(US & Canada) VIDEO | Data Function Cannot Be the Lone Driver of AI Use Cases — Former Fitch Group CDO

Heidi Lanford, Former Chief Data Officer of Fitch Group, Inc. speaks with Amy Horowitz, VP of Data Governance Solution Sales at Informatica, in a video interview about data literacy, AI ownership, aligning use cases with business goals, balancing data democratization with governance, and strategizing to rebrand data governance as an enabler and not a roadblock.

Lanford begins by reiterating that data literacy is everybody’s job and not just the data organization’s. The sense of ownership is shared across the C-Suite, the top-level management, and the rest of the organization.

The data function cannot be a lone driver of AI use cases unless those are generated by cross-functional teams who have vested interests in the outcome, says Lanford. She believes that CDOs experienced in AI and analytics should champion this, but they are executive sponsors along with their peers. Lanford states that AI is pervasive throughout an organization and is not owned by the data function.

Moving forward, Lanford opines that it is critical to understand how the use cases are aligned with business goals and unless that happens, organizations will face a lack of adoption.

When asked about balancing between data democratization and its responsible usage, she stresses building a massive advocacy program throughout an organization. Lanford continues that it will vary depending on the organizational size.

In addition, she recalls having a vibrant data and analytics community at a prior workplace, wherein her role involved making it easier to share data sets. Lanford mentions that many organizations start developing technical expertise at department levels by hiring data engineers, giving them development guidelines, and setting guardrails is critical.

For instance, a “how to” guide is provided if one wants more data sources in the data lake, she adds. While there will be checkpoints, says Lanford, it is ensured that the guardrails and guidelines do not slow down the process or aid in building Shadow IT.

With the increasingly abundant data, a data leader must seed a sense of control and think out of the box to do it in a much more federated and collaborative way. She states that technology for the most part is helpful and it is the time to embrace and use it to the maximum.

To turn data governance from a roadblock to an enabler, a rebranding of the term data governance is necessary, says Lanford. She notes that one has to come up with another term in place of governance.

Furthermore, Lanford mentions following the 80-20 rule. She focuses on how frequently certain tidbits of data can become data elements, data accuracy, and whether the right data is being used. If it becomes a pain point for users, she sticks to telling a story about definitions of new versus dormant customers.

Elaborating further on the story aspect, Lanford explains how in the sales world, there are different definitions around the renewal period of a dormant customer. This affects the organization in ways such as incentive bumps for sales to recruit new customers.

Adding on, she shares how the organization had to pay salespeople more because of the lack of an agreed-upon definition of when a customer became dormant. This also impacted the reports about new customer acquisitions.

As the definitions change based on data sets, Lanford believes that it is critical to resolve the issue and ensure everyone is on the same page.

In conclusion, she states that it is up to the cross-functional team, and having a data stewardship community sitting underneath the data governance function. She is positive that the data stewards can opine and weigh in on the important requirements making data governance more tangible.

CDO magazine appreciates Heidi Lanford for sharing her data insights with our global community.