Someone sent me a column by computer-industry consultant Kathleen Dollard about the pace of change. It started me thinking that there are two great forces that are going to impact our industry. The first is the rate of change in technology. The second is more subtle but just as important — the rate of retirements in our industry. Over the next ten years, our industry will look very different.

Futurist Ray Kurzweil postulates that the rate of change over the next 100 years will be more like 20,000 years of change. We can look at the rapid-version histories of Java and the .NET Framework as examples of the rate of change. New technologies are going to burst on the scene, and we'll need to embrace them. Sometimes external events like September 11 force us to embrace change. In the old days, we all knew the guy who knew every line of code in a particular system written in COBOL or RPGIII. That guy is gone, replaced by teams using several different languages.

If we look at the population changes of the professionals, we can expect that over the next 20 years, a generation will be retiring. Those of us who have been around awhile all know someone who has retired or will retire soon. Companies are starting to appreciate the issue. They are changing policies to encourage people to work longer.

To some extent, the fact that we serve a conservative client base will mediate the need to adopt new technologies. We as professionals should recognize that we'll need to learn new and varied technologies. Our clients will be looking for companies that have experts and can retain them.

For a number of years, the great processor race has been for speed. The speed race is finished. Now the race is to see how many processors can be fit on one chip. The currently available chips can have from two to four processor cores on one chip. A research chip exists that has 80 cores on one chip. If you are using servers with four processor cores, the operating system will handle spreading out the tasks, but you're not really taking advantage of the processing power that exists. We're still writing software for one process at a time. To take advantage, we'll need to reengineer our policy administration systems and adopt parallel programming techniques to move forward.

We're handling policies with multiple risks that are presently being processed single file. One complicated risk will slow down the processing of the entire policy. It seems natural to divide the processing of the risks out and run them simultaneously. Most developers of policy administration systems are not trained extensively on parallel processing; at most, it's one assignment in college. The languages used to develop the policy administration system will need to adapt as well. XML processing would be improved if you could process nodes concurrently.

The entire development process will need to change to meet the coming challenges. We'll need new methodologies to debug our applications. The tools and languages we use will need to change, or we'll have to adopt new tools and languages. We can't just sit on the sidelines and do nothing. In the very near future, we'll have customers that have servers with at least 64 cores running. We had better be able to explain to them how we're taking advantage of the processing power.

A recent article detailing the launch of a new satellite for Google Earth started me thinking about how much the imaging industry has changed. The new satellite will have an accuracy of about 10 to 25 feet. If you checked your house using one of the public mapping services several months ago, you probably saw a photo that was years out of date. A fresh check of the satellite services for my house showed a photo probably taken within the last couple of months. More that just a voyeuristic pursuit, the satellite services have been used recently to look for Steve Fossett.

State and local governments have long been using these services. One of the uses is zoning enforcement. Geographical information systems are fed information about zoning. It is matched with satellite maps. New images are used to determine if any encroachments or new construction has taken place. Some counties use these systems for property tax enforcement, checking for new additions to existing property.

Searching for how satellite information has been used by the insurance industry does not yield much. I was able to find how one company used satellite information in a fraud case. We have to realize that satellite images will be more accurate and current in the future. One day, someone taking an insurance application will view a satellite image on screen after inputting the address. Using that image, basic information could be associated with policy information. Satellite image information could be used to determine new construction just as counties are doing now. Of course, you will no longer have to ask "how far do you drive to work?" You just enter the work address, and a map with distance information will be displayed. It will be interesting to see how policy administration vendors utilize this new source of information.

-Howard E. Kennedy
www.iso.com

Recently, taxi drivers in New York and Philadelphia held a one-day strike to protest the installation of GPS systems in their taxis. They claim that the cost of the systems is excessive, but it has more to do with tracking the location of taxis. Soon, passengers will be able to view their progress on a screen. This is part of a larger trend to include GPS systems in all vehicles. For insurance companies, it's another opportunity to gather information about mileage and driving habits, also a growing trend.

GPS tracking systems have been part of commercial trucking fleets for years. They are now being installed in commercial auto fleets and other commercial vehicles. The information is used to record and analyze driving behavior. It is also used to assist in the recovery of stolen vehicles. This will help to calculate more accurately the risk for a particular commercial fleet.

School districts are also part of the trend. No parent wants to hear from the school district that the bus driver is lost and they have no idea where the bus is. School districts will know if a bus is stopped for an excessive amount of time and will be able to dispatch help. These systems also have the ability to define approved boundaries for a particular bus and create an alert if a bus deviates from its approved route.

