Question: How do you plan for and implement a successfuldata conversion process?

The importance of careful planning and good execution when it istime to convert your data from one system to another cannot beemphasized enough. In fact, data conversion is one of the mostimportant aspects to a successful policy administration systemreplacement or implementation. With the right approach andexecution plan in place, your data can become a strategic assetthat preserves historical experience and leverages customerhistories for future competitive advantage.

First, it is important to clearly articulate the businessrationale for converting data from the current system. Conversionstrategies run the gamut from manually entering policy informationat the point of renewal to fully-automated conversion of allhistorical data. And, these options vary greatly in terms of bothcost and complexity.

Further, it is critical to involve not just IT, but alsobusiness analysts and stakeholders in the planning phases as wellas in the QA and data validation phases. In fact, you will want toinvolve a user from each department that will ultimately need toaccess and utilize the data.

When executing a data conversion project, insurers should keepthe following guidelines in mind: 

• Treat the data conversion effort as its own project, completewith its own stakeholders, project management and governance.
• Analyze and identify the critical and value-add data to bemigrated, and map to target.
• Consider how much history is required, both for regulatory andfuture decision-making or servicing purposes.
• Get early access in the project to the actual source data foranalysis, as there is often the opportunity for discovery ofpotential data quality issues or other unrealized datasources.
• Merging data from multiple sources requires insurers to payclose attention to any structural, semantic and syntacticaldifferences among the data sets.
• Be alert to the possibility of unique application designrequirements for converted data such as validation or underwritingreferral rules that will not be applied to converted policies orgrandfathering.
• Evaluate the requirements of the target policy administrationsystem, as well as downstream systems such as billing, claims, datawarehouse and statistical reporting in the design.
• Define performance requirements up front and allow time in theschedule for performance tuning of the conversion process.
• Conduct multiple dry-runs with large data sets to flush outinconsistent source data, resource leaks or logic bugs.
• Devote separate hardware resources and QA environments to dataconversion, as long-running conversion test cycles may consumeresources that interfere with other development teams.
• Ensure that the design contains a strong audit component tomake certain the process and its outcome can be monitored.

It is no secret that accessing the right data for cross-sellingand up-selling in real-time can prove to be a competitivedifferentiator and turn your data into a powerful businessadvantage. The challenge is how to get the data working for youwhen obstacles such as legacy systems and incompatible data formatsstand in your way.

Working with a provider that has experience with thecomplexities of managing data migration of customer, insurancetransactions and product data can help you achieve the bestbusiness outcome while eliminating unnecessary costs and reducingthe risk of common pitfalls so you can keep that historical dataworking for you and your business.  Without this, yourbusiness challenges will persist and all you will have achieved isadding one more system to manage.

Steve Forte is the director of global product management andalliances for OneShield. He can be reached for further comment viaemail at [email protected].