(Editor's Note: This article is a condensation of a longer white paper available as a PDF download at www.soundingline.com.)
SEMCI and industry survival - the vision
History: Before jumping into the current situation and possible futures, it makes sense to look at a little history - where interface came from, what it was intended to do, the origin of the SEMCI concept, recalcitrant obstacles, and so on.
Multiple entry: Once agents and carriers each had their own computer systems, it became obvious that the distribution system was inherently inefficient. Agents would key data into their systems and then print application forms for their carriers. The carriers would receive the forms and then key the information into their systems. On the way back, agents would have to enter information from carrier declaration pages. Duplicate entry wasn't fun and it encouraged errors and inconsistency.
Industry survival: But more was at stake. Multi-entry inefficiency actually put the whole independent agency system at risk, or so we were told. The more efficient captive system was stealing market share each year. Interface, which is the ability for carrier and agency systems to exchange data electronically and without re-entry, was considered a survival issue. The industry needed to cooperate and use interface technology to dig itself out of a competitive hole.
Complexity: But left to its own devices, interface among multiple kinds of management systems and multiple carrier systems could become so complex it could never be accomplished - at least at a reasonable cost. With a dozen management systems and hundreds of carriers with ten or more lines of business, the thousands of management system/ carrier/LoB triplets would be unmanageable.
Standards/network: The answer would be to create standard record layouts for each LoB transaction so that each vendor and each carrier would need to do only one set of implementations. ACORD, the paper forms standards organization, merged with the Institute for Insurance Research to carry on the interface standards effort. And IVANS, a managed network, was founded to carry the interface traffic.
The argument: The industry was at risk because of duplicate entry inefficiency. Establishing interface between agency and carrier systems could eliminate duplicate entry. Interface standards, plus an industry network, would make interface practical and economical to implement. In the mid 80s it looked like a winning combination. Next problem!
New business upload becomes synchronizing download -
the reality
Start with upload: Because upload seemed like the obvious place to start, initial ACORD standard batch interface efforts focused on new business transactions.
By 1985 it was common to see predictions in the trade press about the imminent arrival of standard interface, with tables listing carrier rollout dates.
Problems: In 1985 I visited an agency in Chicago using an AMS Series 1 system to upload new personal lines transactions to The Travelers. "How much time do you save over submitting paper apps?" I asked. "Well, actually it takes longer," came the reply. "Sometimes it takes a week of re-entry and correcting a submission before the carrier mainframe will accept it." A revelation: new business upload wasn't going to be easy. Carrier systems weren't designed to accommodate it, edits were a problem, each carrier interpreted the standards differently, and the once-promising process dragged on and on.
Carrier terminals: At the same time, carriers began to put interactive terminals into agency offices. Standard upload from management systems wasn't working very well, but carrier terminals provided agents with access right into the carrier systems. The carriers could delegate key entry to the agency and thus save some money, but the agency was still confronted with the problem of duplicate entry. And, it turned out, much of IVANS revenue was coming not from standard batch interface, but from supporting carrier interactive terminals. Not a pretty picture for the agency community.
SEMCI: With standard batch interface (new business upload) sputtering, carrier terminals proliferating, and agents not quite able to see the forest for the trees, Shirley Lukens, IIAA agency automation advocate, coined the term "SEMCI" and encouraged agents, carriers, and vendors to rally to the cause. SEMCI was intended to capture the essence of the ACORD/IVANS interface vision in a simple sound-byte. Now, any particular interface system could be recognized as SEMCI (good) or non-SEMCI (bad). And it became clear that SEMCI wasn't imminent.
Perot Study: Agency frustration and an IIAA desire to help agents with interface reached a crescendo in 1990 when the IIAA hired Perot Systems to study the situation and make recommendations. Things weren't moving, someone or something was in the way, and it was time to get it fixed. But the Perot Report wasn't particularly helpful. It didn't provide much new thinking and its principal recommendation seemed to be that Perot could fix everything, given a chance - not a credible claim. SEMCI was crucial to the survival of the industry, but it wasn't happening, and it wasn't obvious what to do about it. It would be necessary to rethink the current new business upload strategy. That happened in two different ways:
1) with the formation of APT to facilitate upload, and 2) the temporary expedient of focusing on download.
APT formation: In that same 1990 watershed year, a number of vendors came together to form APT (Alliance for Productive Technology, now part of IVANS). Initially it looked as if the industry would have a single upload solution, but Applied eventually decided to continue with its own increasingly successful interface efforts, eventually creating WARP, then Transformation Station (more below).
Fixing upload: The APT founders believed that upload was unsuccessful because each carrier and vendor implemented the ACORD standards differently. What was needed wasn't just standards, but a single, sharable implementation as well. In addition, APT would solve the company edits/unique data problem - something the standards didn't address. Carriers would only have one (or two with Applied) new business upload processes to implement so the costs of doing upload would be manageable. But it was obvious that it would take time for the APT (and Applied) upload efforts to bear fruit. So, what could help the SEMCI effort in the meanwhile?
