CAMBRIDGE, Mass.—Data comes in all shapes, sizes and formats and from all parts of a business. But sooner or later that data is going to come in contact with a human, and humans— even data professionals—are imperfect and illogical.
At MIT’s Information Quality Symposium, speakers and attendees referenced the human factor of data management far more than any technology platform or implementation process. From ground-level operators who input data to top-level executives who choose whether to use data-driven recommendations, the key part of any information quality initiative is getting members of an organization to commit.
Dat Tran, Department of Veterans Affairs’ deputy assistant secretary for data governance and analysis, said during his keynote speech that he “socialized” his data quality and governance initiatives with 45 top executives before going to the agency’s leaders for funding. Members of his team worked with the data entry professionals in the VA’s eight different lines of business to explain how the new data governance would benefit them.
“No matter what I do, I have to show the value proposition,” Tran said. “Not just to leaders, but also to the workers.”
The VA provides nine million veterans and their family members with some combination of health care, retirement benefits, life insurance, home loans and education funding, and other services. But there is very little data sharing across organizations, or even across offices in a single organization, Tran said. The organization has more than 11 petabytes of data, but much of that is duplicate data, or data that’s flat out wrong. Most professionals move data with the “swivel chair” method, physically moving to another computer terminal instead of comprehensive data integration and management.
Tran is striving to have a 360-degree view into each customer, so with just a few clicks his organization can understand how it’s touching people’s lives. With an organization so large – the second largest department in the U.S. government behind the Department of Defense – the work to reach that 360-degree view had been slow and arduous.
But each step of the way, Tran said, he’s communicating just why his organization is implementing new standards, changing processes and testing new technologies. He said he works with business stakeholders to view the data to understand why it’s valuable to a business process, not assuming the data is valuable in and of itself. By tracking who uses the data at each step of its lifecycle, he’s able to better show off the value of his comprehensive data management plan.
The People Part of the Job
Maria Villar, SAP’s global vice president of data management and governance for the company’s global customer operation, said the realization that communication with peers was the most important part of her job has come only recently. She and her colleagues are often more comfortable with crunching numbers than marketing their products.
“The hard part of the job is not necessarily the technology, or some of the processes, it’s the people,” Villar said. “I’ve had to learn how to communicate the story better to executives and stakeholders.”
That communication not only helps with getting budget and buy-in for a data quality initiative, but by explaining the benefits it makes the analytical reports and data-driven conclusions more believable for the executives in a position to use them, she said.
When asked what skills she wished her data quality team had more of, she replied: marketing. When she had a marketing colleague embedded with her team for six months as part of a company program, “we learned more from her about how to communicate an idea effectively than she probably learned from us.”