From security needs to rising costs to data governance and beyond, there is no shortage of challenges for today’s Chief Data Officer (CDO). As new risks and opportunities emerge, current and future data leaders need to be ready.
Recently, Wavicle’s Chief Strategy Officer, Sue Pittacora, sat with leaders from PepsiCo, Global Payments, and Great American Insurance Group for an in-depth discussion on navigating today’s challenges and opportunities and preparing for what the future will hold. Their conversation covered data and analytics priorities; the role of emerging analytics trends and technologies; and how future-focused CDOs are using data to drive growth, success, and innovation.
This panel includes:
Explore the panelists’ perspectives on the evolution of the CDO role and more.
Sue Pittacora: The role of the CDO has changed over the years, evolving from a “defensive” position that was focused on protecting and managing an organization’s data assets to now being able to use that data on the “offensive” through data monetization, AI, and the creation of new data products. In today’s business environment, the role of the CDO isn’t only about data governance, security, privacy, and access – today’s data leaders are focused on creating value. How is your organization using data and analytics to drive innovation and growth?
Iwao Fusillo: The eCommerce CPG marketplace will undergo outsized growth for years to come. That means it’s up to our teams to capture our fair share of that market, and our data strategy is a key part of this endeavor. In order to support the company in this growing market, our data team must focus on three critical components: deriving actionable insights from the rapidly expanding ecosystem of retail media networks, sharing data-driven sales insights with our distribution partners to grow our joint business, and creating predictive analytics to inform our supply chain to minimize stockouts and maximize consumer fill rate. In these ways, our data and analytics practices are crucial to driving forward company growth as the ecosystem evolves.
Rob Golden: We’ve seen data open up new opportunities for growth in insurance as well. For example, I recently worked with a business unit leader for commercial auto insurance to pitch a growth opportunity centered around telematics in auto insurance. By making an investment in telematics data, we’re able to identify risky driving behaviors and provide feedback to drivers in real time, which helps ultimately reduce accidents and claims while improving driver safety. This benefits our customers as well as our company but requires a strong data analytics program to support it.
Sue: There is sometimes a tug-of-war between data security and data democratization. On one extreme, some CDOs feel the need to restrict data access aggressively, and on the other, some CDOs take a much more relaxed approach to give access to non-sensitive data to almost anyone in the organization. It is difficult to find a balance between these extremes, drive buy-in, and manage risks in order to deliver the data the business needs. How have you partnered with your business peers to champion the data initiatives you know your organization needs to invest in?
Dipti Desai: We work closely with business units to enable value creation from data and map related profit or benefits from data and analytics projects. The key is bringing business stakeholders together to share knowledge, ideas, and progress on data initiatives. This helps us identify the most impactful problems to solve and avoid duplicating efforts across business units. Leveraging stakeholders as champions for change management and to set an example for other business units is also powerful.
The balance between risks and value is complicated. You don’t want to overshare information and risk privacy, but you also need to drive insight and further the business. Managing data risks in a sustainable and meaningful way requires a holistic partnership with the business units to cover their data use patterns, product and service roadmaps, and appropriate privacy practices. To do so, we established a pragmatic common ground through an enterprise-level Data Lifecycle Management Council, which helps to manage risk; increase engagement in data projects; and enable the business units to identify impactful business problems, share best practices, and manage change efforts across verticals.
Sue: A successful CDO is, in many ways, the ultimate “Influencer” of an organization, guiding how and where data is used. CDOs are responsible for outcomes but do not always have the authority to force change. How are you able to identify, prioritize, and socialize data investments that demonstrate measurable outcomes for your organization?
Rob: Many data and analytics initiatives, like data governance, can be difficult to measure. To show progress, it’s critical to drive understanding of the data initiatives and why they are important. Then, align projects tightly with core business operating metrics and build around those. I also spend a lot of time with stakeholders to agree on what is most important and where to place our bets. This way, there is broad alignment on big investments and less pressure to show ROI because we’re focused on long-term success.
Dipti: Our enterprise data program is still scaling up, so we try to meet each business unit where they are in their data journey. With some, it’s easier to identify ROI and metrics, but with others, it’s more about establishing a foundation. Measuring ROI can be challenging, but it’s a critical feature. We are working to demonstrate the program’s impact by building a quantifiable pipeline of use cases and impact. We need to shift the perception of data being a cost center to show how we are an enabler of profit. Part of this is telling the story of how the data program supports the business in a tangible way.
Sue: An effective CDO is often a rising tide that lifts all boats – meaning the outcomes of a data-driven organization enable efficiency across every department. Yet many organizations haven’t yet figured out what the CDO’s role within the organization is, and so much has changed in the complexity of our data with new data sources, new ways of working, and higher customer expectations. Even with a push towards self-service analytics, we see new risks as not everybody knows how to properly use and interpret their data. Seeing how difficult the role of CDO is to design, it makes sense that only 35% of major companies report that the CDO role is successful and well-established in their organization. Has this been a challenge in your role? If not, how have you been successful despite these odds?