CUGL recommends download: Also in 1990, CUGL (Congress of User Group Leaders) recommended that the industry change its focus from upload to download, a much simpler process to implement and manage for the carriers. Though not the total solution sought with SEMCI, download from carrier systems into agency systems would provide real benefit. Data entered by agencies via carrier interactive terminals could find its way into agency systems without being re-keyed, agencies could populate their management systems with policy detail via download, renewals would come in automatically, and so on. Reasonable consensus was achieved and the industry adjusted its focus.
Classic interface, technology, and business strategies -
circa 1992
Classic interface context: So here's where things stood ten years ago: the original ACORD AL1 standards had been replaced by more practical AL3 standards; there was general agreement that the industry should shift focus temporarily from upload to download; APT (and Applied on its own) would create an infrastructure to make upload more practical (including ways to handle carrier edits and unique data); carrier terminals were migrating to the PC and therefore becoming less onerous; the SEMCI slogan provided an agency rallying point directed against proprietary company terminals. Everyone pretty well understood where things stood with interface, whether or not they agreed with the SEMCI goal. Agents, carriers, and vendors could make decisions and get on with their efforts. Classic interface was on track.
Classic technology/business context: It's important to note that classic interface was part and parcel of a larger technology and business context. Agents and carriers used self-sufficient, independent computer systems, each occupying its own island of computing. For the most part, interface was intended only to electronically move what had been forms data. It was not intended to change the way business was done, for instance, by putting complete point-of-sale, point-of-service capabilities out at the agency. The agency/carrier division of labor was reinforced rather than changed by technology, and most agents were focused only on conventional P&C business. Agents used comparative rating software to get quotes, with some level of integration present (or demanded) between rating and management systems.
Entrenched thinking: Interface and insurance technology concepts evolved during the 80s, becoming stable and then conventional wisdom by 1992. The momentum of the classic model continues to dominate industry thinking, even though the assumptions on which the model was based are no longer operative - at least to the extent they were in 1992. We'll look at the legacy of the classic model, the changes that make it untenable, and a new model below.
A present accounting
Present: Let's zoom forward to the present, ignoring for now Internet and industry changes, looking only to understand what's been accomplished given the classic interface agenda of 1992.
Download: Download is certainly a success. More than 200 carriers and 20,000 agencies are involved, served for the most part by Applied, AMS, ebix, SIS, and DORIS software, mostly using ACORD AL3 and IVANS. Though far from universal, download is well entrenched and has succeeded pretty well given the early 90s agenda for it. Personal lines is common, though commercial lines much less so. The process is reliable. Implementation issues and certification are well understood. Download is a success story.
Upload: New business upload is a more modest success with significantly fewer takers. For example, Applied reports about 30,000 download agency/carrier pairs but only about 8,000 upload pairs. Though promising in 1992, the APT offering never really took off, but with the IVANS purchase and rewrite (to WebSEMCI), the (once APT) multi-vendor approach to upload should now accelerate. According to Anna England, vice president of e-business solutions, IVANS has so far licensed 14 carriers for WebSEMCI. She hopes for 20 total with 7 pilots by year-end and expects to catch up with Applied over time. Because WebSEMCI is separate from the management systems that feed it, carriers who license and support it can get benefit across multiple vendors from one effort. Though the software operates locally, each time it is put to use the software checks the carrier/IVANS site for edit updates and pulls them down into the agency computer before proceeding.
Good progress: So all in all, how has the industry done with the interface agenda set ten years ago? I'm pleasantly surprised at how much has been accomplished, at how well Applied and IVANS are positioned to bring on more upload carriers and agencies, and at the possibility of cross-vendor, SPX-based editing scripts. There are plans afoot as well to go beyond traditional upload/download policy transactions to include queries to bring back loss runs, direct bill balances, and so on.
Policy change missing: Policy change, an important part of day-to-day agency life, is not, practically speaking, part of the upload/download picture today. IVANS reports one early experiment with policy change through WebSEMCI. Applied wants to work on query transactions before jumping into policy change.
Underway: Not the promised land, but not disaster either. The classic interface agenda is still underway, and, in the case of upload through WebSEMCI/IVANS, is more active than it's been in the past. So, that's one part of the story. But the world has changed significantly since the classic interface agenda was set, making interface significantly more complicated - and potentially more useful - than imagined in 1992. It's the right time to think about a new agenda.
Real-time interface and carrier-sponsored agency portals
Real-time interface and carrier Web sites: Two new developments are especially important relative to classic interface: 1) real-time transactions, and 2) carrier Web sites. Classic interface conceived of agency/carrier transactions as batch and generally overnight transactions. Real-time transactions, on the other hand, look for immediate response and perhaps even ongoing computer-to-computer conversation. ACORD XML rather than AL3 provides the vehicle for real-time transactions.
Agency portals: Carrier Web sites, or agency portals, resemble carrier proprietary interactive terminals in the sense that they provide functionality and information to agencies via carrier software. But comparatively, carrier portals via browser access seem to be much easier to learn and use than traditional terminals (even if PC-based).
