Technology to the Rescue of Insurers Featuring Nick Walker (Private)

Many insurers have embraced connected technology to better understand risk in the vehicles they insure. Others have thought about it but not yet acted. But very few are taking advantage of the opportunity to use the data they get to build better customer experiences and relationships to build loyalty. Join insurance industry expert Nick Walker and VANTIQ’s Blaine Mathieu for this webinar which will take a look at the solutions available and relevant to insurers and show how they can be deployed to build customer service and loyalty and grow profits.

Transcription

Blaine: Hello everyone and welcome to the VANTIQ webinar “Technology to the Rescue of Insurers”. I’m Blaine Mathieu, the Chief Marketing Officer at VANTIQ. We have a special guest today, Nick Walker, who owns management team GMBH, and will be providing much of the content for this insightful webinar. You will also see a great demo by Brett Rudenstein, our VP of Sales Engineering.

The agenda today will be Nick discussing the needs of insurers and the overall emerging technology landscape that exists within. Brett will be providing a live demo of a real-time insurance application, and then we will be answering your questions. While the webinar is going on, please type your questions into the Q&A box at any time. And now, I will turn it over to Nick Walker.

Nick: Thank you, Blaine. Thanks very much and welcome everybody. My name’s Nick Walker. I’m [in] a consultant management team, reasonably small, but efficient management consulting business based out of Germany. We’re around 20 consultants, and we are heavily involved in the deployment and project management off IoT programs. A number of industries, one of the main customer bases, is insurance.

Myself, I have spent the last many years, too many years to even tell you, in IoT. [I] worked in connected car industries. [I] ran the connected car program at the RAC in the UK and also ran European-wide business for a telematics business called Master North for several years in Germany.

Without further ado, let’s move to the beginning of the webinar here. I’m going to start with a question: What do these items have in common? Let’s bring up a few pictures here: an umbrella, a fire extinguisher, and a first aid kit. One of the answers is they’re all red. But actually, the answer I am looking for is we all buy these things in the hope that we’re really not going to need them at al. It’s just a characteristic of people. People like to avoid risk, and buy things that they’re not going to need.

One of the things that falls into that category is an insurance policy itself. Insurance policies are all about risk or mitigating risk. So you pay an insurance policy in the hope that you’re never going to make a claim. We all know that at times we do have to make claims, but that’s just the nature of the business.

One thing I would really like to look at it is, what are the characteristics and the challenges of the insurance companies that face a challenge like that? Basically, selling something that nobody really wants to use at all. Moving along to the next slide here, this is about the characteristics and challenges of insurance.

The first thing that I’d like to refer to in this is the fact that the insurance business has become commoditized. Commodity means that it is sold predominantly on price. In fact, we are all very used to going on to compare sites now when we want to buy insurance. We just look for the best deal. We don’t look too much about what the business is doing behind it. We choose fundamentally on price unless someone is just repeat buying. Commoditization is ride in the industry.

A few other characteristics that are in this business: it’s very low touch. Done are the days when the insurance guy comes and knocks on your door to collect your money. You actually don’t even see the face of an insurer anymore. The business is done very much over the internet or through walking in to an agency filling out forms.

The touch points are signing up for a policy and maybe making a claim, and that’s pretty much it. There is no further opportunity for touch points which really leads to a pretty poor relationship between the insurer and the customer. It’s forcing the business to be very transactional which means that basically, you make your choice, you sign up for it, you pay your money, and that’s it. It’s an extremely transactional, very low touch, no relationship.

Which begs the question: is this an exciting business? I would hazard to say that it’s not. in fact, talking to many insurers, they consider themselves to be in a very unexciting business because the customer interaction is so low.

Furthermore, we buy this product because we’re obliged to do so. There are laws in any countries that you must be insured for your car. You must insure your home. When you take out a mortgage, a loan for your house, the loan provider very often wants you to insure the building. Very often, you are forced to buy insurance even if you don’t want to. The obligation element really makes it a very uninteresting area for consumers to buy.

Also, digital age? Digital age is all about this is a digital business, and as I said about low touch, we are now pushed to a digital point to interact with insurers over the internet and sign up for policies over the internet. Which means that the distance between the insurer and the end customer has actually gotten further away than it should in terms of moving forward and being in close relationships. So, the digital age, at the point that it’s being used by insurers today, is actually causing more problems in terms of challenges than it is solving.

The low touch element in terms of insurers also links to fraud because there is no relationship, because there is no personal interaction, because of the very transactional business, the fraud element is growing. Fraud in the UK on claims alone in the insurance business is now expected to be greater than 2 billion pounds per year. That’s a massive number.

And a lot of the cause of this is that the insurer knows very little about the business of the end customer and the characteristics of the end customer. So, there’s fraud at the point of signing up. The problem is fraud at the point of claim because of lack of knowledge. That’s becoming a much larger problem year on year.

It’s a reactive business because by nature it has to be. The end insured customer has an accident or has something happen to him, and he needs to make a claim. He will contact the insurer. This is not a practice business at all. It’s very reactive.

One of the things that’s happening out there is insurer trends, consumer activities, changing. If we take the example of vehicle ownership just as one example, car sharing is a big trend right now. people are not wanting to own cars. They’re too complicated. They’re too technical. It’s too difficult to part them.

