When I built a new house a few years back, I knew enough to look to the future needs of a growing family. In some cases--such as roughing in the concrete basement floor for future plumbing--I did a good job of planning. But in other areas, I didn't. Specifically, I put too much faith in predictions of the immediate future of the wireless home, eschewing running spare network cables to rooms likely to need them.
Unfortunately, the realities of advertised vs. actual wireless network capabilities, as well as the perils of multistory HVAC interference, soon became apparent. As a result, when I turned the unfinished basement into a media room a few years later, I couldn't connect to the wireless router located two floors higher. It took many hours of additional work and a handful of technological work-arounds to reach a solution that's still not quite as good as it could have been had I just done a better job of future-proofing my design.
When it comes to future-proofing their own investments in technology, insurers focus on two related objectives. First, systems should resist obsolescence for as long as possible and remain adequate to handle future needs of the business.
"Quite frankly my biggest worry is latching on to technology that becomes unpopular in the industry," says John Kellington, chief technology officer and senior vice president at Ohio Casualty Group. He frets if the company chooses the wrong technology direction. "Pretty soon [even] an architecturally sound, well-performing system will become unattractive when recruiting talent," he says.
Second, knowing all investments eventually will become obsolete, carriers should design their systems and environment so that making updates or wholesale changes creates as little disruption as possible. Complicating the achievement of those goals is the fact they are constantly moving targets.
"What we did five years ago supersedes what we did 10 years ago and will be superseded by what happens five years from now," says Mark Hayward, CTO at BusinessEdge Solutions. "You'll never truly future-proof anything, but you can make some sensible bets based on the best percentage of odds of something being sustainable over time."
What are the best bets for insurers? Read on to find out.
It's All About the Architecture
While evolving security threats, scalability and capacity, and infrastructure issues are all valid future-proofing targets, it's critical carriers safeguard the continued viability of the applications that support their business processes.
"Infrastructure is a throwaway. You just keep changing it, but the applications you have must keep going," argues Mike Kirrene, senior vice president and CIO at Republic Indemnity. "The real question is: How do you keep moving into the future with a truckload of the past?"
"Your choices are to predict your needs and build systems now to accommodate them or to build flexibility into your systems to anticipate change," says Jeff Chookaszian, principal at McKinsey & Company, a management consulting firm. "Since you can't do the first, you have to do the latter."
Hayward characterizes a flexible design as an architectural layer cake: a data layer at the bottom, a service layer next, a process layer above that, and at the top, the interface or presentation layer. "All those layers have [future-proofing] challenges, but the one that is most variable and changes most frequently over time has been the interface layer," he says. "In the past, carriers have been stuck with a particular type of application interface and [therefore] not supporting changing business processes or the way people want to connect with applications."
"That [layering] concept has been 'Architecture 101' since the 1990s," Kellington opines. So, why are we still talking about it as a future-proofing strategy insurers should adopt? Because they still haven't learned it.
"The design tools and practices are out there, but do you have the skills and the willingness to use them?" Hayward asks. "Oftentimes people don't take advantage of them [tools and best practices in design] because they're pressured by the time frame or scope of the project or simply the unwillingness to deal with things they don't know or aren't comfortable with. It's amazing to me the need to learn and reinforce basic blocking and tackling of system design hasn't gone away."
In a recent data conversion project at Northwestern Mutual, for instance, it would have been easier for the carrier to use just its existing technical access to the data to complete the first project phase. But instead, it took the time to build a services-oriented access layer to facilitate future conversions and changes and to remove future business applications from having to understand the physical layout of the data.
"This allows us to move the data, which has a longer life cycle, from the applications that change more often," explains Bob Kowalsky, Northwestern Mutual's vice president and chief architect. "Now, the data access layer can maintain stability while allowing us to change both applications and infrastructure [components] more readily, and it provides an increase in data quality."
