Among the leading cloud database providers, product overlaps are no longer rare.
The latest Magic Quadrant dedicated to this market bears witness to that. Most of the “leaders” (5 out of 9) receive a note on this aspect:
- Alibaba Cloud
Overlap between AnalyticDB and Hologres (analytics) just as between DMS and DataWorks (data integration). - AWS
A wide array of DBMS options and integration choices… at the cost of overlaps and conflicts. - Google
Several PostgreSQL solutions (Cloud SQL, AlloyDB, Spanner) where one must balance trade-offs. - IBM
Overlaps on the data warehouse side, between Db2 Warehouse, Neterra and watsonx.data. - Microsoft
Competition among Azure Synapse, Microsoft Fabric, and Azure Databricks.
Managing costs remains a challenge
Another topic broadly shared among the “leaders”: cost management.
It is difficult at AWS due to the lack of unified pricing across services.
It is also challenging for many Databricks customers, despite advances in FinOps tooling.
At Google, it tends to become more complex with the addition of new features.
For Oracle, customers still complain about pricing and contractual rigour, even if the trend eases with cloud adoption and its consumption-based model.
For Snowflake, Gartner adds a more nuanced view: the user-friendly aspect could foster a “black box” mindset, thereby potentially limiting the ability to optimize workloads.
Several of these providers had already been called out on this topic a year ago, in the previous edition of this Magic Quadrant.
Databricks, due to the difficulty of predicting costs with the consumption-unit model.
Google, because expense tracking could prove tricky, especially when interfacing to databases of services billed by consumption units.
Oracle, perceived historically as a provider with expensive offerings.
Alibaba, where the variety of pricing models, combined with decoupled billing for certain resources in the name of flexibility, could be hard to master.
20 providers, 9 “leaders”
Year after year, the criteria remained largely the same. It was still necessary to cover at least one use case among:
- Transactional
- “Light” transactions (high-volume competition and low latency)
- Application state management
- Data warehouse
- Lakehouse
- Event analytics
Once again, Gartner evaluated only managed offerings delivered in public or private cloud. It did not take into account databases hosted on IaaS.
The 20 ranked providers are the same as last year. And the 9 “leaders” from the previous year remain in place. In alphabetical order: Alibaba Cloud, AWS, Databricks, Google, IBM, Microsoft, MongoDB, Oracle and Snowflake.
On the “Execution” axis, reflecting the ability to respond to demand, the situation is as follows:
| Rank | Provider | Year-over-year change |
| 1 | AWS | = |
| 2 | = | |
| 3 | Microsoft | + 1 |
| 4 | Oracle | – 1 |
| 5 | Databricks | = |
| 6 | Snowflake | + 1 |
| 7 | MongoDB | – 1 |
| 8 | IBM | + 2 |
| 9 | Alibaba Cloud | – 1 |
| 10 | InterSystems | – 1 |
| 11 | Huawei Cloud | = |
| 12 | SAP | = |
| 13 | Teradata | = |
| 14 | Cloudera | = |
| 15 | Couchbase | + 3 |
| 16 | SingleStore | + 1 |
| 17 | EDB | + 3 |
| 18 | Redis | – 3 |
| 19 | Neo4j | – 3 |
| 20 | Cockroach Labs | – 1 |
On the “Vision” axis, reflecting strategies:
| Rank | Provider | Year-over-year change |
| 1 | = | |
| 2 | Databricks | + 3 |
| 3 | Microsoft | – 1 |
| 4 | Oracle | – 1 |
| 5 | AWS | – 1 |
| 6 | Snowflake | + 2 |
| 7 | Alibaba Cloud | + 3 |
| 8 | IBM | – 1 |
| 9 | SAP | – 3 |
| 10 | Teradata | – 1 |
| 11 | MongoDB | = |
| 12 | Cloudera | = |
| 13 | InterSystems | + 2 |
| 14 | Neo4j | = |
| 15 | Huawei Cloud | + 1 |
| 16 | EDB | + 4 |
| 17 | Couchbase | = |
| 18 | SingleStore | = |
| 19 | Redis | – 6 |
| 20 | Cockroach Labs | – 1 |
Alibaba Cloud, distinguished for its “data + AI” approach…
The main Alibaba Cloud offerings in this market are PolarDB and ApsaraDB (transactional), AnalyticDB and MaxCompute (analytics), Tair and Lindorm (key-value).
Last year, the group was praised for its substantial and differentiated sector presence, the development of its partner ecosystem, and the weight of its open source community.
This year, Gartner appreciates pricing, deemed attractive. As well as the reliability of its serverless architecture. Now extended to all DBMS, it stands out for its architecture that decouples compute, memory and storage in a hybrid environment. A strong point also for the “data + AI” approach, enabling development and deployment of applications using only Alibaba Cloud technologies.
