Survival Skills for Data Leaders in 2025, and Beyond
Former Mastercard Chief Data Officer JoAnn Stonier summed up the feeling of many as they settled in for the latest Datalogz Future Frontiers event in New York on Feb. 13.
“Welcome to 2025, where every month is its own year.”
Former Mastercard Chief Data Officer JoAnn Stonier summed up the feeling of many as they settled in for the latest Datalogz Future Frontiers event in New York on Feb. 13.
Luckily, attendees received plenty of insight to help data leaders navigate the ups and downs of the year.
Stonier joined former Fortune 500 data executive Guy Lehman and Datalogz CEO Logan Havern to discuss Survival Skills for Data Leaders in 2025, and Beyond.
The wide-ranging fireside chat included discussion on the challenges of managing BI environments, the growth of AI, the importance of governance, and keys to success as a CDO.
Here’s a look at some of the top takeaways:
Explosive Growth in Analytics Means BI Sprawl
While working at JetBlue, Havern saw how quickly BI environments were growing.
Every month, there were more and more reports, dashboards, and other data products. But this environment was unmonitored, presenting all sorts of issues around cost, governance, security, and scalability. Users had questions and issues, and there was no way to get answers to them quickly.
The BI Sprawl that resulted was a major challenge, and Havern started Datalogz to end it — for everyone.
“There was no tool on the market that focused on this problem for BI and analytics,” Havern said. “So we decided to build something to solve it.”
Today, the Datalogz Control Tower integrates with large BI providers like Tableau, Power BI, Qlik, and Spotfire, and identifies issues in BI environments through automated reporting and alerts.
By focusing on the consumption layer, Datalogz is solving problems where most business users interact with data.
From Auditor to CDO of Mastercard
Stonier has had a variety of roles, but data has always run through it.
As an auditor with PricewaterhouseCoopers, she worked with flow charts, controls, and finance.
After earning her law degree, she went to work at American Express Company, where she worked on M&A before rising to become the Chief Privacy Officer.
Stonier understood the company’s operations, and privacy sat at the intersection of three key areas of her background: law, computer science, and data.
Stonier later moved to Mastercard, where she served as EVP Chief Information Governance & Privacy Officer, and built the privacy program, including data processes following the implementation of GDPR.
Like many companies, Mastercard was undergoing a massive transformation for a payment processing company to a data and technology that offered solutions to merchants, banks, government customers, and social impact businesses, while viewing data as an asset. Stonier became the Chief Data Officer, and led development of business intelligence systems, data governance, and data management, among many other roles.
Today, she serves as a Mastercard Fellow, where she applies her expertise in data and AI to advise the firm and engage with external stakeholders.
Along the way, Stonier has established strongly-held principles that served her at each stage.
“I don't think we talked about data back then as its own separate asset, but certainly as a financial professional, I've always thought about assets,” she said. “I've always thought about efficiency. I've always thought about, How can things be done better?”
The Value of Data
When Lehman started his career in data management, it was buried in IT and relegated to a category similar to “janitorial work.” Today, data is essential to companies’ competitive edge, and has a C-suite role in many organizations.
Lehman has been on the forefront of the rise of data and analytics within the Fortune 500, as the Chief Data Officer at Honeywell, leader of Enterprise Data and Enablement at Walmart.
His focus? How to apply the data to help the business.
“I’ve been focused on creating value from data throughout my career,” Lehman said, including, “how to make that happen at scale, and in a way that’s sustainable.”
To Lehman, communicating the value of data is just as important as building it.
“In data, we have many stakeholders. You've got IT, marketing, finance, supply chain, product, legal,” Lehman said. “If you're in a role where you want to transform and make change, you've got to be able to meet people where they are.”
Put data in a business context, communicate in the language of stakeholders, and take opportunities to show how data is revenue generating activity.
One Eye on the Present, One Eye on the Future
To create value, a deep understanding of an organization’s data and a priority list of problems are required.
Enterprises and large organizations must also create data products that provide sustained value over time. To do that, a data leader can’t just build for today, Stonier said. She has to build for tomorrow, and solve problems that people may not even sense yet.
“You have to build things that are going to endure, not for just today. You have to build it so that it's relevant for tomorrow,” Stonier said. “That's a really hard skill set to create, but it will make you so valuable to your firm and your organization. It's closer to what a CEO has to do than people realize.”
Cycles are getting faster. What was once a 3-5 year plan is now a two-year plan. Data programs don’t have to be perfect from the start, but they must have the capacity to adapt. That’s where applying an agile mindset can help.
From Big Data to AI
The words change fast in the data world.
Stonier remembers when the cutting-edge term in data and analytics was big data. Soon, she was offering feedback to a startup at a WeWork coworking space to discuss an AI solution that harnessed natural language processing for document reviews.
Thirty years ago, Lehman was involved with building an analytics system to arbitrage airline ticketing records. It was a classic startup example of smart and curious folks operating without permission to solve a problem they experienced. With success, the team was met with a cease-and-desist from an airline, but the real question they got back was, “How are you doing this?”
“Analytics has been around for a really, really long time,” Lehman said. “The concept of Gen AI is newer and has a lot more capabilities. We've got access now to tools that are extremely powerful. Now, I’m looking at AI and saying, ‘how do I re-engineer my business, and how do I insert these tools in the right place to make that effective?’”
AI could enable massive growth in business intelligence. Havern pointed out that analytics and BI platforms are beginning to add embedded AI tools for the end user.