The trend includes personal vehicles with such systems as OnStar® or other GPS options in higher-end vehicles. The availability of the systems will only increase, becoming standard equipment in all personal vehicles. Some insurance providers are already offering discounts to drivers that provide data from their systems. A few states are mandating a discount for vehicles that have GPS tracking systems, a trend that is sure to increase in the future.

For policy administration system providers, this presents both opportunities and challenges. The first opportunity is to manage this information and provide it to carriers as they need it. The second opportunity is to model the data to show trends. One of the challenges is to quarantine data when a vehicle changes ownership. Another challenge is that as GPS tracking moves from the commercial realm to the personal, privacy concerns will need to be addressed. Finally, we will have to adapt and respond to any regulatory changes. Navigating these waters can be tricky but will surely provide added value to insurers from policy administration system vendors.

By Howard E. Kennedy
www.iso.com

In my previous post, G-Man commented that SOA was a sorely overused acronym. This may be true.

Almost all the RFPs going out have at least one question regarding support for SOA. Most, if not all, policy administration system vendors list SOA compliance as a feature of their systems. Both carriers and policy administration vendors are locked in this dance, a dance that we've done before.

SOA is the latest architecture to hit the scene. The benefits being touted are the same that were promoted with past architectures. Having gone through the millenium revolution, I'm seeing the promises being made again. With '00 we had the opportunity to fail at a small scale; SOA lets us fail at a large scale. I seriously doubt SOA will improve the ability to do an out-of-sequence endorsement. It will allow information to be interchanged among different systems. It's still too early in the adoption cycle to see if SOA will live up to the advertised benefits.

By Howard E. Kennedy
www.iso.com

Two stories that hit the news lately were the leaking of the last book in the Harry Potter series, Deathly Hallows, and some of the fall TV pilots on the Web. Listening to the experts discuss the leak of the Harry Potter book, they all agree that it was someone who had access.

Experts are unsure as to whether the leak of the fall pilots was a publicity stunt or internal leak. In the P&C industry, we do not have access to multi-million dollar blockbusters but we do manage sensitive information within our policy administration systems.

In this industry, we do a good job of creating security within our applications. But what are we doing about securing the systems themselves? How are we helping our clients to secure their systems? At a security conference that I attended, one of the presenters told of a client that glued all the USB ports on all company computers to prevent the loss of information. This is an extreme solution to the problem.

There is no easy answer to the problem. If you use a cryptographic file system, that will prevent unauthorized users from gaining access to the files. Authorized users can still view the information and the files will be automatically de-crypted when copied. It appears as though new operating system releases will help provide means to better restrict access to the system. There are some third party tools out there to manage file transfers but there is no silver bullet to solve this problem.

- By Howard Kennedy
ISO Rating Service Technical Liaison/Technical Architect
ISO-ITS

The other day I received an e-mail invite to a webinar, "SOA & the Insurance Industry: Voice of the Experts." The event was organized by InfoWorld, and included speakers from the TowerGroup, BEA, and Zurich Financial Services. Naturally, my interest was piqued.

I watched the webinar, and there was a lot of good information. I noticed there was someone missing from the discussion. I have been working in the financial services industry for twenty-two years, and the biggest challenge has been involving business users. We do not hear enough from the business user.

If you do a search for SOA and insurance on the web, you will find plenty of articles and presentations by information technology people. You will be hard-pressed to find anything by the business user. A policy administration system is the heart of an insurance company. It is much more than connecting some services or being a presence on the web. It is the embodiment of a set of business rules and processes.

What makes a successful SOA project is the connection between business users and information technology users, not the way the services are connected.

- By Howard E. Kennedy
ISO Rating Service Technical Liaison/Technical Architect
ISO-ITS

It may be true that insurance is sold, not bought, but the real foundation of the business is the policy -- the contract that represents what often is a lifelong relationship between customer and carrier. The terms, pricing and administration of the policy reflect not only the needs and financial circumstances of the policyholder, but also the performance and compensation of an agent or distributor; the capabilities of the insurer, in terms of underwriting, product development, billing and customer service; and the regulatory environment. No wonder there’s so much debate about when, how and why to replace legacy policy admin systems and migrate to new systems. The stakes are high, and the proverbial risks and rewards are huge for companies that make the right –- or wrong –- choices about solutions, vendor and integrator partners, and platforms/operating environments.

To that end, Insurance & Technology is launching the Policy Administration Resource Center, an online forum for analysis, debate and education about policy administration and the next generation of solutions and technologies that can help insurers maximize their investments in this area.

The goal is to provide insights into the product news, research and leadership that is shaping policy issuance and administration in today’s more competitive financial services environment that is characterized by higher risk, more demanding and information-sensitive customers and multi-channel distribution.

Key to the Policy Admin Resource Center will be open discussion among industry leaders –- CIOs and other technology executives, line-of-business heads, analysts, consultants and solution providers.

