[00:00:00.00] [MUSIC PLAYING]
[00:00:00.15] SPEAKER 1: But if there's one thread that's gone through this whole history of the industry, it's going to be what I'll call the permutation and combination problem of carriers, lines of business, and vendors in company unique data. How can we get to a point where you can get scale, trying to handle all these unique extensions to the basic data for each of the lines of business? And that's a problem that's carried right up through this day.
[00:00:25.92] [MUSIC PLAYING]
[00:00:28.62] SPEAKER 2: This is the Insurance Technology podcast, where we bring interesting people from across the insurance ecosystem to discuss and debate technology's impact on the industry. Join us each episode for insights and best practices from industry stewards and tomorrow's innovators. Now, here's your host, Reid Hallsworth.
[00:00:46.08] [MUSIC PLAYING]
[00:00:48.21] REID HALLSWORTH: Hello, everybody. Welcome to the Insurance Technology podcast. I'm your host Reid Hallsworth. In this episode, I'm going to be interviewing David Rowe. David Rowe is an amazing individual.
[00:00:58.41] Not only did Dennis Jackasion recommend that David speak on this podcast, but I also have known David for quite a while. I first met David a number of years ago when he sat on the advisory board for Velocity, which now has been acquired by Salesforce. David also sat as a CEO of AMS which became Vertafore for 17 years. David has done a lot in the industry.
[00:01:23.78] In this episode, we go back to when interface was born, as well as the birth of AL3. We're going deep here guys. And how a consortium of seven carriers built AMS, which became Vertafore.
[00:01:42.11] And how the PC, literally how the PC gave rise to another industry mainstay, Applied Systems that we all know today. Really great episode. Stay tuned. You're really going to enjoy it. Welcome, David.
[00:01:59.95] DAVID ROWE: Well, thanks Reid. So great to be with you again and always fun to explore the things in the industry with you.
[00:02:07.03] REID HALLSWORTH: [CHUCKLES] So for the listeners, David, if you could start by just telling people how you got into the industry, initially. Everybody has a story around this.
[00:02:18.34] DAVID ROWE: Well, this is really a history-type story. It was in 1978 when I had left Texas Instruments and at TI, I had been involved with data exchange type systems. And I went to work for Commercial Union Insurance Company.
[00:02:31.81] And this was a time when all the insurance carriers were really trying to automate their relationships with agents. But the way they did it in the '70s was to extend the company terminals into the agent's office, which gave the agents CSI's the problem of having to deal with a different type of terminal for every one of the carriers they had to support.
[00:02:53.20] Maybe they averaged seven or eight different kinds of terminals in their office. That enough that was one kind of a problem. But at the same time, the agents started to use minicomputers to automate their own policy information, their own customer information for their own prospecting and servicing needs.
[00:03:10.39] Which now created the problem of not only do I have to go from one company terminal to another, but I need that same data stored in my own database and my own minicomputer system at the same time. And that was a real frustration.
[00:03:23.86] When I joined a Commercial Union, another coincidence was that their CEO was a former agent, Lawson Swearingen, and he listened to agents here about this issue and he said, you know, we're going to do something different. We're not going to extend our company capabilities into the agent's office with terminals.
[00:03:40.00] We'll use the agent's own system as a source to do that and create a single-entry multi-company, that's where the word SEMCI came about, capability, so that agents could, with one terminal, go to many different companies with their transactions.
[00:03:55.84] And so my kind of focus was to not work on the inside of the insurance company, but to work outward with the vendors, which led me to meet a guy who I'll talk about a little bit later, was one of the great contributors to the agency automation arena, Rob Thompson from Redshaw.
[00:04:13.24] And Rob had a very strong thesis, and that is that you could build a policy database inside an agency. For years before, this was just accounting and billing-type functions that an agency may have.
[00:04:28.60] But also his thesis was that you can build this function with a common policy database that go across all the lines of insurance, whether it's personal lines or commercial lines. One policy database could handle that in some way.
