Connected Insurance Podcast

Michael Jans: Tony Tarquini, how are you?

Tony Tarquini: Hi, I’m very well, thank you, Michael. Nice to speak to you.

Michael: Well, good to speak to you. I’ll put this in some perspective. You and I do share some common experiences, co-authors of the Insure Tech book. That isn’t what originally brought me or attracted me to this conversation. What attracted me was an article, and I honestly I apologize I don’t recall where I see it, in one of the online trade magazines, where you were making for developments and ensure tech 2020 predictions. Predictions for this year in regards to what insure tech was. That set me on my usual path of learning more about you reading what you were doing, and then finally reaching out because I thought you could make a contribution.

Not my first podcast with overseas or abroad. Always a pleasure when I do, because I think that when we do that we get an unusually broad perspective, which is a real opportunity to step back and take a look. Tony, what I’m hoping, and I’m certain that we’ll be able to do during this conversation is a clear-eyed view on the turbulence or the change, the transition of the industry and the forces and trends that are making that happen.

First, tell us a little bit about Tony Tarquini, because you do have an interesting personal history and professional history.

Tony: Yes, well 33, 34 years of that been in insurance technology.

Michael: As I was, but you were even longer than I were. You were an insure tech professional before the word existed.

Tony: Yes, indeed. It’s been coined over the last five or 10 years predominantly for the startups.

Michael: Honestly, I think when I started Agency Revolution as an early insure tech nobody had thought up the word.

Tony: Yes, absolutely. Actually, there’s been insure techs pretty well since technology came into insurance. We used to call them software houses or technology companies, but I think it’s a good term, particularly at the moment because there’s so much exciting stuff happening in regards to technology. 33, 34 years of doing this stuff.

Michael: You’re currently with?

Tony: I’m currently with Pega systems based in Boston or Cambridge, Massachusetts.

Michael: Right. For our listeners, tell us a little bit about what Pega systems is and where that fits into the industry.

Tony: Okay, so Pega is a 30-year-old insure tech company founded by a guy called Alan Trefler, who’s still in place as a CEO and a major shareholder. The company is several thousand people globally. We deal with international organizations, large corporates, predominantly around CRM, and customer engagement. We also do a lot around artificial intelligence and one-to-one custom personalization. Also, we’re probably the world leaders in case management technology, and digital Process Automation or Intelligent Automation as it tends to be called these days.

We work with all the major banks, the banking industry is where we started. We now work with pretty well every major insurance company around the world. Similarly, government departments, telcos, life sciences, and even Walt Disney Corporation. We have a large spread of organizations we work with. For my scenes, I have responsibility for the insurance side of things in Europe. I look after our strategy and I go to market in the European…

Michael: How would you describe your professional expertise in the industry?

Tony: Well, I started off in broking systems. Like everybody who I think comes into the insurance industry, happened by mistake.

Michael: Well, I know that’s true here. It happens by mistake or happens by birth.

Tony: Nobody ever grows up thinking like, “Burning ambition to work in the insurance industry.”

Michael: Baseball or insurance. Okay.

Tony: Actually, my son is an underwriter. He’s been writing business at Lloyds of London last month, but he’s now moved to Singapore. He famously said to me when he left university, “I would prefer to stick needles in my eyes than work in something as boring as that.”Long story short, he ended up as a qualified underwriter as well.

I mean, I fell into it like everybody else. I was in technology. I ended up working for an insurance technology software business. They were very much in the broking market, particularly commercial lines insurance broking. They headed about 120 of probably the largest commercial lines insurance brokers in the UK. That’s where I cut my teeth for the first few years of my career, and preferably learn what I needed to know about the software business.

Broking systems, not easy. In fact, broking systems are far more complex than insurance company systems. It was an interesting start to my software career.

Michael: Now, let me ask you to pause here for a moment. I want to make sure we’re speaking the same language. Broking systems in the UK, what does that mean in the US?

Tony: In the US that would be an independent agent system.

Michael: Like an agency management system?

Tony: An agency management system. Like I said the terminology is the thing that was always the differentiator between our two sides, and working for an American company, I’m constantly having to remember that. Yes, it’s an agency management systems within the UK, particularly in commercial lines. I did that for six, seven years. Then I went to work for CSC, Computer Science Corporation, very strong in both the broking side and in the insurance company side, and more and more drawn into the insurance company systems market.

I’ve had a variety of roles over the years, working both in technology and in management consultancy. I had a period working in pure management consultancy, which was actually extremely valuable in terms of understanding technology because it wasn’t technology-based, it was a very management consultancy. It reminded me that as much as technologists like to think that the world revolves around technology, actually, it doesn’t. In the insurance industry, the business comes first and technology is the tool. It might be an important tool, but it’s still just a tool.

Michael: In fact, technology revolves around the world, not the other way around.

Tony: That’s right. Yes, I was in management consultancy up until I got a tap on the shoulder to come and join Pega about seven or eight years ago. My role within Europe is looking after, I go to market and our strategy for insurance within Pega.

Michael: Tony, you and I do have a shared common experience to as we are both co-authors of the Insure Tech book, which really was originated where you come from, not from the US. Just for the fun of it, and then we’ll dive into the trends and forces that are affecting the industry. For the fun of it, tell us a little bit about what was your experience there? What was the topic that you addressed in that book?

