The Data Ecosystem Is a Bunch of Legos — Retail/E-commerce Data and Analytics Leader

(US & Canada) Retail and e-commerce data and analytics leader Veronika Durgin speaks with John Dubois, Consumer and Retail Data and AI Leader at EY, in a video interview about operationalizing data products, adapting to market dynamics, knowledge graphs, and ways to mend the disconnect between data scientists and technologists.

When it comes to operationalizing data in an agile fashion or as data products, there are different methodologies to do that, says Durgin. Highlighting the fight for the right definition of data products, she remarks that both agility and data products are mindsets to her.

Elaborating, Durgin states that when business stakeholders come up with a problem to be solved and the value it can bring, the data team builds a data solution that can solve a multitude of issues in addition to the existing problem.

She gives a “rice cooker” analogy, wherein the rice cooker is a product that can cook multiple other things in addition to rice, and to her, that is like a mindset. Durgin notes that while the deliverable of a data product could be anything based on business needs, the goal is to comprehend the value of solving the business problem.

Next, Durgin concurs with the interviewer when he says that data products should be called “information products.” Stating the difference between data and information, she says that while data is just facts stored somewhere, information is data in an actionable context.

Speaking of the ability to adapt to market dynamics, Durgin states that the field is constantly changing, which is exciting, but keeping up is challenging. She maintains that the data ecosystem is a bunch of Legos with numerous foundations and components.

Further, Durgin stresses that while the foundation of data has not changed, the way data can be used has evolved. With the power of Nvidia, Google, and LLMs, one knows how to use the data better. But, she says, having a solid data foundation makes it easier to apply the new tech advancements to business.

When asked about getting traction around knowledge graphs, Durgin affirms that while she understands it theoretically, she has yet to figure out its organizational implementation. According to her, a knowledge graph is more than just data assets; it is a digital copy of the organization. This makes it challenging to implement it in a way that benefits everyone.

Moving forward, Durgin sheds light on the disconnect between data scientists and technologists. She states that the disconnect can be mended if both parties speak the common language. Adding to that, she urges data scientists and technologists to do it right with generative AI to avoid project failures. She asserts that engineers must find a way to speak to non-engineers.

Furthermore, Durgin suggests data people understand non-technical peers because, to deliver a dashboard, one must know what information people want out of it. It is critical to be able to translate the solution to people who do not care about the chosen ML model, but they do when it predicts an actionable trend.

Thereafter, Durgin recommends data people learn best practices from other groups of people, such as those in marketing or customer service. Consequently, they can serve multiple functions within the data organization without reinventing the wheel.

Summing up, Durgin states that while the cognitive overload of generative AI keeps her up at night, the speed of technological advancements excites her. Concluding, she shares an impactful sentence from a blog that changed her approach to everything. It read, “Do you want to be right, or do you want the best outcome?”

CDO Magazine appreciates Veronika Durgin for sharing her insights with our global community.