companies.
Bob: We keep it tight with our customers because they're innovating. A lot of the time, they're smaller and more aggressive and we want them to have that secret sauce. We're starting a ‘voice of the customer’ video series next month, which will expose those customers a little bit more.
We're launching a permit data set. We’re going to launch it at the end of April, starting primarily in Florida and then nationwide by the end of the year. We’re also launching our portfolio risk estimator to move more into a portfolio level analysis of risk, rather than case by case. It's something that we're excited about and we have a new nonparallel water model. It's an exciting time to be part of HazardHub, that's for sure.
Matthew: We’re delighted to have HazardHub as a corporate member of InsTech London. It would be great to hear what motivated you to come and join us?
Bob: InsTech London is the voice of London insurtech and we wanted to be part of that voice. Companies have to choose who they align themselves with carefully. We've chosen to align ourselves with the best and that's Instech London.