The Next Disruptive Technology
"I have a 23-year-old son who's an agent, and I love talking to him about how he does his job and the things he's frustrated with," quips Doug Allen, director of IT at State Auto. Those frustrations deal with not being able to go from client to client with the full, connected capabilities of in-office systems.
"The same goes for our claims people," Allen adds. "You hear stories about people finishing adjusting a claim and then finding out they had another claim to adjust within five minutes of where they just were. We need to make those workers more agile via better mobility, so they don't need to retrace where they were, and that mobility needs to be taken into the integration layer we design."
That mobility hinges on ubiquitous broadband wireless, and carriers should be preparing for just that. Chad Hersh, senior analyst at Celent, claims insurers in general were slow to the Web-enablement game, but those that don't commit to Web-enable all critical business applications do so at their peril.
"Carriers always seem to find an excuse not to extend business applications to the field," he says. "First, the issue was laptops were too expensive, then it was agents didn't want to dial up an Internet connection when they were in somebody's home, or wireless modems were slow, or you could never find an Ethernet jack. Today, carriers do have more apps Web-enabled for producers and field staff, but they're still focused on using those apps in connected mode only while [those staff are] in the office and in offline mode on the road because, outside of individual hotspots in cities, it's very difficult to get a decent connection anywhere else."
That's poised to change--and sooner rather than later. It's difficult to find a laptop without wireless support already built in today, and more cities are announcing plans to deploy wireless mesh networks for communitywide Wi-Fi access. (For what it's worth, Macedonia recently announced plans to blanket the whole country--all 1,000 square miles of it.)
Equally important are other wireless protocols, such as WiMAX (world interoperability for microwave access), which offers the same connection speeds as Wi-Fi but claims a range of up to 30 miles. The WiMAX Forum announced in November that 150 fixed WiMAX networks have been deployed, and an 8,000-square-mile WiMAX network was announced in Canada. Mobile WiMAX is scheduled for availability in late 2006, and Intel has thrown its support behind the technology, as well.
"WiMAX is poised to become the next 'disruptive' [wireless] technology," Hersh believes. "With WiMAX, you have truly blanket coverage of an entire area. An agent can go out to a site, submit an application, and bind coverage on the spot. An adjuster could go out to inspect an industrial site and adjudicate the claim without having to rely on offline systems or cell modems. The efficiency gains are tremendous."
So, for the always cautious insurance industry looking for proven viability and future assurance before investing in technology, now is the time. "Carriers haven't adequately prepared for what this [wireless ubiquity] will mean, in terms of competitive advantage, retention, efficiency, and the ability to quote and even bind policies that require some sort of on-site inspection," Hersh asserts. At a minimum, he recommends Web-enabling any remaining applications for the field and piloting new wireless applications by using existing Wi-Fi hotspots to work out security and connectivity issues.
"To be able to be the first carrier out there in whatever areas get WiMAX first is key," predicts Hersh.
Commit to R&D
Sure they're important, but research and experimentation usually are treated as "nice to haves"--adjunct activities to all the tactical issues that fill up the daily calendars of IT staff at most carriers. Insurers should commit staff--or at least some dedicated staff time--purely to researching newer technologies.
"Companies talk about how much training they provide, but letting staff members work on what's interesting to them can have more of a future-proofing impact than training can," maintains Hersh. "There's a double benefit. One, you'll foster innovation. And two, you attract a better caliber IT staff."
At Ohio Casualty, R&D is a delineated responsibility of the architectural team, which Kellington says focuses not just on keeping software well designed and utilizing organizational standards but on investigating newer technologies. "It's ingrained in the development organization," he says.
Also, insurers should commit to evaluating technologies--and technology vendors--objectively from a future viability standpoint. "Part of the [future] risk is we purchase systems from companies in the business of selling technology, and we have no control over the vendors or that marketplace," says Northwestern Mutual's Kowalsky. "If somebody goes out of business, we have to consider how that impacts us. Particularly with all the consolidation in the vendor community that's going on today, you need to have a clear vendor relationship and management strategy."
