Vous êtes sur la page 1sur 4

From the Backroom to the Boardroom:

The Evolution of Data Quality in the Insurance Market

Mark B. Gorman
CEO and Founder, The Gorman Group Insurance Consultancy
January 2010

Executive Summary
Sponsored by
From the Backroom to the Boardroom: The Evolution of Data Quality in the Insurance Market Executive Summary

Perception is often more important than objective measures when considering data quality. Users care less
whether a data element is accurate 99 percent of the time than they do about the 1 percent inaccuracy that
caused problems with reconciling a financial report, responding to a customer or adjusting a claim decision.

This report is based on a research questionnaire in Insurance Networking News to which 75 insurance
organizations responded, followed with in-depth interviews of thought leaders from 18 organizations.

The findings reveal that it is critical to have a program in place to enhance users’ trust in data and its meaning.
The two most common choices the research found were:

• Fixing data quality errors as they are encountered (the traditional strategy).

• Establishing a data quality program to apply quality metrics as information is captured, transformed, derived,
stored and delivered (the emerging strategy).

The chosen direction is fundamental to whether an organization takes a reactive or a proactive approach to data
quality, because the choice determines the level of resource investment required.

Governance is another consideration when improving data quality. At the enterprise level, this requires three
levels of oversight:

• A senior management body responsible for funding, guiding and supporting a data quality effort.

• A dedicated enterprise unit responsible for setting standards, defining controlled processes and procedures
and monitoring data quality metrics.

• A data steward group resolving data quality issues and applying data quality standards at the line of business
or functional level.

Often, organizations focused on fixing data quality typically have only the senior management and data steward
levels in place and are missing the mid-level oversight that provides critical strategic and procedural direction.

Lastly, it’s important to remember that data for an insurance firm has intrinsic value. Its importance is not subject
to the applications creating or using it, nor is it dependent on the infrastructure supporting it. Both applications
and infrastructure do change. Data needs to be treated as a long-term corporate asset to be protected, invested
in, maintained and improved for sustained ROI.

All these practices will enhance the insurance organization’s intelligence about itself, its customers and the
marketplace and will bolster user confidence that the data is relevant and accurate.

Understanding the Impact of Data Quality

As a senior vice president at an insurance company was being interviewed for a white paper, he reacted strongly
when he heard data quality characterized as a mundane subject.

“Oh, not to me,” he interrupted. When asked why he felt so strongly, he responded, “While I used to see this
as a pain in the neck and just one more thing to deal with during the day, I now understand that this is my
opportunity to assure business managers are sharing meaning on information important to our business.”

2
From the Backroom to the Boardroom: The Evolution of Data Quality in the Insurance Market Executive Summary

His comments were the genesis of this paper. During the research for a previous white paper, Best Practices for
Leveraging Business Analytics in Today’s and Tomorrow’s Insurance Sector, a number of interviewees asked
what others were doing to ensure the quality of insurance data, especially as it pertains to promoting expanded
use of predictive analytics. While visiting with senior managers among insurance carrier clients, the topic of data
quality came up again in a variety of contexts including changes to state and bureau stat reporting requirements,
errors in commission reporting, impact on building predictive models for additional product lines, opportunities
for new product development and discussions about which business segments to pursue for profitable growth.
The issue has clearly expanded beyond discussions and debates over the accuracy of numbers in reports.

The research for this paper uncovered these current topics:

• Issues driving a more widespread discussion of data quality among both business and IT senior managers.

• The definition of data quality and whether this definition has changed or evolved recently.

• Metrics used to measure the impact of poor data quality, the contribution of excellent data quality and how it
relates to gaining support for funding data quality initiatives.

• Activities regarding ongoing data quality issues and ideas for sustaining efforts during multiyear initiatives.

• A roadmap and insight into the future of data quality.

This paper is the outgrowth of two research projects. The first was a research questionnaire in Insurance
Networking News to which 75 insurance organizations responded. This was followed with more in-depth
interviews of thought leaders from 18 insurance organizations during November and December 2009 and early
January 2010. These thought leaders represented business people, IT professionals and people from the grey
area in between often called business intelligence, data management, reporting and analytics.

