Among the many reasons cited for converting from legacy to modern administration systems, improving access to historical data tends to appear low on the list if it is mentioned at all. Some argue that this is as it should be. Others suggest that ignoring existing data could put legacy conversion projects at risk, and recommend a ‘data first’ approach is warranted. What do you think?
Is Legacy Data Really That Important?
We have noted that Administration Systems replacement projects are frequently sold on the basis of supporting growth and profitability by:
- reducing time to market for new products,
- improving underwriting results, and
- reducing costs by creating internal efficiencies.
Conspicuous in its absence is reference to data quality and ease of access to historical information.
Indeed, not everyone believes that full, automated conversion is a requirement. For example, Novarica’s Chad Hersh, writing in PropertyCasualty360.com, thinks that an incremental, semi-automated approach might serve the needs of P&C insurers: “while conversion is indeed critical for life insurance core system replacements due to the long tail of life insurance policies, annuities, etc. … for P/C insurers there are significant opportunities to minimize conversion that many carriers should take advantage of.”
According to Hersh, the most common approach is manual conversion on renewal. A second is to triage out only the bare bones of policy, claims, and billing data and move the balance to a data warehouse where is can be accessed as required.
Historical Data Will Not Be Easily Ignored
Anthony O’Donnell, writing in Insurance & Technology, summarized the opposite side: “The rap against the modern-technology systems is that they can’t compare with the functionality of well established vendor systems that have adapted to market needs, potentially over decades.”
O’Donnell cites the experience of Canadian Automobile Association South Central Ontario (CAA), who turned to insurance solution provider, iTer8, to fill in the gaps it found with its implementation of Guidewire’s suite of administration systems. iTer8’s David Gallagher notes that data conversion is required in the current era where insurers differentiate themselves on the basis of their ability to provide access to historical customer information and analyze historical and current data.
Approaching Modernity ‘Data First’
While post-hoc conversion and integration may meet the needs, some vendors are suggesting that putting data at the front end of the modernization process could reduce cost and risk. Writing in PropertyCasualty360.com, Bruce F. Broussard Jr., vice president, business development and strategic relationships with Insurity, suggests such a shift in the paradigm has been used in the past by large insurers, “that enjoyed the luxury of a deep pool of resources and the ability to take on complex custom development and long term maintenance of proprietary data and integration structures.”
The strategy is to begin the modernization process with the creation of a central data repository which will “be leveraged as the singular source for reporting and integration throughout the duration of the modernization effort and long afterwards.” Broussard notes this is a change in the order of work, not additional effort, as “data mapping and integration are part of any core modernization project.”
Broussard believes that the data first approach is now available to a insurers of all sizes, due to “several developments over the past few years (which) have changed the insurance data landscape, most notably the maturity and acceptance of industry standard ACORD data structures.”
Insurity is not alone in this thinking. Oracle, with its partner Universal Conversion Technologies (UCT), have recommended a similar approach for its insurer clients. UCT notes that it strongly supports ACORD Standards.
As we noted in this space two years ago, ACORD Standards have been the basis for data designs since the 1980s for Agents/Brokers and since the 1990s for insurer data conversions. (As we previously reported, ACORD and its Canadian counterpart, CSIO, have confirmed their respective intent to maintain compatibility of their standards.)
What Do You Think?
Is data conversion a critical component of a move to modern administration technologies, or can it be dealt with as an adjunct? If critical, does the ‘Data First’ approach make sense? We would appreciate the data (old or new) you can share.