Tony: Yes, so I wrote a chapter on robotics. Actually, the reason I chose robotics is that I think there’s a lot of misconceptions about robotics in the insurance industry. It’s seen as a silver bullet that will solve so many problems, but actually, there is an element of snake oil about that. We have to be very careful to not think that because you’re going to use robotics, all your problems are going to be solved. It’s a great tactical tool but it has to be taken in the context of what it is. What I tried to do within the article that I wrote was to help people to understand when and when not to use robotics. To understand the typical benefits that they will be able to get from it, what they wouldn’t be able to get from it. Then the different types of robotics that are out there, and how and when to use them.

Michael: A word on that. When you use the term robotics, what does it mean, how do you define that? I’m thinking of the Jetsons.

Tony: Very much sneaky about software brokers. From an insurance perspective, we’re clearly not manufacturing a physical product, robotic arms coming out producing something.

Michael: Building insurance.

Tony: While insurance is intangible. It’s very much about how you improve processes. There’s two kinds of robotics. When you’re talking about pure robotic, you’re not talking about AI or anything like that, pure robotics, there’s unattended RPA robotic process of automation, and attended robotic process automation. The unattended tends to sit on servers, it does end-to-end processes, and when it’s working it probably results in achieving close to 100% improvement in the area where it’s working, but it’s all much smaller groups of workers, and it’s simple processes, like a fraud detection or an MLA check, something relatively small and self-contained.

Then you have desktop or attended RPA. That tends to be smaller task automation that can address change, and it results in 10% to 20% improvement, but it’s across much larger numbers of workers because you’re doing that process much more regularly. So it’s about finding the right RPA for the situation that you’re trying to address.

Michael: Okay. So as long as we’re on the subject. In this podcast series, I’ve interviewed insure techs with chat bot technology. Does that fit within your definition of robotics?

Tony: Yes it does. There’s an element of AI within that. Yes, that’s right

Michael: Obviously AI capability. So it’s not just worker to worker, it can be retail to public.

Tony: The key thing about a chat bot is that it should always be integrated into a live chat facility. You’ll have heard of the Turing test?

Michael: Yes.

Tony: The Turing test is the point at which somebody can work out there’s a robot that’s speaking to them not a human being. In insurance terms, if you can keep that going for as long as possible using the chat bot, you’re going to be able to assist people hopefully using an automated system which doesn’t involve expensive human beings. However, if you get to the point where that that’s not coping, it needs to be able to say, “That is something a human being needs to deal with,” and seamlessly move that into a human being taking over the chat. You can read what’s gone before, understand immediately and say, “All right, okay. Let me help you with that.” Chat bots are fine, but chat bots in isolation probably don’t work most of the time.

Michael: In conjunction or integrated with human service?

Tony: Absolutely.

Michael: Okay. Does that fit within the– You had mentioned this is an attended robotic 10% to 20%?

Tony: What it’s about process automation rather than an interaction automation. So there’s two quite different things. Once interacting with the customer tring to establish what their need is and to be able to understand that. They’re actually asking to change their address, so directing them to the right place maybe a self-service portal that they can be able to do that or they want a document, and pointing them to a URL which would give them the document they want or they want to see the policy document. You go, “Click on this URL you’ll be able to see a policy document,” those things. Those things can be dealt with robotic. It’s a an AI type of capability rather than a pure robotics one, but the key part of it’s got to be integrated as part of an overall conversation.

Frankly if somebody starts with a chat pop and it moves to a human being, and then needs to speak to somebody in the contact center, and then maybe needs to go, get off their off the laptop and go onto their phone and start doing on the phone. As they’re walking down the street, you should be able to do that completely seamlessly. Unfortunately only about 4% of insurers can do that.

Michael: All right. So let’s put this in some historical perspective. Obviously, I want to get to your perception, your observation about the major trends and forces in insure tech that are affecting the industry, and particularly of course our interest is going to be on the retail side. Again let’s put it in some perspective that term as we talked about earlier, the term insure tech didn’t exist but there has been insurance technology for ever. You have been in this space for over 30 years. What do you see is the major changes in insurance technology historically over the past 30 years?

Tony: Since I started? Well, there’s the generic ones, the introduction of the IBM PC. Sadly I go back that far. I can remember when that came out and how we moved to personal computing, as opposed to everybody’s sitting with a green screen, character based green screen. Obviously, the invention of the Internet has made significant change in technology. A apart of that is email, an endemic part of the insurance industry today, it runs on email, and smartphones more recently the implication of that. So those are the generic stuff.

I think multi-channel communication and distribution. So we’ve seen insurance no longer just sold at a personal level by an agent going out talking to their client, but increasingly we have the wave of contact centers, and then we have websites, we’ve now mobile capabilities, we mentioned chat bot earlier. A whole range of different means of communication and actually distribution of product, and servicing of product through those technologies.

I think specifically around the way the insurance has worked with technology, what I’ve seen is a significant move forward over 30 years within efficiency and effectiveness.