… but not for PolarDB configuration
Last year, Gartner had flagged, beyond cost management, geopolitical risk associated with Alibaba Cloud. It also noted the limited availability of its services outside Asia (fewer regions and availability zones than competitors).
This year again, the limited presence outside Asia is highlighted. It may translate into fewer third-party tool integrations and English-language resources (documentation, training, support). Buyers should also watch PolarDB’s configuration, deemed complex for new users, particularly regarding cost/performance balance and multi-layer storage management. Add to that the product overlaps discussed above.
AWS has an unmatched catalog…
Aurora, Redshift, DynamoDB and SageMaker are among AWS’s core offerings in this market.
Last year, Gartner praised AWS’s functional coverage and its ability to bind its solutions. It also noted the breadth of partnerships and geographic presence.
That point remains valid and is complemented by a solid track record of infrastructure availability and a proactive approach to client dialogue for cost optimization. Overall, AWS offers an unparalleled catalog in this market, with SageMaker as the central governance hub for data/AI.
… but dependencies remain for hybrid orchestration
The integration between AWS services can be complex, Gartner noted last year. The firm also observed that hybrid/multicloud deployments were limited despite native connectors and support for engines like Spark (customers tended to use third-party orchestrators).
This finding remains accurate: many customers rely on third-party solutions for hybrid/multicloud orchestration. Added to that are the two previously mentioned issues: difficult cost management and product overlaps.
Databricks, rapid for rapid innovation…
Beyond the Data Intelligence Platform (which includes Unity Catalog), Databricks offers data warehousing with Databricks SQL, transactional capabilities with Lakebase, as well as integration and engineering with Lakeflow.
Last year, Gartner praised investments in GenAI (including MosaicML acquisition), reflected in the development of its own LLM. It also highlighted the Unity catalog (recently open-sourced) and the Delta Lake format (competing with Iceberg).
This year, Databricks is lauded for its “lakehouse” vision, though it is no longer alone in this space. It is also praised for its speed of innovation—with Agent Bricks delivering significant features almost monthly, the Tabular acquisition supporting Iceberg across the portfolio, and the introduction of low-code capabilities in Lakeflow. It is also commended for its commitment to open standards (Delta Lake, Iceberg, Spark, PostgreSQL…) which promote portability.
… but not that easy to get started
Last year, Gartner pointed to a lack of UI intuitiveness, frequent changes, scant documentation, and limited low-code capabilities. It also flagged FinOps as a concern, noted above.
This year, the firm tempers its openness narrative: some customers worry about potential lock-in at the orchestration and Delta Live Tables (now Lakeflow Spark Declarative Pipelines) level. It notes that customers tend to require a high level of technical proficiency to use the solution. At the same time, FinOps remains a valid concern (see above).
Google, well positioned on AI…
Among other products positioned in this market, Google offers Spanner, BigQuery, AlloyDB, Cloud SQL, Firestore, Memorystore and Bigtable.
Last year, Gartner praised open-source contributions (notably to PostgreSQL). It also highlighted progress in GenAI (Gemini integration and cross-cutting support for vector search via LangChain) and the unified data/AI foundation with Dataplex for governance.
This data/AI foundation again earns high marks; broadly, the capability of Google’s DBMS offerings to cover AI agentic use cases. Gartner especially appreciates the breadth of data models supported by Spanner (relational, key-value, graph, vector).
… but less strong on data sharing
The partner network still needs to mature, Gartner noted last year. It also pointed to FinOps and noted Google offered fewer native options for app integration and master data management than its rivals.
This year, in addition to cost management and the overlaps mentioned above, a point of caution concerns the data marketplace and data-sharing capabilities. They appear less advanced than some competitors, despite improvements in data clean rooms and cross-cloud interoperability.
IBM expands its multi-cloud presence…
The main IBM DBMS offerings in the cloud are Db2 (transactional + analytical) and watsonx.data (lakehouse).
Last year, Big Blue stood out for its sector-specific strategy (solutions tailored to governance, security and compliance). It also highlighted its ability to combine open-source and data-management expertise for hybrid deployments. Its offering is well suited to mission-critical applications, Gartner added.
This year again, the sector-specific strategy is praised. The extension of its cloud footprint is also noted (Db2 available with hyperscalers and the acquisition of DataStax, which has a strong multicloud footprint). A positive point is also IBM’s clearly defined approach to integrating DBMS into data-management frameworks.
… but still struggles to get its message across
IBM has difficulty differentiating itself in communications, and the messaging varies across sales teams, Gartner noted last year. It also reminded readers that geographic deployment of the offering did not yet match that of other hyperscalers.