In fact, a good reason for using that is actually when you own a car, you find that if you look how often you use it, it’s much less than 10 percent of the time that you own it, you are actually sitting in the car and driving it. In fact, it’s significantly less than 10 percent.

You think about journeys to work and back: the car is going to spend 8, 9 hours sitting in a car park outside of your office. That’s a lot of time that a car is not being used. The trend is not to own cars anymore. It is to share cars or to rent cars when you need them. This is changing people’s view of how they then insure them, and it’s changing the way that insurers need to interact with customers to insure those assets.

Moving along here, I tried to put these challenges into buckets in terms of how they are being dealt with or how they can be dealt with. The first area I want to look at it how can you defend yourself against these challenges as new clients come into the market that try and address all those challenges I covered on the previous slide?

The defensive move is let’s try and understand the customers, let’s work to reduce risk, let’s avoid being disrupted, and let’s try our damnest to retain customers. In a commoditized world, the only real way you can do that is through being leaner and meaner and just doing the best you can to be profitable. That’s what certain insurance companies are doing in the face of the insuretek world coming along. Others are embracing it.

Moving up the chain here, we see there is a middle ground which is called maintenance; maintaining a position in this market as an insurer. Rather than reducing risk, let’s manage it. That’s getting to a much more proactive state where we can learn and manage the risk, improve loss ratios through the management of risk which means we pay out less than we take in. In doing so, we become proactive, differentiate, and we increase customer touch points. The standard way of doing that is through connected techniques where you get to know your customer better.

There is a level further that you can go. And that is where I say it’s the leading position which is where you’re in a business where, not only are you looking at the risk level on a one a year or twice a year basis, you’re actually looking at risk in real time. You’re looking at a dynamic risk assessment which means things are not the same, even during the hours of the day, let alone during the days of the week or the months of the year. We have seasons. if he’s thinking about vehicles again, people drive differently in winter to when they’re driving summer. People drive differently in the morning to how they’re driving around midday. Traffic conditions are different, and things change in a very dynamic way.

It seems quite illogical to apply a single risk to a very dynamic environment. Leading-edge insurers are coming into the area now of dynamic risk assessment by knowing much more about what’s going on, even predictive in a way. We’ve been able to look at what’s likely to happen rather than what is happening or what’s happened in the past. And this is a basis on which you can communicate meaningfully with customers because you know much more about what’s going on.

So in this leading sense, you actually get to a point where you can communicate and build a relationship with your customer which goes totally against the trend of being a commodity. It changes the shape of insurance dramatically.

We talked a little bit about IoT. I just want to just touch on what IoT is all about. This is one of the solutions that comes along to move an insurer from a defensive to a maintenance mode at very least. IoT, for those who may not be familiar with it, it’s about having sensors embedded in physical objects, sensors sensing the way things are, doing things, sensing where things are, telematics in vehicle security systems, in homes, fit bits that you wear on your wrist are all forms of sensors for things that are going on in anything that is insured.

These sensors are linked by or through wired or wireless networks to a central platform, and they communicate either with the environments, they communicate with the platform. So the platform knows much more about what that device or that object where the sensor on it is doing. It produced a fundamental change in the way we act with our surroundings.

We think about it today, there’s many things that are connected, many things that have become digital, that have changed things. One consequence is that huge volumes of data are being built and this all gives an opportunity for many disruptive business models. Just moving into a couple of facts here: since as long ago as 2008 which is now 10 years ago, there’ve been more connected devices on this planet than people, which is a quite a startling fact. In 2016, there were 7.2 billion people on the planet and 20 billion connected things, and the prediction is by 2020, which is only two years away, there’ll be an additional million devices added every hour into this connected world.

So this isn’t just something new. This is something that’s become very commonplace and very acceptable in many many many ways. So what does IoT do in insurance? And this looks like a very busy slide, so I’m gonna try and condense this. There’s two streams of opportunity there: one is the ability to reduce cost and the other one is the ability to add revenue.

So if we look at that from a cost point of view, you’ve got the opportunity to optimize resources and leverage data insights. If we take an example of that, just think about a crash in a vehicle, sense the ability to optimize resource. If you get an automatic crash notification rather than waiting for a piece of paper to come in with a claim, you can optimize the resources around a crash process or using a crash notification to settle the claim. So there’s a basis there for automation which couldn’t happen unless you had a sensor which knew that there’d been a crash.

And moving all that chain, there is, once you know what that asset is doing, you’re able then to look at how you can prevent accidents or prevent things going wrong which would remove the need for a claim in the first place. Of course also, when you know more about your asset that you’re insuring, you’ve got the opportunity to reduce fraud. Just think about that 2-billion-pound number that is the fraud number in the UK alone, I dread to think what the what the global number of fraud is.

And moving to the top stream, you’ve got lots of opportunity to add new revenues. The very fact that you’re able to interact better with customers, you can build more frequent interactions with those customers and you can then start to manage new services. Connected mobility gives the opportunity for emergency services and others, which would allow interaction and new business models. You’ve got a whole area there where you can monetize the usage of the data, but also based upon new data, you can often use services. And as we go through this webinar, I’m going to come to a few examples on a customer journey as to how that could come into play.