That’s why I’m launching my I&T blog in conjunction with the Resource Center. I’ll be weighing in regularly with my observations and questions about what’s happening in our industry, and I hope you’ll join in the conversation with your own opinions and insights. What are you seeing in the policy administration solutions segment or legacy systems upgrade space that’s impressive, infuriating, ridiculous or confusing? What needs to happen -– specifically in terms of policy admin, or more generally in terms of IT strategy or corporate culture –- for carriers to become more profitable, productive and competitive? This is your forum to “tell it like it is.”

Posted by Kathy Burger, Editorial Director

There is a strong trend among property and casualty (P&C) carriers to adopt a new breed of policy administration systems. These component-based solutions offer the opportunity to improve customer management, expand distribution channels and upgrade enterprise information management.

Life and annuity (L&A) insurers also are exploring ways to improve policy administration. Some carriers are embracing new-breed systems wholeheartedly. Others are seeking to use service-oriented architecture (SOA), business process automation and other tools to improve the performance and capability of legacy systems.

An additional issue facing those P&C and L&A companies that are exploring new policy administration solutions is whether to select an integrated solution suite for policy administration, claims and billing, or invest in leading solutions for specific functional components. The choice will depend on the company's particular circumstances.

Whether adopting a new policy administration solution, extending the functionality of legacy systems, or choosing between an integrated suite and separate components, a company can only achieve its business objectives if it has a sound implementation and deployment approach. Unfortunately, many implementation projects fail to deliver the anticipated results. Here are several pitfalls in system and process implementations that insurers can encounter as they equip themselves to meet changing business demands:

Pitfall 1: Failing to Define Project Requirements Adequately and Specify Key Transformation Objectives

Companies that want to transform policy administration systems and processes must first set clear objectives and carefully define project performance metrics for the implementation process. Failure to do so can lead to delays, expansion of project scope and even outright implementation failure.

Pitfall 2: Neglecting Customer Data Conversion Requirements

One of the most challenging issues facing insurance companies as they transform core systems and processes is improving the quality of data they have in hand. Companies are much more likely to achieve their implementation goals if they start a separate data cleanup project that proceeds in parallel with the system implementation. As part of the data cleanup, an insurer should strengthen its legacy systems sufficiently to support the edits and other improvements needed to upgrade data.

Pitfall 3: Developing New System Tunnel Vision

We have found that, during the implementation process, companies often become so focused on the new solution they are adopting that they develop a kind of tunnel vision. They neglect the importance of the ancillary systems and processes that are important to effective company performance. As discussed above, data conversion is a critical step. System integration is equally important because core systems and processes typically must be integrated with 20 to 30 other business systems when they go live.

Pitfall 4: Failing To Create a Governance Model That Solves Problems, Removes Barriers and Has Ongoing Senior Executive Support

Insurance companies sometimes underestimate the size and complexity of an implementation project. As a result, they try to adopt the new process or system without creating a truly representative project management team charged with overseeing implementation. We have found that effective implementations are guided by special governance teams that represent all of the company stakeholders.

Pitfall 5: Not Recognizing That Change Management Is More Than Training

Even with an excellent governance team in place, an insurance company is much more likely to achieve the results it wants if it recognizes two things. First, implementation is a change management process. Second, change management is more than training. Preparing the organization starts at the beginning of the project and continues through training and post-implementation support.

Pitfall 6: Not Using Prototypes and Pilots to Create Measurable Deliverables

In the traditional waterfall, or sequential, method of delivering a software solution, a team will spend months defining requirements in Word and Excel. The team then divides up the work among sub-teams, which spend months more building individual parts of the solution. But, when the sub-teams come together at the end, they may realize that the components do not work together or do not meet the actual business needs. Today's agile development methodologies are designed to provide both transparency of the development process and constant progress toward business objectives.

Pitfall 7: Failing To Define and Manage Scope

One of the most common causes of implementation failure is "scope creep." That is, a company fails to set, or more frequently fails to stick to, a well-defined and manageable scope for the implementation project. No change is free, and if the team decides to broaden the scope, it should only do so with a clear understanding by all stakeholders of the implications of the expansion for the schedule, cost and risk of the project.

Avoiding these pitfalls will help insurers address the complexities in transitioning to a more flexible, cost-effective framework that ultimately can support such capabilities as straight-through processing, automatic assignment, assisted underwriting, self service and more.

ABOUT THE AUTHOR

Paul McDonnell serves as senior vice president, managing director and Insurance segment leader for BearingPoint. He has more than 20 years of experience in the financial services industry, both domestic and international. He specializes in assembling and managing high-performance teams to assist financial services clients in transforming their business operations and realizing the benefits of IT investments.