Communication difficulties persist, leading to some brand unfamiliarity in the segment. At the same time, IBM is still perceived as a “legacy” vendor, which can deter buyers. Gartner adds, as noted above, the overlaps between certain products.
An exhaustive offering from Microsoft…
Microsoft’s presence on this market includes Azure SQL Database, Azure Database for PostgreSQL and MySQL, and Azure Cosmos DB.
Last year, Gartner praised the breadth of the offering and the level of integration with other Microsoft services. It also appreciated the opportunities for AI usage in data management. And the progress on multi-cloud management, exemplified by the Azure-Oracle interconnection as well as OneLake shortcuts for federated analytics.
This year again, the breadth of the offering is a strong point, capable of supporting almost all data models and industry use cases. Microsoft’s commitment to PostgreSQL is also praised, along with innovations in the AI area (in-database embeddings, vector indexing, Copilot-Fabric integrations, etc.).
… but the Fabric offering still lacks maturity
Product overlaps were flagged last year, in addition to concerns about the durability of Azure Synapse Analytics and Azure Database in the face of Microsoft Fabric. The latter still lacked maturity, Gartner explained: integration, governance and metadata management capabilities were not as robust as those of other leaders. Deployment could also be complex, particularly for DR, security and cost management.
Beyond the overlap of certain products, Gartner again notes the lack of maturity of Microsoft Fabric. Customer concerns touch both data warehouse functions and governance, including sovereignty, resource sizing, price, metadata management and data quality. Also watch the investments to integrate transactional data into Fabric: in the short term, they may create performance issues.
MongoDB remains a standard for the document model…
Beyond its community edition and on-premises product (Enterprise Advanced), MongoDB offers its Atlas DBMS on AWS, Google and Microsoft.
Last year, Gartner praised a well-regarded offering for high-volume processing, elasticity and the flexibility of the schema. It also highlighted the ease and speed of deployment, contributing to developers’ popularity.
This remains true and continues to yield a large pool of skills. Added to that is the richness of deployment options, reinforced by a robust partnerships program. MongoDB has, more broadly, established a de facto standard for those seeking a document-oriented model.
… but lacks a storytelling around transactional-analytical convergence
If MongoDB combines transactional and analytical capabilities, its offering remains non-relational, Gartner noted last year. Competitors include DBMS providers that bundle document-oriented approaches with other models, as well as those offering MongoDB compatibility.
This observation on rising competition remains valid. The firm adds the learning curve required to master MongoDB’s model. It also notes the lack of a complete storytelling for integrating transactional and analytical workloads.
Oracle, praised for its functional richness…
Autonomous AI Lakehouse, Autonomous JSON Database and Exadata Database Service are among Oracle’s cloud DBMS offerings.
Last year, Gartner lauded the breadth of the offering (functional richness + support for data models and the lakehouse architecture). It also highlighted strong multicloud management (Database@ + interconnection with major hyperscalers) and the ability to rapidly deliver innovations (GenAI, low code, RAFT consensus).
This year again, the functional richness is praised (distributed databases, vector search, agentive framework…). The diversity of deployment options is also noted. Oracle’s suitability for mission-critical applications is reaffirmed.
… but limited adoption for lakehouse deployments
Oracle remains perceived as expensive and has work to do in “cloudifying” its customer base, Gartner noted last year. It also urged buyers to carefully interpret the “one database for everything” approach and what it implies for feature delivery.
This note is reiterated: be vigilant about this approach, which runs counter to architectures that combine DBMSs with data-management systems. The price issue—reiterated above—remains sensitive, and customers continue to favor competing products for lakehouse deployments.
Snowflake has improved its functional coverage…
Last year, Snowflake stood out for its UI adapted to different user profiles, its support for multiple formats on the storage layer, and its expansion of the lakehouse architecture with Iceberg and Polaris.
This year again, Gartner gives a thumbs-up to the UI. It also notes the expanded functionality (advanced data engineering via Openflow, ML/AI with Snowpark and Cortex AI, PostgreSQL support via the Crunchy Data acquisition). It highlights improved scalability with second-generation data warehouses (a better cost-to-performance ratio than gen 1 for complex workloads).
… but remains focused on batch and analytics
Last year, Gartner pointed to limited support for hybrid scenarios. It added concerns about data sharing across Snowflake’s customer organizations and usability challenges stemming from on-premise storage integration via external tables.
Both issues persist. On one hand, performance differs between external tables and native storage or Iceberg tables. On the other hand, sharing requires careful planning of permissions, re-sharing, and regional restrictions. Gartner adds FinOps again (see above). And the architecture remains focused on batch and analytics rather than transactional or real-time workloads (even though there are hybrid tables and advanced PostgreSQL integration).