Michael: That’s an interesting division and it’s also one that I tend to make. The way I look at it is, the efficiency is like the division in arithmetic is represented by the division sign. We can be 10%, 15%, 20% more efficient. The effectiveness is sometimes a multiplication sign. So in other words, and I’ll use my own personal history from this with Agency Revolution, we facilitated the transition from communicating through email, one person at a time to thousands or tens of thousands of person, people at a time. So the efficiency and the effectiveness that’s the bedrock of what software supposed to do. May seem fairly obvious but it seems to be a useful division.

So yes, in those 30 years history, we’ve seen these generic transitions, which probably have driven the more specific or enabled the more specific

Tony: Enabled, yes.

Michael: Clearly, we couldn’t have created a marketing automation software without email. So the timing of the horizontal or the generics allow for the innovation of the specific verticals.

Tony: Actually, from a business perspective Micheal, what I think I’ve seen is, 30 years ago, huge volumes of people working in the back office. The work was being interfaced in the front office and passed to the back office to do the work. Over the last 30 years or so, I think increasingly work that was previously done in the back office has moved to the front office, and work that was done in the front office has moved to a self-serve environment. There are less people involved in the industry because it’s more efficient, but also the processes are more effective. They’re getting it right first time more often as opposed to, “Yes, I wanted to do an address change, but you haven’t quite got the address right. So I’ve come back around and do that loop again, create that processing and go through it again.” Actually the effectiveness piece of doing it well, I think is also something that I’ve seen in the industry.We’re much better at doing what we do and we can also do it faster.

Michael: That perhaps takes us to now. I think you would agree this is a time of some- on the emotional side- some consternation and confusion. Then on the objective side, what may be perceived of as chaos because just the influx of change. I mean we’re seeing now remarkable billions of dollars being invested into insurance technologies. Whatever word we want to call it, disruption and transformation we certainly can call that some jostling There is a jostling of the industry, which does tend to cause some confusion. Certainly my clients, my listeners, they want confidence and it and it comes about with some clarity. What’s going on? what what do you see as– what are the major trends and forces that are affecting the industry now? What does that mean to the independent insurance agent or broker, and what do we need to do about it?

Tony: Clearly, one of the things that’s happened over the last 30 years is a build-up, massive buildup of legacy IT. Sadly, the insurance industry is more burdened with legacy IT than most other industries. Part of that is that if you think about it, the insurance industry is, everybody in the insurance industry is trained in risk management. Technology by its very nature is risky. People are risk averse when it comes to updating and changing their technology.

I sadly heard from an insurance company a couple of years ago who were rejoicing in having got rid of their, replaced their administration system, which had been going for 40 years. 40 years it has been in place. That’s a fairly extreme position but there’s still a lot of it around.

Michael: You’re presenting a context I want to touch on for a moment here, Tony, in fact, I will reference a headline to one of your articles where I think it said, “Be wary to not thrust change on the insurance industry” and because of that, it’s a by nature a risk-averse industry we’re trained in that it’s, it does seep deeply into the culture. We can consistently see other industries, transforming before the insurance industry transforms. We can look at virtually almost every industry, and see fairly major transformation since, let’s say 2000, the emergence of the internet and now there’s a great deal of catch up that’s going on.

I interview a lot of people who come in like insure tech CEOs or startups who come into this industry somewhat surprised that that problem hasn’t been solved in insurance yet, because it was solved years ago in banking or other industry. It presents a problem, and I’m curious what your thought is on this. On one hand, yes, perhaps there’s some caution about being wary about not thrusting change. On the other hand, clearly in order to thrive, the savvy insurance Professional manager in my case, the agency principle, they need to have some willingness and some capacity to embrace change. It’s not as if– The change is happening and so what-

Tony: It’s a balance. I think it’s very much a balance. The thing you’re referring to is more about regulatory change.

Michael: Yes, I saw that.

Tony: But nevertheless, sorry, we have unsustainable lighting here.

[laughs]

Michael: Okay, for the audience, I am on a video with Tony who’s in London. He’s working late and the office just turned the lights off.[laughs]

Tony: If I don’t move too much. I need to move my arms around a little bit.

Michael: There you go. Okay. It’s green building for you.

Tony: It’s very much a green building. The fact is that change is being forced upon the insurance industry now. We got the legacy build up, yes, we have to deal with that, but that’s not the big issue. For 350 years since the Lloyd’s coffee shop and modern insurance was started, the premiums of the many pay for the claims of the few has been the mantra of the insurance industry. It’s been about risk mitigation. Something goes wrong, you’ve already put some money in the pot in the middle, you’re allowed to take some money back out again, to get started again.That’s a great premise and it’s worked very well for 350 years, but the world’s moving on now.

I think it’s a very, very exciting time in the insurance industry, particularly from a technology perspective, because the industry is moving away from risk and risk mitigation to IOT based, an Internet of Things based risk management and protection. We’re no longer sitting back and saying, “We’re going to calculate how many times we think you’re likely to have a claim and will quote the premium accordingly.” What we’re saying is, “We will actively try to stop you having a claim by helping you to not be in situations where claims will arise.”

