As one of the world’s leading mobility companies, Michelin operates in over 170 countries and produces more than 180 million tires annually. The French multinational is driving innovation through a powerful data and AI strategy embedded across every layer of the business.

In this second installment of our three-part interview series, Ambica Rajagopal, Group Chief Data and AI Officer at Michelin Group, speaks with Julian Schirmer, Co-Founder of OAO, about the next phase of Michelin’s transformation: building enterprise-scale data products that unlock business value, at speed and with precision.

While part one explored Michelin’s core priorities and how AI is reshaping customer engagement and internal operations, this conversation goes deeper into execution. Rajagopal outlines how Michelin is redefining data as a business lever — moving beyond buzzwords to build reusable, scalable data products tied directly to enterprise goals. From fostering a culture of empowered data ownership to enabling cross-functional collaboration through a data mesh, Michelin is setting a standard for turning data into impact.

Edited Excerpts

Q: Can you start by explaining what a typical data product looks like at Michelin and how that ties into your broader data and AI strategy?

When a particular technology is high on the hype cycle, the deeper questions tend to get less airtime. AI that is core and useful to any enterprise and for it to work, the enterprise needs to have data of sufficient volume, variety, and quality to train the models. Beyond that, to move past the proof-of-concept phase and create a broader impact, the data must be available at scale.

For distributed enterprises like Michelin, with multiple lines of business, global presence, and various ways of interpreting the same data sets across diverse systems and applications, this can become a daunting task. So here’s the approach we’ve taken:

We’re now thinking about data as an operational lever for business leaders. A key part of this is introducing the concept of a data product. So, a data product is a data set that can serve multiple purposes or generate multiple insights.

For example, we might look at volume and price data from a specific geography. This data includes actual demand, the types of products sold, seasonal sales fluctuations, and associated price variations. Taken together, this becomes a data product that enables relatively accurate demand forecasting.

In partnership with the business, we’ve developed a strategy centered on identifying such critical data products, ones the business sees as necessary to generate the right insights over the next few years.

Q: Often, organizations talk a lot about data, but it stays confined to the tech domain and disconnected from business strategy. How have you addressed this at Michelin to ensure data is seen as a business priority?

We’ve changed how data is perceived over the last few years. Traditionally, data is viewed as a technical topic, and the questions being asked are: How much data are we collecting? When will it be available in the data lake? Will it be in real time?

While these are all important and valid questions and we’ve made the investments to support them technically, the more fundamental question is: Is this the right data for the use I have in mind? Will it give the kind of strategic insight a business leader needs to steer their operations, understand their customers, or deliver quality?

These are the kinds of questions we’ve brought to the forefront. We’ve worked closely with business leaders at Michelin to define their core objectives for the next 3 to 6 years. From there, we trace backward to identify the data sets and data products that will make those insights possible.

This change in thinking has allowed us to build targeted data products and it’s clear that data-driven transformation is well underway.

Q: One common issue is that business leaders hand over ownership to data teams too early, before prioritizing what’s important or understanding what qualifies as a valuable data product. How do you make sure the business stays fully engaged until something truly meaningful is delivered?

What you’re describing is really about the link between the business and technical sides of data. It’s an organizational question: Where should this link live, and what role should it play?

To address this, we’ve built and matured a network of data owners. These are empowered, strategic roles responsible for identifying the right data products within a particular business domain, and for ensuring these products are built and maintained to serve evolving business needs.

I strongly believe in the power of community, bringing people together to work through complex challenges like building a data culture. Empowered participants in this network collaborate, connect, and co-create solutions.

This approach is essential to complement any top-down strategy. Our data owner network has become the engine for driving data product development and building the broader data culture at Michelin.

Q: Putting data product owners close to the business is clearly critical. But how do you avoid fragmentation and ensure that use cases are scalable across the organization?

This is exactly where the network effect becomes valuable.

At Michelin, our data owners are not working in silos. They’re part of an organizational network that builds the data mesh, a shared structure of interoperable data products. But beyond just creating the mesh, we’ve also invested in clearly defining what a data product is.

Anyone at Michelin can go to a central platform and view the data products already available in specific domains. In addition, data owners continuously exchange information, collaborate, and build connections across teams.

Over time, this has strengthened the pathways for knowledge sharing, insight generation, and organizational alignment. It’s become the way we work and it’s how we ensure data product consistency and scalability across the enterprise.

CDO Magazine appreciates Ambica Rajagopal for sharing her insights with our global community.