Real-time struggles: Applied and several carriers have been working on and are beginning to generate real-time, multi-carrier quoting right out of the management system. Unlike comparative rating software, the returned quote comes from the carrier's own rating system. Therefore, the quotes that come back are accurate to the penny, and are generally an improvement over third party rating. Were all carriers to support real-time quoting, traditional comparative rating software might become irrelevant. On the other hand, real-time quoting hasn't yet garnered the level of commitment from carriers that Applied would like to see. It may be that real-time quoting is a lot like upload was early on - just too difficult to lead off with. Applied is considering augmenting real-time quoting with real-time query transactions, which are much simpler to implement and of great interest to agents.
POS progress: An increasing number of carriers are providing comprehensive services out to their agents through the carrier Web site, including: quoting, submission, binding, policy change, billing inquiry, claims inquiry, and so on - sometimes for one line and sometimes for several. Some of these sites are very basic and perhaps even inconvenient to use. Others are well designed, with high usability, and make it very easy for agents to do business with the company. Agency portals allow carriers to support agents who don't have management systems, a minority more sizable than one would expect given that we're a quarter century into the process. The most interesting carrier sites provide complete point-of-sale/point-of-service support to agents for their customers. That allows carriers to finally change the insurance process, putting the agent in charge (though within the confines of carrier editing and underwriting). The carrier saves money by taking its people out of the loop. The agent gets more control and eliminates the need for follow-up, since the transaction is completed right now. And, the consumer gets immediate service - a triple win!
SEMCI and real-time: Real-time interface, while not envisaged by SEMCI, is consistent with its philosophy; namely, that an agent should be able to get done whatever is required from within the management system - that is, within one consistent user interface and workflow and without needing to re-enter information. On the other hand, real-time interface appears less likely for widespread adoption than carrier Web sites, though it doesn't follow that it's an either/or situation.
Carrier Web site issues: Carrier Web sites can be useful when they allow an agent to complete a sales or service transaction right now, but they create a certain amount of frustration as well. Site to site, agents find themselves having to deal with a variety of logon/security schemes, timeouts, workflows, screen layouts, and so on, which means that there is more to learn for an agency staff already deluged by technology. And perhaps worst of all, the data in the management system must be re-keyed into the carrier site forms. Carrier Web sites resemble and have some of the problems associated with carrier proprietary interactive terminals.
Hybrid solution: Carriers and management system vendors are beginning to experiment with introducing a certain amount of cooperation between the management system (or comparative rating) and the carrier site. So, for instance, information about the management system context is passed to the Web site, and the Web site pre-navigates to the appropriate area and populates the Web form (when relevant) with data from the management system. A management system/carrier Web site hybrid environment could overcome some of the objections agents have to carrier Web sites. Were logon, navigation, and data flooding handled automatically as the user was moved from management system to carrier site area, the result would be reasonably convenient, though not have quite the transparency agents might want.
Carrier site advantages: Carriers that host agency portals sometime claim, with reason, that they facilitate fast and convenient sales and service transactions in ways management systems simply cannot. In addition, a carrier site can provide ready access to product and other information that wouldn't be visible from within a management system. Given a well-designed hybrid environment, it could have some advantages over upload or real-time approaches for both the agent and the carrier.
Competing via technology: One common complaint about carrier Web sites is that they're blatant attempts by carriers to use technology for strategic advantage. Carriers, we're told, should compete on the basis of product, service, price, and so forth, but not via technology. Or, if they do try to use technology to compete, whatever they do will be copied shortly, the advantage will disappear, and so why bother?
Why not? It seems to me that carriers should compete in ways that suit their resources, circumstances, and culture. If a carrier (or agent for that matter) wants to (try to) use technology to compete, so be it. If the carrier doesn't do a good job, it will suffer the consequences. Competition should provide agents with more choices and greater value than they'd have without it.
The new Web interface model
Classic model: The classic interface model is based on standard, batch upload and download, assumes an environment of classic agency technology (management system, comparative rating, office productivity), presupposes an agency focused on property and casualty, and looks to a conventional distribution of labor between agency and carrier.
Change: But, as we noted above, that's not the way things are anymore - at least for many agencies and carriers. Agencies are using carrier Web sites and a great number of other Web-based resources (including perhaps their management system). And carriers have become more eager to change the division of labor, giving agents more control and the tools to get the job done quickly. More and more agents find themselves living in a web of interconnections with customers and suppliers and becoming less of a self-sufficient, isolated island of computing.
Pulling the threads together
A new view: So let's try to pull these threads together and see what we've got. I hope I've made the point that classic interface thinking needs to be refurbished. I also hope I've shown that though the current situation is much more complex than it was in the past, it's a lot more interesting and promising. And it's also comprehensible (with some study, thought, and talk). So, what are the important elements of the new view on interface and insurance computing generally?
Gratuitous advice
Agents: What might agents do to understand and profit from this new environment?
Carriers: What might carriers do to serve themselves and their agency sales force?
Vendors: What might vendors do to help their agency customers within the wider view of interface?
Trade Associations: ACT, user groups, producer associations - how to help.
See www.soundingline.com to download the complete white paper from which this article was condensed.
© Copyright 2002 by Sound Internet Strategy. All rights reserved