I think insurers are going to take a much larger role in helping clients to manage their risk, to monitor their risk, and as the number of Internet of Things devices starts to grow in the world, being able to use the data from there in real-time, to help clients to be able to say, “You’ve just walked out of your house, you’ve left one of the windows open, go back, close the window, your house will be safe.” Or, “Your machine on production line number three, is one month beyond the point where it should be maintained. Somebody’s overlooked the maintenance procedure for that. If it overheats, it might explode, it has dangerous chemicals in it and you’re likely to have to close down the entire plant for a month if something like that happens.” Being able to look at that information, and using artificial intelligence to be able to understand what the implications of that is, and then automatically be able to go back to a client and say, “There’s a risk event happening here. You need to be aware of it. We need to be able to deal with it.”

How you reply to the client is really important. There’s no point sending them a letter to tell him to close his window because he’s walked off the street.

[laughs]

Michael: Or lets say in commercial lines, you’re shipping a semi-truck of lobsters from the east coast, from Maine to Chicago, and you need to maintain the temperature exactly within this margin, andn the Internet of Things is going to communicate to you immediately if you exceeded that margin?

Tony: Yes, absolutely. That’s mitigation. It’s moving from mitigation, “Give you some money back if it all goes horribly wrong,” to, “Let’s stop it happening together.” It’s actually makes insurance more relevant to people, because let’s be honest about it, how do people and companies see insurance? They see it as a necessary evil. We’re talking about auto insurance, you have to have it by law because you can’t drive a car otherwise. That aside, it’s something you just have to buy because if you don’t, get in trouble but they see insurance companies as big corporations, and the agents as well as organizations that take money off them every year and give nothing back. I have colleagues who sit with me in my company say, they asked me about insurance all the time as you can imagine. “I pay these guys every year, and I never get anything back from them.”

[laughs]

Tony: Yes, but what happens if you crash your car on your way home tonight? Are you prepared to fund all of the repayment on your car and any damage you might have done to somebody else and somebody else’s vehicle?”

Michael: Then, they may get grumpy about the claims process.

Tony: I think the ability to say, rather than just taking money off you at the beginning of the year and saying, “Please don’t call me, I don’t want to hear from you for a year until it’s time to buy a new one and pay more money next year.” That’s the model that insurance has had. If you get to the point where insurance is saying, every 30 seconds, we’re going to be monitoring the data on your risk and we’re going to estimate. If everything’s okay, you won’t hear from us. The moment there’s a problem and you need to know about it, we’re going to let you know so that you can do something about it. We’re going to protect you. Once insurance gets into that position, you’re completely changing the model and we’re starting to see that sort of thing happening in the auto market.

The auto insurers are starting to build into their vehicles, intelligent systems, which are constantly monitoring everything. When something goes wrong, they’re doing something about it. They’re telling you with a warning light or a message or whatever it may be. We’re going to see that in all kinds of different things and insurance is going to embrace that. I have a client, an insurance company in Italy. The Italian market is the leading market in Europe for telematics insurance. They have black boxes in cars monitoring what’s going on. These guys are the leaders in doing that in Italy, so pretty well the world leaders in telematics insurance, and are amassing huge amounts of data, and they’re helping their customers to understand what’s happening on real time basis.

Michael: What has been the consequence or the implication of that?

Tony: They are able to reduce their premiums. If you look at the way that the likes of McKinsey are viewing autonomous cars, because that’s where we’re heading is autonomous car.At the moment, it’s warnings on getting too close to the vehicle in front, automatic braking, if you’re about to collide. All things kinds of things which are being built into modern cars to ensure that claims don’t happen.

As you get closer and closer to the point where you take your hands off the steering wheel and it drives itself with all of those things around it, you get into autonomous cars. Once you get into autonomous cars, who’s liable? There will be no longer a requirement, unless you put your hands back on it, a requirement for personal lines auto insurance policy. It’s going to be provided by the manufacturer. It’s going to be commercial lines, commercial liability policies. I think it’s Forbes that reconed a 75% reduction in premiums as the risk diminishes and collisions stop happening.

 Michael: First of all, who did you say made that prediction?

Tony: Forbes magazine.

 Michael: Can you put a timeframe on that?

 Tony: When autonomous cars happen?

 Michael: 70% is going to get the attention of everybody who’s listening to this conversation. Part of my job is to recognize that everybody has a strategic horizon. In other words, there’s, that period from now to that point when they don’t care anymore. For an agency principle that might be their perpetuation, perhaps plus a few years, and so that horizon needs to be– [phone rings] I do apologize for that. A horizon needs to be considered strategically by the agency principal. It may be in a large part depending on their age. For a millennial to hear what you’re saying, is something, admittedly yes, there’s no way to say that’s great news for the retail insurance broker but the smart broker simply wants to know what’s going on, so that they can make appropriate strategic decisions.

 Tony: Right. I completely get that. Having started off in the industry market, I completely understand those guys. How those guys are thinking. In the urban markets, big cities we’re already seeing millennials don’t have cars.

 Michael: Right. That’s a different trend that’s affecting that market, right?

 Tony: It is, except they’re getting used to the idea that they will never own a car. They’re very comfortable with that. They may not even take their driving test. In America, having a driving license is a right of passage, because it’s your identification document or whatever. In a lot of places that’s not the case. If they don’t have a driving license, it doesn’t really matter too much. They take the view, “Why would I do that? I might pass my test, but I’m not going to drive for the next five years, because I live in the center of London, I can take public transport, so I don’t have to worry about that.” When they get to the point where they need to drive, they just hire a car for the day. They hire a car for a short period of time. Those short periods of time very soon, are not going to be, “I’m hiring a car that I can drive to somewhere.” It’ll be, “I’m hiring a car which will take me to somewhere.” We’re already expecting that.

