Deriving Value from Data: Insights from Three Data Leaders

Tamr’s Chief Product Officer Anthony Deighton led a lively discussion with three top data leaders: Dan Whitacre, Senior Director, Kroger Labs and Technology Transformation; Todd James, Senior Vice President, Intelligent Automation at Fidelity Investments; and Brett Starr Vice President, Data & Machine Learning Engineering, Intelligent Automation, & Business Intelligence at The Cincinnati Insurance Companies. The topic of the conversation: finding value in data. They discussed today’s top data challenges and trends and provided insights into what’s on the horizon.

The group agreed that every company at its core is a data company, and it’s therefore imperative to get the most value out of data. “Kroger has for a long time been a data company. Ours is a highly competitive industry, and if you want to create an advantage your competitor can’t copy, it has to be hidden from them. Data is our competitive advantage. Any competitor can step into our stores and look at our assortment, but they can’t see your data,” said Whitacre.

Added Starr, “Our #1 asset is our people and our culture and #2 is data. Our biggest challenges are leveraging data from third parties and the explosion of data. We have to be careful about pockets of data that are building up throughout the organization. We all need to be aware of when we need to batten down the hatches and go from being offensive to defensive.”

Fidelity is a data company as well. “We are continuously asking how we can leverage data to improve the customer experience and the workforce experience. It’s incumbent on our analytics and data groups to do it efficiently. We do that by focusing more on the customer need and less on how we’re organized internally,” said James.

Why did data teams become responsible for automation?

Deighton then asked why automation landed with the data team. Starr noted that many of the labor-intensive tasks that needed automation were related to how the insurance company was creating analytics. “Automation and analytics ended up being different sides of the same coin,” he said.

How did data get so messy?

Deighton wondered, if data is so important, then why is it such a mess? Whitacre noted that the expense of storage in the 1980s held the data explosion back, but today, it’s affordable for everybody to have their own data and their own version of the data. “The problem is that there may now be 10 teams using the same data. We’ve ended up with huge amounts of data and enterprise amnesia. It is easy to lose track of what data there is and whether it is fit for purpose,” he said.

James added that whether we’re happy with the state of data or not, the current situation is a product of multiple decisions made across time, to address prevailing needs with the best capabilities and insights available. “If you look back at the problems at hand and the capabilities at the time, everybody was doing the best they could,” he said. Whitacre added, “To some extent, a thousand good decisions added up to a bad one, and in some cases good ones. People are going to be saying the same thing about us in 5 years!”

Whitacre summed it up: “Data is a mess, but so what? The past is past. What’s important is what you do to start managing data as an asset. It’s not about buying new technology; instead, it’s really a cultural campaign to get everyone to care about data and understand why they need to.”

What does the future hold?

Finally, Deighton asked what technologies will make things better (or worse) moving forward.

Kroger is harnessing data in motion and fusing IoT data,  transactional systems, and more in busy store environments.  At Fidelity, a major push is on enabling business to drive analytics and data initiatives, not just IT. “We started with use cases, then moved to systems of use cases, and now we’re getting to capability patterns and ways that we can scale and manage AI. We all need to get to a point where we’re managing sets of reusable capabilities. That’s what is going to have a tremendous impact on the success of analytics and data initiatives.”

Whitacre cautioned that everybody in data must keep risk in mind. “What’s going to go wrong? Will people trust an algorithm, and will the algorithm stay trustable? Will we need to start explaining our algorithms? Will the government start saying that algorithms must be explained?”  Added Starr, “Data is valuable, but it can also be a liability. If unchecked, it becomes a risk and a potential breach that could ruin your good name. Information Lifecycle Management needs to come out of the shadows and back into the forefront.”

Everyone agreed that data must be managed as an asset, especially in light of the fact that all businesses are data businesses.