Finally, "keep scanning the vendor environment for more integrated solutions," recommends Steve Discher, vice president at Robert E. Nolan Company. "The insurance industry doesn't have an ERP [enterprise resource planning] equivalent, such as SAP [does] for manufacturing or the energy sectors, yet. Time will tell whether this transpires for the insurance sector."
Take Time to Tinker
In addition to research, there are some specific technologies insurers should be making more use of--or at least experimenting with--today to avoid falling behind the industry in the near future. The first is open source. "Combined with service-based architecture, open source is tremendously important for future-proofing," says Nathaniel Palmer, chief analyst at Delphi Group. "Probably the most important aspect of this [strategy] is the adopter is not beholden to a certain vendor."
Of course, just because a technology is open source doesn't mean it's future-proofed or even well documented, "but in general, open source, due to its transparency, component orientation, and having peer review, presents a very powerful case for its use," Palmer argues.
"Open-source software tends to evolve constantly," Hersh adds. "They continually look for ways to improve it, secure it, make it bigger, better, faster. That's a change from when insurers bought software from vendors that didn't update the software and/or went out of business."
There's a good chance insurers already are using open source somewhere in their company whether or not they know it, Hersh points out. "Most carriers know whether or not they're using Linux, but they might not know deep in the bowels of IT they're using Apache Web server, because commercial vendors they're dealing with are using it in their stack," he explains. "Now, we won't see policy administration systems written in open source, but I think we'll see those administration systems run on open-source operating systems that are based in languages that are open source themselves, such as PHP, or that run on open-source databases, such as MySQL."
Ohio Casualty has Linux deployed as an application server and also utilizes Apache Web server. Northwestern Mutual has some deployments of Linux and Apache and uses the Eclipse open-source framework for a handful of Internet applications. "We're very pragmatic in our use of open source," Kowalsky says. "We're concerned with some of the licensing issues, but we're looking at all of it for opportunity."
The second technology insurers should be tinkering with now is grid computing. "With grid, you're future-proofing your infrastructure because you're creating a massive amount of computing power held together by what probably will be open standards in the future," Hersh reports. While today there is a limited set of applications carriers can run, as well as constraints on grid that prevent it from being used throughout the enterprise, such as input/output lag times and bandwidth issues, those limitations will diminish.
"Internal bandwidth is getting increasingly fast to the point where installing a one-gigabyte intranet won't be a big deal much longer," says Hersh. "Disk drives are faster, processors on PCs are faster, and over time the grid will become more and more like a giant mainframe that will process smaller transactions effectively. Getting used to grid concepts and standards now is critical if companies are to take advantage of that down the road without falling behind."
While there have been a few published reports of larger grid deployments in insurance, most of the grid developments in the industry have been experimental or limited in scope. Kowalsky, for instance, reports Northwestern Mutual currently has a small grid running in actuarial that brings the PCs in that department together to run financial modeling, which cuts processing time on running the model from days to hours. "We constantly are looking for new opportunities for grid," he says.
Back to the Future
What these different strategies all have in common is an emphasis on flexibility. "The key to future-proofing is not a specific technology. In fact, it's not a specific anything," says Hersh. "It's the idea whatever you build should be able to change as the business and IT worlds change around it."
While no system ever can be completely future-proofed, Republic Indemnity's Kirrene says much progress has been made toward achieving this flexibility and, in turn, toward extending the lives of IT investments.
"The infrastructure technology developments over the past 10 years have allowed us to improve continuously the utility of our aging applications systems by improving security, connectivity, responsiveness, and accessibility. Continuously improved Web tools and technologies have enabled the enterprise's reach to external customers and business partners without requiring a reengineering of the underlying business applications," he says. "Bottom line: The 'futuring' of these technologies has improved dramatically."