I have three sons between early 30s and mid 20s. They’re all saying the same thing, “If I don’t have to drive, then I’m cool with that. I’m quite happy, if I couldn’t call up on my app today, rather than Uber, or Lyft, or one of the others, I would be quite happy to have an autonomous car take me somewhere.” They’re already in that situation where it’s cost effective for them to get a taxi somewhere, so they’re not going to drive. They are quite happy with the idea that one day that taxi driver will not turn up, it will just be the car. In urban areas, that’s already figuring in their thinking. Rural areas, slightly different situation. You need a car to get around. I understand that.

Over time, that’s going to be the top end. It’s going to be all the community who are saying, “Actually, I’m not capable of driving anymore, I’ve got to 70 I don’t do that. Instead of moving to an urban area, because I have to, or have somebody drive me, a car’s going to turn up, it’s going take me to do my shopping or see my friend, whatever I need to do. Increasingly, that’s going to be the issue, people will not need personal lines insurance. It will be provided by the manufacturer as part of that journey.

In fact, there will be more associated with that. If you really decide that you want have a car, the manufacturer will provide it and they will be liable if something goes wrong with the software on that journey of getting you to somewhere. That will be part of their commercial liability insurance. Along the way, they’ll also have insurance for the maintenance of the car, the servicing of the car, and all the other things that go with it, a complete end to end proposition out of a car manufacturer. People won’t be selling personal lines also insurance anymore. Time scales, you want any time scales?

 Michael: Okay. After that deep breath. [laugh] Okay, that’s a significant trend. That one largely seems brought about by outside the industry technologies affecting the industry.What else do you see happening in the next 10 years predictions?

 Tony: I think in terms of the way that insurance is sold. We touched earlier on about the channels of distribution that we’ve seen. At the moment, there’s a very disjointed multi-channel capability with so many organizations, “If you want to contact my contact center, that’s great. We’ll deal with your problem. If you want to go onto the website, that’s fine, you can do that. If you want to use the mobile app, that’s fine.” If you start using a mobile app, when you’re on your way home, and then you decide that you’re going to move to the contact center, the guy in the contact center says, “I have no visibility on the stuff that you’ve been doing. Can you start from scratch? Because we have a different system in the contact center.” You don’t have a seamless omni-channel service going all the way through the different channels, whether you’re selling a policy, whether you’re servicing a claim or you’re just doing a general customer service journey, insurance companies, and to a degree, larger brokers are not in a position to be able to have that full span of channels completely integrated.

I said before 4% of insurance companies have true omni-channel capability, where you can literally move from that to that to being on the phone talking to the contact center. The journey does not have to be repeated, 4%. Now within, and certainly we’ve done a Pega, within five years 86% of insurance companies expect to be able to provide that service.

 Michael: [laugh] Right.

 Tony: Good luck with that. It going from 4% to 86%, in five years [crosstalk]

 Michael: My experience in software development is that you’re always under budget and ahead of schedule, right?

 Tony: The fact is that other industries are driving us to do that kind of thing. We have to be there. We will get there over time, but it’s taking a bit of time. The whole omni-channel communication and distribution piece, I think is important. To be honest, the biggest technology that’s going to have any kind of effect going forward is artificial intelligence. Artificial Intelligence is going to completely change the insurance industry.

 Michael: What’s it going to do to the insurance industry? They’re infinite applications for this. Let’s pull the curtains back on that one.

 Tony: Yes. I’m not really talking about the kind of- AI is a lot of different technologies- it is not the kind of technology that is going to teach you how to play chess. It’s not the kind of AI that’s going to teach you the meaning of life. At the moment of interaction with a customer, how do you make that interaction and engagement better? That’s in insurance terms all about real-time analytics and decisioning.

The best way to describe that is it’s the ability to be able to take everything you know about your client, we’ll take as the example, the contact center, but it equally applies to any other distribution and communication channel. Somebody rings into a contact center, you should be able to recognize who they are. You should probably have a pretty good idea of why they’re calling. When they speak to you and they tell you what the problem is and confirm it, you should be able to give the best possible answer that the company can give. As an example, it’s a bit like being able to freeze time as soon as he’s told you what you need. You can go off and you can speak to anybody in the company, from the CEO down and say, “This is the client. Let me tell you everything about him. What should I do next?” They all get together and they say, “Actually, the right thing to do now is solve that problem for him or to try and cross-sale or up-sale him,” or, “Actually, we need to do some retention work here to make sure that he renews again in three months’ time.”

Michael: He’s got a 71% vulnerability rating or attrition problem

Tony: Be able to do that in an instant. I’m talking about– We have a bank in Australia. He was doing 20 million of these executions of the algorithm a day. They’re doing it in sub-second times so that when someone comes through and they’re talking, the thing is going up on the screen, “Say this next. Do that next.” The best way to look after this customer is to do this.