[00:04:44.14] And that became so important. We actually created the first agency company interface transmission, that's where the word interface was born with Rob and Commercial Union at the time, such so that when the project was done, many of the carriers started to come to Commercial Union saying, how do we do this?
[00:05:04.33] But Rob and Commercial Union, to their honor said, this needs to be not just Commercial Union's or Redshaw's. We need to give this to the industry. And that really became, it was contributed to what was then the IIR, the Institute for Insurance Research, which became the basis for the ACORD standards.
[00:05:22.39] If you look at the ACORD standards, the original AL3, inside the wreckage structure there, you will find Rob's wing MVP record file format for his actual minicomputer is actually mimic in the standards.
[00:05:37.82] So that's where the standards were born. That's where the word agency interface was born through that kind of experience. And that's kind of kept me in this industry for the next 43 years.
[00:05:47.68] REID HALLSWORTH: So going back to that, so are you saying that when you were at Commercial Union, that's really when SEMCI came about. The first thesis of it. Right?
[00:05:57.34] DAVID ROWE: The first agency company interface was done with Redshaw and Commercial Union with some agents in Georgia using personal lines with this file structure, this exchange structure which had a header record, had a basic policy record, had a name and address record, all created by Rob that was used.
[00:06:20.50] And that's where the word interface was coined. But the major thrust of that was Commercial Union, because of Lawson Swearingen care about the distribution channel and the agents, never wanted to make this a proprietary system for them. He offered this to the industry. And that's where AL3 kind of came.
[00:06:39.76] REID HALLSWORTH: Wow, that's amazing. Wow. So the birth of AL3, there it is. So then what happened after that? So you're commercially in doing that, what was next in your career?
[00:06:53.51] DAVID ROWE: Well, what happened after that is all in the industry, this became such a hot topic that other carriers wanted to do interface, but they started to become frustrated that the agency vendors couldn't keep pace with them because you had so many lines of business, so many unique underwriting requirements, and they had to be--
[00:07:12.76] REID HALLSWORTH: When you say agency vendors, what do you mean by that?
[00:07:15.01] DAVID ROWE: These were vendors like today is Vertafore Applied, then it was ARC and Redshaw and Intranet, Safeco, et cetera. There were like 50 different vendors at that time.
[00:07:26.30] And so when you look at the number of carriers and the number of lines of business like 30 and the number of vendors to actually go in and program the unique requirements of each carrier for each line of business and each vendor, became an unmanageable type situation.
[00:07:41.21] So what did the carriers initially do? They bought up all the vendors. Each of the major carriers bought one vendor or another. And Commercial Union bought AMS from the policy management system.
[00:07:55.96] REID HALLSWORTH: Sorry. By them purchasing those vendors, did that mean that they would only build their products, lines of business, whatnot specific to that carrier for that specific program?
[00:08:09.55] DAVID ROWE: The way the game was played in that day is the carriers would give away the minicomputer software and hardware to the agent in exchange for a premium commitment.
[00:08:19.15] It wasn't a total lockout of other people. But the premium business was going to go to those who sponsored the computer system. Commercial Union did AMS to create an open system, a nonbiased system.
[00:08:33.43] And they were the first to create a concept of not just one company owning the system, but opening up a consortium. And that's where a consortium of eventually seven carriers with CNA and Fireman's Fund and AIG all joining Commercial Union around this AMS-type entity.
[00:08:51.49] After they acquired AMS, I did some of the development there, and when the consortium was formed, I was asked to be the CEO. And so for the next 17 years, I kind of was either the CEO or chairman of AMS.
[00:09:07.36] REID HALLSWORTH: So for the listeners, AMS became Vertafore?
[00:09:11.27] DAVID ROWE: Correct. That's after I had left.
[00:09:13.75] REID HALLSWORTH: Yeah. So wow, so you were there from the beginning then, from the birth of AL3 to really the first agency management system that was really built for the industry.
