How to build an analytics-driven organization

BOSTON--What’s the best word to describe building an analytics-driven organization? “The word I would pick is strategy, as you have to be aligned with the corporate strategy or the whole system breaks down,” said Inderpal Bhandari, PhD, senior vice president and chief data officer at Cambia Health Solutions, speaking at the Medical Informatics World Conference on April 29. 

Cambia Health Solutions is a large nonprofit total health solutions company, which is a conglomerate of 22 companies that reaches 100 million people. 

“You can build the best recommendation engine in the world and the best possible algorithm but if the company wasn’t ready to execute all that, it would all be for naught,” Bhandari said.

Thus, strategic alignment between the analytics department and the corporate strategy is critical. From there, a company can develop specific objectives, success metrics and performance metrics. “Then you have the whole organization pushing behind it, and that’s how you do it to scale.”

Bhandari advised the following three approaches to changing corporate culture, which entails “going from a data-as-a-byproduct culture to data-as-an-asset culture”:

• Top-down sponsorship, or executive support and buy in
• Bottom-up stewardship, meaning staff from all departments within a company “have to become stewards of the data"
• Governance to execute strategy

To get there, the first step is reliable reporting (e.g., knowing medication adherence rates among a certain patient population). The second step is action by reaching out to the noncompliant patients to help ignite behavioral change.

Responding to data in such a way is not enough. “If you think about it, the horse already has left the stable. The patient is already non-adherent,” he said.

The organization must take a third step: being proactive. “If you can predict who is going to be non-adherent by the end of the year and act on it, that’s proactive,” Bhandari said. “The whole game is to make the whole thing actionable.”

Challenges like scarcity of talent and interdepartmental tension can thwart effective analytics strategies, however. To counter the cultural challenge, he said, "You have to have a noble cause so people can reflect back on that when making individual decisions.”

Finding the workforce to carry out analytics remains a challenge, as healthcare organizations often need individuals with mastery in three buckets of knowledge: data science, behavioral science and strategy.

“It’s hard to find talent that spans those three easily,” he said, adding that some institutions should structure an educational program to encompass all three areas.