That’s called real-time decisioning. That means that your best operator, every operator is as good because they all know the same thing. They’re all being advised the same way. They can all empathize with the customer. At that moment when the customer is really upset and trying to get a problem solved, tries to upsell him on a policy or vice versa. Just doing the wrong thing at the wrong time, they do the right thing at the right time. They have got the right things to say and they know this is a happy customer, not a happy customer. This is a customer we really want to keep because he’s got a lot of policies, he’s been with us for a long time, very rarely claims, we very rarely hear from him.

He’s a great customer that we want to keep as opposed to a guy that doesn’t pay his bills on time, he’s always ringing us, he’s only been a customer for six weeks and we’ve had four calls from him. You can look at the propensity to want to keep that customer. We have an insurer in Holland, in the Netherlands, who have done a significant amount of that work with Pega. They have been able to increase their new business conversion rates by 27% by using this technology.

Michael: Oh my.

Tony: They regularly check it. They run their rules against various people coming in and they say, “What’s the conversion rate there?” Then they switch ti off. They try it without the rules and they switch it on and they switch it off, they switch it on, switch it off, all the time testing what the rules are. They know that by running the rules consistently, they get 27% more conversions their new business. That’s the first thing.

The second thing is retention. They monitor people going on to the website. They know what the behaviors are for people who are about to cancel their policy because they’ve modeled it and they know. If somebody looks like they are about to cancel a policy, they send a message to them saying, “Will you please contact the contact center? We want to help you and make sure that you’re getting value out of your policy.”

41% of the people who receive that message go to the contact center and ring them up. Of that 41%, 85% are saved. That’s a 38% improvement in their retention of people who were going to leave. Let’s not forget that’s 38% of the ones they want to keep. If you’re a really bad customer who’s always claiming and you never pay your bills on time, then you probably won’t get the message. The implications for that at Pegasystems are actually quite significant because they know they aren’t going to win business from that insurance company. The only business there I’m going to win is the bad business they want to lose. They’re picking up all the bad business and not winning any of the new business.

In the US we have a very mature market and there’s no more new business to win, you just win from everybody else. If you’re only ever winning the bad business, these guys are keeping all the good business. You know what’s going to happen.

Michael: Okay. My understanding of artificial intelligence is that its capacity trends are a very steep curve, so that it can start our relatively clunky. For example, my Google home has limited capacity and I can’t give it, let’s say, sequential commands. and Teresa might say, “Gosh, I wish it could do such and such.” I’ll say, “Just you wait because someday you’ll be scared about how much it can do.”

Tony: That’s another issue we should come on to in a moment. Transparency.

Michael: Okay, transparency. I thought of that earlier in the conversation when you were talking about monitoring my data every 30 seconds. There is that issue of proprietary data, transparency and data sharing. In any case, it would seems that artificial intelligence capacities may grow really quite exponentially fast. At some point Tony, before we’re done, I want to bring this down to the street level to the distribution channel, the Independent Insurance Agency and Broker and begin to take a look, “Wow, okay. What’s the impact for us?” Recognizing that your client base is primarily going to be the editorial level. My client base is at the distribution level and the implications are profound. Already, you’ve spoken about things with profound implications. Let’s circle back to that in a moment. Talk to us about transparency.

Tony: Transparency in AI is about to become a really big thing. Organizations are starting to talk about it and starting to recognize, particularly in a regulated industry like insurance, how important that is going to be. Let me give you an example. Regulators have the legal ability to go into any insurance organization and do what we would call in the UK an arrow visit. We’re coming to look at the books. We’re looking at how you do things, how you’re running your business in accordance with regulatory requirements.

If they look at a particular case and they say, “Okay, explain to me how you came to the decision to recommend that particular policy from that particular insurer to your client.” If you’re a broking organization, an agent, and you’ve got multiple agencies and you’re broking that business, you’ll say, “I chose that insurer because that was the best cover for the price that we could get, most appropriate for my client. I’ve got it all documented. I can show you.”

You start doing that using AI, you have to have an audit trail and you have to be able to say how you came to that conclusion. There are a significant number of AI systems out there which are what we would call black box solutions. That means you put the data in, you get the result out, no idea how you came to that conclusion. From a regulatory perspective, that is not acceptable in insurance. They will say to you, “You have to be able to tell me how you came to that conclusion.” There are myriad examples of AI which is self-learning algorithms and it’s picking up bias and they’re doing things wrong, and I’m sure you’ll see things in  the press or whatever.

You’ve got to be able to say, “I know that 18 months ago when I sold that policy to that particular client, we were running version 12345 of our ruleset. Here’s the full audit trail and that’s how we came to the conclusion to sell that one. My AI system recommended it but that’s how we came. You can look at the rules within that and work out how it executed the rules. I can explain to you how we came to that conclusion. If you can’t do that, transparency, you are basically breaking the regulatory compliance rules. I think over the next few years, there are likely to be some very big fines handed out to large insurance organizations who cannot do that because they have these black box solutions. They have no idea what’s going on in there.

The transparency bit is likely to become a big monetary issue if people are not able to demonstrate their transparency and their audit trail.

Michael: All right. That is worth pondering. It would seem that AI doesn’t, one perhaps share our better values because it doesn’t have any values or two, it may have got some of our lesser values. If we can reflect and perhaps magnify biases and move towards– it could conceivably discriminate on stereotypical grounds. Right?

Tony: It could be that it was recommending the right policy. If you can’t show how you came into that conclusion-

Michael: That’s right. [laughs]

Tony: – regulator doesn’t care because you are not in charge of your business, you’re not compliant.

