Eric Hansen, finanace business manager in the Financial Shared Services Center at Nationwide Insurance
With a drumbeat that harks back to what "lean" once held for manufacturers, the word "governance" has landed toward the top of agendas of large businesses across every industry. In updated lingo, data and information governance are already being translated to reflect a new kind of risk management sweeping into Global 2000 management suites with gravitas you'd expect of global warming.
Though the structure and mechanisms of governance remain unclear to many, the underlying stakes are not, since information quality and timeliness are increasingly linked to competitive advantage and the realities of a tighter regulatory environment. But if we define governance as the processes, policies, roles and controls that support the creation of information as an asset, more likely we'll find it's people more than technologies that fulfill its existence.
And therein lies the rub: as it goes with important technologies, critical and influential human assets are often expensive, scattered, siloed and self-important. Critics are already griping that data governance will create new overhead and layers of bureaucracy in business. "A lot of executives understand the need for governance but know they're going to have an uphill battle and four-letter word conversations convincing everyone else," says Forrester Research Principal Analyst Rob Karel.
In retrospect, this was once the case at Nationwide Insurance, the Columbus, Ohio-based carrier of more than $160 billion in statutory assets. A Fortune 110 company with 36,000 employees, Nationwide has already moved its standard of governance with a one-off strategy built on a tripartite model of checks and balances not unlike our own federal government. That the plan succeeded not on immediate calls for ROI, but on long-term benefits of risk containment and control makes it that much more interesting.
Nationwide fits the description of a large corporation that has grown and expanded into multiple industry segments, presenting a great governance challenge. The organization is made up of businesses that sell property and casualty, life, retirement and specialty insurance products across most of the 50 states. With a diverse portfolio and product mix, the company needs to meet a variety of business unit reporting obligations to match different growth plans. Externally, for example, property and casualty insurance require very different SEC reporting than does life insurance.
Four years ago, Nationwide set out to end the apples and oranges of uneven financial rollups and product definitions by consolidating financial data from every part of the business onto one reporting platform. That's when data governance became the elephant in the room. "In those days, most of our businesses operated in silos with their own systems and their own little space to play in," says Eric Hansen, finance business manager in the Financial Shared Services Center at Nationwide Insurance. "They could do pretty much whatever they wanted, and they really weren't talking across the fence."
The risk overhang was equally unacceptable because the insurance business is regulated at the state level with strict financial data requirements. "The end users of our data are internal financial analysts and accountants," Hansen says. "The quality and integrity of financial data has to meet regulatory reporting requirements but also support accurate decision-making for Nationwide's strategic direction." Because external and internal reporting come from the same data and systems, it made sense to construct a governance program to serve the finance compliance group as well as the strategic thought leaders who decide how to allocate Nationwide's capital in order to expand and grow the business.
The Tripartite Model
Hansen and a colleague were assigned to kick off the project. But among other analysts and managers dedicated to the project, there was already agreement that a central structure of rules and roles was the best hope for herding Nationwide's information resources into a cohesive structure. "We kept saying, 'enterprise, enterprise, enterprise,' and 'consistency, consistency, consistency,'" Hansen recalls.
It's a common governance goal that often goes unmet. "Big insurance carriers often use one strategy and one installation of technology per product line, so they pay 10 times for the same thing," says Karen Pauli, insurance analyst at TowerGroup. "Nationwide has been more enlightened than some other top-10 carriers in understanding the cost of this."
At Nationwide, consistency meant that even simple terms such as "premium" meant the same thing across all businesses and ensuring that the plans of one company would not break the standards of another. So, taking the term literally, Nationwide's governance structure was modeled loosely on our own federal government, with Judicial, Executive and Legislative branches to prescribe roles, enact laws and limit the boundaries of authority.
The Judicial Branch
At a high level, the Judicial branch translates data governance into business terms where leadership and sponsorship determine high value. This group is made up of the board and approximately 15 others, VP level and above, who meet for about two hours each month. Relatively new to the program, the Judicial branch has experienced good buy-in from members who, not coincidentally, are the primary beneficiaries of high quality enterprise data. Alignment of thinking at this level gets things moving within the financial systems environment, providing more speed and confidence to the very people who sign off on the financials and the integrity of the data within.