This paper describes what is driving insurance carrier organizations as they enhance their focus, increase their
efforts and improve their outcomes relative to the quality of the important data within their organization. The
subject here is not data quality tools available in the market, nor is it about a set of tactics to manage data quality
within an initiative. Interviewed leaders responded with a surprising amount of passion, insight and enthusiasm
for this topic. While drawing heavily on the information contained in this paper, the interpretation and conclusions
are solely those of the author.

Conclusion

During the course of the research, we visited with a number of people who have had extensive experience and
expertise regarding enhancing the quality of data available to their organizations. No one we visited with was
satisfied. Comments like, “we’re neophytes right now, but we’re getting better,” “I’m worried we’re not going
fast enough,” and “we’ve made progress but we’re far from where we need to be,” were common. The broader
market appears to agree. When surveyed regarding their perception of their organization’s approach to data
quality, more than 55 percent of respondents to the research questionnaire rated themselves in the bottom two
out of five levels identified below.

3
From the Backroom to the Boardroom: The Evolution of Data Quality in the Insurance Market

Percentage of “Yes” Response

5. Senior management concerned and has authorized


5 funding of tools and systems to proactively monitor,
identify, analyze and resolve issues.
4. Senior management aware and has endorsed
4 a business-led, organization-wide initiative to
monitor data processes and procedures.
3. Senior management aware and has endorsed
3 a business-led, organization-wide initiative to
monitor data processes and procedures.

2. Senior managers aware, but issues handled


2 reactively as they arise.

1. Many people not aware of data quality issues.


1 Data quality handled ad hoc by IT.

Perception of an organization’s approach to data quality

Obviously significant opportunity exists to improve the quality of data available in most organizations. Corporate
IT continues to play a pivotal and critical role in attempting to assure the organization can rely on quality data.
But IT alone is not enough. The level of leadership provided by senior management, the amount of focus
provided by business subject matter experts, the quality of the processes and procedures applied and the
level of automation funded will all impact the amount of organizational angst endured in pursuit of quality
information. Understanding and planning for the needs of the “downstream” user will enhance trust in the data
captured, transformed, delivered and analyzed. And trust in what the data means will enhance an organization’s
intelligence about itself, its customers, its competitors and the markets in which it operates.

What recommendations emerged from the research, and how does an organization get started? While this of
course is unique to each firm, reviewing the results generated three high-level recommendations:

• Pick the appropriate approach. Either focus on solving data quality issues or focus on establishing a data
quality program. This decision drives the following two recommendations.

• In balancing the three-legged stool, first, find the right people. Identifying the blended beings can be difficult.
Once in place, let them set up the policies and procedures appropriate to the chosen approach. Finally, allow
them to determine the appropriate level of technology needed and then buy it.

• Pick the right governance structure. If the approach is to fix data problems, a two-tiered governance structure
(the top and bottom tiers) may work. If setting up an enterprise-level data quality program, the three-tier
approach is recommended. Regardless of the approach, the governance structure needs to empower,
equip and support the people implementing the program, and if it is too bureaucratic, expect significant
organizational resistance.

The pursuit of data quality is not a project. As one interviewee said, “We finally realized that data quality is not a
result, it’s a pathway.” Insight and intelligence form the basis for success and survival as the insurance market
continues to change. What is critical is that the corporation evolves beyond just fixing data problems to actually
establishing, supporting and maintaining a data quality program.

To access the entire research report, please go to: www.sas.com/gormanreport

SAS Institute Inc. World Headquarters   +1 919 677 8000


To contact your local SAS office, please visit: www.sas.com/offices
SAS and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of SAS Institute Inc. in the USA
and other countries. ® indicates USA registration. Other brand and product names are trademarks of their respective companies.
Copyright © 2010, SAS Institute Inc. All rights reserved. 104408_S53095.0310

Vous aimerez peut-être aussi