[00:09:24.88] DAVID ROWE: Well, no. Redshaw had really the first one built for the industry. I was the, at the early days, the first customer of AMS itself. All the other vendors. Yes.
[00:09:38.19] One of the questions that I thought of is, what were some of the drivers of those days that made the technologies go? And this is a topic I'll probably talk about a little bit more.
[00:09:50.72] But if there's one thread that's gone through this whole history of the industry, it's going to be what I'll call the permutation and combination problem of carriers, lines of business, and vendors in company unique data?
[00:10:06.28] How can we get to a point where you can get scale, trying to handle all these unique extensions to the basic data for each of the lines of business? And that's a problem that is carried right up through this day.
[00:10:16.60] REID HALLSWORTH: Oh, it's still a big problem. Yeah.
[00:10:18.32] DAVID ROWE: What allowed the minicomputer formation to happen was not only Rob proving that there was a policy system you could put in an agency. That never happened before. All the other vendors started out with just an invoice entry.
[00:10:32.59] But it was based on this idea that you could build the generic policy system, not one for every line of business or one for every carrier. Rob was brilliant in that. But another person who's made a huge contribution to the industry is Larry Wilson, the person who created PMSC.
[00:10:49.54] And Rob and Larry actually met each other and they actually exchanged ideas. And PMSC success was building one common policy system that was extensible for all the unique aspects of each of the carriers.
[00:11:03.76] He had a concept called GTAM, Generalized Table Access Method, which was way ahead of its time in terms of the things you hear today about configuration tools, so to speak.
[00:11:13.52] But the fact that you could have a policy system that could actually support all the lines of business, not one each, provided an efficient foundation to create the agency minicomputer business.
[00:11:28.31] So the minicomputer was born, but it was sitting on this idea that a policy system could be in the agency and it could be a generic policy system across all lines of business. So what else kind of followed on to help that? One was the consortium era. That's when other consortium started to form.
[00:11:44.74] Then as a result of that, IVANS was-- people started to say we've got to try to find some way to connect all this together, to handle all this uniqueness. Not only was the data unique, but a lot of the carrier systems were different, custom at that point.
[00:11:59.02] So IVANS was created to be a middle intermediary to kind of facilitate that. And I'll tell you, there was some great leadership at that time. When I think of some of these names, Jim Kutro from Crum & Forster who was the senior executive that really put it all on the line to give birth.
[00:12:15.91] Bob Bahram, another, no longer with us, but Bob Bahram was the Ex-IBMer who came in and became the first CEO of IVANS, that really started this.
[00:12:26.32] With a big emphasis on that V, I'm going to come back to that V later on I think. But adding value to the industry process. Not just doing pure transparent connection, but adding value was always a key thing. And it was needed because of the differences of all these vendors and the differences of all these carrier systems.
[00:12:43.67] Another thing that helped accelerate this was 1983, was the birth of the PC. And that's where applied systems jumped into the marketplace. All the other vendors are having to drag the weight of the minicomputer, older technology. And Applied was very agile and coming in.
[00:12:58.90] Their first system was based on a PC. Never having to deal with the minicomputer. So they designed something that was efficient and fast and really took market. That's what launched them in terms of market advantage.
[00:13:11.02] And then finally, what's accelerated this is obviously the internet. When I look back though and say let's look at AMS's, what's motivated AMS, what were the factors for AMS in terms of growth?
[00:13:24.37] There were two things. One was the agents. The agents were the ones that went to the carriers and say let's do this kind of interface. Let's create this capability. Let's cooperate. Let's work with standards. Let's form IVANS.
[00:13:36.43] The agents, one of them, Wade Dunbar past, was no longer with us either, a great leader of the AMS user's group, led what he used to call Wade's brigade, a number of agents that would call on all the CEOs of all the carriers saying, get on board, do the interface, et cetera.
[00:13:54.85] And I can't in any conversation about what were the drivers for AMS for instance, without Dennis Trecassien at CNA. He was the majority owner during this whole time.
