Personnel Choices

Whether efforts to improve data quality and consistency are happening in reaction to regulation and standards mandates, or of their own volition, the work is generating value for data management at a time when the volume of data to be managed is increasing and the resources to handle data are more scarce.

As data managers try to derive value from data quality and consistency initiatives (whether those are driven by regulation or not), they are finding that choices concerning personnel and resources are becoming key to coping with ever-increasing amounts of data to be processed through methods set in new initiatives.

Contending with rising data volumes is more than just a technology problem, explained Brian Miller, senior vice president, brokerage technology at Wells Fargo in St. Louis, speaking on the first day of the Sifma TechExpo this week. Staffing and processes must also "clearly" be part of the response, he said. "Do we have the right roles in the organizations to manage the data? That can be anything from data integrity managers and data stewards to the technology people who implement those processes."

Considering how to organize and deploy data staff requires "thinking differently," said Miller, echoing Apple's landmark ad campaigns. "Having the ability, the courage and wherewithal to undo everything your firm grew up with allows you to free up the resources to do it the right way," he said. One example of such an effort, given by Dilip Krishna, a director at Deloitte & Touche, is taking apart multiple data stores set up to serve different purposes, and then re-investing the resulting savings in a new, consolidated method.

Regarding the personnel piece, Miller cited Wells Fargo's distributed model. "It's not only for data talent but being able to use that talent within the financial services industry, which is the real challenge," he said.

David Kowalski, an information architecture executive whose most recent role was in the financial services industry, sees a federated approach to data management also being used. "That puts a lot of thought into finding a balance between figuring out what you really want, what kind of behavior you wanted to incent, and what kind of data and metadata needs to be reported to the top of the house," he said.

Whether your data management and personnel models are distributed widely or federated, willingness to depart from traditional approaches is proving increasingly necessary, as Miller and Kowalski say.