Eight tips to avoid a data migration migraine – Xceedance

Global data volume is expected to reach roughly 180 zettabytes by the year 2025 – an unprecedented growth rate that poses challenges for insurers in managing and governing the data. Only 46% of data migration projects are delivered on time, and 74% face budget overruns, according to Experian.

As a result, many companies have started considering and investing in a robust data and analytics strategy to derive real value from data.

Why and when you need data migration in insurance

A business driven by data insights is nearly twice as likely to grow revenue and reduce costs, according to Forrester Research. But to truly become an insights-driven business, an organisation must improve the way it captures data. Adopting the latest, cloud-enabled applications is essential in achieving that goal.

Data migration allows insurance companies to leverage new technologies, improve data analytics and reporting capabilities, and ensure systems comply with industry standards and regulations.

Insurers use data migration for several reasons: storage migration to transfer customer records or policy information from an outdated system to a more modern one; database migration to transfer data between different databases; application migration to transfer data from an old policy management system to a new platform to integrate automated claims processing, customer relationship management and risk assessment; business migration to move data related to underwriting processes, claims management, business rules, or customer service operations; and/or cloud migration to move data applications and other systems to the cloud.

Technology is constantly evolving with new solutions such as artificial intelligence, and insurers need to keep up. Companies should regularly update their homegrown applications or adopt suitable off-the-shelf ones tailored to insurance to improve data retrieval speeds and accessibility, data security, data processing and analytics capabilities, and to reduce costs.

Common risks and challenges in data migration projects

For insurance companies, which deal with massive amounts of data, migrating large volumes from an old policy or claims system to a new one is hugely challenging. It’s essential to identify risks early in the process and devise a mitigation plan to address risks.

Data quality issues are difficult to deal with unless a thorough analysis is done before migration begins. Missing values, incorrect and incomplete values, and duplicate and non-standard values must be addressed during the migration process.

Data breaches and regulatory violations can be avoided with correct data compliance. Proper governance mechanisms and security protocols are essential for handling sensitive personally identifiable information (PII) while migrating customer and policy data. Business disruption can be avoided or reduced with initial analysis of data mapping and quality.

Best practices

Here are the top eight best practices that help increase the success rate of data migration:

1. Data Context. Migrating and transforming data requires utilising the right insurance knowledge to ensure data quality and avoid downstream misinterpretation of values.

2. Proper Planning. Draw a detailed plan that covers the need for migration, what needs to be migrated, the time available for migration, potential risks and challenges.

3. Testing Strategy. Testing migrated data to the new system involves various aspects such as reconciliation, validation, and overall data accuracy.

4. Execution Strategy. Whether you need a big bang migration, a phased migration, or a hybrid approach depends on the use case and the requirement.

5. Manual vs Automated Tool. Data migration involves activities such as data profiling, extraction, transformation and cleansing, loading, testing, and validation; the absence of an automated tool can increase the chances of error and reworking.

6. Data Back-ups. A simple but crucial step is to securely back up all the data before starting data migration.

7. Data Assessment and Classification. Studying source and target data structures and formats and classifying data into the correct categories helps build a detailed mapping document.

8. Data Cleansing. Moving data from an old structure to a new structure can be daunting, and can be made worse by insufficient data quality. Assess data issues while analysing the source systems and address them during data transformation.

The importance of migration strategies, processes and tools

Having a strategy for fixing data quality will resolve issues early and avoid reworking. It’s also important to have a data execution strategy to decide between big bang and phased migration, depending on the complexity and volume of data.

Data migration tools such as Informatica, Microsoft SQL Server Integration Services (SSIS) are available, but they are generic and require a great deal of configuration. A more thorough approach is to have a migration framework that can connect to one or more sources, stage the data from the sources, and cleanse, standardize, and transform it before migrating it to the final system.

However, data migration is not all about technology, it’s about having the right people. A skilled, experienced data migration team with insurance knowledge is essential for successful data migration projects.

By Jarno Seegers, vice president, business development at Xceedance