As part of the recently published Fund Technology & Data, North America 2017 report, we interviewed Julia Bardmesser, Global Head of Data Integration at Deutsche Bank about 'Exploring Data Quality Management Implementation Framework for the Asset Management Firms'

David Grana, Head of North American Media, Clear Path Analysis: Can you please explain why Data Quality Management is so important in the asset management industry?

Julia Bardmesser: Data Quality Management is very important to any company that has data. It is a huge part of data strategies in the pharmaceutical industry, manufacturing, retail, energy, transportation, and the list goes on. In today’s world, Data Quality Management is important to essentially all companies.

Data Strategy is driven by various goals the company wants to achieve. In the financial sector, the driver for data management discipline is decision making. In other words, companies want to know what to buy, what to sell and what to hold. They need to be up-to-date on what is happening in the capital markets. Asset managers want quick and trustworthy analysis on the impact various changes may have on the company. In order to achieve this, companies need high quality data that are available and ready for analytic consumption. Data quality management is especially important nowadays since the amount of data any typical company may have has multiplied significantly compared with just a decade ago. In summary, quality of data is vital to making good business decisions and setting up the strategic direction for the organization.

David: How do global organizations have data quality meet all standards simultaneously?

Julia: Most companies within the financial industry are large multi-national corporations. Such organizations cannot meet all the data standards for all data right away. This is where the need for the overall data management discipline comes in. One of the major tasks of data management strategy is to figure out how to define importance and criticality of data for the company.

The company has to have the ability to prioritize and begin with the data of the most importance. Data priority is determined by the strategic direction of the company. The firm has to start with identifying the classes of data that are of greatest concern and relevance for the company strategy. For most financial companies, the first area is usually customer (counterparty) data.

That said, data management is an expensive undertaking. One should not expect quick fixes and results. Good data management is about frame of mind, not about one specific project with a finite end date. Good data management practices involve systematic processes to get each class of data under control, beginning with the most business critical one. Getting data under control means creating data semantics, uncovering data lineage, measuring data quality at different points of the data flow and many other activities.

David: What role does technology play in data management?

Julia: Good technology enables data management. However, it is not about what technology the company has, but rather, what the company does with available technology. It is important to fully utilize available technological capabilities. There are, of course, many tools on the market that support various data management disciplines, such as data governance, metadata management, data lineage, data quality, etc. These tools are frequently stitched together within companies to form enterprise data platforms. These platforms are better suited to support good data management practices. Platforms where tools are integrated with each other are more effective than stand-alone tools that cannot be integrated with other tools.

Finally, if the enterprise makes technology decisions based on a data-centric architecture, as opposed to systems/applications-based architecture, then data management will be more effective and less costly.

David: How does a data manager coordinate and manage with the various departments of a financial institution?

Julia: The typical route for financial institutions is to establish a data policy and then use the policy framework to create data governance and accountability structures. The very first thing that any company embarking on a data management journey must do is to create an accountability structure. This structure defines who is accountable for the data and how decisions are made. Many companies move towards creating data councils, which are responsible for data management oversight activities. An effective data manager must use established accountability structures to move forward.

In the financial industry, it is quite common to have a central organization called the Chief Data Office (CDO). The CDO manages the standards and policies centrally, enterprise-wide. In addition, many financial companies create a structure where each of the divisions have their own people in charge who are responsible for moving data management forward within their respective areas. All divisions work together to get decisions made on the enterprise level.

David: Do you feel that the role of data manager is more reactive or proactive or a bit of both?

Julia: It depends on the specific role of the data manager within the organization. Some data manager’s roles are very proactive.

And example of a proactive role is that of a Data Steward. Data Stewards drive data policy implementation. Another example of proactive roles are people creating good data architecture, people creating capabilities to measure data quality and establish data lineage capabilities The above capabilities are needed in order to prevent data issues from occurring in the first place.

However, even in organizations with the best data management practices, data issues still arise and someone needs to resolve them. This is usually the job of a Data Quality Issue Analyst. A Data Quality Issue Analyst is an example of a reactive role - a data manager who is only responsible for analysis on data problems that have been discovered.

David: What could be improved in the area of technology to make the job of data manager easier?

Julia: A big help to data managers would be making the technical architecture of the company data-centric (i.e. utilizing and promoting proper data semantics, supporting metadata collection, retention, and curation). This type of technology would enable an understanding of the data and data needs, both from the consumer, as well as from the data producer’s perspective. Unfortunately, that type of architecture is rarely implemented.