[00:14:07.18] Not only did he have the commitment to think long term to do the right thing for the distribution system, keeping it open, but he was a great visionary in terms of adding other value saying, David, why don't we get into rating? Why don't we add this? Why don't we do this?
[00:14:21.52] So that combination of people and some technology changes. But fundamentally, starting the whole thing was this kind of platform that was a generic policy system, which really started playing into some major industry issues as well, which I'm sure we'll get to.
[00:14:38.00] REID HALLSWORTH: So all that said, why have we still not automated?
[00:14:44.01] DAVID ROWE: OK. Well, first of all, we laid a good foundation by having a generic policy system, as I just described. When this all started, there were the two organizations-- ACORD and IIR.
[00:14:59.58] And ACORD was an organization oriented around forms. That were 30 ACORD forms. So people from ACORD rightfully said, why don't we create a standard for each of the lines of business? Which means you are going to have to do renewal processing and data processing 30 different times and 30 different ways.
[00:15:18.67] It was Rob and Larry with this generic approach who said no, let's stop. Let's use a hierarchy that there's a subset of this industry data, the name and address, the basic policy field, the billing field, the additional interest field that crosses all the 30 lines of business.
[00:15:35.40] When we get down to the exposure level and the coverage level, that's where you need to be able to kind of have an extension to support this idea of company data, which is going to play into this whole story.
[00:15:48.33] For a while, there were debates which one do we use? But finally people would go to the various industry meetings with buttons, it's the hierarchy stupid, and they finally supported what was really Larry and Rob's generic hierarchy across lines of business.
[00:16:03.93] But then became the problem OK, let's accept that. But how do we get rid of company data? So for several years, the industry started thinking of what's called high church standards.
[00:16:15.18] And what high church standards meant at that time was to legislate away company data. It never happened. There was no way to do it. Everyone needed a different set of extension information for every exposure or coverage from one to the other.
[00:16:30.45] And so trying to find a way to handle transaction processing, connectivity, whatever way you want to call it that had the scale of common information across lines of business that was rooted in Larry and Rob's thinking, but at the same time, handled company uniqueness, really has kind of put us on the path that it is.
[00:16:51.37] So you say, why haven't we been able to do more interface, et cetera? The AL standard itself, because it's based on this generic model and it's been moved into XML and even into APIs, is still good because it handles that hierarchy. But how do you handle the company unique information is really been the major challenge.
[00:17:14.97] It doesn't seem to be a challenge on a download. And the reason is the agency system is not the system of record and doesn't need to maintain all the unique extensions of the underwriting unique to any one company.
[00:17:28.45] REID HALLSWORTH: It's a point.
[00:17:29.37] DAVID ROWE: On Upload, you have to collect all that information to be able to process it. And that's where you run into this problem of how do you collect the information in a efficient scalable way? So what you're seeing in the industry is you're seen download work very well in structured areas like personal lines and small commercial.
[00:17:48.45] When you get out into the larger line, separate from the line of business issue, you get into needing other kinds of information like risk management surveys and photos, et cetera, and you don't have a once and done or straight through processing SEMCI kind of a model, you have more of a document exchanged between the broker and the underwriter in that kind of situation, for very, very large lines.
[00:18:10.86] So for that reason, interface hasn't penetrated the large commercial lines because the model doesn't handle all these, no model could have a handle the supplemental documents you would have to do the submission on. But then the question became, for the upload capability, how can we get scale?
[00:18:30.66] I remember my days at Commercial Union, we would have to contract one vendor for one state to go add distance from the fire hydrant to a homeowner's policy, but do it again with Redshaw, again with ARC, again with Intranet.
[00:18:44.19] And then go from another state, there's another underwriting criteria, do it again. And that's just one line of business. It was not scalable. It was done programmatically inside each of the vendor systems.
[00:18:56.52] So through time, and this is something that could still be an opportunity today, I don't think a lot of people realize this, but Applied and AMS came together and created a joint venture called the Alliance for Productive Technology.