Dipti: For anyone to be successful in a CDO role, there must be a constant partnership and readiness to stay on the pulse of the business. It’s not always easy to identify the right problems and to keep moving forward. Fostering data literacy and educating key stakeholders helps us continue to increase data programs’ relevance to the business and build success for our programs.
Iwao: The challenge of building success often comes back to the difficulty of measuring ROI. And when we try to measure ROI, sometimes we see diminishing returns. It’s often because of the complexity – we’re seeing more data, in real-time, across many channels. In my past roles at American Express, the NFL, and most recently, General Motors, centralizing data and analytics with a modernized tech stack served us well, but increased data complexity has made centralized data analytics ecosystems too complex for any single team of professionals to drive for the enterprise.
What we’re increasingly seeing is more of a data mesh approach to connect rather than collect data and a democratization of analytics. In recent months, in the industry and at conferences, self-service analytics has been a top conversation point, and this is more of the approach we’re taking today.
Rob: Self-service analytics are increasingly popular, but there are downsides to self-service analytics as well. There are risks for bad code, faulty data, and relaxed quality checks that can lead us to drive faster bad decisions with self-service analytics. This is another place where we have to balance risk and value. You need data governance in place, clear business rules, and training for users. Without a data foundation and data quality operations, self-service analytics can break down quickly.
Sue: As we know, everyone is talking about generative AI and its impact on businesses – both from a standpoint of the opportunities it presents, as well as the risks. It’s top-of-mind for every c-suite and board, and many are asking or pressuring their organizations to lead the chase in generative AI. However, it is also imperative for companies to first establish their policies, guardrails, and procedures for how to use these tools. How are you helping your organization capture the value of generative AI while simultaneously addressing the risks it presents?
Dipti: Requests, use cases, and interest are springing up across all different verticals and partnerships – this is a common theme across industries. Plus, everybody feels like they must do something – nobody wants to miss out on the possibilities AI opens up. But it’s equally important to understand where the risks are and build a foundation for success while experimenting with use cases.
Iwao: I see using this capability in two broad categories. First use cases that drive direct business impact such as optimizing product descriptions or images to increase sales without increasing marketing spend. And second, to drive efficiency by summarizing content, developing consumer insights, and automating code.
There are a lot of companies jumping in with both feet, but others aren’t doing anything at all for fear of what we don’t understand about confidential and internal data. There are a lot of questions about what datasets to use, comfort levels using public and internal data, building your own model, and the related costs. I think there is a necessary middle ground here to at least start experimenting with public data to help data scientists understand the capabilities of generative AI and its limitations while working with legal to build a governance framework.
Rob: There is a lot more non-technical work than I was expecting in this space. We knew there would be technical challenges to leveraging generative AI, but extensive work with legal and information security teams is also necessary for success. Long term, we have to focus on guiding our organizations toward building sustainable foundations. This is a place where creating strong relationships between data, information security, legal, and technical leaders has been critical – we must have data ethics, compliance, and privacy in mind when we consider generative AI.
Sue: Analytics talent over the past several years has been difficult to attain and you have to keep those people challenged and motivated as you get them up to speed. What approach are you taking to attract, retain, and develop your talent base?
Iwao: In the past, I doubled the size of General Motors data and analytics team to approximately 250 people in just 10 months without the use of external recruiters, with a core hub of talent in Michigan while also expanding across the US and offshore. We achieved a 97% retention rate and improved employee engagement scores.
There were three key components to our talent strategy. First was PR and branding. I do a lot of public speaking and personally learn so much from these experiences, but my primary purpose is often to elevate the brand of GM – and now PepsiCo – to be an employer of choice among data and tech professionals. I’ve seen this help acquire and retain talent. Second is a multi-geographic footprint, which I have built in multiple roles. This and hybrid work models are huge accelerants on the talent front and help access and attract the best data talent. Last but certainly not least is compensation. Data and tech talent at the early career point are often being compensated in part with equity, and this is something they look for.
Dipti: There are also internal talent sources to consider. Our data literacy and organizational change management efforts have helped both to establish recruiting processes for data talent and to attract and develop internal resources. We aim to identify individuals in other parts of the business who understand the data ecosystem and then provide them with the right training to build the subject matter expertise from within.
Sue: Clearly, there is a lot on every CDO’s mind with the increasingly complex data environment, constantly evolving data landscape, and challenge of demonstrating success. Thanks to CDO Magazine for providing a venue for this discussion and to Iwao, Dipti, and Rob for sharing their insights and experiences.
Check out the full panel discussion here: