Nationwide ensures unified view into financials

With 30,000 employees and many semiautonomous divisions across geographic and line-of-business boundaries — including several acquired but only partially digested business units — Nationwide Insurance found it increasingly hard for its central financial and executive team to get a comprehensive, up-to-date view of the company.

“Independent groups continued to do things the way they always have,” says Vikas Gopal, director of enterprise financial applications at Nationwide. “The pain of trying to get a consistent view of finances across the enterprise led to the recognition of the need for a transformation.”

Led by its CFO, the insurance giant resolved in 2004 to create a unified view of company financials, integrating data from 240 source systems into a central, “golden copy” repository from which central analysis and reporting could be conducted.

Dubbed the Focus program, the combined IT and finance team tasked with figuring out how to unify the financials across the company soon realized it needed to have a formal data governance approach, Gopal recalls. It wasn’t enough to simply update the point-to-point connections among various financial systems and the overview system used by the CFO to ensure that they used the same data definitions. “Even if you took the time for consistency, there was no assurance that it was repeatable,” he says. Instead, there had to be a way of reviewing changes and enforcing the data architecture going forward.

Rather than change various front-end systems, the Focus team centered on what data was needed for the enterprisewide roll-up financial reporting and analysis. That had two benefits, Gopal notes. One, it made the project scope manageable. Two, it allowed the various business units to keep their highly customized financial systems that reflected the specific regulatory, market, and product requirements that each deals with day in and day out.

Even product definitions vary from business to business, resulting in different “logical units” at the core of their financial models, Gopal says. For the enterprisewide system, each unit “wanted to make sure that their view [of logical units was] deliverable in their context,” he adds. So the Focus team spent six months during the initial project stage figuring out exactly what was needed for enterprisewide analysis and how it mapped to the local units’ views. The result was a set of standard data transformations — what Gopal calls the master data definitions — that all local financial data goes through before they are imported into the enterprisewide financial system. With the data model in place, Nationwide was then able to select the technologies to transform and manage the enterprisewide data, in addition to providing the dashboards that executives could use to monitor it.

But the data model also enables a key governance benefit: control over the data going forward to ensure that enterprisewide data — and the assumptions underlying it — doesn’t change unwittingly.

If a local business unit needs to change its data, a central data governance group must approve the revisions to the master data definitions before the data transformations can be updated. Any unauthorized changes to those standard transformations get kicked back to the source.

The final system involves a raft of technologies, including Informatica and Kalido tools to transform the data and perform the ETL work — including detecting unauthorized changes to the source data formats and metadata — before loading the regularized data into a PeopleSoft ERP system, which the CFO’s organization uses to analyze and generate reports using PeopleSoft analysis and Essbase and IBM reporting tools.