Source: Tom Werner/DigitalVision via Getty images.
Results from a survey conducted by 451 Research from S&P Global Energy Horizons reveal that a significant transformation is underway, driven primarily by the integration of generative AI. Organizations are no longer looking to simply store and retrieve information — they want converged platforms that can simultaneously handle operational transactions, power real-time analytics and integrate seamlessly across heterogeneous environments, all while enabling AI workloads.
This survey reveals a market at an inflection point. The traditional boundaries between operational and analytical systems are dissolving, as 87% of organizations now prioritize platforms that support both workload types. Meanwhile, GenAI has rapidly evolved from experimental technology to an essential capability, with 85% of organizations rating GenAI capabilities as critical to their platform selection.
The Take
The database market is rapidly evolving, integrating and catering to AI and agentic-centric workloads and applications. As organizations evaluate data platform systems, they are no longer evaluating databases as isolated persistence layers but as intelligent, integrated platforms that unify transactions, analytics and search while incorporating GenAI functionality. GenAI is driving this evolution, with many organizations stating they are using GenAI to make strategic business decisions as well as for numerous database-related activities, from query generation to database tuning to data analytics. Indeed, we are moving from a database mindset for AI to an AI database mindset. This change, however, continues to place pressure on database vendors to provide architectural flexibility — particularly by supporting open table file formats that leverage cloud object storage, which is a high-level organizational requirement. Overall, these results point to a data platforms market coalescing around platforms that value data accessibility, AI-driven interaction, interoperability and workload aggregation as key design elements.
Summary of findings
GenAI access leads data platform evolution. GenAI capabilities are moving from experimental to essential platform features and represent a fast-moving technology trend that is fundamentally reshaping how users interact with data platforms. About 80% of organizations rate enabling access to large language models and embedding models as important (35% cite as very important; 45% as somewhat important). Additionally, support for vectors that underpin semantic search, retrieval-augmented generation and other GenAI workloads is considered very or somewhat important to nearly three-quarters of respondents. Organizations that fail to integrate GenAI capabilities risk lagging competitors that can leverage these technologies for enhanced productivity and insights.
GenAI is being integrated into and becoming integral to data platform workflows. Specifically, organizations use GenAI to assist with tasks related to their data platform workflows. About a third of organizations report that data analytics is fully integrated, with a mean score of 3.65 on a 0-5 scale. This is followed by 30% of organizations that report full integration of GenAI for data quality, with a mean score of 3.5. GenAI-related activities with a mean score below 3.5 include using GenAI for query generation (3.45), database troubleshooting/tuning (3.37) and database schema design (3.31).
Security remains top of mind. Security continues to be a critical factor for database platform evaluation, with organizations also indicating that security is both a priority and a challenge in their data platform environments. About 57% of organizations rate it as critically important (5 on a 0-5 scale), with an additional 24% rating it very important (4). Despite this priority, 75% of organizations report moderate-to-significant ongoing security challenges in operating their current database platforms.
Unified operational-analytical workloads are broadly accepted. The convergence of operational and analytical workloads represents a fundamental architectural shift in data platforms, suggesting near-universal importance among organizations and indicating that hybrid processing is not a niche requirement. About 87% of organizations consider platforms supporting both operational and analytical workloads important (44% rate this capability as very important; 43% as somewhat important). Only 13% view this hybrid capability as unimportant.
Organizational search is having a moment. Organizational search is undergoing a bit of an evolution, driven largely by the integration of GenAI, which combines semantic and lexical search, also known as hybrid search. More than half (51%) of organizations report having hybrid search operational within their organizations, with another 29% stating they plan to deploy within the next 12 months. Organizations report using it most often for search and discovery tasks in security analytics (such as security information and event management). Leading benefits include fast access to data and search relevance, along with cost and security; however, cost and security are also noted as leading challenges.
Adoption of open table formats accelerates. Support for open table file formats (Apache Iceberg, Delta Lake, Apache Hudi) has become a significant platform requirement. Three-quarters of organizations consider this capability important (33% rate it as very important; 42% as somewhat important). Only 25% view open table file format support as unimportant. A fundamental shift is underway in the market, with organizations favoring vendor-neutral, interoperable data architectures that reduce lock-in and enable greater flexibility. This suggests that organizations are prioritizing portability and avoiding proprietary formats that could limit future technological choices.
Cloud object storage integration is becoming the standard. The ability to store, access and process data in cloud object storage has become a baseline platform requirement, with 87% of organizations rating this capability as important (45% very important; 42% somewhat important). Only 13% consider cloud object storage integration unimportant. Cloud object storage is increasingly seen as an expected feature rather than a differentiator among many organizations. Its near-universal importance indicates that cloud-native architectures are now the default assumption, and platforms lacking robust object storage integration are effectively disqualified by most organizations. The low cost and flexibility in data access are often the key reasons organizations favor this capability.
Nuances emerge for data platform cloud deployment trends. Where organizations choose to run their data platform systems is less of a binary choice. Organizations consider several factors, such as workload type, where those workloads reside and whether the organization is considering moving existing or net-new workloads. Where workloads run (either in the cloud or on-premises) is a deciding factor for organization when deciding whether to move existing or net-new workloads. Roughly half of all organizations plan to keep workloads in their current location, which is consistent for both operational and analytical workloads. However, for both operational and analytical use cases, moving workloads, including net-new workloads, to a cloud environment is still the preferred strategy.
What is an AI Datacenter?
Want insights on data analytics trends delivered to your inbox? Join the 451 Alliance.