And way out there, you’ve got the rethinking pricing models. Pay by use: pay how you use things. There are examples in the industrial world where jet engine manufacturers have started now not the charge for the asset, to charge the capital investment of the jet engine, it’s charged by maintenance and usage. So is a monthly model based upon how much the engine is used rather than the cost of the asset in the first place, which is a much more acceptable business model for for the plane industry. So the impact of IoT is very far-reaching in business models in many businesses.

So let me here try and explain what I mean about a proactive or a predictive model. This is an automotive example where an insured vehicle is involved in an incident. And the reactive model, I’m taking this through four phases from reactive to predictive, the reactive model is today’s model. Somebody has a crash, they will contact their insurer and say, “I need to make a claim.” The insurance answer to that is, “Okay tell me what happened.” because there’s absolutely no connection between the two organizations. So, there’s now a dependency upon the claimant or the third party to report the incident. That’s how the reactive world works.

Intelligent reactive is where IoT starts to become involved where the claimant now is calling. You’re saying, “I need to make a claim”. The starting point is exactly the same, but now the insurer is a little bit more intelligent. The insurer is now saying, “Yes we’re just got a notification your vehicle was an accident. Are you okay?” So now you can answer much more intelligently, you don’t need to be told what’s happened because you know something’s happened, and you’re much more intelligent in the ability to deal with the claim. That one is called intelligent reactive.

You then move to proactive. So now we’ve got a case where there hasn’t been a crash. What we’ve got here well that has been a crash, but the indicators come through the insurer where the insurer is now making a proactive call saying, “We’ve seen some indications your vehicle may have been involved in an accident. Can we help?” So now you’re being proactive. You do not have to wait for a claim at all, you’ve seen a notification come in, and you can quickly be proactive. And the response normally is, Yes I was about to call you.” Because signal coming through to say that there has been a crash is actually quite reliable.

So the answer there is quite correct, but then there’s a predictive model where the insurer is actually calling the driver saying, We’re noticing your driving pattern is increasing, and your risk and potentially, your insurance cost. [It would] be good if we can discuss we could work with you on that.” So this is absolutely a predictive thing, so there hasn’t been a crash, there hasn’t been a claim at this point. This is an intervention looking at the point where someone might be driving in a way where they could potentially have a crash. And the answer is, “Hey that’s great so thanks for the call. If you can prevent me having an accident, that’s really good.”

Now, IoT covers many many industries and, you can imagine that same model in preventive maintenance machines, preventive maintenance on vehicles themselves. This one is just looking at how somebody drives. Most current telematics systems or connected car systems stop at intelligent reactive or it’s moving a little bit into proactive. But the place to get to is being predictive where you can actually prevent accidents happening in the first place.

I want to give an example of this because this all sounds very grand and very futuristic. There’s an example of a real-life case that we were involved in from my company. We were asked to look at a fleet in the UK of 3,500 vans that were on the on the verge of becoming uninsurable because they had so many accidents. They were actually paying a premium of 5.2 million pounds in 2015, which is a pretty hefty sum of money. We fitted out this fleet of vehicles with telematics in 2016 to be able to monitor and feedback driver behavior to the drivers. It also was capable of detecting crashes ,and the data was used to talk to drivers, feedback information to drivers, and coach them on driver style. Fascinatingly enough, within the first 12 months, the crash level was reduced by 75%, not to 75%, by 75%.

The knock-on effect of that was of course there was an insurance premium reduction and the driver risked reduced on average by 43 percent. So, we made that fleet of drivers better drivers, less likely to have accidents, much lower claims level. One step further was careful drivers mean less maintenance on vehicles. So we ended up with a very happy fleet customer because their downtime, their servicing on vehicles, and maintenance was also reduced by a further five million. So their benefit in this tense was absolutely massive. This isn’t fantasy. This is absolutely real, and that is a great example of how this whole thing can work.

Let’s look at the infrastructure that makes it happen. So what you see in this diagram, this is really an insurance system overview. It’s a very simple diagram. On the left-hand side, you have sensors. So in this case, there’s a few Fit Bits up there, there’s some security equipment at the bottom, there’s some telematics devices in there, some cameras. [It} could be anything, anything that is a sensor which is taking live data from somewhere. It could be the asset data. It could be indicators could talk about usage.

One thing about it is it’s very asynchronous, things all happening in real time as they as they happen with those devices. It’s not all synchronized. So that comes into a platform which is a large cloud, in this case. It’s just a platform with some massive amount of computer power and storage in it. That needs to be then compared with a lot of historic data which is typically the data that actuaries would use to assess risk, typically done in an insurance policy once per year.

Now, you’ve got data coming in every few seconds, every few minutes, so in real time and really just building massive amounts of data. So that then needs to be processed, and the output is actionable information, so things you can do things with, information you can act upon which will allow you to be predictive or proactive. So, going back to the diagram I had before with the actions on there were the four stages of being predictive, this can be fed into CRM. So somebody in a call center could make an outgoing call. You could manage risk. You can call someone or you can do something about it  to manage that risk or even prevent risk in the first place.

So actually information can be delivered by a human being, through an app, whatever. Now this level looks very easy. It’s not because now, and this is where most insurers want to be because this is where you can really control your business much better, you’ve got a whole new set of challenges and these are written nice and big on the next slide here. Now we’ve just put down that the future of insurance depends on data, lots of it. It’s a real time business, which it never was before.