Defining key financial metrics for balance sheets and internal decision support consistently across the enterprise also means supporting individual business unit requirements for internal and regulatory needs. "The key is to understand the best method for segregating the data into the different views and aggregating the data in a meaningful way to support each view," says Hansen. "It absolutely can be done, and we really do support all of our reporting within a single environment." Segregating, tagging and rolling up data means that some data in the system will apply only to management reporting and will not be included in statutory and legal reporting. Other entries are booked for purely statutory reports and don't apply to management needs.
Process re-engineering supports Judicial mandates to grease the wheels beneath the data. "We have a culture around continuous improvement, and we are constantly reviewing processes to make them more efficient and streamlined," says Hansen. The quality initiative is homegrown and blends concepts from Six Sigma, TQM and Kaizen in Nationwide's own nine-step program used by the process experts who lead continuous improvement projects.
The Executive Branch
The primary branch of Nationwide's governance structure is the finance data governance group, which executes the processes and policies handed down from the Judicial branch. These 18 full-time business analysts (see Figure 1) are the "doers" or "executors" who implement the many changes that affect the financial reporting systems. Reporting up through the business side, this organization manages governance throughout the entire environment. It's their job to get changes into, through and out of the environment properly end to end.
Within the Executive branch, a business rules and interfaces team normalizes and transforms mismatched data from multiple source systems across 300 mapped interfaces. A second team of four analysts provides about 300 standard reports for all of Nationwide's finance professionals. This group needs to ensure that any and all changes to the system appear correctly in the proper reports. "If we start selling a new product, say it's identity theft insurance, we need to know that resulting reports will give the leaders the insight they're looking for," explains Hansen.
The MDM team is responsible for the execution of all new or changed accounts. "We might need to track financial activity differently for regulatory reasons, maybe our current breakouts aren't detailed enough, so we add new accounts for those things," says Hansen. "If Nationwide sells a new product, for example, it's the MDM team's job to get the necessary details added to 1,100 reference data tables, and that's a lot of work. Those guys are always heads down, very analytical and just plowing through."
Demand management and change control are handled by two analysts who act as a clearinghouse for requests, track changes and approvals, and route work to the teams that need to be assigned. This gives a central point of contact from the business unit's perspective. The metadata function is filled by an analyst who works across the financial systems to make sure that transaction codes map to accounts that appear on the financial statements from any source system.
The last function is standards and policies, backed by documented corporate policies. "When your friend Bob comes into your office and requests a new code for this or that, we need fair criteria to evaluate those requests, and that comes from policies and standards," Hansen says. "We can fairly say, 'I'm sorry, this request does not align with our policies,' or 'Yes, your request meets our standards so we'll move forward.' It gives you a leg to stand on."
The Legislative Branch
The Legislative branch is made up of data stewards, the business unit subject matter experts who sit outside of the data governance team and act as liaisons. "Data stewards are sort of the arms and legs of policies that the governance committee puts into place," says Gartner analyst Ted Friedman. "To achieve success you want day-to-day stewardship processes that tackle front-line needs first."
At Nationwide, 15 stewards spend a quarter of their day managing requests from their own constituencies of business users and help to resolve production issues. "These are people who really understand data and financial activity and can credibly represent their area," says Hansen. "They are financial analysts, accountants, some policy-type folks and process consultants."
Governance Means Business
At Nationwide, the data governance functions sit among analysts on the business side of the house, where Hansen feels they belong. "We're the business users, so we should be able to control changes to our data. I will tell you that people on our team are kind of a hybrid that can understand an income statement and a balance sheet but also understand the technology side." A walk along these blurred lines means organizations are wise to look internally, recruit and retain such high-demand people.
While tangible benefits are to be expected, "It's not easy to tee up a large data governance program and assign an immediate return on investment," says Friedman.
In Hansen's analysis, it gets back to the purpose of having a data governance program in the first place. "For us, it boils down to value creation and financial controls. The price you pay is a little bit of overhead, a little bit of process work, but the benefit is data quality, data integrity and better control over your financial data. To us, that's the investment and the selling point."