[00:19:11.13] And the idea of that, and we brought all the other vendors together to cooperate on this, including the Big I and IVANS and everything.
[00:19:19.44] REID HALLSWORTH: When was this? Like what time?
[00:19:21.30] DAVID ROWE: It was in the late 80s. Here was the idea. The idea was to say the carriers would post to, let's call it the cloud or to a central repository, all their unique extensions. And they would add it and tag it to the core data model under security.
[00:19:42.57] Now instead of each of the agency vendors implementing through custom programming those unique extensions, let's call the distance to the fire hydrant is probably a bad example, but rather than doing that, they would go and when they want to supplement their releases out to the agency, they would go to this repository and draw from that uniquely.
[00:20:06.31] And if Redshaw wanted to prompt the distance from fire hydrant at the end of a screen in beautiful Redshaw red, but AMS wanted to prompt the distance for the fire hydrant in the middle of the screen in beautiful AMS blue, that was all stylistically what the vendor had the right to do.
[00:20:21.82] But they didn't have to program the distance to the fire hydrant with programs, they just had to get the parameters of the table entries that were posted by the carriers to do that. It was to try to break through this company unique type situation.
[00:20:35.05] Unfortunately for a number of reasons that we could go into long on this interview, it fell apart based on some of the competitive natures of the agency vendors at that time.
[00:20:45.55] And eventually, some of the software was contributed to IVANS. But after a couple of years of trying to do this, the cooperation broke down. And as a result of that, when the internet now got born after that, carriers didn't have that as an option. And they went right to putting a portal in front of their agency system.
[00:21:08.02] REID HALLSWORTH: Wow, and that's when the big everybody started, we're just super investing, investing billions if not, tens and hundreds of millions if not billions on their portals.
[00:21:20.42] DAVID ROWE: And when you think of it--
[00:21:21.43] REID HALLSWORTH: Which is actually kind of going backwards if you think about it, because you're just going into the terminal again with all this.
[00:21:27.25] DAVID ROWE: I was just going to say that that's exactly a virtual version of having a unique term.
[00:21:30.97] REID HALLSWORTH: Yeah, you're right. Wow.
[00:21:32.81] DAVID ROWE: And so that's what happened. And so even today with different systems that do real-time interfaces into those portals, they can create even a multi-comparative rating capability, you can create a similar look and feel across multiple carriers for a given line of business and it will take you through 80% to 90% of the entry in one common way across all the carriers.
[00:21:55.81] But to get the actual bendable quote, you have to drop off into the carrier system and actually fill in the remaining unique underwriting fields to actually complete it. People have tried to mask that, but it's still a problem.
[00:22:09.79] So this area of company unique extensions have always been a challenge. But if you had never stopped and done the hierarchy to begin with, if you hadn't done what Rob and Larry had done, putting a thesis forward for a generic policy system, you only would not have given birth to the marketplace in the agency business.
[00:22:31.45] The next generation of policy systems that are going out there, all have been built on a quasi interpretation of the ACORD hierarchy. And so you wouldn't have been able to even do that. So this had a tremendous impact on the whole automation industry, literally starting back in the early '80s.
[00:22:51.81] REID HALLSWORTH: Wow, that was awesome. What a great perspective from someone that was there from the beginning, literally at the birth of AL3. It was also interesting to hear how the industry has really relied on forms then and now and the effect that it's had on standardization across our industry. Hmm, something to solve. Right?
[00:23:21.24] In the next episode, we're going to ask David, what is today's biggest tech influence? And where does David see the industry moving in the future? Really interesting stuff. Stay tuned. It's going to be a great episode.
**
[MUSIC PLAYING] SPEAKER: The Insurance Technology Podcast is a production of IVANS. Visit insuretechpod.com to contact us, suggest a topic or guest for an upcoming show, and subscribe to be notified when our latest podcast is available. You can find all our episodes in your favorite podcast app. It's where you can also leave us a rating and a review that helps other people find the show. Thanks for listening.
[MUSIC PLAYING]