Michael: All right. Let’s bring this down to street level here. We’re looking at a world fairly fast-moving and significant trends. We’re not talking about the quaint days of yesteryear, where we’re really looking at a future that we have perhaps, some glimpse into but certainly some uncertainty. It is perhaps a challenging time from that perspective to be an agency principal that needs to be the strategist for their agency or for their brokerage. What do they need to be thinking about right now and how do you think they then need to behave differently?

Tony: I think there’s two aspects from a technology perspective. One is the technology that they’re going to be using and the second is the technology that their carriers are going to be using. Taking the second first, they’re going to be looking to align themselves with carriers who are forward-thinking in terms of the products and services that they’re going to be providing going forward. If you’re aligning yourself with insurers who want to stay in the risk mitigation space and they want to deliver pieces of paper to clients and you say “if all goes horribly wrong, we’ll give you some cash, don’t worry, it’ll be fine after the event.” There’s an agent down the road who’s saying, “You know what? I can sell you for your car, for your home, commercial lines, whatever it may be, prevention systems which are going to help you to run your business.”

Who’s the client going to listen to? Not to have a claim as a client, I prefer to prevent that happening. If this guy is going to help me to avoid our business closing down or my house from beeing flooded or my car being in an accident, I’m going to go with that guy. The agents need to think about who they’re aligning themselves with, and what kind of products out there. The second thing is they’re going to think about their own technology. Omnichannel has a prerequisite. If your agency system cannot allow you to have that conversation which goes from the mobile cell phone onto the laptop, a call into a contact center, seamlessly, as an agent, you should be able to do that.

Some of the smaller guys, that’s not a problem, they’ve got a relatively small number of people there and they’ll be much more intimate with their clients. They have to have technology which is going to allow them to do that kind of thing because that’s what Millennials would want from them. They’ve also got to be able to interact with their clients on a pretty regular basis. I don’t know how many of your agents for example, send out an email to their clients saying, “Happy birthday.” I know insurance companies don’t do it, they have the date of birth in there, they never use it.

Michael: Well, if they’re a client of Agency Revolution then, there’s some likelihood that they’ll do that, right? [laughs]

Tony: Really, if you’ve got an opportunity to interact with your client, even if it is to wish them a happy birthday or whatever it might be, do it, be relevant. I think from an agency technology perspective, they’ve got to search out those ways in which they can use the data that they’ve got within their systems. They have great sets of data. How do you use that to interact with your clients and become more relevant every day?

There’s one final aspect about that and that’s the insurance is starting to change a bit. We’re seeing things like climate change, we’ve got Davos happening at the moment. If you look at what’s happened over the last 12 months, California fires, what’s happening in Australia with the wildfire out there, and Australia has gone straight from wildfires to floods. There’s significant climate change happening there, that’s going to have an impact on what people are going to buy. If you look at the assets that they’ve got to protect, the levels of insurance they’ve got, the nature of what that insurance would be and how they can be made aware of how they can prevent things happening.

You have a huge wildfire coming at you, there’s only a limited number of things you can do, but maybe think about the implications of that a year before or two years before in terms of, “Where is my house, where is it situated, what are the things around it, am I at risk and is that a kind of house I want to buy in the first place, and if I do buy it what am I going to do around it in order to–” We’ve seen in Australia on the TV coverage, people literally, one house is being decimated as the fire goes by, when another guy is sitting on his veranda filming it saying, “Well, fortunately my house is okay.” How do you work out which one is going to burn down? This data on that being able to get that data and protection and those issues, that’s going to become a big issue as well.

Michael: Okay, Tony, I have one last question for you.

Tony: Okay.

Michael: I frequently ask this of people that have your expertise in insurance technology. The business of buying technology is a challenge in the best of circumstances. If I was selling an enterprise solution to a Fortune 1000 company, they’ve got processes that they’ve developed over the years to help them hopefully make wise decisions. However, it’s not uncommon in the small business world and in our case the distribution sector, not to really have made significant technology purchasing decisions for 5 years, 10 years, 15 years.

They bought their agency management system that’s really the bedrock technology they’ve got, but now, I think they’re in a situation where a day doesn’t go by where they don’t have an inbound phone call from a handful of sales reps and they can’t get online and see one of the insurance trade magazines without seeing ads for this technology or that technology. In the absence of having a decent buying process, it either results in confusion or frustration or wrong decisions and perhaps purchasing the technology that isn’t the best.

There are two levels to this question, one of them is determining which category of technologies should I be engaged in. Two is, within that category, how do I make the decision which vendor or merchant is most appropriate or which product is most appropriate for my agency? Boom, I’ll pass that over to you.

Tony: That’s a really good question. The fact is an asset is greater at the top of the conversation. Agency of broking systems are complicated. Insurance systems are relatively straightforward, they sell policies and they pay claims, that’s it pretty well. An agent is sitting in the middle, they put their clients on the one side and they’re doing all the work up there. They’ve got to keep the insurers happy on the other side and they’ve got to do all the accounting and all of the processes that are associated with being an intermediary.