“Where one BI report creator may have been able to create 10 reports in a month, now we're seeing that they can create 100 reports in a month,” he said.
The growth of AI will only underscore the importance of effectively managing BI environments.
With more data comes more potential for BI Sprawl, as more reports brings more potential challenges with data quality, risk, efficiency, and cost.
AI will make governance more important, not less.
Flipping the Script
As AI brings more data and reporting, there is also more risk of fragmentation in BI environments.
Lehman thinks about a business as an ecosystem, and data and analytics is a supply chain into that.
“Companies can run the risk of creating siloes within that supply chain. You've got data governance over here, you've got the BI organization over here, you've got analytics over here,” he said. “When you think of things like data governance and data management, you should flip the script, and look at it from a point of consumption back.”
Companies can use a tool like Datalogz to plug into BI environments, determine what is used most, and determine which data products should be prioritized. This can become a triggering event to design more optimized flows, and more efficiency around data governance.
As data professionals do every day, start with the output today to make the decisions on the path ahead.
Pay Attention to Your Data
Advances in AI are going to change how we think about what’s possible. But the fundamentals of a data program still matter.
Sound governance. Clean data. Visibility across a data environment. Privacy and risk management.
With more data products, the foundation matters even more.
“ I'm so surprised at how many organizations have not paid attention to their data,” Stonier said. “And then they’re surprised that they haven’t done the basics on their data, and then they’re surprised by the price tag. The logical sequence of getting from A to B is still not clear to many companies.”
AI will introduce even more complexity, even when companies have the building blocks in place. Those who haven’t tackled the first tasks will be two steps behind.
“Do they understand their data? Do they understand the analytics? Do they understand the technology?” Stonier said. “AI governance requires everybody at the table and a certain amount of literacy in an organization that everybody understands the data footprint.”
Before giving everyone a look under the hood, make sure all systems are ready for the word go.
The ROI of AI
What’s the ROI of AI?
At a time of tightening budgets, focus on efficiency, and economic uncertainty, CEOs and boards will be asking about the value of data programs, and taking a closer look.
Stonier offered three key things to consider:
Show the metrics for the AI solutions you’re deploying, and track projects with precision.
Invest in data governance to establish a strong baseline, and understand the origin of all data.
Determine who is responsible for each part of the data program.
At the end of the day, it all has to map to the business.
Domain Knowledge and AI
Generative AI has intriguing capabilities, but it’s not perfect yet. Whether it’s correcting an incorrect citation from ChatGPT or replacing the wrong dataset in a BI report, professionals have a big role to play to make sure the right information gets out.
“Our domain knowledge is more important now than ever,” Stonier said.
An expert can recognize errors, and make corrections.
Professionals also have to prioritize. AI has lots of potential to improve work in many different areas, but it can’t do all of the thinking that is required.
Consider what AI can do to free you up to do analysis, add context, connect trends, and consider what stakeholders will want to see.
At the end of the day, AI is a tool to help you do the work.
Risk Averse or Profit Drivers?
CDOs are walking a tightrope.
They must ensure that data systems run effectively and safely. They must also create value for users, and the business as a whole.
“When we approach a new client or customer at Datalogz, we're very focused on risk averse ROI,” Havern said. “Things like cost savings, How can we clean up tools and make them more efficient? How can we identify your old world compliance risk of GDPR or CCPA? Are there security risks, vulnerabilities in these systems?”
Lehman’s leadership roles at Walmart and Honeywell emerged from needs around risk, but as a C-level executive he quickly identified the importance of aligning with the business strategy. Today, he always pushes to have a revenue target.
The balance is always there, and data leaders should assume that concerns will be raised. That’s what happens when you have a team and stakeholders.
But when she served as CDO of Mastercard, Stonier found that the key to success as a leader is to start from a place of “yes.”
“The best CDOs understand that they are there to navigate the risk, but their job one is to get to that yes. For every person who comes to their office and says, ‘I need data to do this,’ your job is to figure out how and what data you have and whether it's fit for purpose.”
A data leader must determine how to ensure quality, perform analytics, place guardrails and ensure integrity.
Develop strong principles that apply across situations, and the decision will be easier.
Who Owns the Data? No One
“Who owns the data?”
Defining clear roles and ownership of data can be a key task for a CDO, and one reason why it’s a hard job. Today, the average of a CDO is only two years.
Lehman has seen plenty of arguments break around this, where people will declare, “I own the data.”
“No, you don't, because the enterprise owns it,” he said.
“You may be the primary user of the data, you may get first rights in helping me structure it and creating the dictionary, right, you may be the highest ontological user, but you do not own it, the company owns the data. Period. Full stop. It's a company asset.”
There are a lot of reasons that data ownership could be disputed. People may want control, or they don’t want to be held accountable.
But at the end of the day, data is an enabled function. Different people are responsible for its management, and others are responsible for using it.
All of it is a company asset.
Blind Spots
Data blind spots are a major concern for CDOs.
“What can I not see? I can add visibility to data in a lot of places, but when you get closer to the point of consumption, you lose that,” Lehman said.
People add datasets, or they work with unstructured data.
In the process, an organization loses visibility of the data, and it only adds to BI Sprawl.
This can lead to all kinds of problems. That dataset could have stale or duplicated data. It may use a metric that is defined differently elsewhere in the organization. Access may not be properly controlled.
That’s where Datalogz comes in. By integrating directly with tools like Power BI, Looker, and Qlik, organizations can gain greater visibility into all of the data being used across their environment.
“It tells you what is happening at points of consumption that you can't see elsewhere,” Lehman said.