But the problem is that people don’t buy that. They buy services. So to get the data, you’ve got to be able to offer something to your customers which is not just, “I’d like to get more data from you.” Basically, you’ve got to try and balance that. I’ve kind of referred to that as the balancing act. So on the left hand side, you’ve got the insurer who is looking for risk data, lots of it. Actuaries love that they can have that diagram, requires a lot of data to work. Consumers though, look for services. They look for some benefit. They will only give up their right for someone else to look at their information if they get something back.

Typically, in the wonderful world of vehicle telematics insurance, that service or that give-back  has been a discount. The proposition to typically a young driver is you put a telematics device in your vehicle so I can monitor your driving and you will get a cheaper policy because I believe you’re going to drive better. Well that’s great, but what happened there is that the insurer just gave up a chunk of its margin and he’s back to being commoditized again, which is typically what’s happened in the telematics insurance business. The margins may be more controllable, they may be slightly better, but they’re not as good as they could be if the services could be better where the discount element of this whole thing goes away. But it’s necessary to balance the whole thing with services, and I will talk about that in a second.

So what’s going on here is we’re looking at real-time events happening on the left-hand side turned into services on the right-hand end. Maybe, Blaine, I can ask you to talk through the next diagram because this is a very VANTIQ diagram.

Blaine:Yeah that’s right. Thanks a lot, Nick. Sort of more generalizing from the insurance use cases we’ve been talking about – and by the way, some very compelling examples you’ve provided there Nick quite, quite amazing – this notion that you discussed a minute ago about a so-called event-driven application is really what is required in order to make the use cases and scenarios that you’ve been talking about come true.

And at its core, the notion is fairly simple: you have all the data from sensors, from other business systems, IoT devices, car telematics devices, as you said, flowing in. And these are reflecting things that are going on so-called events that are happening. And you see some examples of some of the events on the left and in the diagram here that could be related to an insurance scenario. Now, those events are continually being processed by the real-time event-driven application doing something called complex event processing.

You mentioned, again, how these events, these data flows, are coming in asynchronously. They’re not in a particular order, they’re coming in at any point in time, and the system has to be ready to handle and process them at any point in time. Of course, that handling is done within a particular context. And as you’ll see and I think in the example in a minute, there are speed limits that cars have to operate within. There are other parameters that result in safe driving. So the event processing engine or the application you fundamentally create from it needs to also understand the context.

Most importantly, as you pointed out, the real benefit here is not just about real-time analytics or showing a nice dashboard, but it’s about taking action. It’s about doing something. And in the scenarios we’re talking about here, it may be the insuree who’s taking the action, who’s doing something as a result of these events that are being processed, or it may be the insurer that’s taking action, either in real time or maybe near real time. So at a general level this is the kind of platform that would be necessary to serve the use cases that you’ve been talking about, Nick.

Nick:Exactly. Thank you for that. And and I guess the whole point here is that the amount of data you get from any IoT sensor, the majority of it to an insurer, is absolutely useless. It’s just there, and what you need to do is look for the data that you want to have and the combination of events that happen together that the increase risk or have an effect on risk. And that’s what really this real-time event-driven platform is really all about. It helps sort the data in some kind of filter based on rules that to allow action to be taken.

So, just to put that a little bit more in an insurance sense, this is data adapters’ information coming in, data acquisition, sensor data, time of day, any external factors that typically don’t come from the sense but come from other sensors all brought together to some form of situational analysis sitting in the middle. And assessed in a sure-defined way, looking for true risk assessment, key risk indicators, looking for all of those things, be it dynamic, be it situational, doesn’t really matter. Those rules are defined by the insurer. But the beauty of this is you can look at huge amounts of data and simplify it and turn it into action and collaboration in a in a very simple way.

So, action in this sense being maybe a policy adjustment automatically, push it a message, making a phone call, it’s all relevant. And what you see here is the basis on which you can build those services I talked about because they’re personalized. Every time you get an output out of a platform like this, it is relevant to a particular person or a particular situation. So any action can be entirely customized and personalized to a specific instance or a situation. It’s very different to the way IoT works in insurance today and how insurance has worked in the past. I think, Blaine, we’re gonna look at a build demo now to show how this whole thing works?

Blaine:Yes let’s do that. So, again, Nick, thank you so much for this intro and this very deep background on how insurers can benefit from using these kinds of scenarios. Brett, why don’t you take over now and show us a sample application for how this could actually work in the real world.

Brett:Sure. Thanks, Blaine. So, what you’re looking at on the screen is effectively a digital twin of one particular driver on the road. Our application is basically going to monitor the driver’s behaviors. And so here we see a driver sitting in the talent center here. The driver will get moving and we will monitor the drivers current speed limit and we will monitor the current speed of the driver as well as the speed limit of the road that he’s on. The situation of interest that we’re looking at, the opportunity or the threat in this particular case, is when the driver exceeds the speed limit by 10 miles per hour and then exceeds it for some period of time. Down the bottom, we have a notifications window. Now this notifications window is an application running in our drivers. As you can see, he’s just exceeded the speed limit, as excessive speed can affect your good driver bonus.

And so, what we’ll talk about in the next couple of minutes is how you build an event-based application like this to take these kinds of actions and to integrate with third-party systems like CRM systems. And so there we see the drivers done this. So you may hear these tones in the background as our driver speeds through these backend roads over the speed limit, in this particular case. So how do you build an application like this?