That’s very complicated systems. The truth is you should really be buying a system that’s dedicated to doing that work. Unless you’re a reasonably sizeable organization, buying components and trying to do it yourself is not a good idea. Buying something which is designed to do what you want to do, buying the right technology is key.

That’s the first thing.

The second thing is, who do you buy from? Obviously, you’ve got to have the most suitable technology for what you’re trying to do, you wouldn’t buy a commercial lines agency system if you’re doing personal lines. Because that’s a different marketplace high volume, low value versus high-value low volume. The truth is, once you’ve decided that there is a shortlist of suitable technology vendors, the thing that’s the most important and crucial issue is to understand the culture of the business you’re buying from.

Michael: That’s fascinating. Maybe the first time I’ve heard that response, generally we’re examining the viability, the capacity of the technology itself or examining to the extent that we can and the viability of the company itself because we want to be in a relationship that lasts. Also, clearly there are very, very few serious technologies where you’re not going to be engaging, interacting, getting professional services from the company itself. That culture issue, that’s a significant contribution–

Tony: You made the point, you made this decision maybe every 10 years. Once you made it, it’s really difficult to get out of it. If I was in an agent’s position, I would do what I’ve seen my clients do over the years, is to really delve into what happens when things go wrong because the thing I can guarantee to you about technology, at some point, it’s going to go wrong no matter who’s technology you’re using, it will go wrong.

Michael: I’ve been on both sides of that product equation, Tony, yes. [laughs]

Tony: Always goes wrong somewhere down the line. It’s not a question of whether it goes wrong or not, the question is, how do they handle it?

Michael: Okay. Well said. Very good.

Tony: They’ve got to support you, you’ve got a mechanism for making things right, you’ve got a willingness to make things right. It’s, “Oh, well, If you want that, it’s going to cost you a huge amount of money. Sorry, we haven’t got anybody available at the moment,” with excuses or the guy who sold it to you, he just doesn’t return your calls or whatever it might be. What you want is an organization that you’ve got to look them in the eye and say, “Well, the chips are down and I’ve got a problem. I still have a business to run here. Are you going to be there to look after me?”

Now, you don’t ask them that question. You ask their customers that question. Go out and speak to all the people who are already working with them. What happened when it went wrong for you? Did they respond well? What happened when you needs to make a change or you needed some help or the system went down or whatever it might be? How did they respond? What was the kind of reaction you got? Willingness to help is the key aspect of being in software. You can have great software but when your software fails, and it does always, then you need to make sure that you respond correctly and get your customers, make them whole and allow them to trade their business]. As we said right at the beginning, technology is a tool to support the business. It is not the business.

Michael: A software company is not just an engineering department or a product department. [laughs]

Tony: Really, absolutely. It’s to serve the clients that they work with. That’s very much the manager we have.

Michael: I have one last question. I realize that I’ve said this before. [laughs] One last question, a piece of advice for the retail broker, the agency principal. They do need to keep their finger on the pulse of this rapidly changing industry. Hopefully, if they’re listeners to this podcast, they get some value in regards to this question. Other than that, how would you suggest [clears throat] an agent today keeps their finger on the pulse of the emerging technologies, existing technologies in such that they can make intelligent decisions about it?

Tony: I think there are a lot of people have a lot of opinions in our industry. Some of them they voice very firmly and very loud. The quality of the information shouldn’t be judged on how opinionated and loud those people are.

Michael: [laughs] Okay.

Tony: I think it’s important when you’re a principal within a business like that to be able to find people who you can listen to whether it be your podcasts or somebody who works in the industry that they can ask questions on. It’s what we call influences. Those are genuinely people who know what they’re talking about as opposed to they just want to be fantastic on social media and want to have lots of likes. People who know what they’re talking about, been there done it, who’ll give you an honest answer about things. If anybody tells you everything will be wonderful, be very wary. Don’t go for those people. Go for people who genuinely will tell you warts and all what the situation is.

Michael: Tony, when we started this conversation, remind me, are you in London?

Tony: I’m just outside London, just west of London.

Michael: West of London, when we started, behind you, I was looking at a beautiful late afternoon sky. It gradually turned into a sort of a lovely purple hue. Now it is pitch black and no doubt you have other things to do with the balance of your lights or–

Tony: My suspendable lighting is just about to go again.

Michael: I can see the lights reflecting on the window back into your office. I think you’re the only one left as far as I can tell. You’ve been very generous with your time. I really appreciate the conversation. I will be quite certain that this conversation gets broad distribution both among my regular audience. I’ll share this with other thought leaders in the industry. I think this is a valuable conversation.

Tony: I’ve thoroughly enjoyed it. It’s nice speaking to somebody who understands.

Michael: And a fellow co-author, our first time. Well, this isn’t the first time we’ve met, but the first time we’ve had this wonderful conversation. If somebody does have a question for you, I don’t think you’ll be deluged, but if you are available somehow– Your lights just went out. It is absolutely pitch. How can people reach out to you if needed?

Tony: I’m on LinkedIn. I don’t think there are too many Tony Tarquinis on LinkedIn. Feel free to reach out to me on there.

Michael: All right. Obviously, the spelling of your name will be in the promo in the show notes. It was a pleasure. Look forward to our next conversation.

Tony: Michael, it’s been a pleasure. Thank you very much indeed.

Michael: You’ve been very generous. You bet. Thank you so much.