As organizations invest heavily in data, analytics, and AI, a persistent challenge continues to hold many initiatives back: adoption. While companies may build sophisticated platforms and models, only a small fraction of enterprise analytics initiatives become embedded in everyday decision-making.

For global consumer goods leader Mars, solving this adoption challenge is central to how the organization approaches people analytics.

In the final installment of a three-part interview series, Mars’ Ujjwal Sehgal, Global Head of People Analytics, and Rachel Belino, HR Data Officer, explain how the organization is focusing on usability, trust, and governance to drive meaningful adoption of analytics tools across the business. The conversation is moderated by Shachin Prabhat, Vice President and Business Unit Head at Tiger Analytics.

Part 1 explored the realities of bridging the data-to-insight divide inside one of the world’s largest consumer goods companies. Part 2 examined how Mars is scaling reusable data assets, balancing short-term value with long-term data foundations, and preparing for the rise of agentic AI.

In this final discussion, the leaders address one of the most critical questions facing modern data organizations: how to ensure analytics solutions are actually used.

Why adoption often fails

Despite advances in data engineering and AI, many analytics initiatives struggle to gain traction with business users. According to Sehgal, adoption challenges often stem from multiple recurring issues. Common barriers include:

1. Solutions designed for technical users, not business users

Development teams often build tools that feel intuitive to engineers and data scientists but overwhelming to business stakeholders. “Sometimes, from an engineering, data science, or solutioning perspective, we get so caught up in what we are building. Because we live and breathe that solution, we find it very easy to navigate through it. But a brand new user who’s probably not as technically savvy struggles with it,” Sehgal explains.

2. Complex dashboards that are difficult to navigate

When BI tools require multiple filters, selections, or complex navigation, users may abandon them altogether.

“If I have built a dashboard or a BI product and the user struggles with what filters they have to run or what selections they need to make, it puts them off. They think, ‘This is so complex, it’s very difficult for me to even get the answer I need.’ Accordingly, adoption starts going down,” Sehgal says.

3. Insights that arrive too late

In fast-moving business environments, timing matters as much as accuracy.

“Often, I need to make decisions quickly. If I’m unable to get easy-to-read data on time, adoption can be very low because the business has already moved on,” Sehgal notes.

4. Lack of trust in the underlying data

Even well-designed analytics tools will fail if users do not trust the quality of the data. “From a data quality and accuracy standpoint, it is critical because these are foundational things. Many analytics teams struggle to provide data that is clean, usable, and something the end user can trust. Trust leads to adoption,” Sehgal says.

Sehgal also suggests that emerging technologies may also help address some of these adoption challenges. Generative AI, for example, has the potential to simplify how employees interact with analytics tools by surfacing insights directly with business context rather than requiring users to manually explore dashboards.

Protecting sensitive workforce data

For Belino, trust within Mars’ people analytics function extends beyond data accuracy to data protection, especially when working with sensitive employee information.

“When it comes to data that is sensitive and highly confidential, there could be a challenge in users trusting that they can put this data into an LLM to generate content,” Belino says.

People analytics teams handle information about employees that must be safeguarded carefully. As organizations experiment with AI tools, ensuring that these systems do not expose confidential data becomes a critical governance priority.

“We pride ourselves on our ability, from a data governance perspective, to put the protection of the data first and foremost,” she explains.

Belino urges organizations to establish guardrails to ensure sensitive information remains protected before integrating GenAI into analytics workflows.

“Everything is moving fast, but it’s always best to take a pause and figure out whether you have the security and guardrails in place. Our employees are entrusting us with their data. We all have a responsibility to protect it,” Belino shares.

Building the foundations for AI-ready data

A recurring theme throughout the discussion was the importance of building context-rich data environments supported by common data models and semantic layers. These foundations help ensure that analytics tools and AI systems can interpret data consistently across the organization.

For organizations beginning their analytics transformation, Sehgal outlines several fundamentals to keep in mind:

1. Build teams with both technical and business expertise

Successful analytics initiatives require more than technical talent. Teams must also understand the business context in which insights will be applied.

“First, having the right type of people is important in terms of the technical capability, the functional capability, and the ability to truly understand the business.”

2. Define the problem before building the solution

Many organizations rush into building dashboards or models before clearly defining the problem they are trying to solve.

“Many times, we spend more time on the problem statement than on solution development. The reason is that a well-defined problem statement makes it very easy to identify the solution around it. It also helps you get clarity and pushes the business to think through what they are asking for.”

3. Design solutions with the user in mind

Beyond technical capabilities, Sehgal emphasizes that analytics teams must develop empathy for the end users who will ultimately rely on these tools. Understanding how business leaders make decisions and consume information helps ensure that analytics tools integrate naturally into everyday workflows.

“Have empathy for the end users.”

4. Tell a clear story and communicate value in simple terms

Analytics solutions deliver value only when users understand how to apply them.

“Once I’ve built a solution, how do I tell that story? How do I make users clearly understand what it is and how it will help them? A lot of projects fail because they were unable to clearly contextualize and explain the solution to the end user. The ability to tell that story in a cohesive and easy-to-understand manner is vital.”

CDO Magazine appreciates Ujjwal Sehgal and Rachel Belino for sharing their insights with our global community.