I’ll start quickly by showing you the interface for building the digital twin. VANTIQ provides a client builder interface that simply is essentially a drag-and-drop interface where you can move the various elements. You can basically attach various things to data streams, and that’s really as much as I’ll show you about building the front-end client. The client builder can be used to build web interface and can also be used to build mobile experiences as well. But what we’re really going to take a look at is how did we build the application that is sensing and analyzing those streams of data, combining the speed limit data with the with the speed data, and then taking action on that, which one of the actions, of course, is to notify the driver, the other action potentially being to open up an inquiry or an investigation.

So let me go ahead, I’m going to close this down real quick and I’ll show you. We’ll build a little bit of this from scratch. So let’s go ahead and create a new project, and I’m going to add a new application. I’m going to call the new application the driving demo. When you create a new application in the VANTIQsystem, the first thing you should start with is an activity pattern that basically gives you the ability to intercept some stream of data coming into the platform. And we’ve got a couple here: one is we’ve got device data from our mobile device. Oftentimes this comes from a telematic device. In our example today, it’s coming directly from a mobile device. I’m going to call it the mobile stream and I’ll configure it.

The way this data is coming into the platform in this particular case is a rest call into the platform. It could be coming from a source like an MQTT or a Kafka topic, or could be coming directly from a WebSocket. In this particular case, it’s a rest call on a on a topic called /mobile/data. When I click on OK and then save our application, you’ll notice that we start getting these badges. So this is actually showing an application that’s been built in just a couple of clicks that is taking in the stream of data.

Now, our speed limit data is coming from another service. It could be coming from the car itself. A lot of modern cars actually have the ability to read the speed limit signs or it could be coming from a Google service where you get the speed limit and so I’m going to actually take in that data as well. So let’s go ahead. I’m going to add another event stream. I’ll call this one the car stream. And it’s also reaching the platform via a rest call into the platform and it comes in on a topic called /car/data. So I’ll click on save and now, of course, we now have the two streams of data coming into the platform. Of course, you can hear the real application running in the background. That’s sensing our situation of interest as our driver continues to speed.

Now, the next thing we need to do is to correlate these two events. We’ve got the speed of the driver and we’ve got the speed limit coming in from some third party application, in this case, the car. How do we join those two together in a meaningful way? VANTIQ provides a number of activity patterns, one of them is a joining pattern. And this represents all the complex, asynchronous code that you’d have to write, but you do it in the very codeless fashion. So I’m gonna call this join streams, and of course, I want to link it up with the car stream. So I’ll link this to an existing task and choose join streams. So now what I need to do is correlate these two. How do I know which driver ID goes with which speed limit ID and so forth?

The way that I’ll do that is I’ll provide a constraint. And the constraint in this particular case is that the mobile stream’s device ID matches the car stream’s device ID. The other thing I’ve got to take into consideration here is the time. What if the cars the speed limit came in an hour from the from the actual drivers speed? There’s nothing to suggest that those two values would actually still be even remotely related to each other. So, I’ll not only constraint it by the device ID for the mobile and car stream, but I’ll also specify a duration that this has to be true within. And I’ll say it has to be true within one second of the two events occurring together. Anything else won’t be considered a correlative event.

Now I’ve gone ahead and put these two items together I’ll do one last one and then I’m going to show you the be full application. The next thing we want to do, of course, is look for our situation of interest which is the driver is extending the speed limit over 10 miles per hour for some period of time, or whatever the insurer’s definition is of that particular event. So I’ll link one more task. In this case, I’m going to choose a dwell activity pattern. This is not just going over a threshold but going over it for some period of time. And I’ll call it – let’s call it extended speed maybe over 10 something of that nature – and now once I go ahead and I do this, I’ll configure it with the condition.

The condition, of course, is that I want to take the drivers speed, which I am getting in meters per second, turn it into miles per hour, (that’s the calculation: to turn meters per second into miles per hour). So, speed over the speed limit plus 10 so that is the correlative condition. And I’ll say that this has to occur for more than five seconds at a time. So now you’ve seen just a couple of mouse clicks building an event-driven application that is secure, requires authentication and authorization to the platform, and scales to hundreds of thousands of millions of devices and millions of events per second, whatever is necessary, for the number of devices that are connected up to the platform.

Let’s look at the full application and see what the rest of it’s doing. Here we see this is where we stop that expend its extended speed. This weather enrich is basically getting the weather at the time of the event, of the overspeed event. This is the first place that you see code. It’s a low code environment. It’s been no code in this first section. Down here, the weather enrich, here’s the procedure. It’s essentially a one-liner. It says select from open weather, which is the open weather source. With these query parameters, the latitude and longitude and the units that I want to get things back in. One more line basically turns the GPS coordinates into a physical street address.

And so you see very easily, even when we have to create a small amount of code, we enrich the weather, we basically add more contextual enrichment. This just adds a string to the stream which basically says that it was a speeding event. And then this situation activity represents creating a record, what we call a situation – something that represents opportunity and threat, which triggers our collaboration. So, if this part of the application represents sense and analyze for a situation of interest, this blue part which is represented over here on the left represents the actions that we’re going to take. And so every time you hear that tone in the background, each time the driver actually goes over speed in this particular case, it notifies the driver, that’s the message that’s sent to the driver’s dashboard, and then closes that collaboration.

One other thing that it does is it checks the number of times, this activity pattern, this transformation, basically checks the number of times that this occurred within a period of time. And for demo purposes, I’ve set this at five minutes. If there’s more than four in five minutes, do the second collaboration which opens up an inquiry in a third-party CRM system, in this case. it’s Salesforce and sends a mobile notification. So, if you look at my mobile screen, you’ll see my mobile screen has “review driver habits” and “review investigation”.

If we come over to the Salesforce window, you’ll see that the application has indeed opened up a new record indicating a number of events as they occurred: the drivers ID, the event type (that was extended speed), the time, location, speed, weather, clouds, and broken clouds. All that information has been aggregated for the insurance investigator to take a look at and determine whether or not a good driver bonus or some kind of penalty needs to be incurred in this particular case. So, that was a very quick demonstration in about 10 minutes that essentially shows you the entire platform in terms of taking a situation of interest, identifying that threat or behavior, that threat or opportunity to the underlying organization, and then taking various actions on it.

And so hopefully that gives you a good sense of how easy it is to build applications inside the VANTIQ platform. With that, I’ll turn it back to Blaine and Nick.

Blaine:All right. Thank you so much Brett. That’s very cool. And I imagine that some of the non-technical folks on the line might think this looks a little complicated because of the speed with which we went through this demo, but I would highly recommend when you get a chance, show the video of this webinar which we’ll send you afterward to your technical counterpart or to one of your technical associates, and I promise you they will be amazed at how quickly and easily Brett built, almost before our very eyes, a very complex and very capable application to support one of the use cases we’ve been talking about and telematics related to insurance. So, thank you so much for that, Brett. Very very good stuff.  And Nick, why don’t you take us home before we get to Q&A? Just a reminder to everyone, if you haven’t done it already, please fill the type in any questions into the Q&A box, and we’ll get to them in just a couple of minutes.

Nick:Just to reflect on that demo, I think for an insurance situation, the application build is interesting and gives you an idea of how disparate events can be brought together to produce an alarm or an alert. It’s quite basic in in terms of what it’s comparing, but this can apply to anything. It can apply to any stream of data, any situation which can then be combined to produce action.

And what I’ve got here on the next slide is what you can imagine as a customer journeys through connected services and collaboration, how this can transform the customer experience. Now, before I go through, this just go back to all of those wonderful words I put on the challenges slide about being reactive, about being non real-time, being a commodity, etc. This is now a view of the possible. So, this is this is John and John buys a car. Good ol’ John, he’s one of the guys who’s decided not to do ride share. He’s gonna buy a car, and he buys an insurance policy, he goes on a site, a comparing site, he buys the policy, and off he goes.

Just down the road, his car’s been stolen. So what’s he gonna do? He’s gonna call the police and he’s gonna call his insurance company and say, “my car’s been stolen”. So now he’s without a car. He’s in a bit of trouble, basically. But you know, a few days later, the police find his car, but he really has been damaged. Someone’s driven it into a tree. They’ve really messed up his car, so he’s now gotta claim. Vehicle was stolen, got it back, but the damage is still a claim. So, the insurance company, being a good insurance company, they pay the claim, and now they’ve got a chance to offer John a connected policy. And John, being a good guy, he accepts. He said, “I don’t want the situation again.” This connective stuff sounds great. So, there we go. He’s just become a digital customer in a connected sense.

So now the insurer has the opportunity to offer much more to John. So, down the road, John’s app allows him to share costs in his share a ride to work scheme. So, he’s going to work every day, picking up a bunch of colleagues. They’re going to work. They’re coming back. And what’s he doing?

Well, he shares the cost of the petrol or the fuel, the gas, whichever country you’re. That’s about as good as it gets. But through the app, because we know what journeys he’s taken, the length of the journey, the total cost of the journey, you can build into an app how much John would need from each of these co-passengers to pay for the journey. So, they truly are sharing costs, not only sharing a ride. That can easily be built into a front-end app. It’s not a lot to do with insurance, but it is a real service that helped John to get his money back. And when he’s sharing his nice shiny new car that’s just been repaired with his work colleagues.

Here’s another example: John can now automate his expense claims for the use of his car for work, another service you can provide by knowing what John is doing with his car. It can be built into the app. Some days he’s traveling to work in his car doesn’t count a business expense. Some days he’s going out to see a customer. That is a business expense. Many countries now in Europe, and I’m sure this is going to spread to many other countries in the world, just writing that down with a piece of paper or telling someone how many miles were a business and how many miles were not is just not enough.

You need some kind of verified way of making those journeys a reality. So, through an app, you can get your list of journeys because you know where the vehicle’s been. And you can press on an app and classify them into private and business mileage. You can get the printout of your business mileage and hand that in as your expense claim – saves an awful lot of work at the end of the month. You don’t have to look on Google Maps to find out how far it is from A to B because it’s all in this device. And so the app now is a basis for an additional service, an additional thing to make John happy.

So now John’s driving along, he’s in a traffic black spot. So think of the app that was just built before your eyes, you can now take into that kind of app traffic data, road safety data, known black spots for accidents. Johnny’s now speeding into a traffic black spot. You can enhance that, you can say there’s traffic around the corner. He doesn’t know about it. That’s a time when he’d like to get an alert. That’s a real way of preventing some kind of accident. Oh, John’s actually driving pretty well. Now he gets a discount code for his friends. So, all these share a ride to work friends can get a code because John’s done a great job on behalf of them. And now, the insurance can get a few more customers because they’ve now got a discount code. And because they’re all John’s friends, they’re likely to be good drivers too. If they become digital drivers, then we’re gonna know that for sure anyway.

John decides to go on vacation. He turns up at the port to put his car onto a ferry, but guess what? He hasn’t got a policy for his vehicle that allows him to take his vehicle overseas. So, either that car’s been stolen, or that car is going into a territory where he’s at risk of not being insured and he may not know that. So, a combination of another pair of events will allow the insurer to communicate with John and tell him that he ought to buy an extension to his policy. You can even build that into the app saying “click here to buy.” And he’s now covered as he drives his vehicle on to the ferry.

And of course, at the end of the period, he can get a notification saying “Hey John here’s your renewal. Just click here to accept. Just carry on driving. Don’t worry about those compare sites. The world is great.” So, what you can see there, some of those are realizable, everyday occurrences, some are a little bit fantasized, but I think you get the meaning through this, about how a relationship with a customer can change. And the touch point, without connectivity, would have been John buying a policy, John claiming for damage on his car, and a renewal. And now you’ve built in, close to 10 touch points there, which we were not possible before this technology was around.

So really just moving into what are the takeaways, technology is something to be embraced. It’s an opportunity to change the relevance of the insurer. It can take an insurer from this entirely defensive mode into a leadership role really quite easily without too much disruption. Insurers will definitely need to be able to manage data in real time. This is a real-time world, and unfortunately, it’s not a synchronous world. It’s an asynchronous world. Things happen at all times of day and night, not in synchronicity at all. So, insurers need to be able to manage that in a in a good and well-efficient form without hiring armies of people to wade through lots of data.

There’s new platforms out there. VANTIQ is a great way to enhance the value of that real-time data. Just having data for data’s sake is not gonna solve anything. Just having data without turning it into something you can use and communicate is not going to do anything. It turns it into meaningful collaboration. A meaningful collaboration, at the end of the day, means you’re building relationships with customers, building satisfaction, building loyalty.

Blaine. those are the takeaways and that pretty much concludes what I have to say on the subject.

Blaine:Perfect. Well we definitely have had a few good questions that have come in. And again, thank you so much for leading the great discussion.

Actually, I will start with one of my questions and then we’ll get into the others. Early in the slides you talked about insurer strategies as either defensive – remember you had the red box for defense of, the yellow one for maintenance, and the green one for leading – based on your experience in the industry what’s your estimate of the sort of percentage of insurers today that might be fitting in each of those three boxes? Where are we actually at in the market and how fast is that changing?

Nick:Many insurers are migrating now from the defensive to the maintenance level by deploying connected strategies. Vehicles are leading the way. Homes and health are up there too. But many are going in that direction. I would say that around 50 to 60 percent of insurers have started to migrate that middle ground. There’s still a good chunk, maybe 40% of insurers around, who are still sitting in defensive territory. And of those certain 50 percent that are up in the mid ground, there are those who have taken leadership roles. And typically, these are the insurers who are embracing insure tech type companies or new technologies and really are determined to get into a leadership position but this is probably, today, well less than 10%, probably less than 5%.

Blaine:Interesting. Are there any insurance companies that you’re aware of that are actually doing true, dynamic risk assessment right now, risk assessment in real time?

Nick:I know one, but I can’t tell you who it is on this webinar, I’m sorry.

Blaine:[Laughter]So we know that one does exist, all right.

Nick:Nice try.

Blaine:Great so let me get into a couple of questions that have been submitted. This is an interesting one: so wouldn’t these kind of applications promote distracted driving? We heard the beeping and the alerts and everything. What about distracted driving concerns?

Nick:All right absolutely. It’s something to be really careful of. And this becomes a decision that’s got to be taken by the insurer about how messages are delivered. We got that example, and I did say this is a fairly straightforward and simple example, and normally, what you don’t want to do is distract a driver so you cause an accident. That would just be crazy.

So you’ve got to be very careful about how you do it. This comes down to the collaborative action. Some insurers in the mid-ground are using telematics today for vehicle insurance and sensors for home insurance, and they will find ways of informing customers about things that need to be communicated without distraction. But it’s the insurance, it’s an insurer decision. All insurers, being very risk aware people, will not want things beeping in cars every time a vehicle speeds. So great example to show how the code can be generated, but nobody would want a lot beeping going on.

Blaine:I guess the notion about real-time, as far as notifications and assistance to the driver, doesn’t have to be real-time in that very moment. It could be at the end of a trip. It could be at the end of a day, a summary if they if they got an insurance bonus for based on how well they drove that day. There’s lots of ways of making the business more real-time without an immediate real-time notification, I would imagine.

Nick:That’s absolutely right. The key takeaway from this is that you can do something in real time, as the application showed there, you can put this into CRM. So you can keep a record of what’s happening so you can feed that back to a customer. You’ll have customs with no records because they’re really driving well. You’ll have others that are just full of records and you can exchange that on a weekly basis or monthly basis, whatever you feel is the right way to do it. But it’s useful data for the driver and it’s useful data for the insurers.

