From Data Catalogs to Open Standards to Data Products, 2025 Will Be a Year of Change for Data Management

As we move into 2025, there is a clear shift happening in the data and analytics space, building on many of the trends and innovations that defined 2024. Based on the rapid pace of change in the industry, here are my top 4 predictions for the data space in 2025 and some thoughts on how they’ll shape the technology landscape.

1. Iceberg Will Cement Its Role as the De Facto Open Standard

This year will see Apache Iceberg solidify its place as the de facto standard for open data table formats, which was reiterated in Dremio’s annual survey. This isn’t just about Iceberg gaining traction; it’s about a broader shift in the ecosystem. Within 2025, we’ll see the metadata table formats for Iceberg and Delta converge. When that happens, we won’t talk about Delta Lake or Iceberg Lakehouse as separate entities – we’ll simply refer to it as “the Lakehouse.”

This convergence will be driven by the need for interoperability and standardization as organizations tire of vendor lock-in and the friction caused by incompatible ecosystems. With a unified metadata standard, the Lakehouse will truly deliver on its promise of flexibility, performance and cost efficiency across multiple platforms. For enterprises, this means greater interoperability, easier adoption and faster innovation.

2. Catalog Wars Will Intensify

Although the Lakehouse paradigm will unify table formats, the battle for catalog dominance will escalate in 2025. On one side, there are community-driven REST specifications for Iceberg catalogs, and on the other side are proprietary solutions like Databricks’ Unity Catalog. The stakes are high because catalogs are no longer just about metadata storage but also about becoming the central nervous system of modern data architectures.

Governance and business-level metadata are emerging as requirements. With generative artificial intelligence (AI), organizations require more than just technical metadata; they need rich, contextual information that aligns with governance policies and business objectives. This shift will put pressure on catalog technology to evolve. The winners in this space will be those who can seamlessly integrate governance, interoperability and scalability without compromising simplicity.

Instead of data executives lamenting YAC (yet another catalog), there will be a push for convergence as metadata (and its quality) becomes even more valuable than the underlying data.

3. Thanks to AI, Data Products Will Gain Mainstream Adoption

AI continues to dominate the headlines, but the real story in 2025 will be how AI-ready data is reshaping data management (which will include metadata, per the previously mentioned prediction). The concept of data products – packaged, reusable data sets designed for specific business use cases – is gaining momentum. Case in point, AI thrives on high-quality, well-governed data, and data products provide the agility and scale needed to meet this demand.

Interestingly, the push to data products will make DataOps topical again. But this time, data engineers won’t be leading the charge. Instead, analytics engineers will take on this role, bridging the gap between raw data and business insights. Their skill sets – combining technical expertise with a deep understanding of business needs – make them uniquely positioned to create and manage data products effectively. The push toward “self-service” and “autonomous” data management, which makes it simple for the business to manage its own data, will be in full swing by the end of the year.

4. Governed Self-Service Will Be Imperative

The rise of data products brings us to another key trend: governed self-service. In 2025, organizations will grapple with the need to balance agility and governance at scale. AI and data products are driving unprecedented demand for data access, but without robust governance, this can lead to poor outcomes from both risk and revenue perspectives.

Governed self-service will become a cornerstone of modern data architectures because it allows users to access and leverage data independently while ensuring compliance with enterprise-wide governance policies. It’s a win-win: Business users get the agility they need, and data teams maintain control over access and security.

Achieving this balance will require a combination of technology and cultural shifts. On the technology side, we’ll see more integration between catalogs, governance tools and analytics platforms. On the cultural side, organizations will need to foster a shared understanding of data responsibilities (and roles) across teams.

The Path Forward

The data space is never static. It’s an ecosystem that thrives on innovation and adapts to changing demands – 2025 will be no different, and I’m excited to see how these predictions unfold. As we move further into the year, these trends point to a future in which data management becomes more standardized, accessible and aligned with business needs. Iceberg’s dominance, the evolution of catalogs, rise in data products and emphasis on governed self-service are all part of a broader shift toward making data truly AI-ready.

For organizations, the key to thriving in this new landscape will be staying ahead of these trends. Investing in open standards, fostering cross-functional collaboration and embracing roles like the analytics engineer will be critical. Most importantly, it’s about building data architectures that are future-proof, scalable and designed with the end user in mind.