Blaine:I guess at the end of the day, cars are becoming real-time, our real-time system. So figuring out how to work effectively in that ecosystem is a challenge but also a big opportunity. Let me ask you another challenging question here that came up around privacy. So, how should we feel about giving insurers all this data about ourselves?

Nick:[laughter] Well that goes back to the diagram where I said this is a balance. The data is very useful for insurance, but you’ve got to give something back. And the baseline for all of this is if you give your customer enough reason for him to give you his data, he will agree. And these these are all opt-in policies. This is this type of policy you would have with the customer must have an opt-in element where the customer agrees that you’re gonna have this data about him. Otherwise, you’re doing something which is way out of data protection legislation. So, it’s opt-in and the key, though, is what you give to your customer for that. So, if you think about the customer journey we had, all the benefits that were being delivered to John in that journey were real, solid benefits that he wouldn’t normally get unless he was giving up some of his data.

So, if the service and the and the benefit is sufficient, people will opt in to give you their data. That’s well proven. So, that’s the answer. Without opt-in, you can’t do this. But then, nobody else can either. You can’t be destructive with a customer that doesn’t want to give data to anybody. It can only work with customers that are willing to give up data.

Blaine:So I imagine most insurances, as you said on the maturity curve, already have some kind of IoT ecosystem, maybe either that they’re experimenting with or maybe partially implemented. How can you implement a solution or applications like we saw built with VANTIQ into an ecosystem that’s already underway?

Nick:Well, the VANTIQ system, the platform that VANTIQ is developed is an add-on to any IoT environment. IoT environments create the data. You’ve got to think about this as two things: there’s one side which is just creating all of this data but then you get into overload. There’s too much data there to wade through. And the value of data diminishes with time. Just take the element that I touched on there with a crash. The value of a crash occurrence really tails off with minutes actually after the crash has occurred. And if you’ve gotta wait a month to do something about it, then you’ve lost any data around the circumstances of that crash. So what VANTIQ is doing is its allowing an insurer to wade through that data, look for the things that he needs, and then act on it.

So, on one side, you’ve got data gathering. On the other side, where VANTIQ sits, it’s turning all of that data into something which is a basis to collaborate with the customers.

Blaine:That makes a lot of sense. Only two questions left. Actually this one’s gonna be a question for Brett. So, one of the questions that was asked, that came up was Nick showed one of his last slides with that journey of the customer and all the different kinds of applications or use cases that could be supported in this ecosystem we’re talking about. How long would it take to actually build this this kind of solution?

Brett:I assume the question is how long would it take to build something like this inside VANTIQ? I think building something like that outside of the system is, generally speaking, a several months’ proposition, just based on what I saw on the slide. And VANTIQ, this is probably a couple weeks’ worth of effort. And you know the majority of that being one or two weeks for the part that is internal, all the things that are happening in the VANTIQ platform itself. Those sources, when you talk to anything outside of VANTIQ, we call those through something called the source, which is an abstraction of how to quickly talk to another system. But it does mean that you do have to at least define the source and the rest endpoint. So depending on the complexity of connecting those sources, a couple more weeks. I would suspect, as I mentioned, a couple weeks to build something of that nature.

Blaine:Interesting. I think many of the listeners to this webinar were probably expecting an answer more in terms of years, not weeks, frankly based on probably how long they figure it’s taken them to implement other interesting, digitally transforming initiatives in the past so that’s pretty revolutionary.

Brett:The one thing that’s probably important to keep in mind is: you don’t have to boil the ocean with every situation you can come up with. It’s designed to allow you to iterate on it and iterate on it quickly so that you can build pieces and obtain value in fundamentally real time as opposed to trying to create one, monolithic app that you release a year and a half later.

Blaine:Very good point. Very good point. Final question was about who would actually be building these applications. [He] says “I don’t have someone in my org that has the, even though it looks simple, I don’t have someone in my org that has the bandwidth or a team in my org that has the bandwidth to do this. How could I actually do it or who could implement it?

And maybe I’ll take that one. So VANTIQ often works with systems integrators or consulting partners or others who specialize in actually building VANTIQ applications on behalf of clients so that you can quickly and easily get into market with these applications, if your team is too busy or tied up to be able to do this kind of work themselves.

So we’re at the top of the hour. I just want to thank Nick and Brett for joining us – a really insightful presentation and a great demo, Brett, as well. You can find VANTIQ obviously at vantiq.com, contact us through that and to get more information. And you can see the URL for nick’s company, nw4c.com at the bottom of the slide. And thanks to all of you for joining us for this webinar. Reach out any point in time for more. Bye-bye

 

 

Featured Resources
White Paper
VANTIQ - Enabling the Digital Enterprise
Read the vision of Marty Sprinzen, CEO of VANTIQ, on how VANTIQ truly enables the real-time digital enterprise.
New Gartner Report - Free!
Innovation Insight for Event Thinking
Read Gartner’s latest report on the event-driven application revolution PLUS new product tour videos, VANTIQ FAQ, and interview with VANTIQ co-founders Marty & Paul.
Demo Video
Real-Time Application Overview and VANTIQ Demo
A detailed overview of why real-time, event-driven applications are required to truly transform your business operations. Includes a demo of how to rapidly build and deploy one.

Follow Vantiq